Cytokines, Stress, and Depression

CYTOKINES, STRESS, AND DEPRESSION ADVANCES IN EXPERIMENTAL MEDICINE AND BIOLOGY Editorial Board: NATHAN BACK, State U...

Author: Robert Dantzer | Raz Yirmiya | Emmanuelle E. Wollmann

65 downloads 965 Views 29MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

CYTOKINES, STRESS, AND DEPRESSION

ADVANCES IN EXPERIMENTAL MEDICINE AND BIOLOGY Editorial Board: NATHAN BACK, State University of New York at Buffalo IRUN R. COHEN, The Weizmann Institute of Science DAVID KRITCHEVSKY, Wistar Institute ABEL LAJTHA, N. S. Kline Institute for Psychiatric Research RODOLFO PAOLETTI, University of Milan

Recent Volumes in this Series Volume 454 OXYGEN TRANSPORT TO TISSUE XX Edited by Antal G. Hudetz and Duane F. Bruley Volume 455 RHEUMADERM: Current Issues in Rheumatology and Dermatology Edited by Carmel Mallia and Jouni Uitto Volume 456 RESOLVING THE ANTIBIOTIC PARADOX: Progress in Understanding Drug Resistance and Development of New Antibiotics Edited by Barry P. Rosen and Shahriar Mobashery Volume 457 DRUG RESISTANCE IN LEUKEMIA AND LYMPHOMA III Edited by G. J. L. Kaspers, R. Pieters, and A. J. P. Veerman Volume 458 ANTIVIRAL CHEMOTHERAPY 5: New Directions for Clinical Application and Research Edited by John Mills, Paul A. Volberding, and Lawrence Corey Volume 459 IMPACT OF PROCESSING ON FOOD SAFETY Edited by Lauren S. Jackson, Mark G. Knize, and Jeffrey N. Morgan Volume 460 MELATONIN AFTER FOUR DECADES Edited by James Olcese Volume 461 CYTOKINES, STRESS, AND DEPRESSION Edited by Robert Dantzer, Emmanuelle E. Wollman, and Raz Yirmiya Volume 462 ADVANCES IN BLADDER RESEARCH Edited by Laurence S. Baskin and Simon W. Hay ward Volume 463 ENZYMOLOGY AND MOLECULAR BIOLOGY OF CARBONYL METABOLISM 7 Edited by Henry Weiner, Edmund Maser, David W. Crabb, and Ronald Lindahl Volume 464 CHEMICALS VIA HIGHER PLANT BIOENGINEERING Edited by Fereidoon Shahidi, Paul Kolodziejczyk, John R. Whitaker, Agustin Lopez Munguia, and Glenn Fuller

A Continuation Order Plan is available for this series. A continuation order will bring delivery of each new volume immediately upon publication. Volumes are billed only upon actual shipment. For further information please contact the publisher.

CYTOKINES, STRESS, AND DEPRESSION Edited by

Robert Dantzer INRA-INSERM Bordeaux, France

Emmanuelle E. Wollman CNRS Paris, France

and

Raz Yirmiya The Hebrew University of Jerusalem Jerusalem, Israel

Kluwer Academic / Plenum Publishers New York, Boston, Dordrecht, London, Moscow

Proceedings of an International Congress on Cytokines, Stress, and Depression, held May 15-16, 1998, in Roscoff, France ISBN 0-306-46135-8 ©1999 Kluwer Academic/Plenum Publishers, New York 233 Spring Street, New York, N.Y. 10013 10

9 8 7 6 5 4 3 2 1

A C.I.P. record for this book is available from the Library of Congress All rights reserved No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise, without written permission from the Publisher Printed in the United States of America

PREFACE

Cytokines had been characterized in the early eighties as communication molecules between immune cells, and between immunocytes and other peripheral cells, such as fibroblasts and endothelial cells. They play a key role in the regulation of the immune response and the coordination of the host response to infection. Based on these biological properties, nobody would have predicted that one decade later cytokines would burst upon neurosciences and permeate into several avenues of current research. In neurology, the connection between cytokines and inflammation, and the demonstration of a pivotal role of some of these molecules in cell death by apoptosis, prompted the investigation of their involvement in several neurological diseases involving an inflammatory component, including multiple sclerosis, brain trauma, stroke, and Alzheimer's disease. This movement started in the late eighties, and the corresponding field of research, known as neuroimmunology, is presently booming. In psychiatry, however, the relationship between cytokines and mental disorders was much less evident and took longer to materialize. The first indication that cytokines might be involved in psychopathology came from cancerology and internal medicine. Clinicians who were injecting purified or recombinant cytokines to patients afflicted with cancer and hepatitis B observed a rapid induction of flu-like symptoms, followed in a vast majority of the patients by psychiatric disorders, in the form of acute psychosis or major depressive episodes. These side effects of cytokine therapy obliged clinicians to reduce doses and find safer routes of administration than the intracerebral and intravenous routes used initially. However, they left the world of psychiatry relatively indifferent, especially since, with the possible exception of schizophrenia, the relationship between immunity and mental disorders was seen more as a coincidence than a causal chain of events. Even if depression, because of its commonality, was sometimes viewed as a form of brain flu, and antidepressant drugs as centrally acting inhibitors of prostaglandins synthesis (Leonard, 1987), there was not much evidence to go beyond metaphor. In 1991, Smith proposed for the first time what he called the "macrophage theory of depression". Based on the potent brain effects of proinflammatory cytokines such as interleukin-1, and the anecdotally reported association between pathological states of immune activation and depression, he claimed that excessive secretion of interleukin-1 and other macrophage products causes depression. Since the discovery of proinflammatory cytokines, many reports had repeatedly demonstrated that administration of these molecules to experimental animals and

vi

Preface

volunteers induces most of the symptoms of depression, including activation of the hypothalamic pituitary-adrenal axis, depressed mood, lack of interest in daily activities, suppression of food intake, psychomotor retardation, sleep disorders, fatigue, confusion, and alterations in cognition. The problem was, however, that there was no evidence that depression was associated with immune activation. On the contrary, the predominant view was that depressed patients were immunodepressed, as evidenced, for example, by reduced natural killer cell activity and lymphocyte proliferation in response to mitogens. It took several years after publication of Smith's provocative hypothesis to collect sufficient data for accepting the possibihty that mental depression is not merely associated with immunodepression but also with an imbalance in the functioning of the immune system, with an activation of the monocyte/macrophage arm of the immune response, and a relative depression of lymphocyte functions. Evidence for immune activation in depressed patients was first collected by Maes, and was mainly based on the measurement of acute phase proteins and cytokines in the plasma of patients suffering from major depressive disorders and treatment resistant depression (Maes, 1995). If cytokines are involved in the pathogenesis of depressive symptoms, antidepressant drugs should attenuate cytokine production and action, and conversely, cytokine antagonists should alleviate symptoms of depression in animal models and in patients. Studies testing these predictions have been initiated during the recent years, and the first demonstration that chronic, but not acute, antidepressant treatment alleviated the anhedonic effects of cytokines in laboratory rodents was provided by Yirmiya in 1996. In addition, a few reports indicated that cytokine expression could be experimentally induced not only by immune stimuli, but also by psychological stressors, which have long been regarded as contributors to the development and maintenance of depression. The converging evidence between the brain effects of cytokines and the immunological correlates of depression prompted the editors of the present volume to bring together for a discussion meeting most, if not all, of the scientists working in this field from either a clinical or an experimental perspective. This meeting was carried out as part of the activities of the European Community Programs on Neurobiology of Interleukin-1 Receptors (TMR, CT97-0149) and the Expression and Action of AntiInflammatory Cytokines in the Brain (Biomed2, CT97-2492). It took place in Roscoff, France, on May 14-17, 1998, and was made possible thanks to the generosity of the Association pour la Neuropsychopharmacologie and the Institut de Recherches International Servier. The present volume contains the information presented at this meeting. It aims at providing a review of the current knowledge on the role of cytokines in depression by discussing the brain actions of cytokines, their role in the stress response, their relationship to depression, and their sensitivity to antidepressant treatment. Although there is some overlap between the contents of the different chapters, nothing has been done to reduce or even to hide it. It is the editors' opinion that its mere existence, despite the heterogeneity of disciplines that are represented, reflects very well the converging clinical and experimental evidence in favor of an intersection between cytokines and depression. Whether this intersection will prove to be of practical and theoretical importance in the pathophysiology of depression is still difficult to determine, but hopefully its delineation will encourage further work aiming at elucidating the role of cytokines in psychopathology.

Preface

vii

REFERENCES Leonard, B. E. (1987). Stress, the immune system, and mental illness. Stress Medicine, 3, 257-258. Maes, M. (1995). The interleukin hypothesis of major depression. Progress in NeuroPsychopharmacology and Biological Psychiatry, 19, 11-38. Smith, R. S. (1991). The macrophage theory of depression. Medical Hypotheses, 35, 298-305. Yirmiya, R. (1996). Endotoxin produces a depressive-like syndrome in rats. Brain Research, 7U, 163-174.

CONTENTS

Section I. Depression and Immunity 1. Immune Correlates of Depression Michael Irwin 2. Major Depression and Activation of the Inflammatory Response System Michael Maes

1

25

3. Cytokine Production in Depressed Patients Andreas Seidel, Matthias Rothermundt, and Lothar Rink

47

4. Indicators of Immune Activation in Depressed Patients Anna Sluzewska

59

Section II. Brain Effects of Cytokines 5. Mood and Cognitive Disorders in Cancer Patients Receiving Cytokine Therapy Christina A. Meyers 6. Mechanisms of the Behavioural Effects of Cytokines Robert Dantzer, Arnaud Aubert, Rose-Marie Bluthe, Gilles Gheusi, Sandrine Cremona, Sophie Laye, Jan-Pieter Konsman, Patricia Parnet, and Keith W. Kelley 7. Effects of Cytokines on Glucocorticoid Receptor Expression and Function: Glucocorticoid Resistance and Relevance to Depression Andrew H. Miller, Carmine M. Pariante, and Bradley D. Pearce 8. Effects of Cytokines on Cerebral Neurotransmission: Comparison with the Effects of Stress Adrian J. Dunn, Jianping Wan, and Tetsuya Ando

75

83

107

117

X

Contents

9. Inflammation and Brain Function under Basal Conditions and during Long-Term Elevation of Brain Corticotropin-Releasing Hormone Levels Astrid C E. Linthorst and Johannes M. H. M. Reul

129

Section III. Effects of Stress on Cytokine Production and Actions 10. Dynamic Regulation of Proinflammatory Cytokines Linda R. Watkins, Kien T. Nguyen, Jacqueline E. Lee, and Steven F. Maier 11. Cross-Sensitization between Immune and Non-Immune Stressors: A Role in the Etiology of Depression? F J. H. Tilders and E. D. Schmidt

153

179

Section IV. Effects of Cytokines and Cytokine Antagonists in Animal Models of Depression 12. Anhedonic and Anxiogenic Effects of Cytokine Exposure Hymie Anisman and Zul Merali

199

13. Stress, Learned Helplessness, and Brain Interleukin-lf3 Steven F. Maier, Kien T. Nguyen, Terrence Deak, Erin D. Milligan, and Linda R. Watkins

235

14. Stress, Depression, and the Role of Cytokines B. E. Leonard and Cai Song

251

Section V. Effects of Antidepressants on Cytokine Production and Action 15. Is There Evidence for an Effect of Antidepressant Drugs on Immune Function? Pierre J. Neveu and Nathalie Castanon 16. Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs Raz Yirmiya, Joseph Weidenfeld, Yehuda Pollak, Michel Morag, Avraham Morag, Ronit Avitsur, Ohr Barak, Avraham Reichenberg, Edna Cohen, Yehuda Shavit, and Haim Ovadia

267

283

Cytokines, Stress, and Depression: Conclusions and Perspectives Robert Dantzer, Emmanuelle E. WoUman, Ljubisa Vitkovic, and Raz Yirmiya

317

Contributors

331

Index

335

IMMUNE CORRELATES OF DEPRESSION Michael Irwin Departments of Psychiatry San Diego VA Medical Center, and University of California, San Diego La Jolla, CA 92093

1. INTRODUCTION Substantial evidence has shown that major depression is associated with immune variations. In a meta-analytic review of over thirty-five independent study samples conducted through 1991, depression was found to be associated with reliable alterations in several enumerative measures and in functional assays such as mitogen induced lymphocyte proliferation and natural killer (NK) cell activity (Herbert & Cohen, 1993a). Since 1991, many additional studies on the relationship between depression and immune measures have been published which has led to a recent reevaluation of the immune findings in depression (Zorrilla, Luborsky, McKay, Rosenthal, Houldin, Tax, McCorkle, Seligman, & Schmidt, 1998). While both reviews observed similar relations between depression and several immune measures including major immune cell classes, lymphocyte proHferation, and NK activity, conflicting results were obtained for many other measures. An enormous heterogeneity in the findings is suggested by the discrepancies between these two large meta-analytic reviews. The role of several moderating factors that might contribute to this variability in the link between depression and immunity has been explored in a series of separate studies as will be reviewed below. Moreover, in the meta-analyses conducted by Zorrilla and colleagues (1998), the role of several moderator variables including age, gender, ambulatory status, and depressive symptom severity was tested. Only part of the heterogeneity could be explained by these factors. Further investigations are clearly needed to understand what variables correlate with immune alterations in depression and whether there are subgroups of depressed subjects who are at risk for immune alterations. The present chapter will review and summarize results of studies on immune alterations in depression. In view of the complexity and heterogeneity of the immunological findings in depression, we will also consider the possible characteristics related to depressed subjects or to the nature of their depressive disorder which might Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

1

2

M. Irwin

moderate the association between depression and immune alterations. Identification of such moderators are particularly important as converging evidence suggests that immune alterations in major depressive disorder are not specific correlates of this disorder. Rather, similar kinds of changes in natural and cellular immunity occur in association with stress and other psychiatric disorders which supports the possibihty that certain shared behavioral or biological characteristics underlie and/or mediate the immune changes. An additional challenge of immune studies in depression is to move beyond a description of immune variations and to evaluate whether alterations of the immune system in depression have clinical implications. A third and final section of this chapter will review ongoing investigations of specific immune responses in depression that are related to outcome.

2. IMMUNE RESULTS IN DEPRESSION Three general domains of immune measures have been examined in studies of depressed subjects: enumerative measures, functional measures, and markers of immune activation. Each domain will be briefly described with a review of the respective immune findings in depressive disorder.

2.1. Enumerative Measures The enumerative measures provide a measure of the total number of circulating immune cells and quantify the absolute and relative numbers of the major immune cell classes such as neutrophils, lymphocytes, and monocytes. In addition, enumerative measures provide an assessment of the number and percentage of cells expressing cell surface markers related to B cells and T lymphocyte subset phenotypes. One of the first immunological findings identified in the study of depressed subjects was change in the total number of white blood cells and numbers and percentages of neutrophils and lymphocytes (Sengar, Waters, Dunne, & Bover, 1982; Schleifer, Keller, Meyerson, Raskin, Davis, & Stein, 1984; Kronfol, Turner, Nasrallah, & Winokur, 1984; Albrecht, Helderman Schlesser, & Rush, 1985; Schleifer, Keller, Siris, Davis, & Stein, 1985; Irwin, Smith, & Gillin, 1987; Kronfol, & House, 1989; Schleifer, Keller, Bond, Cohen, & Stein, 1989; Irwin, Caldwell, Smith, Brown, Schuckit, & Gillin, 1990a; Irwin, Patterson, Smith, Caldwell, Brown, Gillin, & Grant, 1990b; Maes, Lambrechts, Bosmans, Jacobs, Suy, Vandervorst, Jonckheere, Minner, & Raus, 1992a). Along with an increase in the total number of white blood cells, increases in the number of neutrophils and decreases in the number of lymphocytes have been reliably demonstrated (Herbert & Cohen, 1993a; Zorrilla et al., 1998). In regards to numbers of monocytes, a relative increase in the number of monocytes has been found in many studies of depressed subjects (Maes et al., 1992a), whereas other depression samples have shown decreases or no differences in the absolute or relative numbers of monocytes (Irwin et al., 1987; Irwin et al., 1990a,b).The two existing meta-analyses have also yielded discrepant results regarding change of monocyte numbers in depression. Cellular enumeration of lymphocyte subsets by the quantification of phenotypic specific cell surface markers has been widely used to evaluate alterations of the immune system in relation to diagnostic depression. Depression is reported to be negatively related to the number and percentages of lymphocytes that are B cells, T cells, T helper

Immune Correlates of Depression

3

cells, and T suppressor/cytotoxic cells (Herbert & Cohen, 1993a). A decrease in circulating number of cells that express the NK phenotype has also been reported which in part is moderated by gender; a decline of NK cell numbers is found in male but not female depressed subjects as compared to gender matched controls (Evans, Folds, Petitto, Golden Pedersen, Corrigan, Gilmore, Silva, Quade, & Ozer, 1992). However, multiple discrepant findings have also been reported and, in one of the largest study samples of depressed subjects, no difference in the number of peripheral blood lymphocytes or T lymphocyte subsets was found between depressed patients and controls (Schleifer et al., 1989). Indeed, with accumulation of a larger sample of studies and subjects, it is increasingly apparent that heterogeneous results predominate and there appears to be no consistent change in the number of circulating B-, T-, or NK cells in depression (Zorrilla et al., 1998).

2.2. Functional Measures Function of the immune system has been typically evaluated in depressed subjects by assay of nonspecific mitogen-induced lymphocyte proliferation, mitogenstimulated cytokine production, and NK cytotoxicity. Lymphocyte proliferation assays examine how well lymphocytes divide in response to a nonspecific mitogen. This assay along with the release of cytokines assumes that with more proliferation and more cytokine release, the lymphocytes are functioning more effectively. The purpose of the third functional assay, NK cell cytotoxicity, is to determine the ability of NK cells to lyse tumor cells in culture. 2.2.1. Mitogen-induced lymphocyte proliferation. A reliable association between depression and lower proliferative responses to the mitogens PHA, ConA, and pokeweed has been found in each of the two meta-analytic reviews (Herbert & Cohen, 1993a; Zorrilla et al., 1998). Some of the first observations evaluating depression and mitogen responses showed reduced proliferation in depressed subjects as compared to controls (Kronfol, Silva, Greden, Dembinski, Gardner, & Carroll, 1983; Schleifer et al., 1984; Kronfol & House, 1985). However, subsequent studies failed to replicate these observations, raising questions about the reliability of this immune alteration in depression (Schleifer et al., 1989). Nevertheless, with over a dozen studies now conducted on lymphocyte proliferation in depression (Albrecht et al., 1985; Syvalahti, Eskola, Ruuskanen, & Laine, 1985; Schleifer et al., 1985; Kronfol, House, Silva, Greden, & Carroll, 1986; Calabrese, Shwerer, Barna, Gulledge, Valenzuela, Butkus, Subichin, & Krupp, 1986; Lowy, Reder, Gormley, & Meltzer, 1988; Maes, Bosman, Suy, Minner, & Raus, 1989; Altshuler, Plaeger-Marshall, Richeimer, Daniels, & Baxter, 1989; Darko, Gillin, Risch, Bulloch, Golshan,Tasevska, & Hamburger, 1989; Kronfol & House, 1989; Cosyns, Maes, Vandewoude, Stevens, DeClerck, & Schotte, 1989; Bartoloni, Guidi, Antico, Pili, Cursi, Carbonin, Gambassi, Rumi, Di Giovanni, & Menichella, 1990; McAdams and Leonard, 1993; Birmaher, Rabin, Garcia, Jain, Whiteside, Williamson, Al-Shabbout, Nelson, Dahl, & Ryan, 1994), it appears that an impairment in the response of lymphocytes to all three non-specific mitogens predominates in studies of depressed subjects. 2.2.2. Natural killer cytotoxicity. A reduction of NK activity is considered to be one of the most reliable and reproducible alterations of ex vivo immune function in depression (Stein, Miller, & Trestman, 1991). Irwin and colleagues first reported a decline of NK activity in depressed subjects patients as compared to age-, and gender

4

M. Irwin

matched comparison controls (Irwin et al., 1987), and ten subsequent independent samples have replicated this observation (Kronfol, Nair, Goodson, Goel, Haskett, & Schwartz, 1989; Nerozzi, Santoni, Bersani, Magnani, Bressan, Pasini, Antonozzi, & Frajese, 1989; Irwin et al., 1990a; Irwin et al., 1990b; Bartoloni et al., 1990; Caldwell, Irwin, & Lohr, 1991; Maes, Stevens, Peeters, DeClerck, Scharpe, Bridts, Schotte, & Cosyns, 1992c; Shain, Kronfol, Naylor, Goel, Evans, & Schaefer, 1991; Irwin, Lacher, & Caldwell, 1992a; Evans et al., 1992; Birmaher et al., 1994). Although there are some discrepant findings (Mohl, Huang, Bowden, Fischbach, Vogtsberger, & Talal, 1987; Schleifer et al., 1989), Herbert and Cohen found a fixed effect size of (-0.266 to -0.254), a result that was confirmed by Zorrilla et al. with an accumulation of over 660 subjects. The factors that might moderate and/or mediate the effects of depression on NK activity will be discussed below.

2.3. Markers of Immune Activation Most previous studies have suggested that depression results in reductions of nonspecific cellular and natural immunity. However, Maes has argued that major depression is associated with immune activation reminiscent of an acute phase response (Maes, Smith, & Scharpe, 1995). Evidence has been accumulated showing increases in levels of cells bearing activation markers such as HLA-DR+, CD25+ (interluekin-2 receptor) (Maes et al., 1992a), and increases in humoral factors or plasma proteins associated with the acute phase of the immune response (al-acid glycoprotein, a l antitrypsin, and haptoglobin) (Maes, Scharpe, Van Grootel, Uyttenbroeck, Cooreman, Cosyns, & Suy, 1992b). In addition, cytokines such IL-6 that are typically associated with an inflammatory process are reportedly elevated in depression and there are also reported increases in the circulating concentration of the soluble IL-2 receptor that is released with immune activation (Maes et al., 1995). While these findings that an inflammatory process may occur in association with depression primarily emanate from one research group, there are additional preclinical data that suggest that macrophage activation might occur in depressed subjects and could be a key factor in the observed impairments of T cell proliferation (Maier, Watkins, & Fleshner, 1994). In a putative animal of depression (uncontrollable shock), concomitant monokine release occurs, and this macrophage activation produces an impairment in T cell mediated proliferation. Replication of clinical data concerning immune activation in depression is clearly needed, but there have been few independent studies separate from the work of Maes. Indeed, in the meta-analytic review conducted by Zorrilla et al., there were only three to five independent study samples identified depending on the immune activation marker. An increase in the circulating concentration of haptoglobin was reliably associated with depression, but the random effects model did not indicate reliable associations between depression and other measures of immune activation including number and percentage of cells expressing HLA-DR or CD25 or circulating levels of ocl-acid glycoprotein or al-antitypsin (Zorrilla et al., 1998). To further evaluate the relationship between depression and immune activation, we have recently completed a study of 25 pairs of depressed subjects and age- and gender matched comparison controls in which serum levels of haptoglobin, al-acid glycoprotein, al-antitypsin, interleukin 6, and soluble interleukin-2 receptor were measured. None of these measures of immune activation differed between the carefully matched groups. The lack of finding elevated levels of soluble interleukin-2

Immune Correlates of Depression

5

receptor in depression is consistent with prior observations from our laboratory (Rapaport & Irwin, 1996). In addition, none of these measures of immune activation correlated with severity of depressive symptoms as measured by HDRS scores. However, correlations were found between cigarette smoking and increases in haptoglobin, al-acid glycoprotein, and al-antitypsin, suggesting the importance of assessing tobacco consumption in future immunological studies of depressed- and other psychiatric populations, as will be discussed below.

2.4. Summary of Results The relationship between clinical depression and measures of immunological function shows considerable variability with a predominance of heterogeneous findings for many of the immune parameters. However, two separate meta-analyses have now shown rehable associations between depression and several immune measures. It appears that depressed patients are likely to show an increase in total white blood cell counts with a relative neutrophilia and a relative lymphopenia. In addition, depressed subjects are likely to show impairments in functional assays of mitogen-induced lymphocyte proliferation and NK cytotoxicity. However, with the exception of haptoglobin, heterogeneous findings predominate in the characterization of immune activation in depression and future research is needed to replicate these results.

3. MODERATORS OF THE DEPRESSION-IMMUNE LINK It is not known what accounts for the studywise heterogeneity in the relationship between depression and immunity. Stein et al. (1991) proposed that alterations in the immune system in major depression are not specific biologic correlates of this disorder, but rather occur in association with other variables that characterize depressed subjects such as age, hospitalization stress, and severity of depressive symptoms. Indeed, in the two meta-analyses conducted by Herbert and Cohen (1993a and b) and Zorrilla et al. (1998), subject characteristics such as age, gender, and symptom severity were evaluated as possible moderators of the relationship between depression and immunity. However, these moderator variables have been found to explain only partially the heterogeneity observed. The role of other factors in moderating the effects of depression on the immune system has not been systematically explored. However, several exploratory studies have been conducted which indicate the possibility that additional subject characteristics and/or depression status variables may moderate the relation between depression and immunity. In the present review, the role of several moderator variables will be considered in addition to age, gender, and symptom severity. For example, life stressors have been found to be associated with changes in the immune system (Herbert & Cohen, 1993b), and such stress-induced immune alterations are similar to the kinds of changes found in depressed subjects. Thus, we will discuss whether the subacute stress of hospitalization and/or chronic life stress often reported in depressed patients contribute to the immune changes found in diagnostic depression (Caldwell et al., 1991; Irwin et al., 1990b). Second, symptom severity has been suggested as an important correlate of immune changes in depressed subjects, but what specific symptoms account for this

6

M, Irwin

relation requires further study to know whether there are certain subgroups or subtypes of depressed subjects who are most at risk for immune changes. Some evidence suggests that depressed patients with neurovegetative symptoms are more Hkely to show immune changes (Cover & Irwin, 1994) and a number of observations impHcate sleep in the regulation of some aspects of immune function (Irwin, Mascovich, Gillin, Willoughby, Pike, & Smith, 1994; Irwin, McClintick, Costlow, Fortner, White, & Gillin, 1996; Irwin, Smith, & Gillin, 1992b). Finally, co-morbidity in depression has emerged as a major concern in biologic research in psychiatry with recent epidemiological data showing, for example, that as many as 50% of depressed patients are co-morbid for an anxiety disorder such as panic (Stein & Uhde, 1988; Murphy, Oliver, Sobol, Monson, & Leighton, 1986). Moreover, the prevalence of alcohol and tobacco dependence is high in depressed subjects (Breslau, Peterson, Schultz, Chilcoat, & Andreski, 1998; Schuckit, 1986), but few studies have examined the contribution of alcohol intake or cigarette smoking on alterations of immune function in depressed subjects despite the well-recognized effects of these substances on immune parameters. The following will consider the effects of comorbidity for anxiety and/or substance dependence in the relation between depression and immunity.

3.1. Subject Characteristics 3.1.1. Age. Age has been proposed as a key moderator in the association between depression and immunity. In three reviews (Weisse, 1992; Stein et al., 1991; Herbert & Cohen, 1993a), it was concluded that older depressed patients are more likely to show immune differences as compared to younger depressed subjects. The findings of Schleifer et al. (1989) are most representative of this conclusion; controls showed an age-related increase in CD4 numbers and mitogen responses, whereas depressed patients did not show such changes in either T cell numbers or lymphocyte proliferation with advancing age. However, specific age group comparisons of depressives vs. controls have not been conducted and it is not known, for example, whether elderly depressed subgroups show greater immune differences than adolescent- or young adult samples. Furthermore, in studies of depressed patients, age, and hospitalization status are correlated, and as cautioned by Herbert and Cohen (1993a), the separate effects of age and hospitalization status on depression-immunity associations can not be determined without carefully matching depressed inpatients and outpatients with controls on both age and hospitalization status. Of course, another strategy would be to design a study that was restricted to either inpatient- or outpatient depressed subjects and then to evaluate the effects of age within that hospitalization status subgroup. For example, Irwin et al. (1990b) focused on inpatient depressed subjects and found that these subjects had lower NK activity than controls. However, within the inpatient depressed subjects, there was no effect of age or an age by depressive symptom interaction on NK activity. Zorrilla et al. (1998) also recently demonstrated using a random effects moderator analyses that age did not demonstrate a significant moderating effect with any of the immune variables found to be rehably altered in depression. 3.1.2. Gender. The stability of the depression immune relationship across gender also deserves further attention due to the well-recognized effects of reproductive hormones such as progesterone to antagonize the suppressive effects

Immune Correlates of Depression

7

of glucocorticoids on cellular immune function (Dietrich, Chasserot-Golaz, Beck, & Bauer, 1986). Many studies have included only male depressed subjects and eliminated the possible contribution of gender (Irwin et al., 1990a; 1990b). Other studies have included male and female depressives and evaluated effects of gender using correlational analyses (Schleifer et al., 1989). The latter have typically not found any effects of gender on either enumerative or functional immune measures in depression, a finding consistent with the conclusion reached by Herbert and Cohen (1993a). However, only a few studies have stratified the sample and made direct comparisons between depressives and controls matched on the basis of gender. Miller, Asnis, Lackner, Halbreich, & Norin (1991) found in outpatient depressives that female subjects had higher NK activity than controls, whereas male depressives were similar to controls. In contrast, Evans et al. (1992) studied both inpatient and outpatient depressed subjects and found that men with major depression had a marked reduction in number of NK cells (Leu-11) and in NK activity compared with levels in controls. By contrast, depressed women did not differ significantly on any of these immune measures. Similar comparisons on the basis of gender have not been conducted for other immune measures, although there is evidence that studies with a higher percentage of females are more likely to show greater deficits in Con A responsiveness along with smaller decreases in NK cell cytotoxicity (Zorrilla etal., 1998). 3.1.3. Stress. Both life stress and depression have been associated with similar changes of immune measures (Herbert & Cohen, 1993a and b), and it is possible that these two conditions together might interact and influence immunity more than the presence of life stress or depression alone. Depressed subjects are more likely to experience severe life events during their depressive episode due in part to the functional impairments associated with depression (Brown, Harris, & Hepworth, 1994). For example, a depressive episode may be associated with deterioration in work performance such that the person loses his/her job. Likewise, impairment in social interactions during depression may affect a spousal relationship leading to threatened divorce or separation. To our knowledge, the joint contribution of threatening life events and depression on immune function has only preliminarily been examined in one study. In a sample of 36 subject pairs of depressed patients and comparison controls, the presence of severe threatening life stress was assessed using the objective methods of Brown and Harris (1978). Subjects in both groups were classified as having low- or high stress and NK activity was measured. Findings demonstrate a reduction of NK activity in subjects with major depression as compared with persons who are neither stressed nor depressed. Second, individuals who are undergoing severe life stress show a similar reduction of NK activity even though they do not have clinically significant depressive symptoms. The combination of severe stress and depression did not result in any further reduction of NK activity. Thus, the immune changes found in life stress parallel those found in depression. However, the changes found in depressed subjects can not be explained simply by the occurrence of severe life stress (Irwin et al., 1990b). In other words, life stress influences cellular immunity, but it does not appear to moderate the association between depression and immunity. Similar conclusions were also reached in the evaluation of the effects of acute hospitalization stress on depression related impairments of NK activity (Caldwell e t a l , 1991).

8

M. Irwin

3.2. Depression Status 3.2.1. Ambulatory Status. A decrease of mitogen responses in hospitalized patients with major depression but not in ambulatory patients with depression suggests that alterations of immunity in depression are related to hospitalization status or to severity of depressive symptoms (Schleifer et al., 1984; 1985). While subsequent study of both inpatient and outpatient depressed subjects failed to identify an effect for hospitalization status (Schleifer et al., 1989), the meta-analyses conducted by Herbert and Cohen (1993a) and by Zorrilla et al. (1998) showed that hospitalization status indeed moderated the effects of depression on mitogen responses but had no effect on other immune measures. Suppression of mitogen-induced lymphocyte proliferation in depressed subjects tends to be greater in inpatient than in outpatient samples. However this meta-analytic result requires caution. The set of studies that was used included hospitalized subjects who were more severely depressed than the ambulatory subjects. Thus, the meta-analytic effect of ambulatory status may be more explained by the moderating effects of depressive symptom severity than by hospitalization status per se. 3.2.2. Depression Severity. To examine the contribution of depressive symptom severity on immune measures, Schleifer et al. (1989) measured severity of depressive symptoms (total HDRS scores) and assessed B, T, and T, cell subset numbers, mitogeninduced lymphocyte stimulation and NK activity in a sample of 30 inpatient and 61 outpatient depressed subjects. Severity of depressive symptoms was found to be associated with a decline of ConA and PHA lymphocyte responses independent of the effects of gender, age, and hospitalization status. However, neither the enumerative measures or NK activity was correlated with severity of depressive symptoms in this large sample. Evans and colleagues also found that neither NK cell numbers nor NK activity correlated with severity of depressive symptoms, even though diagnostic depression was associated with a decline in numbers of NK cells and NK activity as compared to levels found in comparison controls (Evans et al., 1992). While these negative findings contrast with those of Irwin and colleagues who found that severity of depressive symptoms is a distinct correlate of NK activity in three independent samples (Irwin et al., 1987, 1990a, 1990b), it appears that the moderating effect of severity of depressive symptoms on imnmne measures is small. In the recent meta-analyses conducted by Zorrilla et al. (1998), total score on the HDRS was not reliably correlated with either of the functional measures, lymphocyte proliferation or NK activity. Moderator analyses also did not reveal correlations between severity of depressive symptoms and other immune measures such as total white blood cell count or percentages of neutrophils and lymphocytes. The lack of relationship between depressive symptom severity and measures of lymphocyte proliferation and NK activity is surprising considering the robust association between these immune measures and diagnostic depression. However, the discrepancy between these two observations raises an important question concerning the link between depression and immune changes. If there is no association between depressive symptom severity and immune alterations in depressed subjects, then it is possible that immune alterations might persist even into remission and be indicative of a trait rather than a state marker of depression. To our knowledge, one longitudinal case-control study has been conducted to examine state vs. trait issues in the depression-immune link (Irwin et al., 1992a). This longitudinal design also provided an opportunity to test the temporal relationship

Immune Correlates of Depression

9

between symptom severity and immunity beyond the prior correlational analyses conducted in cross-sectional samples. Assessment of NK activity was obtained in the second week of hospitalization and repeated at six months after discharge from the hospital in 20 depressed subjects. Matched comparison controls were studied at the same interval and on the same day as their depressed subject pair. The depressed subjects showed a decrease in HDRS and an increase of NK activity from intake to follow-up. There was no change in either depression scores or NK activity in the controls. The temporal relationship between depressive symptom severity and reduced NK activity contrasts with the meta-analytic findings of Zorrilla et al. (1998) who found no association between depressive symptom severity and NK activity. It might be that the state of current depression is actually a higher order variable. Depression is correlated with symptom severity, but the association between acute diagnostic depression and altered immunity is related to an interaction and/or combination of depressive symptom severity with other correlates of the state of depression. 3.2.3. Depression Subtype. That a constellation of subject characteristics and depressive symptoms explains the decline of cellular immunity is depression is also consistent with findings generated in the study of immune alterations in melancholic depression. Melancholic depressed patients are typically older, hospitalized, and have more severe depressive symptoms (Parker, Hadzi-Pavlovic, Boyce, Wilhelm, Brodaty, 1990; Mitchell, Hickie, & Eyers, 1990). While none of these characteristics alone have been found to moderate the association between depression and immunity, these characteristics together with the diagnosis of melancholia have considerable predictive significance in determining impaired cell mediated immunity in depression. Melancholic depressed subjects are reported to be at increased risk for immune alterations and to show lower levels of mitogen-induced lymphocyte proliferation and of NK activity as compared to depressed subjects without melancholia (Cosyns et al., 1989; Maes et al., 1989; 1991; 1995). In addition, patients with melancholia show impaired in vivo immune responses measured by delayed type hypersensitivity skin responses whereas non-melancholic depressed subjects have responses similar to controls (Hickie, Hickie, Lloyd, Silove, & Wakefield, 1993). Moreover, the decrement of delayed type hypersensitivity responses was predicted by the diagnosis of melancholia independent of the separate effects of age, hospitalization status, and depression severity. 3.2.4. Neurovegetative Symptoms. Rather than the sum of depressive symptoms as reflected in total HDRS scores, certain symptoms may be relatively more important in moderating the association between diagnostic depression and immunity. Indeed, the association between melancholic depression and impaired cellular immunity may be due in part to these patients' increased prevalence of neurovegetative symptoms such as sleep disturbance, appetite disturbance with associated nutritional impairments, and weight loss, and/or increased psychomotor retardation. Thus, to examine whether certain depressive symptoms account for some of the immune changes in depression. Cover and Irwin (1994) extended previous analyses that examined the relationship between depressive symptoms and immunity by evaluating the contribution of clusters of depressive symptoms. Six symptom clusters, defined by factor analysis of items from the HDRS, were identified: anxiety/somatization, weight loss, cognitive disturbance, diurnal variation,

10

M. Irwin

retardation, and sleep disturbance (Cleary & Guy, 1977). In a sample of depressed inpatients, no correlation was found between total HDRS scores and NK activity. Likewise, there was no correlation between four symptom clusters (anxiety/somatization, weight loss, cognitive disturbance, or diurnal variation) and NK activity. However, two symptom clusters, retardation, and sleep disturbance, were correlated with NK activity, and together these two symptom variables accounted for over 16% of the variance in NK activity (Cover & Irwin, 1994). 3.2.5. Nutritional status. In regards to the lack of relationship between weight loss and NK activity as described above, Schleifer et al. (1989) also found that statistical control for weight and recent weight loss was not associated with depression-related declines in lymphocyte proliferation. However, further investigation of the potential effects of nutritional factors in the depression-immune relation is clearly needed before excluding nutritional status as a variable that might account for immune alterations in depressed subjects. 3.2.6. Sleep disturbance. Insomnia is one of the most common complaints of depressed subjects, but its role in moderating and/or mediating immune alterations in depression has been relatively unexplored. However, with evidence that subjective insomnia correlates with NK activity in depression (Cover & Irwin, 1994), the hypothesis emerged that disordered sleep may be a distinct factor accounting for some of the observed immune alterations found in depression. Consequently, a series of studies have now been conducted to test more carefully the role of sleep in the modulation of multiple aspects of the immune system and to determine the moderating effects of sleep disturbance on immune measures in depressed subjects. First, observations regarding subjective complaints of insomnia and immunity were extended by assessment of disordered sleep by EEG (Irwin et al., 1992b). Many aspects or parameters of sleep are assessed during all-night EEG studies, thus the initial approach was to identify those measures of EEG that are altered in association with insomnia or subjective sleep disturbance and then to characterize the relationship of these EEG measures to immunity. Self-report of sleep disturbance or insomnia is characterized by disturbances of EEG sleep continuity measures with increases of sleep latency (the interval from lights out to onset of sleep) and decreases in total sleep time and sleep efficiency (the ratio of sleep to the amount of time in bed) (Benca, Obermeyer,Thisted, & Gillin, 1992). Considering the relation between subjective sleep disturbance and NK activity, we hypothesized that measures of sleep continuity measures would correlate with NK activity in depressed subjects. In addition, if sleep has a distinct role in the modulation of immune function independent of diagnostic depression and other depressive symptoms, then similar correlations between sleep and NK activity should also be identified in control subjects who differ in the amount and quality of their sleep. Indeed, both total sleep time and sleep efficiency were positively correlated with NK activity in separate groups of depressed subjects and controls who underwent all night EEG study with assessment of NK activity in the morning upon awakening (Irwin et al., 1992b). These data were some of the first to indicate that sleep amounts are associated with immune function and to suggest that disordered sleep may be a key variable in understanding the behavioral mechanisms underlying the link between depression and immune alterations. Recent studies in bereaved subjects have replicated this correlation between EEG sleep and immunity and shown by way of causal statistical analyses that disordered sleep also mediates the relationship between

Immune Correlates of Depression

11

severe life stress and a decline of NK responses (Hall, Baum, Buysse, Prigerson, Kupfer, & Reynolds, 1998). A second line of studies to investigate the contribution of sleep on immune function has focused on the immunological assessment of a group of insomniac subjects who report disordered sleep but who are not depressed or suffering from some other psychiatric disorder. Primary insomnia is diagnostically characterized by the presence of subjective sleep difficulties that persist unrelated to another mental disorder or a known organic factor, such as a physical condition, psychoactive substance use disorder, or a medication (American Psychiatric Association, 1994). In a recent study of such primary insomniac patients who have no current or lifetime history of another mental disorder, EEG sleep and immune measures of NK activity, NK cell numbers and lymphokine activated killer (LAK) cell activity were compared with sleep and immune measures obtained in depressed subjects and controls. Both psychiatric groups, insomniac and depressed subjects, showed significant disturbances in sleep continuity with prolonged sleep latency and decreases in total sleep time and sleep efficiency as compared to the controls. Furthermore, the insomniac and depressed groups showed similar alterations in NK and LAK responses. The insomniac subjects showed a reduction of NK activity and LAK activity as compared to levels in controls. The declines of NK and LAK responses in insomniac subjects were similar to those observed in depressed subjects (Irwin & Gillin, 1998). A third strategy used to evaluate the relationship between disordered sleep and immunity has involved an experimental approach, sleep deprivation. In an effort to mimic the kind of disordered sleep found in depressed subjects, the effect of partial night sleep loss on lymphocyte function and NK and LAK responses has been tested. Depressed subjects often report symptoms of early or late insomnia with loss of sleep during only parts of the night. Thus, it was thought that a partial night sleep deprivation paradigm would more closely parallel the kinds of sleep continuity disturbance reported in depression. In contrast, previous studies have examined the effects of prolonged, total sleep deprivation on immunity (Palmblad, Petrini, Wasserman, & Akerstedt, 1979; Moldofsky, Lue, Davidson, & Gorczynski, 1989; Dinges, Douglas, Zaugg, Campbell, McMann, Whitehouse, Orne, Kapoor, Icaza, & Orne, 1994). The effects of partial sleep loss on immunity had not been previously explored. In two separate studies, one that examined sleep loss during the late part of the night (i.e. awake from 3 am to 7 am) (Irwin, et al., 1994) and the other that tested the effects of sleep loss during the early part of the night (i.e., awake from 11 pm to 3 am) (Irwin et al., 1996), substantial alterations of functional immune measures occurred in association with partial sleep deprivation. This modest sleep loss, typical of depressed subjects and other psychiatric and non-psychiatric groups who report sleep disturbance, was associated with declines of NK activity (Figure 1), LAK activity (Figure 2), and stimulated IL-2 production (Figure 3). The reduction of NK and LAK responses appeared to be due to impairments in the activity of these cells, rather than a selective redistribution of circulating populations of NK cells and LAK precursors. For example, calculation of NK lytic activity per number of NK cells (CD16,56) and of LAK cytotoxicity per number of LAK precursors (CD16,56 + CD25) revealed impairments of individual effector cell function. Furthermore, partial night sleep loss induced a marked decline in the stimulated production of IL-2 by peripheral blood mononuclear cells. Interestingly, this effect of sleep loss on the release and/or utilization of IL-2 was due to effects on both lymphocyte and adherent, antigen-presenting cell populations. In experiments that involved mixing adherent and nonadherent cells obtained after base-

M. Irwin

12

2

Sleep

PSD-E

Recovery

Baseline Figure 1. Effects of early night partial sleep deprivation on natural killer cell activity (mean ± SEM lytic units) in 38 healthy men. A repeated measures ANOVA demonstrated a significant time effect (F = 6.3; df = 5,185, p < 0.001) in which the mean value of NK activity following early night partial sleep deprivation (PSD-E) was significantly (p < 0.001) lower than values after control baseline-, sleep laboratory baseline, and recovery nights. The stippled area displays the mean ± SEM of the three control baseline values.

line and partial sleep deprivation nights, sleep loss was found to alter either cell type with an associated decrement of stimulated IL-2 production (Figure 3). Others have also shown that sleep is critical in the regulation of IL-2 production. Sleep onset is associated with a marked enhancement in the production of IL-2 by T cells (Born, Lange, Hansen, Molle, & Fehm, 1997; Uthgenannt, Schoolmann, Pietrowsky, Fehm, & Born, 1995). 3.2.7. Circadian Phase Shift. Circadian rhythms are often disrupted in major depression (Steiger & Holsboer, 1997), but the possible effects of abnormal diurnal variation on immune measures obtained in depressed subjects has received little attention. Most studies examining immune measures in depression have standardized the time of blood sampling in the depressed subjects and controls. However, this strategy with typically only one assessment of immune numbers and/or function does not consider the possibility of a phase shift between the two groups. Thus, immunological differences between the depressed subjects and controls may not be related to changes in either enumerative or functional measures, but rather may be an experimental artifact of a phase difference between the groups. There is now considerable evidence that counts of all major immune cell classes and lymphocyte subsets show circadian rhythmicity. Kronfol and colleagues found that the absolute number of lymphocytes and lymphocyte subsets (CD3, CD4, and CDS) peaked around midnight with a nadir about 10 am (Kronfol, Nair, Zhang, Hill, & Brown, 1997). Moreover, there is a diurnal rhythmicity of human cytokine production (Petrovsky & Harrison, 1997) and Born et al. (1997) have shown that this rhythmicity is preserved during conditions of sustained wakefulness. NK activity also shows a circadian rhythm with peak activity around noon and a nadir around midnight (Kronfol et al., 1997).

Immune Correlates of Depression

13

1^ Baseline

PSD-E

B 0.5.

25 eg O

U +

20. 15-

in

0.4.

O

0.3.

CO

to

a o

o

10(J

^^1 0-

0.2-

5^

^^1 Qaselins

0.1, 0.

PSC -E

Baseline

PSD-E

S 150

«

? _l

If] CO

100

n

>~ n o > w < ix: u

3 t Baseline

PSD E

Figure 2, A. Effects of early night partial sleep deprivation on lymphokine activated killer cell activity (mean ± SEM lytic units) in 18 subjects. Interleukin-2 stimulated lymphokine activated killer cell activity was significantly (two tailed t = 5.3, df = 36, p < 0.001) lower following a night of early partial sleep deprivation compared to baseline levels. Unstimulated cytotoxicity against NK resistant Daudi was not different (t = 1.9, df = 18, p = 0.06). B. Effects of early night partial sleep deprivation on counts of LAK precursors co-expressing CD16,56 and CD25. PSD-E had no effect on either the percentage or number of cells expressing CD 16,56 and CD25 (two tailed t = 0.9, df = 8, p = 0.41; t = 0.9, df = 7, p = 0.37). C. Effects of early night partial sleep deprivation on LAK activity per LAK precursor. Data illustrated are LAK lytic units per number of LAK precursors; logio transformation of these values demonstrated reduced LAK cell activation following sleep deprivation (two tailed t = 3.0, df = 7, p < 0.05).

14

M. Irwin

c o O

^

k->

c

Monocytes: T Cells:

Baseline Sleep Baseline Sleep

Baseline Sleep PSD-E

PSD-E Baseline Sleep

PSD-E PSD-E

Figure 3. Effects of mixing adherent (monocyte) and nonadherent (lymphocyte) cell populations from the baseline- and PSD-E conditions on interleukin-2 production (mean ± SEM ng/ml) in 24 subjects. Interleukin2 production in the four cell mixture groups (monocyte baseline—lymphocyte baseline; monocyte baseline— lymphocyte PSD-E; monocyte PSD-E—lymphocyte baseline; monocyte PSD-E—lymphocyte PSD-E) was signficantly different (F = 7.3, df = 3,69, p < 0.001). As compared to monocyte baseline—lymphocyte baseline values, mixtures of monocyte PSD-E—lymphocyte PSD-E, monocyte PSD-E—lymphocyte baseline, and monocyte PSD-E—lymphocyte PSD-E were significantly (p < 0.05) lower.

The diurnal variation of NK cell and B-lymphocyte measures appears to be disrupted in patients with major depression (Pettito, Folds, Ozer, Quade, & Evans, 1992; Pettito, Folds, & Evans, 1993). As compared to the diurnal rhythm found in controls, depressed patients show blunted diurnal variation in levels of NK cells, NK cytotoxic activity, and B cell numbers. However, despite marked differences in diurnal variation, mean levels of the immune measures were not statistically different between patients with major depression and comparison controls at either the morning or evening times of assessment. Thus, it does not appear that a phase shift accounts for the observed changes in immune measures between depressed subjects and controls. Nevertheless, conclusions from these data remain limited, as it is not possible to determine chronobiological rhythms in which only two time points are obtained to assess diurnal variation. Future investigations are clearly needed to address the possible effects of circadian phase on immunity in relation to depression. Such investigations should use multiple time intervals every 3 to 4 hours with a time series analyses to evaluate possible chronobiological group differences throughout the full circadian cycle in patients with major depression as compared to controls.

3.3. Co-Morbidity 3.3.1. Anxiety disorder. Few psychoneuroimmunology studies of depression have considered the possibility of immunological differences between diagnostic subtypes of major depression. There are even fewer studies that have evaluated

Immune Correlates of Depression

15

immunological differences between major depressed patients who differ in diagnostic co-morbidity. One exception is the work of Andreoli and colleagues who compared 51 pairs of major depressed subjects and age- and gender-matched controls on immune measures of T cell numbers and function (Andreoli, Keller, Rabaeus, Zaugg, Garrone, & Taban, 1992). The depressed subjects were further stratified on the basis of comorbid panic disorder to determine whether this additional diagnosis contributed to immune variability. Depressed subjects with panic disorder (n = 28) were found to have greater numbers of T cells and lymphocyte proliferation in response to PHA as compared to depressed subjects without panic disorder. These co-morbid depressed subjects also showed increased PHA and Con-A induced lymphocyte proliferation as compared to controls. In contrast, depressed subjects without panic disorder showed similar levels of T cells and lymphocyte proliferation as compared to controls. While replication of these findings is an important next step, these data suggest that a diagnostically distinct subgroup of depressed subjects may correspondingly show differences in psychobiology. Whether identification of other subgroups of depressed subjects may explain some of the heterogeneity of the depression-immune link is not known. 3.3.2. Alcohol dependence. Alcohol use and alcohol dependence are associated with alterations in cell-mediated immune function (ChiappelH & Gottesfeld, 1995). However, in most clinical depression-immune studies, the possible moderating effects of alcohol have not been assessed. Moreover, in many studies, it is often not known whether the depressed subjects are free of current alcohol abuse or dependence, although it is presumed that with the diagnosis of a major depressive disorder that the depressive symptoms are not due to the physiological effects of alcohol abuse. Data regarding the influence of prior alcohol dependence on immune function in acutely depressed subject are even more limited, despite epidemiological data showing that as many as 30% of depressed patients are co-morbid for alcohol dependence (Cadoret, 1981). To test the possible joint contribution of alcoholism and depression on immune parameters in depressed subjects, we compared white blood cell counts and NK cell activity in depressed patients with and without histories of alcohol abuse or dependence, alcoholics with and without secondary depression, and controls (Irwin et al., 1990a). The presence of the dual diagnoses of alcoholism and depression produced a further decrease of NK activity as compared to the separate effects of alcoholism and depression alone (Figure 4). Depressed subjects with histories of alcoholism had lower NK activity compared with depressed subjects without such histories. In addition, alcoholics who have secondary depression showed a further decrease in cytotoxicity compared with alcoholics who are not clinically depressed. The influence of alcohol abuse on these immune measures is particularly striking considering that this result reflects the effects of past consumption of alcohol. Both the depressed and alcoholic subjects were free of alcohol for a minimum of two weeks, and thus the decline of NK activity was not due to a direct pharmacological effect of alcohol. Consequently, in studies of depressed subjects, it may not be reasonable to assume that the influence of alcohol abuse on the immune parameter has dissipated simply due to a washout period lasting days to weeks. Rather, more systematic and careful assessment of current alcohol use along with dependence histories is needed in future studies of the relation between depression and immunity in order to reveal the effects of depression independent of variations in alcohol consumption.

M. Irwin

16

60n 50O O

40

o o 30'o a. CO 20-

100 40:1 20:1 10:1 5:1 Eff ector:Target Cef! Ratio Figure 4. NK activity across the four effector-to-target cell ratios in controls subjects (x), primary alcoholics (closed circles), primary alcohohcs with secondary depression (open circles), depressed patients (closed triangles), and depressed patients with histories of alcohol abuse (open triangles).

3.3.3. Tobacco dependence. In view of the prevalence of tobacco dependence in depressed subjects (Breslau et al., 1998) and the possible influence of cigarette smoking on immune function (Person, Edwards, Lind, Milton, & Hersey, 1979), it is also important to examine whether cigarette smoking moderates the relationship between depression and immunity. However, only a few studies concerned with immune alterations in depressed subjects have assessed smoking histories (Irwin et al., 1990b; Andreoli, Keller, Rabaeus, Marin, Bartlett, & Taban, 1993). While neither study found that smoking histories correlated with immune function in depressed subjects, the lack of a relationship between quantity of cigarette use and NK activity is not surprising due to a rather restricted range of cigarette use in the depressed subjects with most reporting moderate (1-2 packs per day) consumption. To our knowledge, no study has evaluated whether there is a possible interaction between depression and smoking on enumerative- and functional measures of immunity. We recently examined the influence of current cigarette smoking on total white blood cells, numbers of major immune cell classes, and NK activity in depressed subjects and in controls (Jung & Irwin, 1998). In a large series of previously described subjects (n = 245), depressed subjects, and controls were stratified on the basis of nonsmoking and current smoking status. Values of total white blood cell counts and differential and of NK activity were compared in the four groups. For total white blood cell count, depressed subjects showed on the average higher numbers of total cells than controls, and smokers had elevated total white blood cell counts as compared to nonsmokers. In addition, there was a significant interaction between depression and smoking on total cell counts. Depressed smokers had higher

Immune Correlates of Depression

17

numbers of white blood cells than depressed nonsmokers and control nonsmokers and smokers. No effect for smoking or interaction between depression and smoking was found for percentages of neutrophils, lymphocytes, or monocytes. For NK activity, there was a significant effect for depression. Depressed subjects showed on the average lower values of NK activity than controls. However, pairwise comparisons demonstrated that the difference between depressed subjects and controls was due to the marked reduction of NK activity in the depressed smokers as compared to the depressed nonsmokers. It did not appear that the immunologic changes found in depressed smokers were due to the simple effects of smoking. Controls who were current smokers showed white blood cell counts and levels of NK activity that were similar to those found in control nonsmokers. Furthermore, there was no correlation between amount of smoking and NK activity in either the control- or depressed smokers, a finding consistent with previous studies of smoking populations (Meliska, Stunkard, Gilbert, Jensen, & Martinko, 1995). These data have several implications. First, smoking status is a critical variable to assess in studies examining the relationship between depression and immunity. The combined effects of depression and smoking appear to predict changes in numbers of white blood cells and NK activity independent of the effects of depression and smoking alone. Due to the interaction between depression and smoking on total white blood cell counts and NK activity, inconsistent findings are likely to be reported if different studies have samples of depressed subjects who differ in the prevalence of smoking. Cigarette use alone and/or in combination with depression might also contribute to the suppression of other nonspecific measures of immune function such as mitogeninduced lymphocyte proliferation. Moreover, cigarette smoking is associated with immune activation (Mendall, Patel, Asante, Ballam, Morris, Strachan, Camm, & Northfield, 1997) and it is important to address whether smoking status alters the reported the relationship between depression and increases in serum levels of interleukin-6 and acute phase proteins. The health implications of reduced NK activity in depressed smokers are uncertain. However, there is evidence that depression might interact with other characteristics such as cigarette smoking to impact health, rather than there being a unitary link between depression and cancer. In a 12-year follow-up of 2264 adult men and women, depressed mood was found to interact with cigarette smoking, and together depressed mood and cigarette smoking were associated with a marked increase in the relative risk of cancer (Linkins & Comstock, 1990). As compared to the risk seen in never smokers who were without depressed mood, smokers with depressed mood as measured by elevated scores on the Clinical Epidemiological Scale for Depression had a relative risk of 18.6 for cancers at sites associated with smoking and a relative risk of 2.9 for cancers at sites not associated with smoking. In contrast, smokers who were without depressed mood showed only a relative risk of 4.2 for cancers at sites associated with smoking and no increase in relative risk of cancers at sites not associated with smoking.

4. POSSIBLE CLINICAL IMPLICATIONS OF IMMUNE CHANGES IN DEPRESSION The immunologic consequences of major depression and their possible clinical relevance to infectious diseases have not been delineated (Stein et al., 1991). Major

M. Irwin

18

depression has been associated with alterations in the distribution of T cell subsets and with declines in nonspecific measures of immune function, such as NK activity and mitogen-induced lymphocyte proliferation. The clinical significance of these immunologic findings is uncertain because the in vitro assays employed were non-specific and thus not directly relevant to specific disease endpoint. To address these issues, we have recently examined the effect of major depression on varicella zoster virus (VZV)-specific cellular immunity (Irwin, Costlow, Williams, Artin, Levin, Hayward, & Oxman, 1998). The frequency of cells in the peripheral blood capable of proliferating in response to VZV antigen (VZV-responder cell frequency, VZV-RCF) was determined in patients with major depression and in ageand gender-matched normal controls. In addition, we evaluated VZV-RCF in a group of older adults to determine whether the decline in VZV-RCF observed in major depression was comparable in magnitude to that typically found in older persons who are known to be at increased risk of developing HZ. Comparison of depressed subjects with age- and gender-matched normal controls demonstrated that the mean age and gender distribution were similar in the two groups. The mean age of the depressed subjects was 51.3 ± 15.7 years old and that of the matched normal controls was 51.4 ± 17.2 years. There were 7 men and 4 women in each group. As shown in Figure 5, VZV-specific responder cell frequency was significantly lower in the subjects with major depression than in matched normal controls. To assess the possible cUnical significance of the reduced VZV-RCF observed in the depressed subjects, VZV-RCF values in the depressed subjects were compared to

12

10.

6.

I a 2.

Major

Matched

Depression

Normal Subjects

Figure 5. Varicella-zoster virus responder cell frequency (VZV-RCF) in depressed subjects and age- and gender-matched normal controls; mean ± SD. Depressed subjects: n = 11; mean age 51.3 ± 15.7 years; range 32-77; Normal controls: n = 11; mean age 51.4 ± 17.2 years; range 28-79.

Immune Correlates of Depression

19

those in the normal older subjects. The results are shown in Figure 6. As reported previously (Hayward & Herberger, 1987), VZV-RCF values decline with increasing age. Normal subjects younger than 60 years of age (n = 7, mean age = 40.1 ± 9.3 years, range 28 to 57) had higher levels of VZV-RCF than normal subjects 60 years of age and older (mean age = 71.2 + 6.7 years. Depressed subjects (mean age = 51.4 ± 8.2 years, range 28 to 79) had a level of VZV-RCF comparable to that of normal subjects more than 20 years their senior. These findings suggest that major depression is associated with a marked decline in VZV-specific cellular immunity, as measured by the frequency of peripheral blood mononuclear cells capable of proliferating in response to VZV antigens. Furthermore, consistent with prior observations, the present study documents an age-related decline in VZV-specific cellular immunity in normal adults. While the results reported here do not directly link depression with an increase in VZV reactivation and in the incidence of HZ, comparable declines in VZV-specific cellular immunity observed in older adults have been correlated with a significant increase in the incidence of HZ and its complications (Oxman & Alani, 1993). The levels of VZV-RCF observed in our depressed

14n

f^ o c V

3 D"

s*

LL * n

12^

m a. «o

10-

T—

a.

O k_

m •o

<1> "O

y.

a

c Ui

<0

a:

a

6-

Ul

CJ

^

42-

i_ Major

Depression

Normal Subjects <60 years

>60 years

Figure 6. Varicella-zoster virus responder cell frequency (VZV-RCF) in depressed subjects and in normal subjects <60 and > 60 years of age; mean ± SD. Depressed subjects: n = 11; mean age 51.3 + 15.7 years; range 32-77; Normal subjects <60 years: n = 7; mean age 40.1 ± 9.3 years; range 28-57; Normal subjects > 60 years: n = 35; mean age 71.2 ± 6.7 years; range 60-80.

20

M. Irwin

subjects are similar to those observed in normal adults over the age of 60, in whom the incidence of HZ is more than double that in younger normal adults. Animal models have yielded compelling evidence that behavioral stressors may impact viral diseases, such as herpes simplex, influenza, and coxsackievirus infections, via alterations in immune function (Sheridan, Dobbs, Brown, & Zwilling, 1994). In contrast, there is a paucity of clinical data addressing the confluence of behavioral, immunologic, and outcome variables in the same individual at the same time (KiecoltGlaser & Glaser, 1995). Psychosocial stress appears to be associated with reduced immunologic control of latent herpesviruses (Epstein-Barr virus, herpes simplex virus, and cytomegalovirus) as evidenced by elevated antibody titers (Kiecolt-Glaser & Glaser, 1995). Cohen and colleagues have found that psychological stress is associated with increased rates of respiratory infection after experimental inoculation of common cold viruses (Cohen, Tyrrell, & Smith, 1991; Cohen, Doyle, Skoner, Rabin, & Gwaltney, 1997), and Glaser and colleagues have reported that psychological stress is associated with decreased immune responses to hepatitis B and influenza vaccinations (Glaser, Kiecolt-Glaser, Bonneau, Malarkey, Kennedy, Hughes, 1992; Kiecolt-Glaser, Glaser, Gravenstein, & Malarkey, 1996). Whether these observations generated in psychologically stressed persons will generalize to depressed subjects is not yet known.

5. S U M M A R Y A N D C O N C L U S I O N S The relation of depression to immunological assays is complex and variable. However, meta-analyses have demonstrated that depressed subjects are likely to show changes in several immune assays. Depressed subjects are likely to have changes in major immune cell classes with an increase in total white blood cell counts and a relative increase in numbers of neutrophils. However, the relative number of lymphocytes is likely to be reduced in depressed subjects. Depression also appears to be associated with increases in at least one measure of immune activation, although further investigations are clearly needed to replicate these interesting observations. Finally, depression is reliably associated with a suppression of mitogen-induced lymphocyte proliferation and with a reduction of NK activity. Despite the heterogeneity of findings, the effect sizes in the relationship between depression and lymphocyte proliferation and NK activity are large as compared to those observed in other areas of psychological and medical research. Several moderating factors may explain and account for the heterogeneity that has been found in the depression-immune results. Future immunologic studies in depressed subjects are needed to clarify the effects of gender and reproductive hormones on the relation between depression and immunity. Severity of melancholic symptoms and sleep disturbance appear to moderate the immune changes in depression but the biological mechanisms that account for the link between these neurovegetative symptoms and depression are not yet known. Finally, assessment of co-morbidity in depressed subjects deserves an increased focus. Data generated from our laboratory clearly show that assessment of alcohol- and tobacco dependence is critical in the interpretation of immune changes in depressed subjects. The clinical significance of changes in immune responses in depressed subjects remains an unanswered question. Studies that use immune measures with disease specific endpoints, as has been recently conducted in the study of VZV immune

Immune Correlates of Depression

21

responses, would help identify the possible link between depression, immune system alterations, and health outcomes.

ACKNOWLEDGMENTS The work reported in this chapter was supported by the following grants: NIMH5-T32-18399, NIH-2-P50-30914, NIH-2M01-RR00827, NIAAA-10215.

REFERENCES Albrecht, J., Helderman, J., Schlesser, M., & Rush, J. (1985). A controlled study of cellular immune function in affective disorders before and during somatic therapy. Psychiatry Research, 15, 185-193. Altshuler, L. L., Plaeger-Marshall, S., Richeimer, S., Daniels, M., & Baxter, L. R., Jr. (1989). Lymphocyte function in major depression. Acta Psychiatrica Scandinavica, 80, 132-136. American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. Washington, D.C.: American Psychiatric Press. Andreoli, A., Keller, S. E., Rabaeus, M., Zaugg, L., Garrone, G., & Taban, C. (1992). Immunity, major depression, and panic disorder comorbidity. Biological Psychiatry, 31, 896-908. Andreoli, A. V., Keller, S. E., Rabaeus, M., Marin, P, Bartlett, J. A., & Taban, C. (1993). Depression and immunity: age, severity, and clinical course. Brain, Behavior & Immunity, 7, 279-292. Bartoloni, C , Guidi, L., Antico, L., Pili, R., Cursi, P., Carbonin, P., Gambassi, G., Rumi, C., Di Giovanni, A., Menichella, G., & et ai. (1990). Psychological status of institutionalized aged: influences on immune parameters and endocrinological correlates. International Journal of Neuroscience, 51, 279-281. Benca, R. M., Obermeyer, W. H.,Thisted, R. A., & Gillin, J. C. (1992). Sleep and psychiatric disorders: a meta analysis. Archives of General Psychiatry, 49, 651-668. Birmaher, B., Rabin, B. S., Garcia, M. R., Jain, U., Whiteside, T. L., Williamson, D. E., al-Shabbout, M., Nelson, B. C , Dahl, R. E., & Ryan, N. D. (1994). Cellular immunity in depressed, conduct disorder, and normal adolescents: role of adverse life events. Journal of American Academy of Child and Adolescent Psychiatry, 33, 67\-67H. Born, J., Lange, T., Hansen, K., Molle, M., & Fehm, H. L. (1997). Effects of sleep and circadian rhythm on human circulating immune cells. Journal of Immunology, 158, 4454-4464. Breslau, N., Peterson, E. L., Schultz, L. R., Chilcoat, H. D., & Andreski, P. (1998). Major depression and stages of smoking. A longitudinal investigation. Archives of General Psychiatry, 55, 161-166. Brown, G. W. & Harris, T. (1978). Social Origins of Depression: A Study of Psychiatric Disorders in Women. New York: The Free Press. Brown, G. W., Harris, T O., & Hepworth, C. (1994). Life events and endogenous depression. Archives of General Psychiatry, 51, 525-534. Cadoret, R. J. (1981). Depression and alcoholism. In R. E. Meyer, B. C. Glueck, & J. E. OBrien (Eds.), Evaluation of the alcoholic. Rockville, Md: Nat Institute on Alcohol Abuse and Alcoholism. Calabrese, J. R., Skwerer, R. G., Barna, B., Gulledge, A. D., Valenzuela, R., Butkus, A., Subichin, S., & Krupp, N. E. (1986). Depression, immunocompetence, and prostaglandins of the E series. Psychiatry Research, 77,41-47. Caldwell, C. L., Irwin, M., & Lohr, J. (1991). Reduced natural killer cell cytotoxicity in depression but not in schizophrenia. Biological Psychiatry, 30, 1131-1138. Chiappelli, F. & Gottesfeld, Z. (1995). Introduction to the symposium on alcohol and T-cell immunity [comment]. Alcoholism: Clinical & Experimental Research, 19, 535-538. Cleary, P. & Guy, W. (1977). Factor analysis of the Hamilton Depression Scale. Drugs Experimental Clinical Research, 1, U5-\20. Cohen, S., Tyrrell, D. A., & Smith, A. P. (1991). Psychological stress and susceptibility to the common cold. New England Journal of Medicine, 325(9), 606-656. Cohen, S., Doyle, W J., Skoner, D. P, Rabin, B. S., & Gwaltney J. M. (1997). Social ties and susceptibility to the common cold. Journal of the American Medical Association, 277, 1940-1944. Cosyns, P., Maes, M., Vandewoude, M., Stevens, W. J., DeClerck, L. S., & Schotte, C. (1989). Impaired mitogen-

22

M. Irwin

induced lymphocyte responses and the hypothalamic-pituitary-adrenal axis in depressive disorders. Journal of Affective Disorders, 16, 41-48. Cover, H. & Irwin, M. (1994). Immunity and depression: Insomnia, retardation, and reduction of natural killer cell acXWiiy. Journal of Behavioral Medicine, 17, 217-223. Darko, D. F., Gillin, J. C , Risch, S. C , Bulloch, K., Golshan, S., Tasevska, Z., & Hamburger, R. N. (1989). Mitogen-stimulated lymphocyte proliferation and pituitary hormones in major depression. Biological Psychiatry, 26, 145-155. Dietrich, J. B., Chasserot-Golaz, S., Beck, G., & Bauer, G. (1986). Antagonism of glucocorticoid induction of Epstein-Barr virus early antigens by different steriods in Daudi lymphoma cells. Journal of Steroid Biochemistry, 24, 417-421. Dinges, D. F., Douglas, S. D., Zaugg, L., Campbell, D. E., McMann, J. M., Whitehouse, W. G., Orne, E. C , Kapoor, S. C , Icaza, E., & Orne, M. T. (1994). Leukocytosis and natural killer cell function parallel neurobehavioral fatigue induced by 64 hours of sleep deprivation. Journal of Clinical Investigation, 93, 1930-1939. Evans, D. L., Folds, J. D., Petitto, J. M., Golden, R. N., Pedersen, C. A., Corrigan, M., Gilmore, J. H., Silva, S. G., Quade, D., & Ozer, H. (1992). Circulating natural killer cell phenotypes in men and women with major depression. Archives of General Psychiatry, 49, 388-395. Ferson, I., Edwards, A., Lind, A., Milton, G. W., & Hersey, P. (1979). Low natural killer-cell activity and immunoglobulin levels associated with smoking in human subjects. British Journal of Cancer, 23, 603-609. Glaser, R., Kiecolt-Glaser, J. K., Bonneau, R. H., Malarkey, W., Kennedy, S., & Hughes, J. (1992). Stressinduced modulation of the immune response to recombinant hepatitis B vaccine. Psychosomatic Medicine, 54, 22-29. Hall, M., Baum, A., Buysse, D. J., Prigerson, H. G., Kupfer, D. J., & Reynolds, C. K (1998). Sleep as a mediator of the stress-immune relationship. Psychosomatic Medicine, 60, 48-51. Hayward, A. R. & Herberger, M. (1987). Lymphocyte responses to varicella zoster virus in the t\der\y. Journal of Clinical Immunology, 7, 174-178. Herbert, T. B. & Cohen, S. (1993a). Depression and immunity—A meta-analytic review. Psychological Bulletin, 113, 472-486. Herbert, T. B. & Cohen, S. (1993b). Stress and immunity in humans: a meta-analytic review. Psychosomatic Medicine, 55, 364-379. Hickie, I., Hickie, C , Lloyd, A., Silove, D., & Wakefield, D. (1993). Impaired in vivo immune responses in patients with melancholia. British Journal of Psychiatry, 162, 651-657. Irwin, M. & Gillin, J. C. (1998). The neuroimmunology of normal and disturbed sleep. European Sleep Research Society, 1998, (abstract). Irwin, M., Smith, T. L., & Gillin, J. C. (1987). Low natural killer cytotoxicity in major depression. Life Sciences, 41,2127-2133. Irwin, M., Caldwell, C , Smith, T. L., Brown, S., Schuckit, M. A., & Gillin, J. C. (1990a). Major depressive disorder, alcoholism, and reduced natural killer cell cytotoxicity: Role of severity of depressive symptoms and alcohol consumption. Archives of General Psychiatry, 47, 713-719. Irwin, M. R., Patterson, T. L., Smith, T. L., Caldwell, C , Brown, S. A., Gillin, J. C , & Grant, I. (1990b). Reduction of immune function in life stress and depression. Biological Psychiatry, 27, 22-30. Irwin, M., Lacher, U., & Caldwell, C. (1992a). Depression and reduced natural killer cytotoxicity: a longitudinal study of depressed patients and control subjects. Psychological Medicine, 22, 1045-1050. Irwin, M., Smith, T. L., & Gillin, J. C. (1992b). Electroencephalographic sleep and natural killer activity in depressed patients and control subjects. Psychosomatic Medicine, 54, 107-126. Irwin, M., Mascovich, A., Gillin, J. C , Willoughby, R., Pike, J., & Smith, T. L. (1994). Partial sleep deprivation reduces natural killer cell activity in humans. Psychosomatic Medicine, 56, 493^98. Irwin, M., McCHntick, J., Costlow, C , Fortner, M., White, J., & Gillin, J. C. (1996). Partial night sleep deprivation reduces natural killer and cellular immune responses in humans. The FASEB Journal, 10, 643-653. Irwin, M., Costlow, C , Williams, H., Artin, K. H., Levin, M. J., Hayward, A. R., & Oxman, M. N. (1998). Cellular immunity to varicella-zoster virus in depression. Journal of Infectious Disease, in press. Jung, W. & Irwin, M. (1998). Reduction of natural killer cytotoxic activity in major depression. Submitted. Kiecolt-Glaser, J. K. & Glaser, R. (1995). Psychoneuroimmunology and health consequences: data and shared mechanisms. Psychosomatic Medicine, 57, 269-274. Kiecolt-Glaser, J. K., Glaser, R., Gravenstein, S., & Malarkey, W. B. (1996). Chronic stress alters the immune

Immune Correlates of Depression

23

response to influenza virus vaccine in older adults. Proceedings of the National Academy of Sciences, 93, 3043-3047. Kronfol, Z. & House, J. D. (1985). Depression, hypothalamic-pituitary adrenocortical activity, and lymphocyte function. Psychopharmacology Bulletin, 21, 476^78. Kronfol, Z. & House, J. D. (1989). Lymphocyte mitogenesis, immunoglobulin, and complement levels in depressed patients and normal controls. Acta Psychiatrica Scandinavica, 80, 142-147. Kronfol, Z., House, J. D., Silva, J., Greden, J., & Carroll, B. J. (1986). Depression, urinary free Cortisol excretion, and lymphocyte function. British Journal of Psychiatry, 148, 10-13. Kronfol, Z., Silva, J., Greden, J., Dembinski, S., Gardner, R., & Carroll, B. (1983). Impaired lymphocyte function in depressive illness. Life Sciences, 33, 241-247. Kronfol, Z., Turner, R., Nasrallah, H., & Winokur, G. (1984). Leukocyte regulation in depression and schizophrenia. Psychiatry Research, 13, 13-18. Kronfol, Z., Nair, M., Goodson, J., Goel, K., Haskett, R., & Schwartz, S. (1989). Natural killer cell activity in depressive illness: a preliminary report. Biological Psychiatry, 26, 753-756. Kronfol, Z., Nair, M., Zhang, Q., Hill, E. E., & Brown, M. B. (1997). Circadian immune measures in healthy volunteers: relationship to hypothalamic-pituitary-adrenal axis hormones and sympathetic neurotransmitters. Psychosomatic Medicine, 59, 42-50. Linkins, R. W. & Comstock, G. W. (1990). Depressed mood and development of cancer. American Journal of Epidemiology, 132(5), 962-972. Lowy, M. T., Reder, A. T., Gormley, G. J., & Meltzer, H. Y. (1988). Comparison of in vivo and in vitro glucocorticoid sensitivity in depression: relationship to the dexamethasone suppression test. Biological Psychiatry, 24, 619-630. Maes, M., Bosmans, E., Suy, E., Minner, B., & Raus, J. (1989). Impaired lymphocyte stimulation by mitogens in severely depressed patients. A complex interface with HPA-axis hyperfunction, noradrenergic activity, and the ageing process. British Journal of Psychiatry, 155, 793-798. Maes, M., Lambrechts, J., Bosmans, E., Jacobs, J., Suy, E., Vandervorst, C , deJonckheere, C , Minner, B., & Raus. J. (1992a). Evidence for a systemic immune activation during depression: results of leukocyte enumeration by flow cytometry in conjunction with monoclonal antibody staining. Psychological Medicine. 22, 45-53. Maes, M., Scharpe, S., Van Grootel, L., Uyttenbroeck, W., Cooreman,W., Cosyns, P., & Suy, E. (1992b). Higher alpha 1-antritrypsin, haptoglobin, ceruloplasmin, and lower retinol binding protein plasma levels during depression: further evidence for the existence of an inflammatory respose during that illness. Journal of Affective Disorders, 24, 183-192. Maes, M., Stevens, W., Peelers, D., DeClerk, L., Scharpe, S., Bridts, C , Schotte, C , & Cosyns, P (1992c). A study of the blunted natural killer cell activity in severely depressed patients. Life Science, 50, 505-513. Maes, M., Smith, R., & Scharpe, S. (1995). The monocyte-T-lymphocyte hypothesis of major depression. Psychoneuroendocrinology, 20, 111-116. Maier, S. F., Watkins, L. R., & Fleshner, M. (1994). Psychoneuroimmunology: the interface between behavior, brain, and immunity. American Journal of Psychology, 49, 1004-1017. McAdams, C. & Leonard, B. E. (1993). Neutrophil and monocyte phagocytosis in depressed patients. Progess in Neuro-Psychopharmacology and Biological Psychiatry, 17, 971-984. Meliska, C. J., Stunkard. M. E., Gilbert, D. G., Jensen, R. A., & Martinko, J. M. (1995). Immune function in cigarette smokers who quit smoking for 31 days. Journal of Allergy & Clinical Immunology, 95, 901-910. Mendall, M. A., Patel, P, Asante, M., Ballam, L., Morris, J., Strachan, D. P, Camm, A. J., & Northfield, T. C. (1997). Relation of serum cytokine concentrations to cardiovascular risk factors and coronary heart disease. Heart, 78, 273-277. Miller, A. H., Asnis, G. M., Lackner, C , Halbreich, U., & Norin, A. J. (1991). Depression, natural killer cefl activity, and Cortisol secretion. Biological Psychiatry, 29, 878-886. Mohl, P C , Huang, L., Bowden, C , Fischbach, M., Vogtsberer, K., & Talal, N. (1987). Natural killer cell activity in major depression (letter). American Journal of Psychiatry, 144, 1619 Moldofsky, H., Lue, F. A., Davidson, J. R., & Gorczynski, R. (1989). Effects of sleep deprivation on human immune function. The FASEB Journal, 3, 1972-1977. Murphy, J. M., Oliver, D. C , Sobol, A. M., Monson, R. R., & Leighton, A. H. (1986). Diagnosis and outcome: depression and anxiety in a general population. Psychological Medicine, 16, 117-26. Nerozzi, D., Santoni, A., Bersani, G., Magnani, A., Bressan, A., Pasini, A., Antonozzi, I., & Frajese, G. (1989). Reduced natural killer cell activity in major depression: neuroendocrine implications. Psychoneuroendocrinology, 14, 295-302.

24

M. Irwin

Oxman, M. N. & Alani, R. (1993). Varicella and Herpes Zoster. In T. B. Fitzpatrick & et al. (Eds.), Dermatology in General Medicine, (pp. 2543-2572). New York: McGraw-Hill Book Co. Palmblad, J., Petrini, B., Wasserman, J., & Akerstedt,T. (1979). Lymphocyte and granulocyte reactions during sleep deprivation. Psychosomatic Medicine, 41(4), 273-278. Parker, G., Hadzi-Pavlovic, D., Boyce, P,Wilhelm, K., Brodaty, H., Mitchell, P, Hickie, I., & Eyers, K. (1990). Classifying depression by mental state signs. British Journal of Psychiatry, 157, 55-65. Petitto, J. M., Folds, J. D., Ozer, H., Quade, D., & Evans, D. L. (1992). Abnormal diurnal variation in circulating natural killer cell phenotypes and cytotoxic activity in major depression. American Journal of Psychiatry, 149, 694-696. Petitto, J. M., Folds, J. D., & Evans, D. L. (1993). Abnormal diurnal variation of B lymphocyte circulation patterns in Major Depression. Biological Psychiatry, 34, 268-267. Petrovsky, N. & Harrison, L. C. (1997). Diurnal rhythmicity of human cytokine production. Journal of Immunology, 158, 5163-5168. Rapaport, M. H. & Irwin, M. (1996). Serum soluble interleukin-2 receptors and natural killer cell function in major depression. Communications in Biological Psychiatry, 21, 73-78. Schleifer, S. J., Keller, S. E., Meyerson, A. T , Raskin, M. J., Davis, K. L., & Stein, M. (1984). Lymphocyte function in major depressive disorder. Archives of General Psychiatry, 41, 484-486. Schleifer, S. J., Keller, S. E., Siris, S. G., Davis, K. L., & Stein, M. (1985). Depression and immunity: lymphocyte function in ambulatory depressed patients, hospitalized schizophrenic patients, and patients hospitalized for herniorrhaphy. Archives of General Psychiatry, 42, 129-133. Schleifer, S. I , Keller, S. E., Bond, R. N., Cohen, J., & Stein, M. (1989). Major depressive disorder and immunity: role of age, sex, severity, and hospitalization. Archives of General Psychiatry, 46, 81-87. Schuckit, M. A. (1986). Genetic and clinical implications of alcoholism and affective disorder. American Journal of Psychiatry, 143, 140-147. Sengar, D. P., Waters, B. G., Dunne, J. V., & Bover, I. M. (1982). Lymphocyte subpopulations and mitogenic responses of lymphocytes in manic-depressive disorders. Biological Psychiatry, 17, 1017-1022. Shain, B. N., Kronfol, Z., Naylor, M., Goel, K., Evans, T., & Schaefer, S. (1991). Natural killer cell activity in adolescents with major depression. Biological Psychiatry, 29, 481-484. Sheridan, J. F., Dobbs, C , Brown, D., & Zwilling, B. (1994). Psychoneuroimmunology: stress effects on pathogenesis and immunity during infection. Clinical Microbiological Reviews 7, 200-212. Steiger, A. & Holsboer, F. (1997). Nocturnal secretion of prolactin and Cortisol and the sleep EEG in patients with major endogenous depression during an acuted episode and after full remission. Psychiatry Research, 72, 81-88. Stein, M. B. & Uhde, T. W. (1988). Panic disorder and major Depression. A tale of two syndromes. Psychiatry Clinics of North America, 11, 441^61. Stein, M., Miller, A. H., & Trestman, R. L. (1991). Depression, the immune system, and health and illness. Archives of General Psychiatry, 48, 171-177. Syvalahti, E., Eskola, J., Ruuskanen, O., & Laine,T. (1985). Nonsuppression of Cortisol and immune function in depression. Proceedings in Neuropsychopharmacology & Biological Psychiatry, 9(4), 413-422. Uthgenannt, D., Schoolmann, D., Pietrowsky, R., Fehm, H. L., & Born, J. (1995). Effects of sleep on the production of cytokines in humans. Psychosomatic Medicine, 57, 97-104. Weisse, C. S. (1992). Depression and immunocompetence: A review of the literature. Psychological Bulletin, 111(3), 475-489. Zorilla, E. P., Luborsky, L., McKay, J. R., Rosenthal, R., Houldin, A., Tax, A., McCorkle, R., Seligman, D. A., & Schmidt, K. (1998). The relationship of depression and stressors to immunological assays: a meta-analytic review. Brain, Behavior and Immunity, (in press).

MAJOR DEPRESSION AND ACTIVATION OF THE INFLAMMATORY RESPONSE SYSTEM Michael Maes* Clinical Research Center for Mental Health (CRC-MH) Antwerp, Belgium IRCCS, Istituto Fatebenefratelli, Brescia, Italy, and Department of Psychiatry, Vanderbilt University, Nashville

1. I N T R O D U C T I O N Contemporary models of major depression emphasize the role of hypothalamicpituitary-adrenal (HPA)-axis hyperactivity and of dysfunctions in the turnover of serotonin (5-HT) or catecholamines in the etiopathogenesis of major depression. Contemporary models of major depression do not incorporate the effects of the inflammatory response system (IRS), even though the IRS powerfully influences HPA-axis activity, 5-HT and catecholaminergic turnover and even though activation of the IRS may induce depression-like behavior in animals and humans. There is now evidence that major depression is accompanied by a moderate activation of the IRS (reviews: Maes, 1993; 1995; 1997; Maes, Smith, & Scharpe, 1995c; Holden, Pakula, & Mooney, 1997; Maes & Smith, 1997; Connor & Leonard, 1998; Maier & Watkins, 1998). In this paper we propose a concise IRS model of major depression.

2. INDICATORS OF IRS ACTIVATION IN DEPRESSION 2.1. IRS Activation Indicators of IRS activation in major depression are confirmed findings of increased numbers of leukocytes, monocytes, neutrophils, activated T-lymphocytes, increased secretion of neopterin and prostaglandins. Herbert & Cohen (1993) showed.

* Mailing Address: Michael Maes, M.D., Ph.D., Director Clinical Research Center for Mental Health, Antwerp University Department of Psychiatry, 267 Lange Beeldekensstraat, 2060 Antwerp, Belgium, FAX: 32-3-4483265. e-mail: [email protected] Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

25

26

M. Maes

in a meta-analytical review, that major depression is associated with a significant increased number of leukocytes and neutrophils. There are also reports on monocytosis, increased CD47CD8^ T cell ratio and increased numbers of activated T cells, such as CD25^ and HLA-DR^ T cells, in depression (Muller, Ackenheil, & Hofschuster, 1989; Muller, Hofschuster, Ackenheil, Mempel, & Eckstein, 1993; Maes, Lambrechts, Bosmans, Jacobs, Suy, Vandervorst, De Jonckheere, Minner, & Raus, 1992a; Maes, Scharpe, Meltzer, & Cosyns, 1993d; Perini, Zara, Carraro, Tosin, Gava, Santucci, Valverde, & Defranchis, 1995; Seidel, Arolt, Hunstiger, Rink, Behnisch, & Kirchner, 1995; Seidel, Arolt, Hunstiger, Rink, Behnisch, & Kirchner, 1996a; Deger, Bekaroglu, Orem, Orem, Uluutku, & Soylu, 1996; review: Maes, 1997). Increased serum concentrations of the soluble interleukin-2-receptor (sIL-2R) also indicate in vivo T cell activation in depression (Maes, Bosmans, Suy, Vandervorst, Dejonckheere, Minner, & Raus, 1991a; Maes, Bosmans, Suy, Vandervorst, Dejonckheere, & Raus, 1991b; Maes, Meltzer, Bosmans, Bergmans, Vandoolaeghe, Ranjan, & Desnyder, 1995a; Nassberger & Traskman-Bendz, 1993; Sluzewska, Rybakowski, Bosmans, Sobieska, Berghmans, Maes, & Wiktorowicz, 1996a). Other direct indicators of IRS activation in depression are increased concentrations of prostaglandin E2 (PGE2) in serum, cerebro-spinal fluid (CSF), and mitogen-stimulated culture supernatant (Lieb & Karmali, 1983; Linnoila, Whorton, Rubinow, Cowdry, Ninan, & Waters, 1983; Calabrese, Skwerer, Barna, Gulledge, Valenzuela, Butkus, Subichin, & Krupp, 1986; Song, Lin, Bonaccorso, Heide, Verkerk, Kenis, Bosmans, Scharpe, Cosyns, DeJong, & Maes, 1998), and increased serum or urine secretion of neopterin (Duch, Woolf, Nichol, Davidson, & Garbutt, 1984; Dunbar, Hill, Neale, & Mellsop, 1992; Maes, Scharpe, Meltzer, Okayli, D'Hondt, & Cosyns, 1994e; Bonaccorso, Lin, Verkerk, VanHunsel, Libbrecht, Scharpe, DeClerck, Stevens, Biondi, Janca, & Maes, 1998). Neopterin, a low molecular weight notconjugated pteridin, is produced by activated macrophages after stimulation with interferon-y (IFNy) (review: Maes et al., 1994e).

2.2. Acute Phase Response Indicators of an acute phase response in major depression are the confirmed findings on increased serum concentrations of positive acute phase proteins, such as haptoglobin (Hp), al-antitrypsin, ceruloplasmin, al-acid glycoprotein, C-reactive protein, and al-antichymotrypsin; and lowered serum concentrations of negative acute phase proteins, such as albumin (Alb) and transferrin (Tf) (Swartz, 1990; Maes, Vandewoude, Scharpe, De Clerck, Stevens, Lepoutre, & Schotte, 1991c; Maes, Scharpe, Bosmans, Vandewoude, Suy, Uyttenbroek, Cooreman, Vandervorst, & Raus, 1992b; Maes, Scharpe, Van Grootel, Uyttenbroeck, Cooreman, Cosyns, & Suy, 1992c; Maes, De Langhe, Scharpe, Meltzer, Cosyns, Suy, & Bosmans, 1994b; Maes, Scharpe, Neels, Wauters, Van Gastel, & Cosyns, 1995b; Maes, Delanghe, Ranjan, Meltzer, Desnyder, Cooreman, & Scharpe, 1997b; Joyce, Hawes, Mulder, Sellman, Wilson, & Boswell, 1992; Song, Dinan, & Leonard, 1994; Seidel et al., 1995; Sluzewska, Rybakowski, Sobieska, & Wiktorowicz, 1996b; Van Hunsel, Wauters, Vandoolaeghe, Neels, Demedts, & Maes, 1996; Berk, Wadee, Kuschke, & O'Neill-Kerr, 1997; Kronfol, Singh, Boura, & Brown, 1998; Mikova, Stoyanova, & Tanchev, 1998). Increased serum concentrations of elastase in major depression also indicate the presence of an acute phase response (Deger et al., 1996). An acute phase response also occurs in the chronic mild stress model of depression in the rodent and the olfactory bulbectomized rat model of depression (Sluzewska, Nowakowska, Gryska, & Mackiewicz, 1994; Song & Leonard, 1994).

Major Depression and Activation of the Inflammatory Response System

27

2.3. In Vivo Secretion of Cytokines In depression, there are confirmed findings on increased in vivo secretion of proinflammatory cytokines, such as IL-6 and IL-8 (Maes et al., 1995a; Maes, Bosmans, De Jongh, Kenis, Vandoolaeghe, & Neeis, 1997a; Sluzewska, Rybakowski, Laciak, Mackiewicz, Sobieska, & Wiktorowiz, 1995; Sluzewska et al., 1996a; Berk et al., 1997; Frommberger, Bauer, Haselbauer, Fraulin, Riemann, & Berger, 1997; Song et al., 1998) and increased mitogen-induced production of proinflammatory cytokines, such as ILip, IL-6, and IFNy (Maes et al., 1991a; Maes, Bosmans, Meltzer, Scharpe, & Suy, 1993a; Maes, Scharpe, Meltzer, Bosmans, Suy, Minner, Calabrese, Uyttenbroeck, Vandervorst, Raus, & Cosyns, 1993c; Maes et al., 1994e; Seidel et al., 1995; Seidel, Arolt, Hunstiger, Rink, Behnisch, & Kirchner, 1996b). In Wistar rats, an eight-week exposure to mild, unpredictable stress induces a depression-like state with an increased production of IL1 and IL-2 by splenocytes (Kubera, Symbirtsev, Basta-Kaim, Borycz, Roman, Papp, & Claesson, 1996). Since proinflammatory cytokines induce an immune and acute phase response and since we found significant and positive correlations between cytokine production, e.g. IL-6, and indicators of immune activation, e.g. increased numbers of peripheral blood mononuclear cells (PBMC) and acute phase proteins, we have suggested that IRS activation in depression is caused by an increased production of the proinflammatory cytokines, IL-ip, IL-6, and IFNy (Maes et al., 1992a; 1993d; Maes, 1993; 1995; 1997).

3. I N D I R E C T I N D I C A T O R S O F IRS A C T I V A T I O N I N MAJOR DEPRESSION If there is an IRS activation in major depression, it was anticipated to find other indicators of IRS activation, such as lower serum zinc (Zn), and specific alterations in fatty acid metabolism and the erythron.

3.1. Trace Elements IRS activation is accompanied by a fall in serum Zn (Mussalo-Rauhamaa, Konttinen, Lehto, & Honkanen, 1988; Solomons, 1988; Srinivas, Braconier, Jeppsson, Abdulla, Akesson, & Ockerman, 1988). Zn is a trace element, which, in the plasma, is firmly bound to aa-macroglobulin (40%), while the remaining Zn is loosely bound to Alb (55%) or amino acids (5%).The loosely bound Zn fraction provides the Zn delivery to the tissues (Solomons, 1988). There is now evidence that depression is accompanied by lower serum Zn (Little, Castellanos, Humphries, & Austin, 1989; McLoughlin & Hodge, 1990; Maes, Scharpe, D'Haese, De Broe, & Cosyns, 1994d; Maes, Vandoolaeghe, Neels, Demedts, Wauters, Meltzer, Altamura, & Desnyder, 1997d).There are two factors which can explain lower serum Zn in depression. First, because IRS activation results in decreased serum Alb concentrations and Alb is the major Zn binding protein, there is potentially less Zn binding protein available, which could in part explain lower serum Zn (Giroux, Schechter, Schoun, and Sjoerdsma, 1977; Goldblum, Cohen, Jay, & McClain, 1987). However, we found that serum Alb and diagnostic classification had additive effects and independently from each other explained an important part of the variance in serum Zn (Maes et al., unpublished observations). These results suggest that lower serum Zn in depression is in part related to lowered concentrations of its "carrier" protein. Alb, and that another depression-related mech-

28

M. Maes

anism may play a role in lower serum Zn. Second, lowered serum Zn during IRS activation may be secondary to sequestration of the intracellular heavy metal binding protein metallothionein in the liver, which, in turn, may be related to an increased production of the proinflammatory cytokines, IL-1, and IL-6 (Cousins & Leinart, 1988; Morimoto, Murakami, Nakamori, Sakata, & Watanabe, 1989a; Morimoto, Sakata, Watanabe, & Murakami, 1989b; Van Miert, Van Duin, & Wensing, 1990). The decrease in serum Zn during IRS activation is associated with an increase in liver Zn contents (Dunn & Cousins, 1989). Reducing serum Zn concentrations during IRS activation could have an advantage to the host since more Zn could be made available for the increased acute phase protein synthesis in the liver (Solomons, 1988). We found significant inverse relationships between lower serum Zn and markers of IRS activation in depression, e.g. increased CD47CD8* T cell ratio, serum neopterin, and increased serum IL-6 (Maes et al., 1994d; 1997d). Therefore, the relationships between serum Zn and IL-6 in depression, suggests that lower serum Zn is caused by increased production of IL-6 in that illness (Maes et al., 1997d).

3.2. Serum Lipids Other hallmarks of IRS activation are lower serum total cholesterol, high-densitylipoprotein cholesterol (HDL-C), a lower esterified cholesterol: total cholesterol ratio, decreased activity of lecithin: cholesterol acyltransferase (LCAT; EC 2.3.1.43), and specific changes in polyunsaturated fatty acids (PUFAs). There is now some evidence that alterations in fatty acid metabolism and the composition of phospholipids in serum and membranes are involved in the pathophysiology of major depression (Horrobin, 1990; Smith, 1991; Maes, Delanghe, Meltzer, D'Hondt, & Cosyns, 1994a; Maes, Smith, Christophe, Cosyns, Desnyder, & Meltzer, 1996a; Maes, Vandoolaeghe, Neels, Demedts, Wauters, & Desnyder, 1997c; Hibbeln & Salem, 1995; Maes & Smith, 1998; Peet, Murphy, Shay, & Horrobin, 1998). Some (Morgan, Palinkas, Barrett-Connor, & Wingard, 1993; Glueck, Tieger, Kunkel, Hamer, Tracy, & Speirs, 1994), but not all authors (Swartz, 1990; McCallum, Simons, Simons, & Friedlander, 1994) reported an association between lower serum total cholesterol and major depression. Prevention trials designed to lower serum cholesterol levels by diet, drugs, or both were shown to increase the number of deaths due to suicide (review: Maes et al., 1997c). It was shown that men with low cholesterol or serum HDL-C were more likely to have ever made a medically serious suicide attempt (review: Maes et al., 1997c). Depressed patients exhibit significantly reduced esterified cholesterol and lowered HDL-C and HDLC/total cholesterol ratios (Maes et al., 1994a; 1997c). These results suggest an abnormal intake and/or metabolism of fatty acids and a decreased formation of cholesteryl esters in major depression (Maes et al., 1997c). The latter phenomenon may point toward decreased activity of LCAT, since most of the cholesteryl esters in man are formed in serum under the activity of LCAT, which reacts preferentially with free cholesterol of the HDL particles. Since in depression, lower serum HDL-C is related to indicators of IRS activation, such as the CD47CD8^ T cell ratio (negative), and serum Zn and Alb (both positively), we have hypothesized that the changes in HDL-C may be secondary to IRS activation. PUFAs have important effects on inflammatory processes since they are precursors of eicosanoids (C20: 4co6 or arachidonic acid is the most abundant eicosanoid precursor in people consuming a Western diet) and can affect eicosanoid and cytokine formation. It has been shown that an imbalance of co6 to co3 PUFAs may lead to an

Major Depression and Activation of the Inflammatory Response System

29

overproduction of cytokines (Endres, 1993). Supplements of linoleic (C18:2co6) oil increases IL-1 and TNFa secretion (Meydani, Endres, Woods, Goldin, Soo, MorrillLaborde, Dinarello, & Gorbach, 1991; Soyland, Lea, Sandstad, & Drevon, 1994). 0)3 PUFAs, e.g. C20:5(o3 or eicosapentaenoic acid, inhibit the synthesis of eicosanoids such as prostaglandin E2 (Meydani et al., 1991) and reduce the production of proinflammatory cytokines and other signs of IRS activation (Meydani et al., 1991; Espersen, Grunnet, Lervang, Nielsen, Thomsen, Faarvang, Dyerberg, & Ernst, 1992; Endres, 1993; Soyland et al., 1994). Recently, it has been shown that depressed patients show alterations in the CO6/Q)3 ratio, i) Our laboratory was the first to report that depressed patients show an increased C20:4co6/C20:5a)3 ratio in serum phospholipids and cholesteryl esters and significantly decreased total (Ii)co3, C18:3co3 and C20:5co3 fractions in serum cholesteryl esters (Maes et al., 1996a). In a recent study, these findings were replicated (Maes et al., unpublished data): major depression was associated with increased C20: 4co6/C20: 5(D3 and C22:5(o6/C22: 6co3 ratios, but lowered C20:5co3 and C22:5(o3 fractions in phospholipids; and increased C20:4(o6/C20: 5co3 and 2CL)6/ECO3 ratios, but lowered C18:3co3, C20:5CD3, and total (11)0)3 FA fractions in cholesteryl esters, ii) Adams, Lawson, Sanigorski, & Sinclair (1996) detected a significant positive relationship between severity of illness and the C20: 4o)6/C20: 5o)3 ratio in serum phospholipids and in RBC membranes of depressed patients, iii) Peet et al. (1998) reported significantly depleted Xo)3 PUFAs and, in particular C22:6o)3, in cell membranes from RBC of depressive patients. In conclusion, these results show that, in major depression, there is a deficiency of o)3 PUFAs in phospholipids. Therefore, it has been suggested that the increased C20:4o)6/C20:5o)3 ratio and the imbalance in 0)6/o)3 PUFAs in major depression may be related to the increased production of proinflammatory cytokines and eicosanoids in that illness (Maes et al., 1996a; Maes & Smith, 1998). Most importantly, there were significant and positive correlations between serum Zn and C20: 5o)3 and C22:6o)3 fractions in phospholipids; and significant inverse correlations between serum Zn and the i:o)6/So)3, C20:4o)6/C20:5o)3, and C22: 5o)6/C22: 6o)3 ratios in phospholipids (Maes et al., unpublished data). In the rodent, lowered intake of Zn does not affect food intake or weight gain but reduces whole-body accumulation of desaturated and elongated products of linoleic acid (C18:2o)6) and a-linolenic acid (C18:3o)3) (Cunnane, Yang, & Chen, 1993). Desaturase enzymes require Zn as cofactor (Russo, Olivieri, Girelli, Guarini, Pasquallini, Azzini, & Corrocher, 1997). Thus, depletion of long-chain o)3 PUFAs in depression could be related to lower serum Zn, which, in turn is an indicator of IRS activation in that illness. The results suggest that i) there is a metabolic disorder in the elongation and desaturation of fatty acids; and ii) the fatty acid alterations in depression are related to IRS activation in that illness.

3.3. Erython IRS activation is also characterized by alterations in the erythron (Lee, 1983; Fairbanks and Beutler, 1988; Brock, 1994). The specific changes are decreased serum iron (Fe) and Tf and normal or increased ferritin levels, and a reduced number of red blood cells (RBC), lowered hematocrit (Htc), and hemoglobin (Hb) (Lee, 1983; Fairbanks, & Beutler, 1988; Brock, 1994). It has been shown that patients with major depression have significantly lower serum Fe and Tf, and a significantly lower number of RBC, lower Htc and Hb, and a significantly increased number of reticulocytes than normal controls (Maes, Van de Vyvere, Vandoolaeghe, Bril, Demedts, Wauters, & Neels, 1996b; Vandoolaeghe, DeVos, DeSchouwer, Neels, & Maes, 1999). Significant relationships were

30

M. Maes

reported between erythron variables and indicators of IRS activation. For example, there are significant and positive correlations between serum Zn and number of RBC, Htc, Hb (all positive), and serum ferritin (negative), and between serum Alb and RBC, Htc and Hb (all positive). There are also significant correlations between serum Zn and Fe and serum Tf (positive), serum Alb and Fe (positive), serum Alb and Tf (positive), and the a,-globulin fraction and Fe (negative). Finally, there were significant and positive correlations between the number of reticulocytes and number of leukocytes and neutrophils and the arglobulin fraction (Maes et al., 1996b; Vandoolaeghe et al., 1999). Proinflammatory cytokines, such as IL-1 and IL-6 modulate iron metabolism and the erythron through increased storage of Fe, reduced release of Fe from the reticuloendothelial cells, increased Fe incorporation into ferritin, increased ferritin synthesis, failure to deliver Fe to the erythron and a reduction in erythrocyte life span (review: Maes et al., 1996b). The physiological function of the alterations in Fe metabolism during the acute phase response is still a matter of debate. It is interesting to note that one hypothesis is that hypoferremia may serve to increase the activity of IFNy (Weiss, Fuchs, Hausen, Reibnegger, Werner, Werner-Felmayer, & Wachter, 1992), which has been reported to occur in depression. Therefore, we have hypothesized that disorders in the erythron indicate IRS activation in major depression (Maes et al., 1996b; Vandoolaeghe et al., 1999).

4. NEUROENDOCRINE DISORDERS IN MAJOR DEPRESSION AND IRS ACTIVATION IRS activation not only involves specific immune (mentioned under 2.1. and 2.3.) and metabolic alterations (mentioned under 2.2. and 3), but also neuroendocrine changes such as hyperactivity of the HPA axis, and alterations in HP-thyroid (HPT) axis function and in the peripheral and central turnover of 5-HT.

4.1. Neuroendocrine Effects of Cytokines 4.1.1. Hypothalamic-Pituitary-Adrenal Axis. IL-1 and IL-6 exert potent enhancing effect on HPA-axis by stimulating hypothalamic corticotropin-releasing hormone (CRH), pituitary adrenocorticotropic hormone (ACTH), and adrenal steroidogenesis (Berkenbosch, Oers, & del Rey, 1987; Sapolsky, Rivier, & Yamamoto, 1987; Navarra, Tsagarakis, Faria, Rees, Besser, & Grossman, 1991). IL-6 may stimulate the HPA-axis at concentrations known to occur in human plasma (Navarra et al., 1991) and sufficiently to exert its actions on immune cells (Carmeliet, Vankelecom, Van Damme, Billiau, & Denef, 1991). It is reported that IL-l(3-induced ACTH, corticosterone, and IL-6 production is mediated by IL-1 type I receptors (Van Dam, Malinowsky, Lenczowski, Bartfai, & Tilders, 1998). It has been shown that HPA-axis hyperactivity during IRS activation is monokine-mediated and may result from monokine-induced serotonergic and catecholaminergic turnover (Dunn, 1988; Dunn, Powell, Meitin, & Small, 1989). Moreover, It is equally well established that some proinflammatory cytokines, such as IL-1, may induce resistance to the effects of glucocorticoid hormones, by influencing glucocorticoid receptor (GR) expression or GR translocation from the cytoplasm to the nucleus (Miller, Spencer, Pearce, Pisell, Tanapat, Leung, Dhabhar, McEwen, & Biron, 1997).

Major Depression and Activation of the Inflammatory Response System

31

4.1.2. Hypothalamic-Pituitary-Thyroid Axis. A wide spectrum of alterations in HPT-axis function has been observed in patients with IRS activation or systemic nonthyroidal illnesses, caused by infection, sepsis, or injury (Kushner, 1982; Kaptein, 1986; Nicoloff & LoPresti, 1991). Abnormal low total T3 or T4, lowered basal TSH, and increased free T4 concentrations, may be observed in those conditions (Wartofsky & Burman, 1982; Wehmann, Gregerman, Burns, Saral, & Santos, 1985; Kaptein, 1986; Nicoloff & LoPresti, 1991; Hermus, Sweep, van der Meer, Ross, Smals, Benraad, & Kloppenborg, 1992). IL-ip and IL-6 are important mediators of the IRS-related alterations in HPT-axis function, e.g. IL-6 administration may suppress serum basal TSH (Dubuis, Dayer, Siegrist-Kaiser, & Burger, 1988; Fujii, Sato, Ozawa, Kasono, Imamura, Kanaji, Tsushima, & Shizume, 1989; Nicoloff & LoPresti, 1991; Hermus et al., 1992; Spath-Schwalbe, Hansen, Schmidt, Schrezenmeier, Marshall, Burger, Fehm, & Born, 1998). 4.1.3. Serotonin. Proinflammatory cytokines have profound effects on the peripheral and brain serotonergic systems. Peripherally and central administration of IL-lp, IFNy, and TNFa increase extracellular 5-HT concentrations in several brain areas such as the hypothalamus, the hippocampus, and the cortex (Clement, Buschmann, Rex, Grote, Opper, Gemsa, & Wesemann, 1997). IL-ip modulates the activity of the 5-HT transporter (Ramamoorthy, Ramamoorthy, Prasad, Bhat, Mahesh, Leibach, & Ganapathy, 1995), which plays a central role in serotonergic neurotransmision by reuptake of 5-HT. Proinflammatory cytokines, such as IL-1 and IFNy may induce the activity of indoleamine-2,3-dioxygenase (IDO), the first enzyme of the kynurenine pathway, which converts tryptophan, the precursor of 5-HT, to kynurenic acid and quinolinic acid. Induction of IDO, which occurs in infection or inflammation, may be detrimental because it leads to depletion of the plasma concentrations of Ltryptophan and reduced synthesis of 5-HT in the brain (dependent partly of the plasma availability of tryptophan) with concomitant neurological effects like lowering of mood (Heyes, Saito, Crowley, Davis, Demitrack, Der, Dilling, Elia, Kruesi, Lackner, Larsen, Lee, Leonard, Markey, Martin, Milstein, Mouradian, Pranzatelli, Quearry, Salazar, Smith, Strauss, Sunderland, Swedo, & Tourtellotte, 1992). Plasma concentrations of tryptophan and the ratio of tryptophan to the sum of amino-acids known to compete for the same cerebral uptake mechanism (i.e. competing amino acids, CAA) are lower in major depressed patients than in normal volunteers (review: Maes & Meltzer, 1995). Brain 5-HT synthesis depends, in part, on the availability of plasma tryptophan, as indexed by total tryptophan plasma concentrations or the molar ratio of tryptophan to the grouped CAA (Maes & Meltzer, 1995).

4.2. Neuroendocrine Function, Cytokines, and Depression Major depression is accompanied by HPA-axis hyperactivity, HPT-axis alterations, such as lower basal TSH concentrations, and serotonergic disturbances. The most consistently reported signs of HPA-axis hyperactivity in major depression are endogenous hypercortisolemia and the failure to suppress plasma Cortisol with the 1 mg DST (review: Maes et al., 1993a). The most consistent sign of HPT-axis dysfunction in depression is lower basal TSH (review: Maes et al., 1993a). There is converging evidence that disorders in peripheral and central serotonin (5-HT) activity are implicated in the pathophysiology of major depression (review: Maes & Meltzer, 1995): i) dysfunctions in the central presynaptic 5-HT neurons, which are, in part, related to a lowered avail-

32

M. Maes

ability of plasma L-tryptophan; and (adaptive?) changes in postsynaptic receptors such as increased number, affinity or responsivity of postsynaptic 5-HT2 receptors; and downregulated or desensitized postsynaptic 5-HTlA receptors. Therefore, we have hypothesized that, if major depression is indeed characterized by IRS activation, the glucocorticoid resistance in depression, lower serum basal TSH concentrations, and lower availabihty of tryptophan to the brain may be related to indicators of IRS activation. In accordance with this hypothesis we found the following, i) In depressed patients, there is a significant positive correlation between postdexamethasone Cortisol values and mitogen-stimulated IL-6 or IL-1 production, and a significant positive correlation between baseline plasma Cortisol and IL-6 concentrations (Maes et al., 1993a; 1993c). ii) In major depression, but not in normal controls, significant inverse relationships were found between serum basal TSH and plasma Hp concentrations (Maes, Scharpe, Cosyns, & Meltzer, 1994c). iii) Lower availability of plasma tryptophan to the brain was significantly correlated to serum IL-6, serum Hp, the a2 globulin fractions, neopterin and the CD47CD8^ T cell ratio (inversely), and to serum Alb, Fe, Tf, and Zn (all positively) (Maes et al., 1994c; Maes, Verkerk, Vandoolaeghe, Van Hunsel, Neels, Wauters, Demedts, & Scharpe, 1997e). Therefore, we have hypothesized that the most prominent neuroendocrine disorders in depression, i.e. HPA-axis hyperactivity, lower serum basal TSH, and lower tryptophan availability to the brain could be induced by IRS activation.

5. CYTOKINES AND THE ETIOLOGY OF MAJOR DEPRESSION 5.1. Proinflammatory Cytokines Induce Depression-like Effects Another question is whether—beside the possible role of proinflammatory cytokines in generating the IRS response and neuroendocrine changes in major depression—hypersecretion of these cytokines may contribute to the etiology of major depression. It has been suggested that increased production of proinflammatory cytokines may play a role in the etiopathology and symptomatogy of depression (Smith, 1991; Yirmiya, 1996; 1997; Maes, 1997). Acute or repeated administration of LPS, IL-1, and IL-6 to the rodent may induce "sickness behavior", a symptom complex characterized by anorexia, weight loss, sleep disorders, suppression of social, locomotor, and exploratory behavior and anhedonia, all key symptoms of major depression (Maier & Watkins, 1995; 1998; Anisman, Kokkinidis, Borowski, & Merali, 1998; Dantzer, Bluthe, Laye, Bret-Dibat, Parnet, & Kelley, 1998; Linthorst & Reul, 1998). Both IL-lp and tumour necrosis factor-alpha (TNFa) induced "anxiogenic-like" effects on the elevated plus maze, whereas interleukin-2 and interleukin-6 did not (Connor, Song, Leonard, Merali, & Anisman, 1998). Administration of interferons, including IFNy, results in behavioral effects and mood alterations, similar to those observed in major depression, such as anhedonia, anorexia, weight loss, anxiety, social withdrawal, psychomotor retardation, anergy, irritability, sleep disturbances, and malaise (Gutterman, Fein, Quesada, Horning, Levine, Alexanian, Bernhardt, Kramer, Spiegel, Colburn, Trown, Merigan, & Dziewanowski, 1982; Weinberg, Schulteis, Fernando, Bakhit, & Martinez, 1988; Smith, 1991). Valentine, Meyers, Kling, Richelson, & Hauser (1998) reported that the risk factors for development of IFNa-induced depression and cognitive changes include duration of treatment, high-dose therapy, and past or current psychiatric illness. Possible pharmacologic interventions to decrease depressive symptoms associated with

Major Depression and Activation of the Inflammatory Response System

33

IFNa therapy include antidepressants (Valentine et al., 1998). Endotoxin administration to rats decreased the free consumption of saccharin, a model of anhedonia or the inability to experience pleasure (Yirmiya, 1996). Sustained administration of IL-6, produced by infecting healthy MRL +/+, C3H.SW and Balb/C mice with adenovirus vector carrying cDNA for murine IL-6, resulted in high serum IL-6 levels over 5 days, and a rapid decline in preference for sucrose. These findings suggest that altered motivation/emotional states occur after infection, and that sustained IL-6 production is an early mechanism in behavioral alterations during chronic inflammatory conditions (Sakic, Szechtman, Braciak, Richards, Gauldie, & Denburg, 1997). In major depression, we found significant relationships between indicators of IRS activation and the vegetative symptoms of depression, i) Serum Alb (negatively) or the a l and a2 globulin fractions (positively) were significantly related to psychomotor retardation, anorexia, and middle insomnia, ii) Serum Zn was significantly and inversely related to psychomotor retardation, iii) Plasma Hp concentrations were significantly and positively related to psychomotor retardation, anorexia, weight loss, anergy, loss of interest, and middle insomnia (Maes, Meltzer, Scharpe, Uyttenbroeck, Cooremans, & Suy, 1993b). It should be underscored that no significant correlations were found between the affective (e.g. depressed mood, a distinct quality of mood, nonreactivity) and cognitive (e.g. feelings of guilt, suicidal ideation) symptoms of depression and any of the IRS activation indicators in depression. Therefore, we have hypothesized that the vegetative symptom profile of major depression may, in fact, be related to IRS activation, through hypersecretion of proinflammatory cytokines (Maes et al., 1993b).

5.2. External Stress, Cytokines, and the Etiology of Depression External or psychosocial stressors, such as negative life events, and internal stressors, such as trauma, infection, autoimmune disease, and other physical diseases or conditions, are emphasized in the etiology of depression. The IRS activation model of depression provides a mechanism for the psychosocial (external stress) and organic (internal stress) etiology of major depression. External stressors in experimental animals and in human may induce the production of proinflammatory cytokines. 5.2.1. External Stressors and Cytokines in Animals. Rats exposed to electric footshock, physical restraint or a conditioned, aversive stimulus have increased concentrations of plasma IL-6 (Zhou, Kusnecov, Shurin, DePaoli, & Rabin, 1993). Restraint stress significantly enhances the expression of IL-6 mRNA and reduces that of the IL-6R mRNA in the midbrain (Shizuya, Komori, Fujiwara, Miyahara, Ohmori, & Nomura, 1997). Immobilization stress increases plasma IL-6 levels, while the peak levels of stressinduced plasma IL-6 in the animals immobilized for 120min are significantly higher than those in the animals subjected to 30min stress (Takaki, Huang, Somogyvari-Vigh, 6 Arimura, 1994). Exposure to an open-field increases plasma IL-6 activity in rats: plasma IL-6 activity is 40.6 ± 7.2U/ml in control rats, 105 ± 6.8U/ml after 30 minutes exposure to an open-field, and 221 ± 17U/ml after 60 minutes of exposure (LeMay, Vander, & Kluger, 1990). In the rodent, immobilization stress increases IL-1 mRNA expression in the hypothalamus only (Minami, Kuraishi, Yamaguchi, Nakai, Hirai, & Satoh, 1991). 10-hour immobilization rises IL-1 and IL-3 activities in supernatants on days 2,5 and 1,2,4,5, 7 after stress, respectively (Khlusov, Dygai, & Gol'dberg, 1993). It was found that immobilization stress enhances biologically active IL-1 in the hypothalamus and that hypo-

34

M. Maes

thalamic IL-1 plays a role in the regulation of stress-induced responses including elevated monoamine release in the hypothalamus and activation of the HPA-axis (Shintani, Nakaki, Kanba, Kato, & Asai, 1995a; Shintani, Nakaki, Kanba, Sato, Yagi, Shiozawa, Aiso, Kato, & Asai, 1995b). These authors suggested that, since 5min is too short a time for the stress to induce production of IL-1, immobilization stress may augment the effects of preexisting IL-1 in the hypothalamus. In Wistar rats, an eightweek exposure to mild, unpredictable stress induces a depression-like state with an increased capacity of splenocytes to produce IL-1 and IL-2 (Kubera, Symbirtsev, BastaKaim, Borycz, Roman, Papp, & Claesson, 1996). Mild inescapable footshocks in rats causes a marked increase of IL-ip and TNFa secretion upon LPS stimulation in vitro, but no changes in the IL-6 production by isolated alveolar macrophages (Persoons, Schornagel, Breve, Berkenbosch, & Kraal, 1995). After 7 days of tail restraint in rats, a 5-fold increase in IL-ip with no significant changes in ACTH and corticosterone concentrations was found (Mekaouche, Givalois, Barbanel, Siaud, Maurel, Malaval, Bristow, Boissin, Assenmacher, & Ixart, 1994). Inescapable shock produces large increases in brain IL-ip protein in the adrenalectomized rats 2 hours after stress, suggesting that elimination of stress-induced rises in corticosterone may unmask a robust and widespread increase in brain IL-1 (3 (Nguyen, Deak, Owens, Kohno, Fleshner, Watkins, & Maier, 1998). Lev. application of the IL-1 receptor antagonist (IL-IRA) before inescapable shock blocks the subsequent interference with escape learning and enhancement of fear conditioning normally produced by this treatment (Maier & Watkins, 1995). Administration of IL-1 induces effects comparable to stress-induced changes, e.g. IL-1 acts on the hypothalamus, to enhance release of monoamines, such as norepinephrine, dopamine, and 5-HT, as well as secretion of CRH. Recently, evidence for cross sensitization between immune and non immune stressors has been reported (Tilders & Schmidt, 1998). Peripheral administration of IL-1 to rats induces sensitization of the HPA-axis response to a second immune challenge weeks later. When exposed to psychological stress, IL-1-primed animals show exaggerated HPAaxis and autonomic responses (cross-sensitization) (Tilders & Schmidt, 1998). Postnatal stress in mice during the first two weeks of life in the absence of the mother, induces enhanced immune reactivity as assessed by T-cell mitogenesis in comparison with unhandled mice at 60 days of age (Neveu, Deleplanque, Puglisi-Allegra, D'Amato, & Cabib, 1994). These findings suggest that "early life stressors", which are known to play a role in the etiology of depression, may sensitize the organism to respond with exaggerated IRS responses upon subsequent stressors. It has been hypothesized that IRS activation with increased production of proinflammatory cytokines, such as IL-1, may have a central role in the stress responses (Shintani et al., 1995a). In conclusion, there is strong evidence that external stressors increase the production rate of the proinflammatory cytokines, IL-1 and IL-6, in animals. 5.2.2. External Stressors and Cytokines in Humans. In humans, no effects of psychological stress on plasma IL-1 or IL-6 concentrations could be found, although a significant and positive correlation between plasma IL-6 and serum C-reactive protein (CRP) concentrations is found in the stress condition (Dugue, Leppanen, Teppo, Fyrquist, & Grasbeck, 1993). Dobbin, Harth, McCain, Martin, & Cousin (1991) found that academic stress reduces the mitogen-induced lymphoproliferative responses and the production of IFNy in students, but at the same time significantly increases IL-ip production. These authors concluded that psychological stress in humans may have different effects on different cell populations by enhancing responses of monocytes and

Major Depression and Activation of the Inflammatory Response System

35

depressing those of lymphocytes (Dobbin et a!., 1991). In another study, the induction of mild negative emotional changes caused an increase in TNFa, a decrease in IL-2 and IL-3, with no changes in IL-ip and IL-6 (Mittwoch-Jaffe, Shalit, Srendi, & Yehuda, 1995). Recently, we found that academic examination stress in university students significantly increases the stimulated production of proinflammatory cytokines, such as IL-6, TNFa, and IFNy, and of the negative immunoregulatory cytokine IL-10 (Maes, Song, Lin, Gabriels, DeJongh, Van Gastel, Kenis, Bosmans, DeMeester, Benoyt, Neels, Demedts, Janca, Scharpe, & Smith, 1998b). Most importantly, students with a high stress perception (as measured by the Perceived Stress Scale) during the examination period have a significantly higher production of TNFa, IFNy, and IL-6 than students with lower stress perception. The stress-induced changes in perceived stress are significantly and positively related to the stress-induced production of TNFa, IFNy, and IL-6, suggesting that the stress-induced production of these cytokines is sensitive to graded differences in the perception of stressor severity. It has been shown that the ratio of IFNy production to IL-10 production in culture supernatant is of critical importance in determining the capacity of supernatants to activate or inhibit monocytic and T lymphocytic functions (Katsikis, Cohen, Londei, & Feldmann, 1995). Therefore, we have examined the IFNy/IL-10 ratio in these subjects (Maes, Song, Lin, DeJongh, Kenis, Bosmans, DeMeester, Neels, & Scharpe, 1999). We found that the response to psychological stress in humans consists of two different profiles of cytokine production, i.e. a first characterized by an increased IFNy/IL-10 ratio, attributable to a higher IFNy than IL-10 response (labeled IFNy reactors) and a second characterized by a lowered IFNy/IL-10 ratio, attributable to a higher IL-10 than IFNy response (labeled IL-10 reactors). IFNy reactors, but not IL-10 reactors, show a significantly increased stimulated production of other proinflammatory cytokines, i.e. TNFa and IL-6, and of serum IL-IRA, sCD8 and IgA, IgG and IgM concentrations. Thus, the external stress responsivity entails either a predominant proinflammatory cytokine response or a negative immunoregulatory cytokine response. IFNy reactors, but not IL-10 reactors, show significant stress-induced increases in anxiety and depression ratings, suggesting that a proinflammatory response and a lower stress-induced production of the negative immunoregulatory cytokine IL-10 appears to be accompanied by increased stress-induced anxiety and depression. The above findings in animals and humans may suggest that proinflammatory cytokines are involved in the stress response and that external stressors are perceived by the immune system and, through secretion of proinflammatory and negative immunoregulatory cytokines, take part in an integrated psychoneuroendocrine homeostatic response. The discovery that psychological stress in humans can alter the equilibrium between proinflammatory and negative immunoregulatory cytokines has important implications for human psychopathology. Indeed, since major depression may be induced by external stressors (negative life events and early life stressors) and since external stressors induce an inflammatory response in some subjects (i.e those with increased anxiety and depression ratings), we have hypothesized that psychological stress-induced hyperproduction of proinflammatory cytokines may play a role in the etiopathology of major depression (Maes et al., 1999).

5.3. Internal Stress, Cytokines, and the Etiology of Depression Important epidemiological features of major depression are the higher incidence of major depression in the medically ill or in "organic" conditions (i.e. mood disorders

36

M. Maes

due to a general medical condition) and in women, and the increasing rates of depression this century. The IRS activation model is consistent with each feature. The high occurrence of major depression in the medically ill is clearly consistent with the IRS activation model of depression (Maes, 1997; Yirmiya, 1997; Maes & Smith, 1998). Indeed, medical illnesses involving IRS activation, such as infectious (e.g. herpes virus, HIV, Borna virus, influenza) and non-infectious disorders (e.g. autoimmune disorders, multiple sclerosis, Alzheimer's disease), exhibit an increased incidence of depression. Moreover, conditions that are accompanied by an enhanced production of cytokines, e.g. the postpartum period, stroke, and myocard infarction are often accompanied by depressive symptoms or major depression. Thus, autoimmune-, postnatal-, inflammation-, and stroke-induced cytokine hyperproduction provides a reasonable mechanism to account for the significant higher rates of depression linked with these conditions. The elevated rate of depression in women is consistent with the greater immune responsivity in females and the immune activating effects of sex hormones (Washburn, Medearis, & Childs, 1965; Ahmed, Penhale, & Talal, 1985; Knapp, Levy, Giorgi, Black, Fox, & Heeren, 1992). Indeed, it is known that immune responsiveness is greater in women than in men and that sex hormones can alter immune responses (Lynch, Dinarello, & Cannon, 1994). For example, female sex hormones modulate maturation of lymphocytes in the lymphoid tissues, regulate leukocyte entry into peripheral tissues, alter immune cell membrane composition, augment leukocyte adhesion, enhance proliferation of lymphocytes in response to mitogenic stimuli and increase IL-1 production by mononuclear cells (Grossman, 1985; Bagdade & Subbaiah, 1988; Hammerschmidt, Knabe, Silberstein, Lamche, & Coppo, 1988; Athreya, Fletcher, Zulian, Weiner, & Williams, 1993; Cid, Kleinman, Grant, Schnaper, Fauci, & Hoffman, 1994; Lynch et al., 1994). We found that female students who are taking monophasic contraceptive drugs had a significantly greater stress-induced response in the number of leukocytes, neutrophils, and B cells than females without use of contraceptives and male students, suggesting that ethinylestradiol and/or derivatives of progesterone may increase the immune responsiveness to psychological stress (Maes, VanBockstaele, VanGastel, Van Hunsel, Neels, DeMeester, Scharpe, & Janca, 1998c). Also the increased incidence rate of major depression since 1913 may be explained by the IRS activation model of depression. This increased incidence parallels the increasing ratio of (o6 to co3 fatty acids in Western diets the last century (Smith, 1991). This results in high levels of co6 PUFAs in tissues and low levels of long-chain co3 PUFAs in serum and membranes of the Western population (Sinclair, Johnson, O'Dea, & Holman, 1994). As explained above, a high dietary o)6/co3 fatty acid ratio may increase the secretion of proinflammatory cytokines (review: Maes et al., 1996a). It has also been suggested that the increased (o6/co3 PUFA ratio may have contributed to an increased incidence of cardiovascular and inflammatory disorders (Smith, 1991).

6. ANTIDEPRESSANT TREATMENTS, CYTOKINES, AND DEPRESSION 6.1. In Vivo Effects of Antidepressants If increased production of proinflammatory cytokines is at all involved in the etiology of depression one would expect that the various antidepressive treatments

Major Depression and Activation of the Inflammatory Response System

37

have negative immunoregulatory effects. It is generally believed that tricyclic antidepressants have immunosuppressive effects ex vivo as well as in vivo (review: Miller & Lackner, 1989). Subchronic treatment with fluoxetine, a selective 5-HT reuptake inhibitor (SSRI), normalizes the initially increased serum IL-6 concentrations in depressed patients (Sluzewska et al., 1995). Subchronic treatment with tricyclic antidepressants and fluoxetine is able to suppress the acute phase response in major depression (Maes et al., 1997b). SSRIs, such as sertraline and fluoxetine, attenuate the acute phase response in the olfactory bulbectomized and chronic mild stress model of depression in the rat (Sluzewska et al., 1994; Song & Leonard, 1994). In Wistar rats, the antidepressant effect of repeated administration of imipramine is accompanied by a reduction in the mild stress-induced capacity of splenocytes to produce IL-1 and IL-2 following an eight-week exposure to unpredictable stress (Kubera et al., 1996). A single IP injection of lOmg/kg desipramine to naive mice increases the capacity of splenocytes to produce IL-10 (Kubera et al., 1998). Repeated administration of imipramine induces IL-1 and IL-IRA mRNA in widespread area of the rat brain, but it induces a greater effect on IL-lRA mRNA than on IL-1 mRNA (Suzuki, Shintani, Kanba, Asai, & Nakaki, 1996). Since the IL-IRA is a pure antagonist of the IL-IR and, as such, inhibits some biological activities of IL-1, the imipramine-induced changes may result in immunosuppression. The above results show that, in vivo, antidepressants have antiinflammatory effects through downregulation of proinflammatory cytokines and upregulation of the negative immunoregulatory cytokines/receptor antagonists, IL-10 and the IL-IRA.

6.2. In Vitro Effects of Antidepressants Antidepressive agents have also in vitro antiinflammatory effects. Clomipramine, imipramine, and citalopram significantly suppress the secretion of IL-2 by stimulated T lymphocytes and of IL-1 (3 and TNFa by stimulated monocytes (Xia, DePierre, & Nassberger, 1996). The same authors, found that there was a trend toward a significant decreased secretion of IFNy by stimulated T lymphocytes preincubated with the above antidepressants. Since diluted whole blood stimulated with phytohemagglutinin (PHA) plus lipopolysaccharide (LPS) offers the most appropriate and reproducible culture condition in order to measure cytokines, such as IFNy, IL-6, and IL-1 (reviews: De Groote, Zangerle, Gevaert, Fassotte, Beguin, Noizat-Pirenne, Pirenne, Gathy, Lopez, Dehart, Igot, Baudrihaye, Delacroix, & Franchimont, 1992; De Groote, Gevaert, Lopez, Gathy, Fauchet, Dehart, Jadoul, Radoux, & Franchimont, 1993), our laboratory has examined the effects of antidepressive agents on cytokine production by diluted whole blood cultures. The focus of interest was the IFNy/IL-10 ratio, since it is of critical importance in determining the capacity of supernatants to activate or inhibit monocytic and T lymphocytic functions (Katsikis et al., 1995). We found that clomipramine, sertraline, and trazodone, at concentrations in the range of the therapeutic plasma concentrations achieved during clinical treatment, had a significant suppressive effect on the IFNy/IL-lO ratio, which was attributable to a suppression of the stimulated production of IFNy and a significant stimulatory effect on IL-10 production (Maes, Song, Lin, Bonaccorso, Scharpe, Kenis, DeJongh, Bosmans, & Scharpe, 1998a). Thus, antidepressive agents, including SSRIs, tricyclic, and heterocyclic antidepressants, may have negative immunoregulatory effects, since they significantly suppress the IFNy/IL-10 ratio. Since antidepressants decrease the IFNy/IL-10 ratio and since IFNy has depressogenic properties and since IFNy production is increased in depression and in stressinduced depressive and anxious states, it may be speculated that antidepressants exert

38

M. Maes

some of their antidepressant effects through their negative immunoregulatory capacities. The exact mechanisms by which antidepressive agents from such different classes exert their activity on cytokine production are still unknown. These effects might be mediated by serotonergic mechanisms. T Lymphocytes express 5-HT receptors, such as 5-HTlA and 5-HT2A/2C receptors, as well as a high affinity 5-HT transporter. Macrophages have a specific active 5-HT uptake system similar in affinity to that of platelets (Jackson, Walker, Brooks, & Roszman, 1988; Aune, Golden, & McGrath, 1994; Jahnova, 1994; Faraj, Olkowski, & Jackson, 1994). There are now some reports that depletion of intracellular 5-HT stores, increased extracellular 5-HT and/or 5-HT2A/2C receptor blockade may result in negative immunoregulatory effects (Idova & Cheido, 1987; Bengtsson, Zhu, Thorell, Olsson, Link, & WaUnder, 1992; Nordlind, Sundstrom, & Bondesson, 1992; Bondesson, Nordlind, Liden, & Sundstrom, 1993; Jahnova, 1994; Smejkal-Jagar & Boranic, 1994; Young & Matthews, 1995). Thus, part of the immune effects of SSRIs, tricyclic, and heterocyclic antidepressants may be explained by their serotonergic activities, such as depletion of intracellular 5-HT stores, increased extracellular 5-HT and/or 5-HT2A/2C receptor blockade (Maes et al., 1998a).

7. CONCLUSIONS Major depression is accompanied by various direct and indirect indicators of a moderate activation of the IRS. Increased production of proinflammatory cytokines, such as IL-1, IL-6, and IFNy, may play a crucial role in the immune and acute phase response in depression. Lower serum Zn, changes in the erythron, lower serum HDL-C, and depletion of co3 PUFAs are indirect indicators of IRS activation in depression. The reciprocal relationships between IRS activation and HPA-axis hyperactivity, alterations in HPT-axis function and the availability of tryptophan to the brain led us to hypothesize that these neuroendocrine changes are indicators of IRS activation in depression and that a combined dysregulation of the IRS, the turnover of 5-HT and the HPA-axis is an integral component of depression. The IRS activation model of depression provides an explanation for the psychosocial (external stress) as well as organic (internal stress) etiology of major depression. Antidepressive treatments with various antidepressive agents, including SSRIs, tricyclic, and heterocyclic antidepressants, have in vivo and in vitro negative immunoregulatory effects, suggesting that their antidepressant efficacy may be attributed, in part, to their immune effects.

REFERENCES Adams, P. B., Lawson, S., Sanigorski, A., & Sinclair, A. J. (1996). Arachidonic acid to eicosapentaenoic acid ratio in blood correlates positively with clinical symptoms of depression. Lipids, 3I(Suppl), S157-161. Ahmed, S. A., Penhale, W. J., & Talal, N. (1985). Sex hormones, immune responses, and autoimmune diseases. American Journal of Pathology, 21, 531-551. Anisman, H., Kokkinidis, L., Borowski,T., & Merali, Z. (1998). Differential effects of interleukin (IL)-lbeta, IL-2, and IL-6 on responding for rewarding lateral hypothalamic stimulation. Brain Research, 779, 177-187. Athreya, B. H., Fletcher, J., Zulian, F., Weiner, D. B., & Williams, W. V. (1993). Subset-specific effects of sex hormones and pituitary gonadotropins on human lymphocyte proliferation in vitro. Clinical Immunology and Immunopathology, 66, 201-211.

Major Depression and Activation of the Inflammatory Response System

39

Aune, T. M., Golden, H. W., & McGrath, K. M. (1994). Inhibitors of serotonin synthesis and antagonists of serotonin lA receptors inhibit T lymphocyte function in vitro and cell-mediated immunity in vivo. Journal of Immunology, 153, 489^98. Bagdade, J. D. & Subbaiah, P. V. (1988). Influence of low-estrogen-containing oral contraceptives on lipoprotein phospholipid composition and mononuclear cell membrane fluidity. Journal of Clinical Endocrinology and Metabolism, 66, 857-861. Bengtsson, B.-O., Zhu, J.,Thorell, L.-H., Olsson, T., Link, H., & Walinder, J. (1992). Effects of zimeldine and its metabolites, clomipramine, imipramine, and maprotiline in experimental allergic neuritis in Lewis rats. Journal of Neuroimmunology, 39, 109-122. Berk, M., Wadee, A. A., Kuschke, R. H., & O'Neill-Kerr, A. (1997). Acute phase proteins in major depression. Journal of Psychosomatic Research, 43, 529-534. Berkenbosch, F., Oers, J., & del Rey, A. D. (1987). Corticotropin-releasing-factor-producing neurons in the rat activated by interleukin-1. Science, 238, 524-526. Bonaccorso, S., Lin, A., Verkerk, R., VanHunsel, F., Libbrecht, I., Scharpe, S., DeClerck, L., Stevens, W., Biondi, M., Janca, A., & Maes, M. (1998). Immune markers in fibromyalgia: comparison with major depressed patients and normal volunteers. Journal of Affective Disorders, 48, 75-82. Bondesson, L., Nordlind, K., Liden, S., & Sundstrom, E. (1993). Inhibiting effects of serotonin and serotonin antagonists on the migration of mononuclear leucocytes. Immunopharmacology and Immunotoxicology, 15, 243-250. Brock, J. H. (1994). Iron in infection, immunity, inflammation, and neoplasia. In J. H. Brock, J. W. Halliday, M. J. Pippard, & L. W. Powell (Eds.), Iron Metabolism in Health and Disease (pp. 353-390). London: W. B. Saunders Company Ltd. Calabrese, J. R., Skwerer, R. G., Barna, B., Gulledge, A. D., Valenzuela, R., Butkus, A., Subichin, S., & Krupp, N. E. (1986). Depression, immunocompetence, and prostaglandins of the E series. Psychiatry Research, /7, 41-47. Carmeliet, P., Vankelecom, H., Van Damme, J., Billiau, A., & Denef, C. (1991). Release of interleukin-6 from anterior pituitary cell aggregates: developmental pattern and modulation by glucocorticoids and forskolin. Neuroendocrinology, 53, 29-34. Cid, M. C , Kleinman, H. K., Grant, D. S., Schnaper, H. W., Fauci, A. S.. & Hoffman, G. S. (1994). Estradiol enhances leukocyte binding to tumor necrosis factor-stimulated endothelial cells via an increase in TNF-induced adhesion molecules E-selectin, intercellular adhesion molecule type 1, and vascular cell adhesion molecule type 1. Journal of Clinical Investigation, 93, 17-25. Clement, H. W., Buschmann, J., Rex, S., Grote, C . Opper, C , Gemsa, D., & Wesemann, W. (1997). Effects of interferon-gamma, interleukin-1 beta, and tumour necrosis factor-alfa on the serotonin metabolism in the nucleus raphe dorsalis of the rat. Journal of Neural Transmission, 104, 981-991. Connor, T. J. & Leonard, B. E. (1998). Depression, stress, and immunological activation: the role of cytokines in depressive disorders. Life Sciences, 62, 583-606. Connor, T. J., Song, C , Leonard, B. E., Merali, Z., & Anisman, H. (1998). An assessment of the effects of central interleukin-1 beta, -2, -6, and tumor necrosis factor-alpha administration on some behavioural, neurochemical, endocrine, and immune parameters in the rat. Neuroscience. 84, 923-933. Cousins, R. J. & Leinart, A. S. (1988). Tissue-specific regulation of zinc metabolism and metallothionein genes by interleukin-1. FASEB Journal, 2, 2884-2890. Cunnane, S. C , Yang, J., & Chen, Z. Y. (1993). Low zinc intake increases apparent oxidation of linoleic and alpha-linolenic acids in the pregnant rat. Canadian Journal of Physiology and Pharmacology, 71, 205-210. Dantzer, R., Bluthe, R. M., Laye, S., Bret-Dibat, J. L., Parnet, R, & Kelley, K. W. (1998). Cytokines and sickness behavior. Annals of the New York Academy of Sciences. 840, 586-590. Deger, O., Bekaroglu. M., Orem, A.. Orem, S., Uluutku, N., & Soylu, C. (1996). Polymorphonuclear (PMN) elastase levels in depressive disorders. Biological Psychiatry, 39, 357-363. De Groote, D.. Zangerle, P. F, Gevaert, Y, Fassotte, M. F , Beguin, Y, Noizat-Pirenne, F, Pirenne, J.. Gathy, R., Lopez, M., Dehart, I., Igot, D., Baudrihaye, M., Delacroix, D., & Franchimont, P. (1992). Direct stimulation of cytokines (IL-ip, TNF-a, IL-6, IL-2, IFN-y, and GM-CSF) in whole blood. I. Comparison with isolated PBMC stimulation. Cytokine, 4, 239-248. De Groote, D., Gevaert, Y, Lopez, M., Gathy, R., Fauchet, F, Dehart, I., Jadoul, M., Radoux, D., & Franchimont, P. (1993). Novel method for the measurement of cytokine production by one-stage procedure. Journal of Immunological Methods, 9, 259-267. Dobbin, J. R, Harth, M., McCain, G. A., Martin, R. A., & Cousin, K. (1991). Cytokine production and lymphocyte transformation during stress. Brain, Behavior and Immunity, 5, 339-348. Dubuis, J. M., Dayer, J. M., Siegrist-Kaiser, C. A., & Burger, A. G. (1988). Human recombinant interleukin-

40

M. Maes

i p decreases plasma thyroid hormone and thyroid stimulating hormone levels in rats. Endocrinology, 725,2175-2181. Duch, D. S., Woolf, J. H., Nichol, C. A., Davidson, J. R., & Garbutt, J. C. (1984). Urinary excretion of biopterin and neopterin in psychiatric disorders. Psychiatry Research, 11, 83-89. Dugue, B., Leppanen, E. A., Teppo, A. M., Fyrquist, E, & Grasbeck, R. (1993). Effects of psychological stress on plasma interleukins-1 and beta and 6, C-reactive protein, tumour necrosis factor alpha, anti-diuretic hormone and serum Cortisol. Scandinavian Journal of Clininical Laboratory Investigation, 56,555-561. Dunbar, P. R., Hill, J., Neale, T. J., & Mellsop, G. W. (1992). Neopterin measurement provides evidence of altered cell-mediated immunity in patients with depression, but not with schizophrenia. Psychological Medicine, 22,1051-1057. Dunn, A. J. (1988). Systemic interleukin-1 administration stimulates hypothalamic norepinephrine metabolism parallelling the increased plasma corticosterone. Life Sciences, 43, 429-435. Dunn, A. J., Powell, M. L., Meitin, C , & Small, P. A. Jr. (1989). Virus infection as a stressor: influenza virus elevates plasma concentrations of corticosterone, and brain concentrations of MHPG and tryptophan. Physiology and Behaviour, 45, 591-594. Dunn, M. A. & Cousins, R. J. (1989). Kinetics of zinc metabolism in the rat: effect of dibutyryl cAMP American Journal of Physiology, 256, 420-430. Endres, S. (1993). Messengers and mediators: interactions among lipids, eicosanoids, and cytokines. American Journal of Clinical Nutrition, 57, 798-800. Espersen, G. T., Grunnet, N., Lervang, H. H., Nielsen, G. L., Thomsen, B. S., Faarvang, K. L., Dyerberg, J., & Ernst, E. (1992). Decreased interleukin-1 beta levels in plasma from rheumatoid arthritis patients after dietary supplementation with n-3-polyunsaturated fatty acids. Clinical Rheumatology, 11, 393-395. Fairbanks, V. E & Beutler, E. (1988). Iron. In M. E. Shils & V. R. Young (Eds.), Modern Nutrition in Health and Disease (pp. 193-226). Philadelphia: Lea and Febiger. Faraj, B. A., Olkowski, Z. L., & Jackson, R. T. (1994). Expression of a high-affinity serotonin transporter in human lymphocytes. International Journal of Immunopharmacology, 16, 561-567. Frommberger, U. H., Bauer, J., Haselbauer, P., Fraulin, A., Riemann, D., & Berger, M. (1997). Interleukin-6(IL-6) plasma levels in depression and schizophrenia; comparison between the acute state and after remission. European Archives of Psychiatry and Clinical Neuroscience, 247, 228-233. Fujii,T., Sato, K., Ozawa, M., Kasono, K., Imamura, H., Kanaji,Y., Tsushima, T., & Shizume, K. (1989). Effect of interleukin-1 (Il-l) on thyroid hormone metabolism in mice: stimulation by Il-l of iodothyronine 5'-deiodinating activity (type I) in the liver. Endocrinology, 124, 167-174. Giroux, E., Schechter, P. J., Schoun, J., & Sjoerdsma, A. (1977). Reduced binding of added zinc in serum of patients with decompensated hepatic cirrhosis. European Journal of Clinical Investigation, 7, 71-73. Glueck, C. J.,Tieger, M., Kunkel, R., Hamer,T., Tracy, T., & Speirs, J. (1994). Hypocholesterolemia and affective disorders. American Journal of Medical Science, 308, 218-225. Goldblum, S. E., Cohen, D. A., Jay, M., & McClain, C. J. (1987). Interleukin 1-induced depression of iron and zinc: role of granulocytes and lactoferrin. American Journal of Physiology, 252, E27-E32. Grossman, C. J. (1985). Interactions between the gonadal steroids and the immune system. Science, 227, 257-261. Gutterman, J. U., Fein, S., Quesada, J., Horning, S. J., Levine, J. F , Alexanian, R., Bernhardt, L., Kramer, M., Spiegel, H., Colburn, W., Trown, P., Merigan, T, & Dziewanowski, Z. (1982). Recombinant leukocyte A interferon: pharmacokinetics, single dose tolerance, and biologic effects in cancer patients. Annals of Internal Medicine, 96, 549-556. Hammerschmidt, D. E., Knabe, A. C , Silberstein, P T., Lamche, H. R., & Coppo, P A. (1988). Inhibition of granulocyte function by steroids is not limited to corticoids. Studies with sex steroids. Inflammation, 12, 277-284. Herbert, T. B. & Cohen, S. (1993). Depression and immunity: a meta-analytic review. Psychological Bulletin, 7/3,472-486. Hermus, A. R. M. M., Sweep, C. G. J. E, van der Meer, M. J. M., Ross, H. A., Smals, A. G. H., Benraad, T. J., & Kloppenborg, P. W. C. (1992). Continuous infusion of interleukin-ip induces a nonthyroidal illness syndrome in the rat. Endocrinology, 131, 2139-2146. Heyes, M. P., Saito, K., Crowley, J. S., Davis, L. E., Demitrack, M. A., Der, M., Dilling, L. A., Elia, J., Kruesi, M, J. P., Lackner, A., Larsen, S. A., Lee, K., Leonard, H. L., Markey, S. P., Martin, A., Milstein, S., Mouradian, M. M., Pranzatelli, M. R., Quearry, B. J., Salazar, A., Smith, M., Strauss, S. E., Sunderland, T, Swedo, S. W., & Tourtellotte, W. W. (1992). Quinolinic acid and kynurenine pathway metabolism in inflammatory and non-inflammatory neurological disease. Brain, 115, 1249-1273.

Major Depression and Activation of the Inflammatory Response System

41

Hibbeln, J. R. & Salem, N. (1995). Dietary polyunsaturated fatty acids and depression: when cholesterol does not satisfy. American Journal of Clinical Nutrition, 62,1-9. Holden, R. J., Pakula, I. S., & Mooney, P. A. (1997). A neuroimmunological model of schizophrenia and major depression: a review. Human Psychopharmacology, 12, 177-201. Horrobin, D. F. (1990). Essential fatty acids, psychiatric disorders and neuropathies. In D. F. Horrobin (Ed.), Omega-6 Essential Fatty Acids: Pathophysiology and Roles in Clinical Medicine (pp. 305-320). New York: Wiley-Liss. Idova, G. V. & Cheido, M. A. (1987). Stimulation of the immune response during blockade of serotonin receptors by cyproheptadine. Biull Eksp Biol Med, 103, 440-442. Jackson, J. C , Walker, R., Brooks, W. H., & Roszman, T. L. (1988). Specific uptake of serotonin by murine macrophages. Life Sciences, 42, 641-650. Jahnova, E. (1994). The role of serotonin in the regulation of the immune response. Bratisl Lek Listy, 95, 181-184. Joyce, R R., Hawes, C. R., Mulder, R. T., Sellman, J. D., Wilson, D. A., & Boswell, D. R. (1992). Elevated levels of acute phase plasma proteins in major depression. Biological Psychiatry, 32,1035-1041. Kaptein, E. M. (1986). Thyroid hormone metabolism in illness. In G. Hennemann (Ed.), Thyroid hormone metabolism (pp. 297-334). New York: Marcel Dekker. Katsikis, P. D., Cohen, S. B., Londei, M., & Feldmann, M. (1995). Are CD4+ Thl cells pro-inflammatory or anti-inflammatory? The ratio of IL-10 to IFN-gamma or IL-2 determines their function. International Immunology, 7,1287-1294. Khlusov, I. A., Dygai, A. M., & Gol'dberg, E. D. (1993). The adrenergic regulation of interleukin production by bone marrow cells during immobilization stress. Biull Eksp Biol Med, 116, 570-572. Knapp, R H., Levy, E. M., Giorgi, R. G., Black, R H., Fox, B. H., & Heeren,T. C. (1992). Short-term immunological effects of induced emotion. Psychosomatic Medicine, 54, 133-148. Kronfol, Z., Singh, V. K., Boura, J., & Brown, M. B. (1998). Cytokines and acute phase proteins in major depression: are there gender differences? Biological Psychiatry, 43, 98S. Kubera, M., Symbirtsev, A., Basta-Kaim, A., Borycz, J., Roman, A., Papp, M., & Claesson, M. (1996). Effect of chronic treatment with imipramine on interleukin 1 and interleukin 2 production by splenocytes obtained from rats subjected to a chronic mild stress model of depression. Polish Journal of Pharmacology, 48, 503-506. Kubera, M., Holan, V., Basta-Kaim, A., Roman, A., Borycz, J., & Shani, J. (1998). Effect of desipramine on immunological parameters in mice, and their reversal by stress. Journal of Immunophamacology, in press. Kushner, I. (1982). The phenomenon of the acute phase response. Annals of the New York Academy of Sciences, 389, 39-48. Lee, G. R. (1983). The anemia of chronic disease. Seminars in Hematology, 20, 61-80. LeMay, L. G., Vander, A. J., & Kluger, M. J. (1990). The effects of psychological stress on plasma interleukin6 activity in rats. Physiology and Behavior, 47, 957-961. Lieb, J. & Karmali, R. (1983). Elevated levels of prostaglandin E2 and thromboxane B2 in depression. Prostaglandins and Leukotrienes in Medicine, 10, 361-368. Linthorst, A. C. & Reul, J. M. (1998). Brain neurotransmission during peripheral inflammation. Annals of the New York Academy of Sciences, 840, 139-152. Linnoila, M., Whorton, R., Rubinow, D. R., Cowdry, R. W, Ninan, R T, & Waters, R. N. (1983). CSF prostaglandin levels in depressed and schizophrenic patients. Archives of General Psychiatry, 40, 405-406. Little, K. Y, Castellanos, X., Humphries, L. L., & Austin, J. (1989). Altered zinc metabolism in mood disorder patients. Biological Psychiatry, 26, 646-648. Lynch, E. A., Dinarello, C. A., & Cannon, J. G. (1994). Gender differences in IL-la, IL-lp, and IL-1 receptor antagonist secretion from mononuclear cells and urinary excretion. Journal of Immunology, 153,300-306. McCallum, J., Simons, L., Simons, J., & Friedlander, Y. (1994). Low serum cholesterol is not associated with depression in the elderly: data from an Australian community study. Australian New Zealand Journal of Medicine, 24, 561-564. McLoughlin, I. J. & Hodge, J. S. (1990). Zinc in depressive disorder. Acta Psychiatrica Scandinavica, 82, 451-453. Maes, M. (1993). Acute phase protein alterations in major depression: A review. Reviews in the Neurosciences, 4, 407-416. Maes, M. (1995). The Interleukin hypothesis of major depression. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 19, 11-38.

42

M. Maes

Maes, M. (1997). The immune pathophysiology of major depression. In A. Honig & H. M. van Praag (Eds.), Depression: Neurobiological, Psychopathological, and Therapeutic Advances (pp. 197-215). London: John Wiley. Maes, M. & Meltzer, H. Y. M. (1995). The serotonin hypothesis of major depression. In F. Bloom & D. Kupher (Eds.), Psychopharmacology: the fourth generation of progress (pp. 933-944). New York: Raven Press. Maes, M. & Smith, R. (1997). Immune activation and major depression: a hypothesis. Psychiatry, Current Medical Literature Psychiatry, The Royal Society of Medicine, 9, 3-6. Maes, M. & Smith, R. S. (1998). Editorial: Fatty acids, cytokines, and major depression. Biological Psychiatry, 43, 3\9-3\ 4. Maes, M., Bosmans, E., Suy, E., Vandervorst, C , Dejonckheere, C , Minner, B., & Raus, J. (1991a). Depression-related disturbances in mitogen-induced lymphocyte responses, interleukin-ip, and soluble interleukin-2-receptor production. Acta Psychiatrica Scandinavica, 84, 379-386. Maes, M., Bosmans, E., Suy, E., Vandervorst, C , DeJonckheere, C , & Raus., J. (1991b). Antiphospholipid, anti-nuclear, Epstein-Barr, and cytomegalovirus antibodies, and soluble interleukin-2 receptors in depressive patients. Journal of Affective Disorders, 21, 133-140. Maes, M., Vandewoude, M., Scharpe, S., De Clerck, L., Stevens, W., Lepoutre, L., & Schotte, C. (1991c). Anthropometric and biochemical assessment of the nutritional state in depression: evidence for lower visceral protein plasma levels in depression. Journal of Affective Disorders, 23, 25-33. Maes, M., Lambrechts, J., Bosmans, E., Jacobs, J., Suy, E., Vandervorst, C , De Jonckheere, C , Minner, B., & Raus, J. (1992a). Evidence for a systemic immune activation during depression: results of leukocyte enumeration by flow cytometry in conjunction with monoclonal antibody staining. Psychological Medicine, 22, 45-53. Maes, M., Scharpe, S., Bosmans, E., Vandewoude, M., Suy, E., Uyttenbroek, W., Cooreman, W., Vandervorst, C , & Raus, J. (1992b). Disturbances in acute phase plasma proteins during melancholia: additional evidence for the presence of an inflammatory process during that illness. Progress in Neuropsychopharmacology and Biological Psychiatry, 16, 501-515. Maes, M., Scharpe, S., Van Grootel, L., Uyttenbroeck, W., Cooreman, W., Cosyns, P., & Suy, E. (1992c). Higher al-antitrypsin, haptoglobin, ceruloplasmin, and lower retinol binding protein plasma levels during depression: further evidence for the existence of an inflammatory response during that illness. Journal of Affective Disorders, 24, 183-192. Maes, M., Bosmans, E.. Meltzer, H. Y, Scharpe, S., & Suy, E. (1993a). Interleukin-ip: A putative mediator of HPA-axis hyperactivity in major depression? American Journal of Psychiatry, 150, 1189-1193. Maes, M., Meltzer, H. Y, Scharpe, S., Uyttenbroeck, W., Cooremans, W., & Suy, E. (1993b). Psychomotor retardation, anorexia, weight loss, sleep disturbances, and loss of energy: psychopathological correlates of hyperhaptoglobinemia during major depression. Psychiatry Research, 47, 229-241. Maes, M., Scharpe, S., Meltzer, H. Y. Bosmans, E., Suy, E., Minner, B., Calabrese, J., Uyttenbroeck, W., Vandervorst, C , Raus, J., & Cosyns. P. (1993c). Relationships between interleukin-6 activity, acute phase proteins, and HPA-axis function in severe depression. Psychiatry Research, 49, 11-27. Maes, M., Scharpe, S., Meltzer, H. Y, & Cosyns, P. (1993d). Relationships between increased haptoglobin plasma levels and activity of cell-mediated immunity. Biological Psychiatry, 34, 690-701. Maes, M., Delanghe, J.. Meltzer, H. Y, D'Hondt, P., & Cosyns, P. (1994a). Lower degree of esterification of serum cholesterol in depression: relevance for depression and suicide research. Acta Psychiatrica Scandinavica, 90, 252-258. Maes, M., De Langhe, J., Scharpe, S., Meltzer, H. Y, Cosyns, P., Suy, E., & Bosmans, E. (1994b). Haptoglobin phenotypes and gene frequencies in unipolar major depression. American Journal of Psychiatry, 151, 112-116. Maes, M., Scharpe, S., Cosyns, P., & Meltzer, H. Y (1994c). Relationships between basal hypothalamicpituitary-thyroid axis activity and plasma haptoglobin levels in depression. Journal of Psychiatric Research, 28, \23-l34. Maes, M., Scharpe, S., D'Haese, W., De Broe, M., & Cosyns, P. (1994d). Hypozincaemia in depression, yowrna/ of Affective Disorders, 31, 135-140. Maes, M., Scharpe, S., Meltzer, H. Y, Okayli, G., D'Hondt, P., & Cosyns, P. (1994e). Increased neopterin and interferon gamma secretion and lower L-tryptophan levels in major depression: further evidence for immune activation in severe depression. Psychiatry Research, 54, 143-160. Maes, M., Meltzer, H. Y, Bosmans, E., Bergmans, R., Vandoolaeghe, E., Ranjan, R., & Desnyder, R. (1995a). Increased plasma concentrations of interleukin-6, soluble interleukin-6, soluble interleukin-2, and transferrin receptor in major depression. Journal of Affective Disorders, 34, 301-309. Maes, M., Scharpe, S., Neels, H., Wauters. A., Van Gastel, A., & Cosyns, P. (1995b). Total serum protein and serum protein fractions in major depression. Journal of Affective Disorders, 34, 61-69.

Major Depression and Activation of the Inflammatory Response System

43

Maes, M., Smith, R., & Scharpe, S. (1995c). The monocyte-T lymphocyte hypothesis of major depression. Psychoneuroendocrinology, 20, 111-116. Maes, M., Smith, R., Christophe, A., Cosyns, P., Desnyder, R., & Meltzer, H. Y. (1996a). Fatty acid composition in major depression: decreased 0)3 fractions in cholesteryl esters and increased C20: 4co6/C20: 5(o3 ratio in cholesteryl esters and phospholipids. Journal of Affective Disorders, 3S, 35-46. Maes, M., Van de Vyvere, J., Vandoolaeghe, E., Bril,T., Demedts, P.. Wauters, A., & Neels, H. (1996b). Alterations in iron metabolism and the erythron in major, depression: further evidence for a chronic inflammatory process. Journal of Affective Disorders, 40, 23-33. Maes, M., Bosmans, E., De Jongh, R., Kenis, G., Vandoolaeghe, E., & Neels, H. (1997a). Increased serum IL6 and IL-1 receptor antagonist concentrations in major depression and treatment resistant depression, cytokine, 9, 853-858. Maes, M., Delanghe, J., Ranjan, R., Meltzer, H. Y., Desnyder, R., Cooreman, W., & Scharpe, S. (1997b). The acute phase protein response in schizophrenia, mania, and major depression: effects of psychotropic drugs. Psychiatry Research, 66, 1-11. Maes, M., Vandoolaeghe, E., Neels, H., Demedts, P., Wauters, A., & Desnyder, R. (1997c). Lower high density lipoprotein cholesterol in major depression and in depressed men with serious suicidal attempts: relationships to immune-inflammatory markers. Acta Psychiatrica Scandinavica, 95, 212-221. Maes, M., Vandoolaeghe, E., Neels, H., Demedts, P., Wauters, A., Meltzer, H. Y, Altamura. C . & Desnyder, R. (1997d). Lower serum zinc in major depression is a sensitive marker of treatment resistance and of the immune/inflammatory response in that illness. Biological Psychiatry, 42, 349-358. Maes, M., Verkerk, R., Vandoolaeghe, E., Van Hunsel, F., Neels, H., Wauters, A., Demedts, P., & Scharpe, S. (1997e). Serotonin-immune interactions in major depression: lower serum tryptophan as a marker of an immune-inflammatory response. European Archives of General Psychiatry and Clinical Neuroscience,247, 154-161. Maes, M., Song, C , Lin, A., Bonaccorso, S., Scharpe, S., Kenis, G., DeJongh, R., Bosmans, E., & Scharpe, S. (1998a). Negative immunoregulatory effects of antidepressants: increased production of interleukin10 and suppressed production of interferon-gamma. Neuropsychopharmacology, accepted. Maes, M., Song, C , Lin, A., Gabriels, L., DeJongh, R., Van Gastel, A., Kenis, G., Bosmans, E., DeMeester, I., Benoyt, I., Neels, H.. Demedts, P, Janca, A., Scharpe, S., & Smith, R. S. (1998b). The effects of psychological stress on humans: increased production of proinflammatory cytokines and a Th-l-Hke response in stress-induced anxiety. Cytokine, 10, 313-318. Maes, M., VanBockstaele, D.. VanGastel, A., Van Hunsel, F., Neels, H., DeMeester, I., Scharpe, S.. & Janca, A. (1998c). Influence of psychological stress on leukocyte subset distribution in normal Humans: evidence for interrelated immunosuppression and T cell activation. Neuropsychohiology, in press. Maes, M., Song, C., Lin, A., DeJongh, R., Kenis, G., Bosmans, E., DeMeester, I., Neels, H., & Scharpe, S. (1999). Immune and chnical correlates of psychological stress-induced production of interferon-y and IL-10 in humans. In N. P. Plotnikoff (Ed.), Cytokines, Stress, and Immunity. Boca Raton: CRC-Press [In press]. Maier, S. F & Watkins, L. R. (1995). Intracerebroventricular interleukin-1 receptor antagonist blocks the enhancement of fear conditioning and interference with escape produced by inescapable shock. Brain Research, 695, 279-282. Maier, S. F & Watkins, L. R. (1998). Cytokines for psychologists: implications of bidirectional immune-tobrain communication for understanding behavior, mood, and cognition. Psychological Reviews, 105, 83-107. Mekaouche, M., Givalois, L., Barbanel, G., Siaud, P., Maurel, D., Malaval, F., Bristow, A. F., Boissin, J., Assenmacher, I., & Ixart, G. (1994). Chronic restraint enhances interleukin-1-beta release in the basal state and after an endotoxin challenge, independently of adrenocorticotropin, and corticosterone release. Neuroimmunomodulation, 1,292-299. Meydani, S. N., Endres. S., Woods, M. M.. Goldin, B. R., Soo, C , Morrill-Laborde. A., Dinarello, C. A., & Gorbach, S. L. (1991). Oral (n-3) fatty acid supplementation suppresses cytokine production and lymphocyte proliferation: comparison between young and older women. Journal of Nutrition, 121, 547-555. Mikova, O., Stoyanova, V, & Tanchev, O. (1998). Acute phase proteins in affective disorders. Abstract PT09032 XXIst CINP Congress 1998 Glasgow. 1998. Miller, A. H. & Lackner, C. (1989). Tricyclic antidepressants and immunity. In H. Miller (Ed.), Depressive Disorders and Immunity (pp. 85-104). Washington, DC: American Psychiatric Press. Miller, A. H.. Spencer, R. L., Pearce, B. D., Pisell, T. L., Tanapat, P., Leung, J. J., Dhabhar, F. S., McEwen, B. S., & Biron, C. A. (1997). 1996 Curt P. Richter Award. Effects of viral infection on corticosterone

44

M. Maes

secretion and glucocorticoid receptor binding in immune tissues. Psychoneuroendocrinology, 22, 455^74. Minami, M., Kuraishi, Y., Yamaguchi,T.,Nakai, S., Hirai, Y., & Satoh, M. (1991). Immobilization stress induces interleukin-1 beta mRNA in the rat hypothalamus. Neuroscience Letters, 123,254-256. Mittwoch-Jaffe, T., Shalit, F., Srendi, B., & Yehuda, S. (1995). Modification of cytokine secretion following mild emotional stimuli. Neuroreport, 27, 789-792. Morgan, R. E., Palinkas, L. A., Barrett-Connor, E. L., & Wingard, D. L. (1993). Plasma cholesterol and depressive symptoms in older men. The Lancet, 341, 75-79. Morimoto, A., Murakami, N., Nakamori, T., Sakata, Y., & Watanabe, T. (1989a). Brain regions involved in the development of acute phase responses accompanying fever in rabbits. Journal of Physiology, 416, 645-657. Morimoto, A., Sakata, Y., Watanabe, T., & Murakami, N. (1989b). Characteristics of fever and acute-phase responses induced in rabbits by IL-1 and TNF. American Journal of Physiology, 256, 35-41. Muller, N., Ackenheil, M., & Hofschuster, E. (1989). Altered T-cell number and reduced suppressor cell activity in patients with affective psychosis. In J. W. Hadden, K. Masek, & G. Nistico (Eds.), Interactions Among Central Nervous System, Neuroendocrine, and Immune Systems (pp. 385-394). Rome: Pythagora Press. Muller, N., Hofschuster, E., Ackenheil, M., Mempel, W., & Eckstein, R. (1993). Investigations of the cellular immunity during depression and the free interval: evidence for an immune activation in affective psychosis. Progress in Neuropsychopharmacology and Biological Psychiatry, 17, 713-730. Mussalo-Rauhamaa, H., Konttinen, Y. T., Lehto, J., & Honkanen, V. (1988). Predictive clinical and laboratory parameters for serum zinc and copper in rheumatoid arthritis. Annals of the Rheumatic Diseases, 47, 816-819. Nassberger, L. & Traskman-Bendz, L. (1993). Increased soluble interleukin-2 receptor concentrations in suicide attempters. Acta Psychiatrica Scandinavica, 88, 48-52. Navarra, P., Tsagarakis, S., Faria, M. S., Rees, L. H., Besser, G. M., & Grossman, A. B. (1991). Interleukins-1 and -6 stimulate the release of corticotropin-releasing hormone 41 from rat hypothalamus in vitro via the eicosanoid cyclooxygenase pathway. Endocrinology, 128, 37-44. Neveu, P J., Deleplanque, B., Puglisi-Allegra, S., D'Amato, F R., & Cabib, S. (1994). Influence of early life events on immune reactivity in adult mice. Developmental Psychohiology, 27, 205-213. Nguyen, K. T, Deak, T., Owens, S. M., Kohno, T., Fleshner, M., Watkins, L. R., & Maier, S. F (1998). Exposure to acute stress induces brain interleukin-1 beta protein in the rat. Journal of Neuroscience, 18, 2239-2246. Nicoloff, J.T. & LoPresti, J. S. (1991). Nonthyroidal illness. In L. E. Braverman & R. D. Utiger (Eds.), The thyroid: A Functional and Clinical Text (pp. 357-367). New York: J.B. Lippincott Company. Nordlind, K., Sundstrom, E., & Bondesson, L. (1992). Inhibiting effects of serotonin antagonists on the proliferation of mercuric chloride stimulated human peripheral blood T lymphocytes. International Archives of Allergy and Immunology, 97, 105-108. Peet, M., Murphy, B., Shay, J., & Horrobin, D. (1998). Depletion of omega-3 fatty acid levels in red blood cell membranes of depressive patients. Biological Psychiatry, 43, 315-319. Perini, G. I., Zara, M., Carraro, C , Tosin, C , Gava, F , Santucci, M. G., Valverde, S., & Defranchis, G. (1995). Psychoimmunoendocrine effects of panic disorder. Human Psychopharmacology, 10, 461465. Persoons, J. H., Schornagel, K., Breve, J., Berkenbosch, F , & Kraal, G. (1995). Acute stress affects cytokines and nitric oxide production by alveolar macrophages differently. American Journal of Respiratory and Critical Care Medicine, 152, 619-624. Ramamoorthy, S., Ramamoorthy, J. D., Prasad, P. D., Bhat, G. K., Mahesh, V. B., Leibach, F H., & Ganapathy, V. (1995). Regulation of the human serotonin transporter by interleukin-1 beta. Biochememical and Biophysical Research Communications, 216, 560-567. Russo, C , Olivieri, O., Girelli, D., Guarini, P., Pasqualhni, R., Azzini, M., & Corrocher, R. (1997). Increased membrane ratios of metabolite to precursor fatty acid in essential hypertension. Hypertension, 29, 1058-1063. Sakic, B., Szechtman, H., Braciak, T., Richards, C , Gauldie, J., & Denburg, J. A. (1997). Reduced preference for sucrose in autoimmune mice: a possible role of interleukin-6. Brain Research Bulletin, 44, 155165. Sapolsky, R., Rivier, C , & Yamamoto, G. (1987). Interleukin-1 stimulates the secretion of hypothalamic corticotropin-releasing factor. Science, 238, 522-524, 1987. Seidel, A., Arolt, V., Hunstiger, M., Rink, L., Behnisch, A., & Kirchner, H. (1995). Cytokine production and serum proteins in depression. Scandinavian Journal of Immunology, 41, 534-538.

Major Depression and Activation of tiie Inflammatory Response System

45

Seidel, A., Arolt, V., Hunstiger, M., Rink, L., Behnisch, A., & Kirchner, H. (1996a). Major depressive disorder is associated with elevated monocyte counts. Acta Psychiatrica Scandinavica, 94,198-204. Seidel, A., Arolt, v., Hunstiger, M., Rink, L., Behnisch, A., & Kirchner, H. (1996b). Increased CD56+ natural killer cells and related cytokines in major depression. Clinical Immunology and Immunopathology, 78, 83-85. Shintani, F., Nakaki, T., Kanba, S., Kato, R., & Asai, M. (1995a). Role of interleukin-1 in stress responses. A putative neurotransmitter. Molecular Neurobiology, 10, Al-lX. Shintani, F., Nakaki, T , Kanba, S., Sato, K., Yagi, G., Shiozawa, M., Aiso, S., Kato, R., & Asai, M. (1995b). Involvement of interleukin-1 in immobilization stress-induced increase in plasma adrenocorticotropic hormone and in release of hypothalamic monoamines in the rat. Journal of Neuroscience, 15, 1961-1970. Shizuya, K., Komori,T., Fujiwara, R., Miyahara, S., Ohmori, M., & Nomura, J. (1997). The influence of restraint stress on the expression of mRNAs for IL-6 and the lL-6 receptor in the hypothalamus and midbrain of the rat. Life Sciences, 61, PL. Sinclair, A. J., Johnson, L., O'Dea, K., & Holman, R. T. (1994). Diets rich in lean beef increase arachidonic acid and long-chain 0)3 polyunsaturated fatty acid levels in plasma phospholipids. Lipids, 29,337-343. Sluzewska, A., Nowakowska, E., Gryska, K., & Mackiewicz, A. (1994). Haptoglobin levels in a chronic mild stress model of depression in rats before and after treatment. European Neuropharmacology, P-1-18, 302. Sluzewska, A., Rybakowski, J. K., Laciak, M., Mackiewicz, A., Sobieska, M., & Wiktorowiz, K. (1995). Interleukin-6 serum levels in depressed patients before and after treatment with fluoxetine. Annals of the New York Academy of Sciences, 762, 474^76. Sluzewska, A., Rybakowski, J., Bosmans, E., Sobieska, M., Berghmans, R., Maes, M., & Wiktorowicz, K. (1996a). Indicators of immune activation in major depression. Psychiatry Research, 64, 161-167. Sluzewska, A., Rybakowski, J. K., Sobieska, M., & Wiktorowicz, K. (1996b). Concentration and microheterogeneity glycophorms of alpha-1-acid glycoprotein in major depressive disorder. Journal of Affective Disorders, 39, 149-155. Smejkal-Jagar, L. & Boranic, M. (1994). Serotonin, serotonergic agents, and their antagonists suppress humoral immune reaction in vitro. Research in Experimental Medicine, 194, 297-304. Smith, R. S. (1991). The macrophage theory of depression. Medical Hypotheses, 35, 298-306. Solomons, N. W. (1988). Zinc and copper. In M. E. Shils & V. R. Young (Eds.), Modern Nutrition in Health and Disease (pp. 238-262). Philadelphia: Lea & Febiger. Song, C. & Leonard, B. E. (1994). An acute phase protein response in the olfactory bulbectomized rat: effect of sertraline treatment. Medical Science Research, 22, 313-314. Song, C , Dinan, T, & Leonard, B. E. (1994). Changes in immunoglobulin, complement, and acute phase protein levels in the depressed patients and normal controls. Journal of Affective Disorders, 30, 283-288. Song, C , Lin, A., Bonaccorso, S., Heide, C , Verkerk, R., Kenis, G., Bosmans, E., Scharpe, S., Cosyns, P., DeJong, R., & Maes, M. (1998). The inflammatory response system and the availability of tryptophan to the brain of patients with major depression and sleep disorders. Journal of Affective Disorders, 49, 211-219. Soyland, E., Lea, T, Sandstad, B., & Drevon, A. (1994). Dietary supplementation with very long-chain n-3 fatty acids in man decreases expression of the interleukin-2 receptor (CD25) on mitogen-stimulated lymphocytes from patients with inflammatory skin diseases. European Journal of Clinical Investigation, 24, 236-242. Spath-Schwalbe, E., Hansen, K., Schmidt, F, Schrezenmeier, H., Marshall, L., Burger, K., Fehm, H. L., & Born, J. (1998). Acute effects of recombinant human interleukin-6 on endocrine and central nervous sleep functions in healthy men. Journal of Clinical Endocrinology and Metabolism, S3, 1573-1579. Srinivas, U., Braconier, J. H., Jeppsson, B., Abdulla, M., Akesson, B., & Ockerman, P. A. (1988). Trace element alterations in infectious diseases. Scandinavian Journal of Clinical and Laboratory Investigation, 48, 495-500. Suzuki, E., Shintani, F, Kanba, S., Asai, M., & Nakaki, T. (1996). Induction of interleukin-1 and interleukin1 receptor antagonist mRNA by chronic treatment with various psychotropics in widespread area of rat brain. Neuroscience Letters, 215, 201-4. Swartz, C. M. (1990). Albumin decrement in depression and cholesterol decrement in vnama. Journal of Affective Disorders, 19, 173-176. Takaki, A., Huang, Q. H., Somogyvari-Vigh, A., & Arimura, A. (1994). Immobilization stress may increase plasma interleukin-6 via central and peripheral catecholamines. Neuroimmunomodulation, 1, 335-342.

46

M. Maes

Tilders, F. J. & Schmidt, E. D. (1998). Interleukin-1-induced plasticity of hypothalamic CRH neurons and long-term stress hyperresponsiveness. Annals of the New York Academy of Sciences, 840, 65-73. Valentine, A. D., Meyers, C. A., Kling, M. A., Richelson, E., & Hauser, P. (1998). Mood and cognitive side effects of interferon-alpha therapy. Seminars in Oncology, 25 (1 Suppl 1), 39-Al. Van Dam, A. M., Malinowsky, D., Lenczowski, M. J., Bartfai, T., & Tilders, F. J. (1998). Interleukin 1 (IL-1) type I receptors mediate activation of rat hypothalamus-pituitary-adrenal axis and interleukin 6 production as shown by receptor type selective deletion mutants of IL-lbeta. Cytokine, 70,413^17. Vandoolaeghe, E., DeVos, N., DeSchouwer, P., Neels, H., & Maes, M. (1999). Lower number of red blood cells, hematocrit, and hemoglobin in major depression: effects of antidepressants. Human Psychopharmacology, accepted. Van Hunsel, E, Wauters, A., Vandoolaeghe, E., Neels, H., Demedts, P., & Maes, M. (1996). Lower total serum protein, albumin, and beta- and gamma-globulin in major and treatment-resistant depression: effects of antidepressant treatments. Psychiatry Research, 65, 159-169. Van Miert, A. S., Van Duin, C. T., & Wensing, T. (1990). Fever and changes in plasma zinc and iron concentrations in the goat. The effects of interferon inducers and recombinant IFN-alpha 2a. Journal of Comparative Pathology, 103, 289-300. Wartofsky, L. & Burman, K. D. (1982). Alterations in thyroid function in patients with systemic illnesses: The "euthyroid sick syndrome". Endocrinology, 3, 164-217. Washburn, T C , Medearis, J., & Childs, B. (1965). Sex differences in susceptibility to infections. Pediatrics, 35, 57-64. Wehmann, R. E., Gregerman, R. L, Burns, W. H., Saral, R., & Santos, G. W. (1985). Suppression of thyrotropin in the low-thyroxine state of severe nonthyroidal illness. The New England Journal of Medicine, 312, 546-552. Weinberg, S. B., Schulteis, G., Fernando, A. G., Bakhit, C., & Martinez, J. L. (1988). Decreased locomotor activity produced by repeated, but not single, administration of murine-recombinant interferon-gamma in mice. Life Sciences, 42, 1085-1090. Weiss, G., Fuchs, D., Hansen, A., Reibnegger, G., Werner, E. R., Werner-Felmayer, G., & Wachter, H. (1992). Iron modulates interferon-gamma effects in the human myelomonocytic cell line THP-1. Experimental Hematology, 20, 605-610. Xia, Z., DePierre, J. W, & Nassberger, L. (1996). Tricyclic antidepressants inhibit IL-6, IL-ip, and TNF-a release in human blood monocytes and IL-2 and interferon-y in T cells. Immunopharmacology, 34, 27-37. Yirmiya, R. (1996). Endotoxin produces a depressive-like episode in rats. Brain Research, 711, 163-174. Yirmiya, R. (1997). Behavioral and psychological effects of immune activation: implications for depression due to a general medical condition. Current Opinion in Psychiatry, 10, 470-476. Young, M. R. & Matthews, J. P. (1995). Serotonin regulation of T cell subpopulations and of macrophage accessory function. Immunology, 84, 148-152. Zhou, D., Kusnecov, A. W, Shurin, M. R., DePaoli, M., & Rabin, B. S. (1993). Exposure to physical and psychological stressors elevates plasma interleukin 6: relationship to the activation of hypothalamicpituitary-adrenal axis. Endocrinology, 133, 2523-2530.

CYTOKINE PRODUCTION IN DEPRESSED PATIENTS Andreas Seidel,'* Matthias Rothermundt,^ and Lothar Rink' ' Institute of Immunology and Transfusion Medicine ^ Department of Psychiatry University of Liibeck School of Medicine Ratzeburger Allee 160-162, 23538 Liibeck, Germany

I. INTRODUCTION Traditionally, the nervous, endocrine, and immune systeme have been regarded as separate systems both in clinical circumstances as well as in research. A concept of an interrelationship between the immune system and the nervous system, i.e. the psychological state, has been suggested before the first studies in psychoneuroimmunology have been conducted. Several authors implicated that mood disturbances may result in an increased susceptibility to infectious or neoplastic diseases (Ader, Felten, & Cohen, 1991; Crow 1978; King, Cooper, Earle, Martin, McFerran, Rima, & Wisdom, 1985). However, it is now widely accepted that the immune system interacts with the nervous and the endocrine system both anatomically and functionally (Cunningham and DeSouza 1993; Fabry, Raine, & Hart, 1994; Goetzel and Sreedharan, 1992). Lymphatic organs, e.g. the thymus, lymph nodes, and the bone marrow are innervated by nerve fibers (Goetzel and Sreedharan, 1992; Homo-Delarche and Dardenne, 1993). More than thirty different receptors for hormones, neurotransmitters, and neuropeptides have been described for different leukocytes (Goetzel and Sreedharan, 1992; Homo-Delarche and Dardenne, 1993). Furthermore, immune cells are able to produce and release hormones and neuropeptides. The production of ACTH and p-endorphine by mononuclear cells is one example (Homo-Delarche and Dardenne, 1993). Another example for these interactions is the fact that neurons and leukocytes are able to express identical surface antigens, e.g. CD 4 or CD 56 (Homo-Delarche and Dardenne,

*Telephone +49 451 5002840,Telefax +49 451 5002857, E-mail [email protected] Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

47

48

A. Seidel et al

1993). Within the central nervous system (CNS) immunocompetent resident cells such as microglia or circulating immune cells (lymphocytes or monocytes) can be found under physiological conditions (Cserr and Knopf, 1992; Homo-Delarche and Dardenne, 1993). Although there is only weak expression of MHC (major histocompatibility complex) class I and II molecules on cerebral cells, e.g. microglia, under normal conditions, the expression of those molecules will be augmented under influence of cytokines such as IL-2 or IFN-y or due to viral infections (Fabry et al., 1994; HomoDelarche and Dardenne, 1993). Cytokines are a heterogenous group of soluble polypeptides, which are mainly produced by monocytes or macrophages (monokines) as well as lymphocytes (lymphokines). These mediators act as signals on a variety of target cells and act within a complex network. Cytokines have a short half-life usually due to rapid degradation as a way of regulation and are therefore difficult to measure in the circulation. Several cytokines may act on a particular cell, either simultaneously or sequentially, and this combination results in distinct functions (Abbas, Lichtman, & Pober, 1991; Roitt, Brostoff, & Male, 1989). Although the CNS should be viewed as an immunoprivileged organ because of the blood-brain barrier (BBB), investigations during the last decade showed that it is not absolutely separated from the immune system. Alterations of the BBB around the paraventricular regions have been demonstrated as well as capillary structures within the brain and drainage of circulating immune cells from the CNS towards cervical lymphnodes (Cserr and Knopf, 1992). Receptors for cytokines such as interleukin (IL)-l (type I and II), IL-2, IL-3, IL-6, tumor necrosis factor (TNF), and interferons (IFN) have been described in central nervous tissue, e.g. the hypothalamic-pitutary region (Bartfai and Schultzberg, 1993; Fabry et al., 1994; Homo-Delarche and Dardenne, 1993; Hopkins and Rothwell, 1995). Cells of the CNS such as microglia, astrocytes, endothelial cells of cerebral vessels, and pericytes have been shown to produce and secrete several cytokines and, furthermore, to act as antigen-presentating cells (Cunningham and DeSouza, 1993; Rothwell and Hopkins, 1995). Microglial cells express all three types of Fc-receptors for immunoglobulins on their surfaces. After attachment of those immunoglobulins these cells release cytokines, which are able to activate lymphocytes (Ulvestad, Williams, Matre, Nyland, Olivier, & Antel, 1994). Macrophages as well as microglial cells sometimes function as a part of the BBB themselves (Goetzel and Sreedharan, 1992). In former times the term depression (of mood) has been used without specifity, meaning compromised psychological functions. Nowadays it can be defined as a passing feeling with no serious consequences, as a clinical symptom, as a syndrome or as a certain entity of disease (Frombonne, 1995; Kraepelin, 1899; Lehtinen and Joukama, 1994). The international accepted classifications for depressive states of the World Health Organisation (WHO, 1994) and the American Psychiatric Association (1994) provide a better statistic and documentation of affected patients and are helpful for comparative research studies. Since depression is a complex clinical phenomenon, i.e. a heterogenous disorder, the term depression in the following is used throughout merely as a convention.

II. CYTOKINES AND MOOD DISTURBANCES Neuropsychiatric symptoms have been observed in (auto)immune diseases, whereas alterations of the immune system in several neurological and/or psychiatric

Cytokine Production in Depressed Patients

49

disorders have also been reported (Ader et al., 1991). However, there is still need to prove whether those immunological disturbances are the reason for these neuropsychiatric symptoms or/and if neurological/psychiatric diseases are the cause for the observed immunological variations. Alterations of white blood cells, i.e. immunocompetent and cytokine releasing cells, were found before more precise methodology allowed the determination of leucocyte function in detail (Darko, Rose, Gillin, Golshan, & Baird, 1988; Kronfol, Turner, Nasrallah, & Winokur, 1984; Schleifer, Keller, Meyerson, Raskin, Davis, & Stein, 1984). Although Bartrop et al. described a reduced lymphocyte function after severe stress due to bereavement in 1977, it was some years before immunological studies were systematically carried out in patients suffering from depression (Bartrop, Lazarus, Luckhurst, Kiloh, & Penny, 1977). Since immunotherapy has been used in several diseases, e.g. the treatment of hepatitis or morbus behcet with interferons or the treatment of certain neoplastic diseases with interleukins, neuropsychiatric symptoms as side effects of those therapies have become evident (Meyers and Valentine, 1995). Neurostimulatory effects due to cytokines, which can be observed under experimental or therapeutic conditions, demonstrate that peripheral administration of those soluble proteins affect the CNS (Bartfai and Schultzberg, 1993; Denicoff, Rubinow, & Papa, 1987; Fabry et al., 1994; McDonald, Mann, & Thomas, 1987). The cytokine IL-1 is able to induce depressive-like symptoms such as psychomotor and appetite disturbances, alterations of sleep, lethargy, and weakness (Cunningham and DeSouza, 1993; Hopkins and Rothwell, 1995). IL-1 and IL-6 activate the hypothalamic-pituitary axis (HPA) with subsequent release of CRH and ACTH (Bartfai and Schultzberg, 1993; Hopkins and Rothwell, 1995). Alterations of the HPA axis are consistent findings in depressive states (Gold, Goodwin, & Chrousos, 1988). Moreover, cytokines in certain conditions modulate central neurotransmitters and neuropeptides, both directly and indirectly (Rothwell and Hopkins, 1995).Therapeutic administration of IL-2 and IFN-acan induce depression-like episodes (McDonald et al., 1987; Meyers and Valentine, 1995). Recent studies indicate that several immunological disturbances can be found in depression (Maes, 1995; Seidel, Arolt, Hunstiger, Rink, Behnisch, & Kirchner, 1995; Sluzewska, Rybakowski, Bosmans, Sobieska, Berghmans, Maes, & Wiktorowicz, 1996). In 1991 Smith formulated the macrophage theory of depression, which proposes that an excessive secretion of cytokines such as IL-1, IL-6 or IFN-a, might cause depression (Smith, 1991). Smith suggested that at least a certain percentage of depressive patients have immunological abnormalities as primary changes and that behavioural deficits, central neurotransmitter and HPA-axis abnormalities might be secondary.

III. IMMUNE ACTIVATION AND CYTOKINES IN DEPRESSION The first results of immunological investigations in depressive patients revealed a reduced immune function, since the majority of the studies demonstrated blunted in vitro mitogen induced lymphoproliferative response of peripheral blood mononuclear cells (PBMC) and decreased natural killer cell activity in these individuals in comparison to healthy controls (Darko, Wilson, Gillin, & Golshan, 1991; Evans, Folds, Pettito, Golden, Pedersen, Corrigan, Gilmore, Silva, Quade, & Ozer, 1992; Irwin, Caldwell, Smith, Brown, Schuckit, & Gillin, 1990; Schleifer et al., 1984; Stein, Miller, & Trestman, 1991).

50

A. Seidel et al.

Within the last ten years additional evidence has come to suggest that depression is associated by an activation of the immune system (Maes, 1995; Seidel et al., 1995, Sluzewska et al., 1996). Most of the studies showed an increase of leukocytes in the peripheral blood, mainly manifested by a higher number of granulocytes and monocytes (Darko et al., 1988; Maes, Van der Planken, Stevens, Peeters, DeClerck, Bridts, Schotte, & Cosyns, 1992b; Seidel, Arolt, Hunstiger, Rink, Behnisch, & Kirchner, 1996a). Results for the absolute numbers of lymphocytes are still contradictory, although increased numbers of CD4+ lymphocytes or higher CD4/CD8 ratios have been described in depression several times (Maes, Lambrechts, Bosmans, Jacobs, Suy, Vandervorst, DeJonckheere, Minner, & Raus, 1992a; Muller, Hofschuster, Ackenheil, Mempel, & Eckstein, 1993; Seidel el al., 1996a), whereas other studies have failed to reproduce these results (Denney, Stephenson, Penick, & Weller, 1988; Targum, Clarkson, MagacHarris, Marshall, & Skwerer, 1990). Lymphocyte staining by monoclonal antibodies against the IL-2 receptor (IL-2R) and HLA-DR revealed increased numbers of these peripheral lymphocytes in patients with depression (Maes, Bosmans, Vandervorst, DeJonckheere, & Raus, 1990; Maes et al., 1992a). Some studies showed a significant increase of natural killer cells in the acute state of the disease (Ravindran, Griffiths, Merali, & Anisman, 1995; Seidel, Arolt, Hunstiger, Rink, Behnisch, & Kirchner, 1996b). Furthermore, one study showed an increase of CD56+ natural killer cells and a correlation between lymphokine secretion in the supernatants of mitogen stimulated leukocyte cultures (IL-2, IFN-y) and these cells (Seidel et al., 1996b). In addition, many studies showed higher levels of acute phase proteins (APP), e.g. haptoglobin, C-reactive protein, al-antitrypsin or a2-macroglobulin, or neopterin in serum of depressed patients (Joyce, Hawes, Mulder, Sellman, Wilson, & Boswell, 1992; Maes et al., 1990; Maes, Scharpe, Meltzer, & Cosyns, 1993b; Seidel et al., 1995). Acute phase proteins are produced by hepatocytes normally in the early stage of inflammation due to the production of monokines such as IL-1 or IL-6 (Baumann and Gauldie, 1994). Concerning cytokines, within the last ten years evidence evolved that there are remarkable in vivo and in vitro findings in depressive patients (Joyce et al., 1992; Maes et al., 1990; Maes et al., 1992a; Seidel et al., 1995; Seidel et al., 1996a; Sluzewska et al., 1996). Maes et al. demonstrated a higher monokine production in the supernatant of mitogen-stimulated PBMC (Maes, Bosmans, Suy, Vandervorst, DeJonckheere, & Raus, 1991; Maes et al., 1993a). However, in our longitudinal study there was no significant difference in the supernatant concentrations of the monokines IL-ip and IL-6 between depressive patients in the acute clinical state and controls (Seidel et al., 1995). A possible explanation for these different findings comes from the methods used. Whereas other groups assessed cytokine production of isolated PBMC, we used a whole blood assay for cytokine production (Table 1). The whole blood assay has the advantage of determining cytokine production of leukocytes in their natural setting (DeGroote, Zangerle, Gevaert, Fassotte, Beguin, Noizat-Pirenne, Pirenne, Gathy, Lopez, Debar, Igot, Baudrihaye, Delacroix, & Franchimont, 1992; Kirchner, Kleinicke, & Digel, 1982). Since we demonstrated a significant increase of the APP a2macroglobulin in depressive patients as compared to controls, it is conceivable that a certain proportion of the cytokines present was not detected due to binding to this APP. a2-macroglobulin exerts a potentially important influence on cytokines, since the activated form of this protein seems to bind several cytokines, which may contribute to increased systemic activities of bound cytokines (James, 1990; Taylor and Mortensen,

Cytokine Production in Depressed Patients

51

Table 1. Cytokine production in leukocyte cultures of depressed patients in comparison to healthy controls (in v/^ro-studies) Cytokine

Reference/s

Comment

IL-ipT IL-1 i IL-lpn.s. IL-2i IL-2 n.s. IL-3i IL-6 n.s. IL-10 n.s. IFN-Y t IFN-y T

Maeset al. 1991; 1993a Weizmann et al. 1994 Seidel et al. 1995 Weizmann et al. 1994 Seidel et al. 1995 Weizmann et al. 1994 Seidel et al. 1995 Seidel et al. 1995 Maes et al. 1994 Seidel et al. 1995

PBMC, PHA PBMC, LPS, NSAID WBA, PHA PBMC WBA. PHA PBMC WBA, PHA WBA, PHA PBMC, PHA WBA, PHA

Cytokine levels: T = increased, i = decreased, n.s. = no significant difference, PBMC—Peripheral blood mononuclear cells, WBA—Whole blood assay, PHA— Phytohaemagglutinin, LPS—Lipopolysaccharide, NSAID—indomethacin in leukocyte cultures, IL—Interleukin, IFN—Interferon.

1991). In the supernatants of PBMC such an effect cannot be seen due to isolation of the cells from the autologous serum. Maes et al. furthermore reported increased concentrations of the IL-1-receptor antagonist (IL-IRA) in serum of depressive patients. IL-IRA is released in vivo during the acute phase reaction and may limit the proinflammatory effects of IL-1 (Maes, Bosmans, DeJongh, Kenis, Vandoolaeghe, & Neels, 1997). Other studies showed an increase of IL-1, IL-6, and soluble IL-6-receptor in the serum of depressed patients (Frommberger, Bauer, Haselbauer, Fraulin, Riemann, & Berger, 1997; Maes et al., 1997; Sluzewska et al., 1996). These in-vivo findings were mostly accompanied by increased APR One group reported, that initially increased serum IL-1 concentrations in depressive patients returned to normal in those patients who responded to a 12-week treatment of sertraline (Griffiths, Ravindran, Merali, & Anisman, 1996). The study of Frommberger et al. revealed increased serum concentrations of IL-6 during the acute state of depression, which decreased after remission (Frommberger et al., 1997). These findings support the possible importance of monokines in the onset of depressive symptomatology. However, the study of Brambilla and Maggioni (1998) showed no significant differences of serum IL-ip, IL-6, and TNF-a levels of depressed elderly women compared with age-matched controls before and after 30 days of treatment with antidepressants (Table 2). Monocytes, i.e. monokine releasing immune cells, have been shown to be elevated several times in the peripheral blood of depressive patients (Maes et al., 1992b; Seidel et al., 1996a). In our longitudinal study we were able to demonstrate, that patients who recovered well during hospitalization showed a decrease in monocyte counts, whereas those with a slower clinical improvement had an ongoing monocytosis (Seidel et al., 1996a). Moreover, McAdams and Leonard (1993) reported a significantly increased phagocytic activity of monocytes in depressive patients. Interestingly, these alterations reversed upon successful antidepressant treatment. This may additionally support the hypothesis that monocytes play a role in the acute phase of depression or modulate the clinical state of depressive patients. In contrast to these findings one recent published study revealed an unaltered monocyte function in patients with major depres-

52

A. Seidel et al. Table 2. Cytokine levels in the peripheral blood of depressed patients in comparison to healthy controls (in vivo-studies) Cytokine

Reference/s

Comment

IL-1 T IL-1 n.s. IL-IRA T IL-6T

Griffiths et al. 1996 Brambilla &Maggioni 1998 Maeset al. 1997 Sluzewska et al. 1996 Frommberger et al. 1997 Maes et al. 1997 Brambilla &Maggioni 1998 Brambilla &Maggioni 1998

Decrease after improvement 10 eldery depressive women

IL-6 n.s. TNF-a n.s.

Decrease after remission 10 elderly depressive women 10 elderly depressive women

Cytokine levels: T = increased, i = decreased, n.s. = no significant difference, IL—Interleukin,TNF—^Tumor necrosis factor, RA—Receptor antagonist.

sion before and after three months of antidepressive therapy (Landmann, Schaub, Link, & Wacker, 1997). In some studies on lymphokine production elevations of IL-2, IFN-y, and IL-10 were observed in the supernatant of leukocyte cultures from depressive patients (Maes, Scharpe, Meltzer, Okayli, Bosmans, D'Hondt, Vanden Bossche, & Cosyns, 1994; Seidel et al., 1995) (Table 1). Together with the simultaneous increase of IL-2 and IFN-y on one hand, and IL-10 on the other hand, we detected increased lymphokine secretion of THI and TH2 producing cells in the leukocyte cultures from patients upon admission to hospital (Seidel et al., 1995). Interestingly, variations in both lymphokine types showed identical trends during the period of observation (six weeks); in patients all lymphokines decreased significantly during the investigation and reached a lower level than those from the controls at the end of the investigation. Since THI and TH2 released cytokines are usually considered to be antagonists (Powrie and Coffman, 1993), this simultaneous increase of IL-2, IFN-y, and IL-10 suggests an overall activation of lymphocytes in the early stage of illness. These findings underline the importance of an elevated monokine production in depression, since such a pattern of cytokine release might be explained by lymphocyte activation due to monocyte derived inflammatory cytokines. In contrast to these studies, Weizman observed a decreased production of IL-1, IL-2, and IL-3 in mitogen stimulated PBMC (Weizman, Laor, Podliszewski, Notti, Djaldetti, & Bessler, 1994). In these patients the initially decreased cytokine concentrations normalized following clomipramine treatment. Whereas Weizmann stimulated PBMC with lipopoysaccharide to induce IL-1 production, the other studies, which focused on monokine production in vitro used phytohaemagglutinin as stimulant (Maes et al., 1991; Maes et al., 1993a; Seidel et al., 1995; Weizmann et al., 1994). Furthermore, the group of Weizman added indomethacin to PBMC in order to prevent possible inhibitory effects of prostaglandins on IL-1 production (Weizman et al., 1994) (Table 1). However, it has been shown that (in vivo) administration of nonsteroidal antiinflammatory drugs to healthy volunteers, e.g. piroxicam, results in a decreased IL1 production per se (Rosenstein, Kunicka, Kramer, & Goldstein, 1994). The concentrations of the soluble IL-2 receptor (sIL-2R), an indicator for T-cell activation, have been shown to be elevated in depressive patients both in serum and supernatants (Maes et al., 1990; Maes et al., 1992a; Seidel et al., 1995). We found that patients with a sIL-2R concentration in serum of more than 600U/ml on admission revealed an outcome worse than patients without that sign of systemic activation of the

Cytokine Production in Depressed Patients

53

immune system (unpublished data). Such a finding is of scientific and diagnostic interest, because it shows that depressive patients without a systemic immune response are more hkely to have a better outcome. In conclusion, most of the clinical studies carried out in depressed patients reveal signs of immune activation in the acute state of disease, since the concentrations of monokines {in vivo and in vitro), lymphokines {in vitro), and cytokine receptors {in vivo and in vitro) are increased in depressive patients. Furthermore, in our longitudinal study, which addressed cytokine production over a six-week period, cytokine production normalized in most of the patients with improvement of the depressive symptoms (Seidel et al., 1995). This underlines the hypothesis that immune activation, i.e, overproduction and secretion of cytokines plays a crucial role in the onset of depressive symtoms. However, there are still contradictory results in the presently available literature, which could be explained by heterogenous disease entities and severity and/or different immunologic methods.

IV. HOW CAN PERIPHERALLY PRODUCED CYTOKINES ALTER BRAIN FUNCTION? In all conducted studies cytokine concentrations have been measured directly in the serum/plasma {in vivo) or after stimulation of peripheral blood leukocyte cultures with mitogens {in vitro). Either peripherally or centrally produced cytokines can affect the CNS (Connor and Leonard, 1997). Although cytokines are hydrophilic and large molecules, that make a passive diffusion through the BBB unlikely, there are several pathways by which peripherally released cytokines can have central effects (Hopkins and Rothwell, 1995). Cytokines can enter the brain at those sites where the BBB is deficient, e.g. the circumventricular organs. Cytokines bound to glial cells in these regions are believed to subsequently release other cytokines or mediators such as prostaglandins. Thus the peripheral release of cytokines can result in the synthesis and secretion of cytokines in the CNS (Connor and Leonard, 1997; Meyers and Valentine, 1995). Another mechanism is an active transport of cytokines across the BBB. An active transport mechanism has been described for IL-1 andTNF-a (Fabry et al., 1994; Connor and Leonard, 1997). The finding of an elevation of the APP a2-macroglobulin in depression is of particular interest because of its capacity to bind cytokines, since they are protected from proteases in the protein-cytokine-complex (James, 1990; Taylor and Mortensen, 1991). This may result in an increased half-life and systemic activity of bound cytokines (Seidel et al., 1995). There is also experimental evidence for the existence of a neurally mediated communication pathway between peripherally produced cytokines and the brain, although its mechanism is not yet fully understood (Connor and Leonard, 1997). In systemic inflammatory diseases an alteration of the BBB can occur. Cytokines such as IL-1 and IFN-y lead to an increased expression of adhesion molecules on the endothelial cells of cerebral blood vessels (Fabry et al., 1994). IFN-y induces an augmented expression of MHC class II molecules on perivascular macrophages, endothelial cells and pericytes, thus resulting in a further alteration of the cellular elements of the BBB (Fabry et al., 1994). Cultured microglial cells coincubated with IFN-y display an increased expression of adhesion molecules and Fc receptors and

54

A. Seidel et al.

an augmented synthesis and secretion of several cytokines (Patrizo, Costa, & Levi, 1995). Alterations of the BBB such as increased expression of adhesion molecules and MHC class II complexes may result in migration of lymphocytes into the CNS (Hickey, 1991; Homo-Delarche and Dardenne, 1993; Wekerle, Engelhardt, Risau, & Meymann, 1991). An activation of lymphocytes in depressive patients has been demonstrated, since those patients express high amounts of activation markers like CD25 (IL-2R) or HLADR on the cell surfaces (Maes et al., 1990; Maes et al., 1992a). Since in depressive patients signs of a systemic inflammatory process have been found, an alteration of the BBB might be a crucial event in the onset and pathophysiology of depression. Under certain circumstances cytokines induce structural and anatomical changes in the CNS implementing microglial cells and astrocytes in an inflammatory process (Owens, Renno,Taupin, & Krakowski, 1994; Yong and Balasingam, 1995).Thus, an overproduction or a massive non-physiological secretion of cytokines in depression could explain both the onset ("acute depressive episodes") and maintanance ("long lasting episodes", "chronic depressive states") of depressive illness. The systemic activation of the immune system in depression may play a crucial role in the aetiology of depression. Since the hypersecretion of cytokines, which has been shown in depression in several in vivo and in vitro studies, could provide an explanation for behaviour, endocrine, haematologic, and central neurochemical changes, further studies designed to investigate the link between central actions of cytokines and depression will be necessary to confirm and clarify the present findings.

REFERENCES Abbas, A. K., Lichtman, A. H., & Pober, J. S. (1991). Cellular and molecular immunology. Saunders Company, Philadelphia/USA. Ader, R., Felten, D. L., & Cohen, N. (1991). Psychoneuroimmunology, 2nd edition. Academic New York/USA. American Psychiatric Association (1994). DSM-IV Diagnostic and Statistical Manual of Mental Disorders, 4th edition. Bartfai, T. & Schultzberg, M. (1993). Cytokines in neuronal cell types. Neurochemistry International, 22, 435-444. Bartrop, R. W., Lazarus, L., Luckhurst, E., Kiloh, L. G., & Penny, R. (1977). Depressed lymphocyte function after bereavement. The Lancet, 834-836. Brambilla, F. & Maggioni, M. (1998). Blood levels of cytokines in eldery patients with major depressive disorder. Acta Psychiatrica Scandinavica, 97,'309-331. Baumann, H. & Gauldie, J. (1994). The acute phase response. Immunology today, 15, 74-80. Connor, T. J. & Leonard, B. E. (1997). Depression, stress and immunological activation: The role of cytokines in depressive disorders. Life Sciences, 62, 583-606. Crow, T. J. (1978). Viral causes of psychiatric disease. Postgraduate Medical Journal, 54, 763-767. Cserr, H. E & Knopf, P. M. (1992). Cervical lymphatics, the blood-brain barrier and the immunoreactivity of the brain: a new view. Immunology today, 13, 507-512. Cunningham, E. T jr & De Souza, E. B. (1993). Interleukin 1 receptors in the brain and endocrine tissues. Immunology today, 14, 171-176. Darko, D. R, Rose, 1, Gillin, J. C , Golshan, S., & Baird, S. M. (1988). Neutophilia and lymphocytopenia in major mood disorders. Psychiatry Research, 25, 243-251. Darko, D. P., Wilson, N. W., Gillin, J. C , & Golshan, S. (1991). A critical appraisal of mitogen induced lymphocyte proliferation in depressed patients. American Journal of Psychiatry, 148, 337-344. DeGroote, D., Zangerle, P. E, Gevaert, Y., Fassotte, M. E, Beguin, Y, Noizat-Pirenne, E, Pirenne, J., Gathy, R., Lopez, M., Dehart, I., Igot, D., Baudrihaye, M., Delacroix, D., & Franchimont, P. (1992). Direct Stimulation of Cytokines in whole blood. I. Comparison with isolated PBMC stimulation. Cytokine, 3,239-248.

Cytokine Production in Depressed Patients

55

Denicoff, K. D., Rubinow, D. R., & Papa, M. Z. (1987). The neuropsychiatric effects of treatment with interleukin-2 and lymphokine-activated killer cells. Annals of Internal Medicine, 107, 293-300. Denney, D. R., Stephenson, L. A., Penick, E. C , & Weller, R. A. (1988). Lymphocyte subclasses and depression. Journal of Abnormal Psychology, 97, 499-502. Evans, D. L., Folds, J. D., Pettito, J. M., Golden, R. N., Pedersen, C. A., Corrigan, M., Gilmore , J. H., Silva, S. G., Quade, D., & Ozer, H. (1992). Circulating natural killer cell phenotypes in men and women with major depression. Archives of General Psychiatry, 49, 388-395. Fabry, Z., Raine, C. S., & Hart, M. N. (1994). Nervous tissue as an immune compartment: the dialect of the immune response in the CNS. Immunology today, 15, 218-224. Frombonne, E. (1995). Increased rates of depression: Update of epidemiological findings and analytical problems. Acta Psychiatrica Scandinavica, 90, 145-156. Frommberger, U. H., Bauer, J., Haselbauer, P., Fraulin, A., Riemann, D., & Berger, M. (1997). Interleukin-6 (IL-6) plasma levels in depression and schizophrenia: comparison between the acute state and after remission. European Archives of Psychiatry and Clinical Neuroscience, 247, 228-233. Goetzel, E. J. & Sreedharan, S. P. (1992). Mediators of communication and adaptation in the neuroendocrine and immune systems. FASEB, J, 6, 2646-2652. Gold, P. W., Goodwin, F. K., & Chrousos, G. P. (1988). Clinical and biochemical manifestations of depression—part I & part II. New England Journal of Medicine, (part I) 319, 413-420; (part II) 319, 348-353. Griffiths, J., Ravindran, A. V., Merali, Z., & Anisman, H. (1996). Society of Neurosciences, Abstract, 22, 1350. Hickey, W. F. (1991). Migration of hematogenous cells through the blood-brain barrier and the initiation of CNS inflammation. Brain Pathology, I, 97-105. Homo-Delarche, F & Dardenne, M. (1993). The neuroendocrine-immune axis. Springer Sem Immunopathol, 14, 221-238. Hopkins, S. J. & Rothwell, N. J. (1995). Interactions between cytokines and the nervous system. Trends in Neurosciences, 18, 83-88. Irwin, M., Caldwell, C , Smith, T L., Brown, S., Schuckit, M. A., & Gillin, J. C. (1990). Major depressive disorder, alcoholism and reduced natural killer cell cytotoxicity. Archives in General Psychiatry. 47, 713-718. James, K. (1990). Interactions between cytokines and a2-macroglobulin. Immunology today, 11, 163-166. Joyce, R R., Hawes, C. R., Mulder, R.T, Sellman, J. D., Wilson, D. A., & Boswell, D. R. (1992). Elevated levels of acute phase proteins in major depression. Biological Psychiatry, 32, 1035-1041. King, D. J., Cooper, S. J., Earle, J. A. R, Martin, S. J., McFerran, N. V., Rima, B. K., & Wisdom, G. B. (1985). A survey of serum antibodies to eight common viruses in psychiatric patients. British Journal of Psychiatry, 147, 137-144. Kirchner, H., Kleinicke, C , & Digel, W. J. (1982). A whole blood technique for testing production of human interferon by leukocytes, yoarna/ of Immunology Methods, 48, 213-219. Kraepelin, E. (1899). Compendium der Psychiatric, Leipzig. Kronfol, Z., Turner, R., Nasrallah, H., & Winokur, G. (1984). Leukocyte regulation in depression and schizophrenia. Psychiatry Research, 13, 13-18. Landmann, R., Schaub, B., Link, S., & Wacker, H. R. (1997). Unaltered monocyte function in patients with major depession before and after three months of antidepressive therapy. Biological Psychiatry, 41, 675-681. Lehtinen, V. & Joukamaa, M. (1994). Epidemiology of depression: Prevalence, risk factors and treatment situation. Acta Psychiatrica Scandinavica Supplement, 377, 157-167. Maes, M., Bosmans, E., Vandervorst, C , DeJonckheere, C , & Raus, J. (1990). Immune disturbances during major depression: upregulated expression of interleukin-2 receptors. Neuropsychohiology, 24, 115-120. Maes, M., Bosmans, E.,Suy, E., Vandervorst, C , DeJonckheere, C.,& Raus, J. (1991). Depression-related disturbances in mitogen-induced lymphocyte responses and interleukin-ip and soluble interleukin-2 receptor production. Acta Psychiatrica Scandinavica, 84, 379-386. Maes, M., Lambrechts, J., Bosmans, E., Jacobs, J., Suy, E., Vandervorst, C , DeJonckheere, C , Minner, B., & Raus, J. (1992a). Evidence for a systemic immune activation during depression: results of leukocyte enumeration by flow cytometry in conjunction with monoclonal antibody staining. Psychological Medicine, 22, 45-53. Maes, M., Van der Planken, M., Stevens, W. J., Peeters, D., DeClerck, L. S., Bridts, C. H., Schotte, C , & Cosyns, P. (1992b). Leukocytosis, monocytosis and neutrophilia; hallmarks of servere depression. Journal of Psychiatry Research, 26, 125-134.

56

A. Seidel et al.

Maes, M., Bosmans, E., Meltzer, H. Y., Scharpe, S., & Suy, E. (1993a). Interleukin-ip: a putative mediator of HPA axis hyperactivity in major depression?. American Journal of Psychiatry, 150,1189-1193. Maes, M., Scharpe, S., Meltzer, H. Y., & Cosyns, P. (1993b). Relationship between increased haptoglobin plasma levels and activation of cell-mediated immunity in depression. Biological Psychiatry, 34, 690-701. Maes, M., Scharpe, S., Meltzer, H. Y, Okayli, G., Bosmans, E., D'Hondt, P., Vanden Bossche, B. V., & Cosyns, P. (1994). Increased neopterin and interferon-gamma secretion and lower availability of L-tryptophan in major depression: further evidence for an immune response. Psychiatry Research, 54, 153-160. Maes, M. (1995). Evidence for an immune response in major depression: a review and hypothesis. Progress in Neuropsychopharmacology and Biological Psychiatry, 19,11-38. Maes, M., Bosmans, E., DeJongh, R., Kenis, G., Vandoolaeghe, E., & Neels, H. (1997). Increased serum IL-6 and IL-1 receptor antagonist concentrations in major depression and treatment resistent depression. Cytokine, 9, 853-858. McAdams, C. & Leonard, B. E. (1993). Neutrophil and monocyte phagocytosis in depressed patients. Progress in Neuropsychopharmacology and Biological Psychiatry, 17, 971-984. McDonald, E. M., Mann, A. H., & Thomas, H. C. (1987). Interferons as mediators of psychiatric morbidity. The Lancet, 1175-1177. Meyers, C. A. & Valentine, A. D. (1995). Neurological and psychiatric adverse effects of immunological therapy. CNS drugs, 3, 56-68. Miiller, N., Hofschuster, E., Ackenheil, M., Mempel, & Eckstein, R. (1993). Investigations of the cellular immunity during depression and the free interval: evidence for an immune activation in active psychosis. Progress in Neuropsychopharmacology and Biological Psychiatry, 17, 713-730. Owens, T., Renno,T.,Taupin, V., & Krakowski, M. (1994). Inflammatory cytokines in the brain: does the CNS shape immune responses? Immunology today, 15, 566-571. Patrizio, M., Costa, T., & Levi, G. (1995). Interferon-y and lipopolysaccharide reduce cAMP response in cultured gUal cells: reversal by type IV phosphodiesterase inhibitor. Glia, 14, 94-100. Powrie, R & Coffman, R. L. (1993). Cytokine regulation of T-cell function: potential for therapeutic intervention. Immunology today, 14, 210-21 A. Ravindran, A. V, Griffiths, J., Merali, Z., & Anisman, H. (1995). Lymphocyte subsets associated with major depression and dysthymia: modification by antidepressant treatment. Psychosomatic Medicine, 57, 555-563. Roitt, I. M., Brostoff, J., & Male, D. K. (1989). Immunology. Gower Medical Publishing, London. Rosenstein, E. D., Kunicka, J., Kramer, N., & Goldstein, G. (1994). Modification of cytokine production by piroxicam. Journal of Rheumatology, 21, 904-904. Rothwell, N. J. & Hopkins, S. J. (1995). Interactions between cytokines and the nervous system. Trends in Neurosciences, 18,130-136. Schleifer, S. J., Keller, S. E., Meyerson, A. T, Raskin, M. J., Davis, K. L., & Stein, M. (1984). Lymphocyte function in major depressive disorder. Archives of General Psychiatry, 41, 484-486. Seidel, A., Arolt, V, Hunstiger, M., Rink, L., Behnisch, A., & Kirchner, H. (1995). Cytokine production and serum proteins in depression. Scandinavian Journal of Immunology, 41, 434-538. Seidel, A., Arolt, V, Hunstiger, M., Rink, L., Behnisch, A., & Kirchner, H. (1996a). Major depressive disorder is associated with elevated monocyte counts. Acta Psychiatrica Scandinavica, 94,198-204. Seidel, A., Arolt, V, Hunstiger, M., Rink, L., Behnisch, A., & Kirchner, H. (1996b). Increased CD56-t- natural killer cells and related cytokines in major depression. Clinical Immunology and Immunopathology, 78, 83-85. Sluzewska, A., Rybakowski, J., Bosmans, E., Sobieska, M., Berghmans, R., Maes, M., & Wiktorowicz, K. (1996). Indicators of immune activation in major depression. Psychiatry Research, 64,161-167. Smith, R. S. (1991). The macrophage theory of depression. Medical Hypotheses, 35, 298-306. Stein, M., Miller, A. H., & Trestman, R. T. (1991). Depression, the immune system, and health and illness. Archives of General Psychiatry, 48,171-177. Targum, S. D., Clarkson, L. L., Magac-Harris, K., Marshall, L. E., & Skwerer, R. G. (1990). Measurement of Cortisol and lymphocyte subpopulations in depressed and conduct-disordered adolescents. Journal of Affective Disorders, 18, 91-96. Taylor, A. W. & Mortensen, R. F. (1991). Effect of alpha-2-macroglobulin on cytokine-mediated human Creactive protein production. Inflammation, 15, 61-70. Ulvestad, E., Williams, K., Matre, R., Nyland, H., Olivier, A., & Ante!, J. (1994). Fc receptors for IgG on cultured human microglia mediate cytotoxicity and phagocytosis of antibody-coated targets. Journal of Neuropathology and Experimental Neurology, 53, 27-36.

Cytokine Production in Depressed Patients

57

Weizman, R., Laor, N., Podliszewski, E., Notti, I., Djaldetti, M., & Bessler, H. (1994). Cytokine production in major depressed patients before and after clomipramine treatment. Biological Psychiatry, 35, 42-47. Wekerle, H., Engelhardt, B., Risau, W., & Meyermann, R. (1991). Interaction of lymphocytes with cerebral endotheHal cells in vitro. Brain Pathology, 1,107-114. World Health Organisation ICD-10 (1994). Mental disorders: tenth revision of the International Classification of diseases. Geneva: WHO. Yong, V. W. & Balasingam, V. (1995). Cytokines as mediatiors of reactive astrogliosis. Methods in Neurosciences, 24, 220-235.

INDICATORS OF IMMUNE ACTIVATION IN DEPRESSED PATIENTS Anna Sluzewska* Department of Adult Psychiatry University of Medical Sciences in Poznan Szpitalna 27/33, 60-572 Poznan, Poland An interaction of genetic factors with environmental factors, including stress and infectious agents, is assumed to play a causal role in the pathogenesis of major depression. Since the rate of major depression in genetically succeptible populations has continued to increase over the past years, the identification of these environmental factors would provide targets for more effective antidepressant agents. There is now evidence that cytokines, peptide hormons, and neurotransmitters, as well as their receptors, are endogenous to the brain, endocrine, and immune systems (Blalok, 1994). These shared ligands and receptors provide common chemical language for communication within and between the immune and neuroendocrine systems. This organization of communication suggests an immunoregulatory role of the brain and a sensory function for the immune system. Accordingly, the immune system, may be viewed as a sensory organ that recognizes physical and emotional stress (trauma) and relays this information to the CNS and endocrine system via the cytokines. The existence of dysregulations of the stress response in depression could potentially reflect a generalized stress response that has escaped its usual counter-regulatory restraints (Gold, 1996). The stress system is active when the body is at rest, responding to many distinct circadian, neurosensory, blood-borne, and limbic signals. These signals include cytokines produced by immune-mediated inflammatory reactions, such as tumor necrosis factor a, interleukin-1 (IL-1), and interleukin-6 (IL-6) (Chrousos and Gold, 1992). On the other hand, locally, immune and immune accessory cells are activated, and cytokines, lipid mediators of inflammation and neuropeptides are generated (Paul and Seder, 1994). Usually these events are clinically silent, but inflammation occasionally causes an activation of the stress system and systemic symptoms and signs. Among the many mediators of inflammation, the three "inflammatory" cytokines TNFa, IL-la and [3, and IL-6 are responsible for the strongest stimulation of the hypo-

* Dr. Anna Sluzewska died unexpectedly in early 1999. This is the last review she wrote. Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

59

60

A. Sluzewska

thalamic-pituitary-adrenal (HPA) axis that is associated with immune/inflammatory reactions. Hypercortisolism is one of the most consistent findings in depression (Carroll et a l , 1976; Nemeroff., 1996). There is substantial evidence from clinical studies that the hypercortisolemia seen in major depression is due to hypersecretion of corticotropin-releasing hormone (CRH) (Gold et al., 1994; Holsboer et al., 1984). I M M U N E ACTIVATION IN M A J O R DEPRESSION The status of the immune system in major depression has been extensively studied over the last ten years and there is now some evidence that the acute episode of this illness may be accompanied by immune activation/acute phase response (Maes., 1993; Sluzewska et al., 1996b). The acute phase response (apr) is a response of the organism to disturbances of its homeostasis due to factors such as, infection, tissue injury, neoplastic growth or immunological disorders (Heinrich et al., 1990). Within this systemic reaction, that involves the endocrine, immunologic and metabolic system (Kushner and Mackiewicz., 1993), there are also behavioural changes, which are expressed as "sickness behaviour," characterised by psychomotor retardation, sleep disturbances, anorexia, anergy, and lethargy (Kent et al., 1992). Changes in immune response in major depression include activation of the monocytic and T-lymphocytic arm of cell-mediated immunity (CMI), the apr, and autoimmune response. An activation of the monocytic arm of CMI is suggested by: — the production of an increased number of monocytes (Kronfol et al., 1989; Irwin et al., 1990; Maes et al., 1992c; Sluzewska et al., 1994c). — increased plasma concentrations of IL-6, sIL-6R, IL-ip (Maes., 1993, 1995a; Sluzewska et al., 1995a, b, 1996b). — increased plasma concentrations of neopterin (Dunbar et al., 1992; Maes et al., 1994b). — increased plasma concentrations of prostaglandin E2 and tromboxan B2 (Lieb et al., 1993). An activation of the T-lymphocytic arm of CMI is suggested by: — increased number and percentage of T lymphocytes and T helper cells and increased T helper CD4+/T suppressor CD8+ ratio (Maes et al., 1992c, 1994a; Muller et al., 1993; Sluzewska et al., 1994c). — T-cells activation as indicated by increased number and percentage of activated T-cells (CD25+) (Maes et al., 1992c, 1993; Sluzewska et al., 1994c). — increased concentration of soluble interleukin-2-receptor (sIL-2R) and transferrin receptor (Trf) (Maes et al., 1995a; Nessberger et al., 1993; Schleifer et al., 1984; Sluzewska et al., 1996b). — increased plasma concentrations of interferon (Maes et al., 1994b). — increased plasma concentrations of soluble CD8 (sCD8) (Sluzewska et al., unpublished). The presence of apr in major depression is suggested by: — increased plasma concentrations of positive acute phase proteins (apps) such as: alpha-1-acid glycoprotein (AGP) (Nemeroff et al., 1990; Healy et al., 1992; Sluzewska et al.. 1993. 1996a^:

Indicators of Immune Activation in Depressed Patients

61

-1 antitrypsin (AT) (Maes et al., 1992a); -1 antichymotrypsin (ACT) (Joyce et al., 1992; Sluzewska et al., 1995b); haptoglobin (Hp) (Joyce et al., 1992; Maes et al., 1992a, 1993); C-reactive protein (CRP) (Sluzewska et al., 1994c); celuroplasmin (Cp) (Maes et al., 1992a) — downturn in negative acute phase proteins such as: albumin (Maes et al., 1992a, 1995b; Joyce et al., 1992; Song et al. 1993; Sluzewska et al., 1997b) transferrin (Maes et al., 1992a) The presence of autoimmune response in major depressive illness is suggested by: — increased expression of antinuclear antibodies (Gastpar et al., 1981; Maes et al., 1991); antiphospholipid antibodies (Maes et al., 1991) and antibodies against serotonin and gangliosides (Schott et al., 1992; Sluzewska et al., 1997a). During the last 15 years, a large number of studies have focused upon the ex vivo response of lymphocytes stimulated by mitogens. Most authors have observed decreased lymphocyte proliferation which is not due to decreased number of T lymphocytes but rather to the presence of immunosuppressive mechanisms such as: apr, decreased plasma levels of DPP IV (Maes et al., 1993), increased levels of sIL-2R and prostaglandins as well as Cortisol (Maes et al., 1989).

THE ACUTE PHASE RESPONSE AND THE IMPLICATION OF ALPHA-1-ACID GLYCOPROTEIN (AGP) AGP has been implicated in depression as it is an endogenous inhibitor of platelet serotonin uptake and H3-imipramine binding (Abraham et al., 1987); it is also the important binding protein for many antidepressants (Kremer et al., 1988) and an inhibitor of human platelet aggregation (Costello et al., 1979). Specific alterations in glycosylation of AGP occur in many pathophysiological states (Turner, 1992). These alterations are due to variations in the structure of the oligosaccharides present at a given site and have been referred to as the consequence of microheterogeneity. They accompany changes in the serum concentration of AGP, but are independent of its synthesis rate. The different glycophorms of serum AGP have varied immunoregulatory effects. ConA-unreactive variants show significant (50-75%) inhibition of thymocyte proliferation (Pos et al., 1990), and they are also more effective in the induction of release of interleukin-1 inhibitor (Durand., 1989). These glycophorms also decrease polymorphonuclear neutrophil chemotaxis and oxidative metabolism (Vasson et al., 1994). Changes in the serum concentrations of AGP result mainly from the effects of proinflammatory cytokines and other humoral factors upon their biosynthesis by the liver enzymes (Kushner and Mackiewicz, 1993). Changes in the proportions of different AGP glycophorms originate from modifications in the glycosylation mechanisms in the liver, mediated by a number of cytokines (IL-6, IL-1,TNF, interferon) and glucocorticoids (Van Dijk et al., 1994). Crossed-affinity Immunoelectrophoresis (CAIE) with free concanavalin A (Con A) as a ligand is the method to determinate different microheterogeneity glycophorms (Mackiewicz & Mackiewicz, 1986). CAIE reveals four microheterogeneity glycophorms (variants) of AGP: variant A—nonreactive with Con A, variant B—weakly reactive with Con A, variant C—reactive with Con A, and variant D—strongly reactive with Con A. A reactivity coefficient (RC) was calculated according to the formula: sum of

A. Sluzewska

c

Figure 1. CAIE with Con A as a ligand of AGP in sera of: (a) a healthy individual (b) an active rheumatoid arthritis patient (type 11 glycosylation changes) (c) a patient with systemic lupus erthematosus with intercurrent infection (type I of glycosylation changes).

Con A reactive variants/Con A non-reactive variant (B+C+D/A) Three AGP variants were observed in the sera of healthy individuals. In our previous study (Sluzewska et al., 1996a), we have found various alterations of major microheterogeneity of AGP in 75% of patients during an acute depressive episode. Type I changes of glycosylation, consisting of high AGP levels and high values of AGP-RC, similar to those occurring in some acute inflammatory states were found in one third of depressed patients. Patients belonging to this group suffered a longer duration of the illness and of the last episode and were refractory to pharmacological treatment. Type II changes of glycosylation, including high levels of AGP and low values of AGP-RC, similar to those occurring in chronic inflammatory states, were found in patients with a shorter duration of the last episode.

IMMUNE ACTIVATION IN TREATMENT RESISTANT DEPRESSION (TRD) Nearly 30% of depressed patients fail to display an adequate response to psychopharmacological treatment and at least 60% fail to achieve complete remission (Fawcett, 1994; Hornig-Rohan and Amsterdam, 1994). Despite advances in psychopharmacological treatment the mortality and morbidity of this disorder render depression a major public health hazard in most countries. The biological basis of treatment resistance in depression still remains unclear. Several lines of research have suggested that a variety of biological markers may be abnormal in treatment resistant depression (TRD). The study of these indices may enhance our understanding of why some patients do not respond to conventional treatment, and may help identify which therapies might be successful for particular patients or groups of patients. There are findings that suggest the involvement of HPA axis hyperactivity in the pathogenesis of TRD. Christiansen et al. (1989) found poor response to antidepressant treatment in major depressed patients with high spontaneous Cortisol levels. Others (Mc Leod, 1972; Amsterdam et al., 1983, 1994) observed a poor response to tricyclic antidepressants in patients displaying pathological responses on the dexamethasone suppression test (DST). Moreover, the findings of Murphy et al. (1991) and of

Indicators of Immune Activation in Depressed Patients

63

Amsterdam et al. (1994) suggested that steroid suppressive treatment may be of some efficacy in TRD patients with persistent abnormahty of HPA axis. There are a few existing studies which suggest that TRD is accompanied by a very pronounced apr (e.g. Sluzewska et al., 1995b, 1996a, 1997a; Van Hunsel et al., 1996).

ACUTE PHASE PROTEINS IN MAJOR DEPRESSION AND TRD We have studied three positive acute phase proteins (apps) in 201 patients with major depression. The patients were recruited from Department of Psychiatry, Medical Academy in Bydgoszcz (60 patients). Department of Adult Psychiatry University of Medical Sciences in Poznan (81 patients), the Depression Research Unit, University of Pennsylvania in Philadelphia (60 patients). All the patients met DSM IV criteria for major depression. Among them there were 175 Major Depressive Disorder, Recurrent (MDDR) and 26 Bipolar Disorder (BP) (Sluzewska et al., 1996e). In all major depressed patients elevated levels of apps were found and in 75% of the patients there were changes in microheterogeneity. In three groups of patients studied, one-third had type I-glycosylation of AGP and ACT which suggests more reactive variants with Con A and higher values of AGP-RC and ACT-RC. This profile of glycosylation was observed also in patients with burns, rheumatoid arthritis, and lupus erythemetosus with intercurrent infection (Mackiewicz et al., 1987; Pawlowski et al., 1989; Breborowicz and Mackiewicz, 1989). Patients with this profile of glycosylation pattern were all TRD of stage 2 or stage 3, according to the staging of depression based on prior treatment nonresponse (Thase and Rush, 1995). Their duration of illness was the longest (12.6 _ 5 years) and the duration of index episode was the longest as well (11.6 _ 5.3 months). On the other hand, responders to antidepressant treatment had significantly lower concentrations of CRP and had type II of glycosylation which means less reactive variants with Con A and lower values of AGP-RC and ACT-RC. A similar type of AGP glycosylation was observed in chronic inflammatory states, liver cirrhosis and RA (Turner, 1992).

CYTOKINES IN MAJOR DEPRESSION AND TRD Among the immune-inflammatory markers studied in depression there are the pro-inflamatory cytokine IL-6, soluble IL-6 receptor (sIL-6R), and soluble interleukin2 receptor (sIL-2R) as well as soluble transferrin receptor (sTfR). Maes et al. (1995a) and Sluzewska et al. (1995a, 1996b) found significantly elevated concentrations of all the above-mentioned cytokines and soluble receptors in acute depressive episodes. Furthermore, Maes et al. (1995a) observed this too in clinical remission patients in comparison with normal controls. The soluble form of the IL-6 receptor (sIL-6R) is the only soluble interleukin receptor which, after combining with IL-6, intensifies its bioactivity (Bock et al., 1992). As an index of the potentiating effects of plasma sIL-6R on IL-6, the product term: sIL-6R x IL-6 is computed in clinical studies. In our previous study (Sluzewska et al., 1996b), we demonstrated that the value of the product term was three times higher in depressed patients than in healthy controls. It was recently reported by Sluzewska et al. (1995b) that TRD patients have specially elevated concentrations of IL-6, sIL-6R, and the product term in comparison

64

A. Sluzewska

with responders-to-pharmacological treatment. These results correspond with those obtained from 33 TRD hospitahzed patients in our Department during the last 20 months (Sluzewska et al., unpublished data) in whom we have found elevated levels of IL-6, sIL-6R, product term, sIL-2R, and sTfR. There was a significant correlation between AGP and IL-6 in TRD patients (Sluzewska et al., 1995b) as well as between AGP, ACT, and IL-6 in 33 TRD patients, hospitalized in Department of Adult Psychiatry in Poznan during the year of 1995 (Sluzewska et al., 1997b). Depression is also accompanied by significantly increased plasma levels of the IL1 receptor antagonist (IL-IRA) (Maes et al., 1997; Sluzewska et al., 1997e, submitted). IL-IRA is a pure IL-1 receptor antagonist (Dripps et al., 1991) that inhibits the biological activities of IL-1 at target cells (Dayer and Burger, 1994). It is mainly derived from monocyte/macrophage lineage (Dinarello, 1994) and released in vivo during an AP response (Dayer and Burger, 1994). IL-1 and IL-6 are pleotropic cytokines that mediate the primary host response during the apr following infection or injury (Heinrich et al., 1990). These cytokines are predominantly released by cells of the macrophage/monocytic arm of CMI, and in lower amounts also by T and B lymphocytes (Wolvekamp and Marquet, 1990). Changes in the concentration of cytokines (especially IL-6, its soluble receptor and IL-IRA) as well as glucocorticoids during depression may be involved in the regulation of AGP and ACT glycosylation. It has been shown that IL-6 exerts a direct effect on type I and type II glycosylation of apps (Van Dijk et al. 1994). Glucocorticoids modulate the effects of cytokines on glycosylation of apps. Van Dijk and Mackiewicz (1993) found in vivo and in vitro studies indicating that the magnitude and type of glycosylation alterations are dependent on the composition and the amount of interacting cytokines; these effects can be further modulated by treatment with the synthetic glucocorticoid dexamethasone. Morover, Pos et al. (1988) showed that dexamethasone treatment in vivo resulted in the increase of serum concentrations of type I glycosylation forms of AGP and Hp. One of the indicators of activation of T lymphocytic arm of the CMI is the presence of soluble CD8 in plasma. In 33 TRD patients studied, increased plasma concentration of sCD8 was found (Sluzewska et al., unpublished). The sCD8 molecule or suppressor/cytotoxic T cell antigen is secreted by activated T lymphocytes such as CD8+T cells (Tomkinson et al., 1989). A positive relationship between increased levels of sIL-2R and sCD8 is frequently observed in immune disorders (Linker-Israeli et al., 1994). The sIL-2R is released by activated T cells into the circulation and plasma concentrations of sIl-2R correlate well with in vivo IL-2 production (Caruso et al., 1993). Increased concentrations of sIL-2R in depressed patients were reported by Maes et al. (1995a) and Sluzewska et al. (1996b). The TfR is an activation marker (CD 71) for lymphocytes, whose expression is tightly coupled to the IL-2R (Woith et al., 1993). Increased plasma concentration of sTfR in major depressed patients were found by Maes et al. (1995a) and Sluzewska et al. (1996b).

CYTOKINES AND THE HPA AXIS IN MAJOR DEPRESSION AND TRD Several findings suggest the existence of an HPA axis hyperactivity in TRD (Christiansen et al., 1989; Mc Leod, 1972; Amsterdam et al.. 1983.19941 Conseauentlv.

Indicators of Immune Activation in Depressed Patients

65

in 60 major depressed inpatients from the Department of Adult Psychiatry in Poznan, all Caucasians with and without melancholia, the DST test was performed (Sluzewska et al., 1995b). Among those patients with an abnormal DST response, the concentrations of AGP were significantly higher than in the DST suppressors. Pathological results were observed in 33% of the patients and among them 38% suffered from TRD. In our previous study, a significant correlation between morning plasma Cortisol levels and levels of AGP was found in the depressed patients, which could suggest an effect of glucocorticoides on AGP synthesis (Sluzewska and Rybakowski, 1993). Recent data obtained from 33 TRD patients (Sluzewska et al., 1997c) reveal a significant positive correlation between the morning plasma Cortisol levels and the plasma concentrations of IL-6, sIL-2R, AGP, ACT. Further, Maes et al. (1995c) found positive correlations between IL-6, as well as sIL-2R, and Cortisol in major depressed patients, and in combined group of major depressed and healthy controls. From the studies of Navarra et al. (1991) and Tominga et al. (1991), there is evidence that IL-6, at concentrations known to occur in human plasma (Navarra et al., 1991; Naitoh et al., 1988), exerts a potent effect on the HPA axis by stimulating the secretion of CRH and ACTH and steroidogenesis. However, as pointed out by Fukata et al. (1993), the detection of any particular cytokine in the blood does not necessarily means that this entity assumes a role as an immune-derived mediator that causes hyperglucocorticoidemia, even when the entity is known to be effective. The above-presented data may suggest that changes in the concentration and microheterogeneity of AGP and ACT in depressed patients and especially in TRD patients could be related to both HPA axis pathology and increased production of cytokines (especially pro-inflammatory cytokines such as IL-6 and its sIL-6R, as well as IL-1). Futhermore, these abnormalities may be connected with the observed "refractoriness" of the depressive state to the antidepressant treatment.

TOTAL SERUM PROTEIN AND ELECTROPHORETICALLY SEPARATED PROTEIN FRACTIONS IN MAJOR DEPRESSION AND TRD Evidence for an apr that accompanies major depression was first reported by Maes et al. (1992a, b).The apr is reflected, too, in lowered levels of total serum protein (TSP) and changes in the major electrophoretically separated serum protein fractions. Several investigators (e.g. Schwartz et al., 1990; Joyce et al., 1992; Song et al., 1994a; Maes et al., 1991, 1995b) reported lowered levels of serum albumin (Alb) in major depressed patients suffering major depression compared to normal controls. Interestingly, in the olfactory bulbectomized (OB) rat model of depression. Song and Leonard (1994b) found lower levels of Alb. The different apps migrate electrophoretically between Alb and gamma-globulin fractions. Apps such as AGP, ACT, CRP migrate in the alpha-1 zone, Hp and Cp migrate in the alpha-2 zone, fibrinogen and C3 migrate in the (3 zone, and immunoglobulines in the gamma-zone (Ritzmann and Daniels, 1976). This may explain why during apr there are alterations in TSP and in electrophoretically separated serum protein fractions (Putnam, 1984a, b). There are only a few studies examining serum protein fractions in major depression. Song and Leonard (1993) reported increased levels of alpha-1 and alpha-2 globulin fractions in depression and Maes et al. (1995b) found increased levels of alpha-2 globulin and gamma-globulin. Recently, Van Hunsel et al. (1996) reported in 37 major HpnrpNsprl natients of whom 29 nresented TRD. lowered levels of TSP. alb. and B—as

66

A. Sluzewska

well as gamma-globulin. The percentages of alpha-1 and alpha-2 globulin fractions were significantly higher in major depressed and TRD patients than in normal controls, and alpha-2 globulins were lower in TRD patients than in responders. We have recently reported (Sluzewska et al., 1997b) decreased levels of TSP and Alb as well as elevated levels of alpha-1 and alpha-2, and a higher percentage of alpha1, alpha-2, and gamma-globulin in 49 major depressed patients in comparison with normal controls. In the same sudy, 25 TRD patients compared with responders were found to have higher percentages of alpha-1 and gamma-globulin. In the TRD patients, there were positive correlations between AGP, ACT, and alpha-1 globulin. The basal morning Cortisol values were significantly and positively correlated with AGP, ACT, alpha-1, alpha-2, p, and gamma globulin, and negatively correlated with albumin (Sluzewska et al., 1997b). In these patients, a positive correlation was also found between serum levels of IL-6 and alpha-1 globulin and a negative correlation was demonstrated between IL-6 and albumin. These findings of increased serum alpha-1, alpha-2, and gamma-globulin percentages, taken together with the indications of lowered TSP and Alb, may point toward an elevated production of apps migrating in the alpha-1 and alpha-2 and immunoglobuhns (majority of IgG, IgM, IgA) of the gamma electrophoretical zone, associated with elevated protein catabolism. Taken together, these results support the hypothesis that major depression and, in particular, treatment-resistent-depression may be accompanied by alterations of the acute phase response.

ANTIBODIES AGAINST SEROTONIN AND GANGLIOSIDES IN MAJOR DEPRESSION AND TRD Several authors have show the occurrence of antinuclear, antiphospholipid and antiserotonin, antihistone, antibodies in depression (Villemain, 1988; Irwin et al., 1990; Maes et al., 1991; Schott et al., 1992). Some autoimmune disorders have been related to the induction of antihormone and anti-receptor antibodies (Cohen and Cook, 1986). We have recently reported the occurrence of antibodies against serotonin and gangliosides in major depressed patients (Sluzewska et al., 1997d). Antibodies against gangliosides have been shown to be part of the serotonin receptor complex (Fishman, 1988). Patients with major depression showed a significant antibody reactivity of IgG or IgM type to serotonin (55% of the patients) and to gangliosides (50% of the patients). In most patients the antibody against serotonin was associated with the antibody against gangliosides. 73% of the major depressed patients with antibodies against serotonin and 70% major depressed patients with gangliosides antibodies were TRD patients. Among 66% of the TRD patients studied, the antibodies against serotonin were present and in 58% of the cases the antibodies against gangliosides were present too. Patients with antibodies against serotonin and gangliosides showed more elevated levels of CRP, AGP, and ACT. They also presented different patterns of microheterogeneity with higher values of AGP-RC (type I of glycosylation changes), elevated serum levels of IL-6, and decreased levels of serotonin in comparison to normal controls (Sluzewska et al., 1997d). IL-6 may be involved in the production of autoantibodies which, in turn, could be an element of pathogenesis of autoimmune diseases (Cavallo et al., 1994). Antinuclear and antiphospholipid antibodies have already been found in depression (Maes

Indicators of Immune Activation in Depressed Patients

67

et al., 1991). In addition, decrements of serum serotonin levels in major depression are express specifically in TRD; the frequent association of anti-serotonin and antigangliside antibodies, as well as the fact that gangliosides indicate the component of serotonin receptor complex (Fishman, 1988) allow us to speculate that these indicators may be anti-receptor antibodies. Since the concomitant presence of the apr was observed, it may be hypothesized that changes in the serotonin receptor system are related to the apr and that there is an immune (autoimmune) process in depression. Still, there is no evidence, referring to causation, of such a process. One may only speculate that the induction of antiserotonin and antiganglioside antibodies is the consequence of the viral infection or of the stress factors involved in the overproduction of cytokines (eg. IL-6); these factors appear to act within a regulary network, and biological outcome in infection are likely to be representative of the unique, as well as common, action of the "cascade" mediators (Balkweill and Burke, 1989; Balkweill, 1993). Although these immunological data are difficult to interpret, they appear to support the notion of an involvement of (auto) immunity in the pathogenesis of depression.

CONCLUSION In the light of the results presented here regarding the changes in different immunological indices in major depression and TRD one may raise the question: what then is the cause of the acute phase response that is specially evident in TRD? Apr in depression may be induced by a precipitating, psychological stress that may precede or accompany the disorder. In this respect, it is reported that the chronic mild stress model of depression in rodents is characterized by signs of an apr, such as elevated levels of AGP and Hp (Sluzewska et al., 1994a, 1996d). It should be emphasized that various forms of psychological stress may induce the secretion of IL-6 in rodents (LeMay et al., 1990; Zhou et al., 1993). Consequently, it could be argued that the psychological (emotional) stress preceding or accompanying the illness and each acute depressive episode, may be related to increased production or activity of IL-6 and, consequently to the apr in major depression in general and in TRD, in particular. From studies performed on transgenic mice with the expression of IL-6 in the CNS there is evidence that IL-6 can become a potent pathogenetic agent in the CNS producing alterations observed in human neurodegenerative diseases and in depression (Campbell and Chiang, 1995). There is also the possibility that some viral infection is the cause of the acute phase response in major depression and TRD. Pathogenetic role of viral infections (mostly with herpes-type viruses) has long been suggested in the onset or recurrence of affective illness episodes. Several groups of investigators reported that antibody titers to herpes simplex virus (HSV) are increased in unipolar and bipolar depressive psychoses (Lycke et al., 1974; Cappel et al., 1978). The location of the latent herpes viruses in the limbic system suggest their involvement in the pathogenesis of affective recurrences (Stroop, 1986). Zorzenon et al. (1996) found an evidence of active viral multiplication and elevated antibody titers with herpes viruses in 41 % of patients with major depression. In our recent study we have found significantly elevated antibody titers against HSV-1 and HSV-2 in 48 major depressed patients during an acute episode. Patients with higher antibody titers had higher concentrations of interleukin—6 and its soluble receptor as well as higher values of AGP-RC, which corresponds to the

68

A. Sluzewska

inflammatory type of glycosylation (Sluzewska, Rybakowski & Suwalska, 1998, submitted). Reactivation of latent herpes virus has often been attributed to biological or psychological stress (Glaser et al., 1985). IL-6 was recently found to induce transcription factors within the HSV-1 genome which suggests the involvement of IL-6 in reactivation of herpes simplex virus (Kriesel et al., 1997). Human infections with Borna disease virus, with possibly world-wide prevalence have become evident by detection of Borna disease virus (BDV) serum antibodies in patients with mental disorders, including depression (Amsterdam et al., 1985; Rott et al., 1985; Bode et al., 1988). In addition, very recently, the first isolation of infectious human Borna disease virus from two bipolar patients with acute depressive episodes was published by Bode et al. (1996). Persistent BDV infection in animals, can induce phasic behavioral and cognitive changes resembling human bipolar disorder, probably by altering neuronal functions in the limbic system through binding to neurotransmitter receptor (Dittrich et al., 1995; Gosztonyi and Ludwig, 1995). Although these immunological data are still difficult to interpret, they appear to offer a strong support for the notion of (auto) immune involvement in the pathogenesis of depression. It is possible that antidepressant compounds, lithium and antiglucocorticoid agents, such as ketokonazole, have immunoregulatory properties that contribute to their antidepressive action. A clearer understanding of these antidepressive mechanisms together with the interplay between immune, nervous and endocrine systems may lead to dramatic revision of the present conceptualizations of pathophysiology and treatment of major depression and TRD.

ACKNOWLEDGMENTS This research was supported by grants KBN 502-3-326, 501-1-020, 501-2-15-01 from Karol Marcinkowski University of Medical Sciences in Poznan, Poland.

REFERENCES Abraham, K. I., leni, J. R., & Meyerson, L. R. (1987). Purification and properties of human plasma endogenous modulator for platelet tricyclic binding/serotonine transport complex. Bioch Bioph Acta, 923, 8-21. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders. Fourth Edition, American Psychiatric Association. Washington, D.C. Amsterdam, J. D., Winokur, A., Bryant, S., Larkin, J., & Rickels, K. (1983). The dexamethasone suppression test as a predictor of antidepressant response. American Journal of Psychiatry, 80, 43-45. Amsterdam, J. D., Winokur, A., Dyson, W, Herzog, S., Gonzalez, R, Rott, R., & Koprowski, H. (1985). Borna disease virus: a possible etiologic factor in human affective disorders? Archives of General Psychiatry, 42,1093-1096. Amsterdam, J. D., Rosenzweig, M., & Mozley, P. D. (1994). Assessment of adrenocortical activity in refractory depression: steroid suppression with ketoconazole. In W. A. Nolen, J. Zohar, S. P. Roose, & J. D. Amsterdam (Eds.), Refractory Depression: Current Strategies and Future Directions (pp. 199-210). John Wiley & Sons Ltd. Balkwill, F. R. & Burke, F. (1989). The cytokine network. Immunology Today, 10, 299-303. Balkwill, F. R. (1993). Cytokines in health and disease. Immunology Today, 14,149-150. Baumann, H., Richards, C , & Gauldie, J. (1987). Interaction among hepatocyte-stimulating factors, interleukin 1, and glucocorticoids for regulation of acute phase proteins in human hepatoma (Hep G2) cells./oMrna/ of Immunology, 15, 4122-4128.

Indicators of Immune Activation in Depressed Patients

69

Blalock, J. E. (1994). The syntax of immune-neuroendocrine communication. Immunology Today, 15, 11, 504-511. Bock, G. R., Marsh, J., & Widdows, K. (1992). Polyfunctional Cytokines: IL-6 and LIF. Ciba Foundation Symposium 167 Chichester: John Wiley and Sons. Bode, L., Riegel, S., Ludwig, H., Amserdam, J. D., Lange, W., & Koprowski, H. (1988). Borna disease virusspecific antibodies in patients with HIV infection and with mental disorders. Lancet, ii, 689. Bode, L., Durrwald, R., Rantam, F. A., Ferszt, R., & Ludwig, H. (1996). First isolation of infectious human Borna disease virus from patients with mood disorders. Molecular Psychiatry, 1, 200-212. Breborowicz, J. & Mackiewicz, A. (1989). Affinity electrophoresis for diagnosis of cancer and inflammatory conditions. Electrophoresis, 10, 568-573. Campbell, I. L., Chiang, Ch-Sh. (1995). Cytokine involvement in central nervous system disease: Implications from transgenic mices.: In G. P. Chrousou, R. Mc Carty, K. Pacak, G. Cizza, E. Sternberg, W. Gold Ph, & R. Kwetnewsky (Eds.), Stress: Basic Mechanisms and Clinical Implications. Annals of the New York Academy of Sciences, 777, 301-312. Cappel, R., Gregoire, E.,Thiry, L., & Sprecher, S. (1978). Antibody and cell-mediated immunity to herpes simplex virus in psychotic depression. Journal of Clinical Psychiatry, 39, 266-268. Carroll, B. J., Curtis, G. C , Davies, B. M., Mendels, J., & Sugerman, A. A. (1976). Urinary free Cortisol excretion in depression. Psychological Medicine, 6, 43-50. Caruso, C , Candore, G., Cigna, D., Colucci, A. T., & Modica, M. A. (1993). Biological significance of soluble IL-2 receptor. Med. Inflamm.,2, 3-21. Cavallo, M. G., Pozzilli, P., & Thorpe, R. (1994). Cytokines and autoimmunity. Clinical and Experimental Immunology, 96, 1-7. Christiansen, P., Lolk, A., & Gram, L. F. (1989). Cortisol and treatment of depression: Predictive value of spontaneous and suppressed Cortisol levels and course of spontaneous plasma Cortisol. Psychopharmacology, 97, 471^75. Chrousos, G. P. & Gold, P. W. (1992). The concepts of stress and stress system disorders: overview of physical and behavioural homeostasis. yoMrna/ of the American Medical Association, 26H, 200. Chrousos, G. P. (1995). The hypothalamic-pituitary-adrenal axis and immune-mediated inflammation. The New England Journal of Medicine, 332, 1351-1362. Cohen, I. R. & Cook, W. (1986). Natural antibodies might prevent autoimmune disease. Immunology Today, 7, 363-364 Costello, M., Field, B. A., & Gewurz, H. (1979). Inhibition of platelet aggregation by native and disialised alpha-1-acid glycoprotein. Nature, 281, 677-681. Dayer, J-M. & Burger, D. (1994). Interleukin-1, tumor necrosis factor, and their specific inhibitors. European Cytokine Network, 5, 563-571. Dinarello, C. A. (1994). The biological properties of interleukin-1. European Cytokine Network, 5, 517-531. Dittrich, W., Bode, L., Ludwig, H., Kao, M., & Schneider, K. (1995). Learning deficiencies in Borna disease virus-infected but clinicaly healthy rats. Biological Psychiatry, 26, 818-828. Dripps, D. J., Brandhuber, B. J., Thompson, R. C , & Eisenberg, S. P. (1991). Interleukin-1 (IL-1) receptor antagonist binds to the 80-kDa IL-1 receptor but does not initiate IL-1 signal transduction. Journal of Biological Chemistry, 266, 10331-10338. Dunbar, P. R., Hill, J., Neale, T J., & Mellsop, G. W. (1992). Neopterin measurement provides evidence of altered cell-mediated immunity in patients with depression but not with schizophrenia. Psychological Medicine, 22, \05l-\057. Durand, G. (1989). Glycan variations of human alpha-1-acid glycoprotein modulate the biology of macrophages. Progress in Clinical and Biological Research, 300, 247-252. Fawcett, J. (1994). Progress in treatment-resistant and treatment refractory depression: we still have a long way to go. Psychiatric Annual, 24, 214-216. Fishman, P. H. (1988). Gangliosides as cell surface receptors and transducers of biological signals. In R. W. Ledeen, E. L. Hogan, G.Tettamanti, A. J. Yates, & R. R. Yu (Eds.), New Trends in Ganglioside Research: Neurochemical and Neuroregenerative Aspects. Liviana Press, Padova, pp 183-217. Fukata,J., Usui,T.,Naitoh,Y.,& Imura, H. (1989). Effects of recombinant human interleukin-1 alpha,-1 beta, 2, and 6 on ACTH synthesis and release in the mouse pituitary tumor cell line AtT-20. Journal of Endocrinology, 122, 33-39. Gastpar, M. & Muller, W. (1981). Autoantibodies in affective disorders. Prog. Neuropsychopharmacology, 5, 91-96. Glaser, R., Kielcolt-Glaser, J. K., Speicher, C. E., & Holiday, J. E. (1985). Stress, loneliness and changes in herpes virus latency. Journal of Behavioral Medicine. 8, 249-253. Gold, R W., Chrousos, G., Keller, C , Post, R., Roy, A., Augerinlos, R, Schulte, H., Oldfield, E., & Loriaux,

70

A. Sluzewska

D. L. (1994). Psychiatric implications of basic and clinical studies with corticotropin-relasing factor. American Journal of Psychiatry, 141, 619-627. Gold, P. W. (1996). Editorial: Stress system abnormahties in melancholic and atypical depression: molecular, pathophysiological, and therapeutic implications. Molecular Psychiatry, 4, 265-277. Gosztonyi, G., Ludwig, H. (1995). Borna disease—neuropathology and pathogenesis. In H. Koprowski, W. I. Lipkin (Eds.), Borna disease. Current Topics in Microbiology and Immunology, 190, 39-73. Healy, D., Calvin, J., & Whitehouse, A. M. (1991). Alpha-1-acid glycoprotein in major depressives and eating disorders. Journal of Affective Disorders, 22,13-20. Heinrich, P. C , Castell, J. V., & Andus, T. (1990). Review article: interleukin-6 and the acute phase response. Biochemical Journal, 265, 621-636. Holsboer, K, Girken, A., Stalia, G. K., & Muller, O. A. (1984). Blunted corticotropin and normal Cortisol response to human corticotropin-releasing factor in depression. New England Journal of Medicine, 311,1127. Hornig-Rohan, M. & Amsterdam, J. D. (1994). Clinical and biological correlantes of treatment-resistant depression: an overview. Psychiatric Annuals, 24, 220-227. Irwin, M., Patterson, T., Smith, T. L., Caldwell, C , Brown, S. A., Gilhn, J. C , & Grant, I. (1990). Reduction of immune function in life stress and depression. Biological Psychiatry, 27, 22-30. Joyce, P. R., Hawes, C. R., & Mulder, R. T. (1992). Elevated levels of acute phase plasma proteins in major depression. Biological Psychiatry, 32, 1035-1041. Kent, S., Bluthe, R. M., Kelly, K. W., & Dantzer, R. (1992). Sickness behavior as a new target for drug development. Trends in Pharmacological Sciences, 13, 24-28. Kehoe, W., Kwentus, J., Sheffel, W., & Harralson, A. (1991). Increased alpha-1-acid glycoprotein in depression lowers free fraction of imipramine. Biological Psychiatry, 29, 489^93. Kremer, J. M. H., Wilting, J., & Janssen, L. H. M. (1988). Drug binding to human alpha-1-acid glycoprotein in health and disease. Pharmacological Review, 40, 1-47. Kriesel, J. D., Ricigliano, J., Spruance, S. L., Garza, H. H., & Hill, J. M. (1997). Neuronal reactivation of herpes simplex virus may involve interleukin-6. Journal of NeuroVirology, 3, 441-448. Kronfol, Z., Nair, M., Goodson, J., Goael, K., Haskett, R., & Schwartz, S. (1989). Lymphocyte mitogenesis, immunoglobulin and component levels in depressed patients and normal controls. Acta Psychiatrica Scandinavica,80, 142-147. Kushner, I. & Mackiewicz, A. (1993). The acute phase response: An overview. In A. Mackiewicz, I. Kushner, & H. Baumann (Eds.), Acute phase proteins. Molecular biology, biochemistry, and clinical applications. (pp. 3-19). Boca Raton, FL: CRP Press. LaMay, L. G , Vander, A. J., & Kluger, M. J. (1990). The effect of psychological stress on plasma interleukin6 activity in rats. Physiology and Behavior, 47, 957-961. Lieb, J. & Karmali, R. (1993). Elevated levels of prostaglandin E2 and tromboxane B2 in depression. Prost. Leukotr Med., 10, 361-368. Linker-Israeh, M., Hyun, S., Ozeri-Chen, T., Wallace, D. J., Banks, K., & Klinenberg, J. R. (1994). Elevated in vivo and in vitro secretion of CD8-alpha molecules in patients with systemic lupus erythematosus. Journal of Immunology, 152, 3158-3167. Lycke, E., Norrby, R., & Roos, B-E. (1974). A serological study on mentally ill patients with particular reference to prevalence of herpes virus infections. British. Journal of Psychiatry, 124, 273. Mackiewicz, A. & Mackiewicz, S. (1986). Determination of lectin-sugar dissociation constants by agarose affinity electrophoresis. Annals of Biochemistry, 56, 480-484. Mackiewicz, A., Pawlowski,T., Mackiewicz-Pawlowska, A., Wiktorowicz, K., & Mackiewicz, S. (1987). Microheterogeneity forms of alpha-1-acid glycoprotein as indicator of rheumatoid arthritis activity. Clin Chim Acta, 163,185-190. Maes, M., Bosmans, E., Suy, E., Minner, B., & Raus, J. (1989). The impaired mitogen lymphocyte stimulation in severely depressed patients: a complex interface between HPA-axis hyperfunction, noradrenergic activity, and ageing process British Journal of Psychiatry, 155, 793-798. Maes, M., Bosmans, E., Suy, E., Vandervorst, C , Dejonckheere, C , & Raus, J. (1991). Antiphospholipid, antinuclear, Ebstein-Barr and cytomegalovirus antibodies, and soluble interleukin-2 receptor in depressive patients. Journal of Affective Disorders, 21, 133-140. Maes, M., Scharpe, S., Van Grootel, L., Uyttenbroeck, W., Cooreman, W, Cosyns, P., & Suy, E. (1992a). Higher alpha-1-antitrypsin, haptoglobin, ceruloplasmin, and lower retinol protein plasma levels during depression. Futher evidence for the existence of an inflammatory process during that illness. Journal of Affective Disorders, 24, 183-192. Maes, M., Scharpe, S., Bosmans, E., Vandewoude, M., Suy, E., Uyttenbroeck, W, Cooreman, W, Vandervorst, C , & Raus, J. (1992b). Disturbances in acute phase proteins during melancholia: additional evidence

Indicators of Immune Activation in Depressed Patients

71

for presence of inflammatory process during that illness. Progress in Neuropsychopharmacology and Biological Psychiatry, 16, 501-515. Maes, M., Van Der Planken,M., Stevens, J. W.,Peeters,D.,De Clerck, L. S., Bridts, C. H.,Schotte,C.,& Cosyns, P. (1992c). Leukocytosis, monocytosis and neutrophilia: hallmarks of severe depression. Journal of Psychiatric Research, 26, 125-134. Maes, M. (1993). Acute phase protein alterations in major depression: A review. Reviews of Neuroscience, 4, 407^16. Maes, M., Meltzer, H. Y., Stevens, W., Cosyns, P., & Block, P. (1994a). Multiple reciprocal relationships between in vivo cellular immunity and hypothalamic-pituitary-adrenal axis in depression. Psychological Medicine, 24, 167-177. Maes, M., Scharpe, S., & Meltzer, H. Y. (1994b). Increased neopterin and interferon secretion and lower availability of L-tryptophan in major depression; further evidence for activation of cell-mediated immunity. Psychiatric Research, 20, 111-116. Maes, M., Meltzer, H. Y, & Bosmans, E. (1995a). Increased plasma levels of interleukin-6, soluble interleukin2, and transferrin receptors in major depression. Journal of Affective Disorders, 34, 301-309. Maes, M., Wautwers, A., Neels, H., Scharpe, S., Van Gastel, A., Dhondt, P., Peeters, D., Cosnys, P., & Desnyder, R. (1995b). Total serum protein and serum protein fractions in depression: relationships to depressive symptoms and glucocorticoid activity. Journal of Affective Disorders, 34, 61-69. Maes, M., Bosmans, E., & Meltzer, H. Y (1995c). Immunoendocrine aspects of major depression. European Archives of Psychiatry and Clinical Neuroscience, 245, 172-178. Maes, M., Delanghe, J., Ranjan, R., Meltzer, H. Y, Desnyder, R., Cooreman, W., & Scharpe, S. (1997). The acute phase protein response in schizophrenia, mania, and major depression: effects of psychotropic drugs. Psychiatry Research, 26, 221-225. McLeod, W. R. (1972). Poor response to antidepressants and dexamethasone nonsuppression. In B. Davis, B. J. Carroll, & R. M. Mowbray (Eds.), Depression illness: Some Research Stdies. (pp. 202-206). Springfield, IL. Thomas. Muller, N., Hofschuster, E., Ackenheil, M., Mempel, W., & Eckstein, R. (1993). Investigations of the cellular immunity during depression and the free interval: evidence for immune activation in affective psychosis. Progress in Neuro-Psychopharmacol and Biological Psychiatry, 17, 713-730. Murphy, B. E. P., Dhar, V., Ghadirian, A. M., Chouinard, G., & Keller, R. (1991). Response to steroid suppression in major depression resistant to antidepressant therapy. Journal of Clinical Psychopharmacology, 2, \2\-\25. Naitoh, Y, Fukato, J.,Tomiga,T., Nakai, Y.,Tamai, S., Mori, K., & Immura, H. (1988). Interleukin-1 stimulates the secretion of adrenocorticotropic hormone in conscious freely-moving rats. Biochemistry, Biophysiology Research Communications, 155, 1459-1466. Navarra, P., Tsagarakis, S., Faria, M. S., Rees, L. H., Besser, G. M., & Grossman, A. B. (1991). Interleukin-1 and interleukin-6 stimulate the release of corticotropin-releasing hormone from rat hypothalamus in vitro via the ecosanoid cyclooxygenase pathway. Endocrinology, 128, 37-44. Nemeroff, C. B., Krisnam, R. R., Blazer, D. G., Knight, D. L., Benjamin, D., & Meyerson, I. R. (1990). Elevated plasma concentration of alpha-1-acid glycoprotein, a putative endogenous inhibitor of tritiated imipramine binding site in depressed patients. Archives of General Psychiatry, 47, 337-340. Nemeroff, C. B. (1996). The corticotropin-releasing factor (CRF) hypothesis of depression: new findings and new directions. Molecular Psychiatry, 4, 336-342. Nessberger, L. & Traskman-Bendz, L. (1993). Increased soluble interleukin-2 receptor concentration in suicide attempters. Acta Psychiatrica Scandinavica, 88, 48-52. Paul, W. E. & Seder, R. A. (1994). Lymphocyte responses and cytokines. Cell, 76, 241-251. Pawlowski, P., Mackiewicz, S., & Mackiewicz, A. (1989). Microheterogeneity of alpha-1-acid glycoprotein in detection of intercurrent infection in rheumatoid arthritis. Arthritis Rheumatology, 32, 347-351. Pos, O., Van Dijk, W., Ladiges,N., Sala, M., VanTiel, D., & Boers, W. (1988). Glycosylation of four acute-phase glycoproteins secreted by rat liver cells in vivo and in vitro. Effects of inflammation and dexamethasone. European Journal of Cell Biology, 46, 121-128. Pos, O., Van der Stelt, M. E., & Wolbik, G. J. (1990). Changes in the serum concentration and glycosylation of human alpha-1-acid glycoprotein and alpha-lprotease inhibitor in severely burned patients: relation to interleukin-6 levels. Clinical and Experimental Immunology, 82, 579-582. Putnam, F. W. (1984a). Progress in plasma proteins. In: The plasma proteins: structure, function, and genetic control, (pp. 2 ^ 4 ) . Second edition. Academic Press, Orlando. Putnam, F. W. (1984b). Alpha, beta, gamma omega- the structure of plasma proteins. In: The plasma proteins: structure, function, and genetic control, (pp. 46-166). Secon edition. Academic Press, Orlando. Ritzmann, S. E. & Daniels, J. C. (1976). Serum protein electrophoresis and total serum proteins. In: Serum

72

A. Sluzewska

protein abnormalities: Diagnostic and clinical aspects, (pp. 3-25) Little, Brown and Company, Boston, MA. Rott, R., Herzog, S., Fleisher, B., Winokur, H., Amsterdam, J. D., Dyson, W., & Koprowski, H. (1985). Detection of serum antibodies to Borna disease virus in patients with psychiatric disorders. Science, 228, 755-756. Schleifer, S. J., Keller, S. E., Meyerson, A. T, Raskin, D., Davis, K. L., & Stein, M. (1984). Lymphocyte function in major depressive disorder. Archives of General Psychiatriatry, 41,484^86. Schott, K., Batra, A., Klein, R., Bartels, M., Koch, W., & Berg, P. A. (1992). Antibodies against serotonin and gangliosides in schizophrenia and major depressive disorder. European Psychiatry, 7, 209-212. Schott, K. & Batra, A. (1995). No antibody binding to serotonin in serum of patients with paranoid schizophrenia and major depressive disorder. European Psychiatry, 10, 372. Sluzewska, A. & Rybakowski, J. K. (1993). Cortisol levels and immunological indices in depression. Neuropsychopharmacology, 9, 107S. Sluzewska, A., Nowakowska, E., Gryska, K., & Mackiewicz, A. (1994a). Haptoglobin levels in chronic mild stress model of depression in rats before and after treatment. European Neuropsychopharmacology, 4, 3, 302. Sluzewska, A., Rybakowski, J. K., Sobieska, M., & Wiktorowicz, K. (1994b). The effect of lithium, carbamazepine and fluoxetine on alpha-1-acid glycoprotein, and alpha-1-antichymotripsin in depressed patients. Neuropsychopharmacology, 10,202. Sluzewska, A., Wiktorowicz, K., Mackiewicz, S., & Rybakowski, J. K. (1994c). The effect of short-term treatment with lithium and carbamazepine on some immunological indices in depressed patients. Lithium, 5, 41-46. Sluzewska, A., Rybakowski, J. K., Laciak, M., Mackiewicz, A., Sobieska, M., & Wiktorowicz, K. (1995a). Interleukin-6 levels in depressed patients before and after treatment with fluoxetine. Annals of the New York Academy of Science, 762, 474-477. Sluzewska, A., Rybakowski, J. K., Sobieska, M., Bosmans, E., Pollet, H., & Wiktorowicz, K. (1995b). Increased levels of alpha-1-acid glycoprotein and interleukin-6 in refractory depression. Depression, 3,170-175. Sluzewska, A., Rybakowski, J. K., Bosmans, E., Maes, M., Berhmans, R., & Pollet, H. (1995c). The effects of treatment with lithium and carbamazepine on some interleukin and their receptors in depressed patients. Pharmacological Research, 3, 366. Sluzewska, A., Horing-Rohan, M., Sobieska, M., Rybakowski, J. K., & Amsterdam, J. D. (1995d). Changes in acute phase proteins in depressed patients during treatment with fluoxetine and venlafaxine. European Neuropsychopharmacology, 5, 3, 295-296. Sluzewska, A., Rybakowski, J. K., Sobieska, M., & Wiktorowicz, K. (1996a). Concentration and microheterogeneity glycophorms of alpha-1-acid glycoprotein in major depression. Journal of Affective Disorders, 39,149-155. Sluzewska, A., Rybakowski, J., Bosmans, E., Sobieska, M., Berghans, R., Maes, M., & Wiktorowicz, K. (1996b). Indicators of immune activation in major depression. Psychiatry Research, 63, 3, 161-167. Sluzewska, A., Rybakowski, J., & Sobieska, M. (1996c). Immune activation in endogenous depression. Psychiatria Polska, 5, 771-782. Sluzewska, A., Gryska, K., & Mackiewicz, A. (1996d). Acute phase proteins in chronic mild model of depression. Behavioural Pharmacology, 7,1,105-106. Sluzewska, A., Rybakowski, J. K., Sobieska, M., & Amsterdam, J. D. (1996e). Changes in concentration and microheterogeneity of two acute phase proteins in major depression. Glycoconjugate Journal, 13, 891-892. Sluzewska, A., Sobieska, M., & Rybakowski, J. K. (1997a). Changes in acute-phase proteins during lithium potentiation of antidepressants in refractory depression. Neuropsychobiology, 35,123-127. Sluzewska, A., Rybakowski, J. K., & Sobieska, M. (1997b). Total serum protein and serum protein fractions in major depression and treatment resistant depression. Biological Psychiatry, 42, IS, 114. Sluzewska, A., Bosmans, E., & Rybakowski, J. (1997c). Relationships between selected immunological parameters and Cortisol in treatment resistant depression. European Neuropsychopharmacology, 7,2,170. Sluzewska, A., Samborski, W, Sobieska, M., Klein, R., Bosmans, E., & Rybakowski, J. K. (1997d). Serum antibodies in relation to immune activation in major depression. Human Psychopharmacology, 12, 453-458. Song, C. & Leonard, B. L. (1993), Changes in plasma proteins in depressed patients and rodent model of depression. Neuropsychopharmacology, 9,106S. Song, C , Dinan, T, & Leonard, B. E. (1994a). Changes in immunoglobulines, component and acute phase proteins levels in depressed patients and normal controls. Journal of Affective Disorders, 30,283-288.

Indicators of Immune Activation in Depressed Patients

73

Song, C. & Leonard B. E. (1994b). An acute phase protein response in the olfactory bulbectomised rat: effect of sertraline treatment. Medical Science Research, 22, 313-314. Stroop, W. G. (1986). Herpes simplex virus encephalitis of the human adult: reactivation of latent brain infection. Pathology Immunopathology Research, 5,156-161. Swartz, C. M. (1990). Albumin decrement in depression and cholesterol decrement in mania. Journal of Affective Disorders, 19, 173-176. Thase, M. E. & Rush, A. J. (1995). Treatment- resistant depression. In F. E. Bloom, D. J. Kupfer (Eds.), Psychopharmacology, the forth Generation of Progress, (pp. 1081-1098). Raven Press, New York. Tominga, T., Fukata, J., & Naitoh, Y. (1991). Prostaglandin-dependent in vitro stimulation of adrenocortical steroidogenesis by interleukins. Endocrinology, 128, 526-531. Tomkinson, B. E., Brown, M. C , Ip, S. H., Carrabis, S., & Sullivan, J. L. (1989). Soluble CD8 during T cell activation. Journal of Immunology, 142, 2230-2236. Turner, G. A. (1992). N-glycosylation of serum proteins in disease and its investigation using lectins. Clin. Chem. Acta., 208,149-171. Van Dijk, W. & Mackiewicz, A. (1993). Control of glycosylation alterations of acute phase proteins. In A. Mackiewicz, 1. Kuchner, & H. Bauman (Eds.), Acute Phase Glycoproteins: Molecular Biology, Biochemistry, and Applications, (pp. 559-580). Boca Raton, PL: CRC Press. Van Dijk, W., Turner, G. A., & Mackiewicz, A. (1994). Changes in glycosylation of acute phase proteins in health and disease: occurence regulation and function. Glycosylation and Disease, 1, 5-14. Van Hunsel, F., Wauters, A., Vandoolaeghe, E., Neels, H., Demedts, P., & Maes, M. (1996). Lower total serum protein, albumin, and globulin in major and treatment resistant depression: effects of antidepressant treatment. Psychiatry Research, 65, 159-169 Vasson, M. P., Roche-Arvellier, M., Coudrec, R., Baguet, J. C , & Raichvarg, D. (1994). Effects of alpha-1-acid glycoprotein on human polynuclear neutrophils: Influence of glycan microheterogeneity. Cli Chim Acta, 224, 65-11. Villemain, F, Magnin, H., Feuillet-Fieux, MR, Zarifian, E., Loo, H., & Bach, J. H. (1988). Antihistone antibodies in schizophrenia and affective disorders. Psychiatry Research, 24, 53-60. Woith, W., Nusslein, I., & Antoni, C. (1993). A soluble form of the human transferrin receptor is released by activated lymphocytes in vitro. Clinical and Experimental Immunology, 92, 537-542. Wolvekamp, M. C. J. & Marquet, R. L. (1990). Interleukin-6: historical background, genetics, and biological significance. Immunology Letters, 24, 1-10. Zhou, D., Kusnecov, A. W., Shurin, M. R., DePaoli, M., & Rabin, B. S. (1993). Exposure to physical and psychological stressors elevates plasma interleukin-6: relationships to the activation of hypothalamicpituitary-adrenal axis. Endocrinology, 133, 2523-2530. Zorzenon, M., Colle, R., Vecchio, D., Sertoli, M., Giavedoni, A., Degrassi, A., Lavaroni, S., & Aguglia, E. (1996). Major depression, viral reactivation and immune system. European Psychiatry, 11,4, 332.

MOOD AND COGNITIVE DISORDERS IN CANCER PATIENTS RECEIVING CYTOKINE THERAPY Christina A. Meyers Department of Neuro-Oncology The University of Texas M. D. Anderson Cancer Center 1515 Holcombe Boulevard, Houston, Texas, 77030

1. INTRODUCTION The use of cytokines, especially interferon-a (IFN-a) and interleukin-2 (IL-2) to treat maHgnancies and chronic viral infections is broadening, and increasing attention is being paid to the effects of these agents on the central nervous system (CNS). It is now known that cytokine therapy can be associated with the development of cognitive and psychiatric adverse effects that range from subtle impairments of attention and memory undetectable on routine medical evaluation to delirium and psychosis. Because IFN-a is a highly effective therapy in certain diseases, it is important that patients continue on treatment at effective doses. Understanding the nature, incidence, severity, and mechanisms of CNS effects is essential to develop interventions to decrease morbidity while maintaining patients on treatment. The CNS effects of cytokines are separable from those due to other systemic, potentially toxic, treatments, and chronic disease itself. For example, over 50% of cancer patients with widely metastatic solid tumors treated with cytokines have cognitive deficits on neuropsychological testing, compared to 18% treated with chemotherapy agents only (Meyers & Abbruzzese, 1992). The CNS effects are apparent rapidly, even at low doses. A single dose of 0.1 million international units (MIU) of IFN-a reduces subjective alertness in normal healthy volunteers, and slowed reaction time is observed 10 hours after a single dose of 1.5 MIU (Smith, Tyrrell, Coyle, & Higgins, 1988). By way of contrast, the usual doses used for chronic myelogenous leukemia (CML) range from 5 to 10MIU daily, and for hepatitis the doses are generally 3 MIU three times per week. Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

75

76

C. A. Meyers

The effect of cytokines on the mood of cancer patients is often difficult to separate from other factors. Cancer patients may be vulnerable to develop depression, especially those individuals with a past history of depression or substance abuse, those with advanced disease or uncontrollable pain, and those who receive multiple medications (Valentine, Meyers, Kling, Richelson, & Hauser, 1998). However, the diagnosis of cytokine-induced depression is likely when the symptoms develop during treatment and do not appear to be attributable to other causes. In general, cancer patients experiencing cytokine-induced depressive symptoms are most likely to exhibit dysphoria, anhedonia, and helplessness in addition to constitutional symptoms of fatigue, apathy, and mental slowing. They rarely exhibit guilt, delusions, or the acute onset of severe dysphoria with suicidal ideation (Valentine et al., 1998). Depression is not the only psychiatric disturbance manifested by patients treated with cytokines. Although rare, some patients have developed frank symptoms of mania, including euphoria, irritability, decreased need for sleep, pressure of speech, and flight of ideas (Strife, Valentine, & Meyers, 1997).

2. PATTERN OF NEUROBEHAVIORAL DEFICITS SECONDARY TO CHRONIC IFN-a TREATMENT One study compared two groups of CML patients, one on IFN-a and one on chemotherapy (Pavol, Meyers, Rexer, Valentine, Mattis, & Talpaz, 1995). One half of the patients on IFN-a performed in the abnormal range (±2 standard deviations below the normative mean) on a test of verbal memory compared to 22% of chemotherapy treated patients. Patients on IFN-a also performed more poorly on tests of executive function (32% abnormal compared to 1% in the non-IFN group), and motor speed and dexterity (23% impaired versus 0% in the non-IFN group). In addition, 50% had elevated scores on the depression subscale of the Minnesota Multiphasic Personality Inventory compared to 25% of the non-IFN patients. Patients in both groups had no difficulty on tests of attention span, reasoning, language, visual-perception, or other intellectual functions. Similar cognitive abnormalities were seen in a patient receiving IL-2 and tumor necrosis factor (TNF). Single photon emission tomography (SPECT) revealed reduced cerebral blood flow in both frontal lobes (Meyers, Valentine, Wong, & Leeds, 1994). The pattern of abnormal test findings in patients on cytokine treatment (memory, motor, executive, and mood) is suggestive of frontal-subcortical dysfunction and is similar to the pattern of deficits seen in patients with Parkinson's disease and multiple sclerosis (Cummings, 1990). The putative mechanisms of IFN-a neurotoxicity fit well with a model of frontal-subcortical brain dysfunction.

3. RISK FACTORS FOR DEVELOPING NEUROTOXIC SIDE-EFFECTS OF IFN-a 3.1. Dose versus Treatment Chronicity The development of neurotoxic side-effects to IFN-a is related to both to dose and time on treatment. Time on treatment may be the most salient factor, as it accounts for 55% of the variance in cognitive and mood test scores in patients with CML who have been on treatment an average of 26 months (Pavol et al., 1995). Another study

Mood and Cognitive Disorders in Cancer Patients Receiving Cytokine Therapy

77

using high dose induction with IFN-a found fatigue to be the dose limiting toxicity, but the cognitive deficits were related to the number of courses the patient received. All patients experienced memory loss after 6 courses of therapy regardless of dose level (Amato, Meyers, EUerhorst, Finn, Kilbourn, Sella, & Logothetis, 1995). The importance of treatment chronicity in patients receiving cytokine therapy suggests that the incidence and severity of mood and cognitive symptoms may be underestimated if patient follow-up is not long enough.

3.2. Route of Administration Subcutaneous injection of cytokines is generally considered to be a benign route. However, the intraventricular route for the treatment of leptomeningeal metastases (LMD) has proved to be particularly toxic. In one study, 78% of patients developed a wakeful vegetative state after receiving intraventricular IFN-a at a wide range of doses (cumulative dose of 15 to 54MIU) (Meyers, Obbens, Scheibel, & Moser, 1991). It took an average of 3 weeks off of therapy for the patients to recover to their baseline performance status. The toxicity of intraventricular IL-2 for leptomeningeal melanoma appears to be more sporadic, with some individuals tolerating treatment without mood or cognitive side effects, and others developing severe reactions including delirium, obtundation, and death (unpublished data). Delayed effects are also problematic. Intraventricular IL-2 has been reported to cause a progressive dementia in a patient who appeared to be otherwise cured of LMD (Meyers & Yung, 1993).

3.3. Synergy with Brain Irradiation All of the patients developing severe neurotoxicity from intraventricular IFN-a had previous whole brain radiation therapy; the 2 patients who did not develop toxicity did not have previous cranial irradiation (Meyers, et al., 1991). This suggests that patients who have received brain irradiation for cerebral metastases may be at greater risk for developing toxicity.

3.4. Reversibility of Toxicity Mood and cognitive changes generally resolve within 2 to 3 weeks following the discontinuation of therapy (Meyers & Valentine, 1995). However, in some patients neurotoxic side-effects persist for months and years off treatment (Meyers, Scheibel, & Forman, 1991). For example, we evaluated a 54 year old male patient who was diagnosed with CML and treated with 12MIU IFN-a daily for 5 years. At the time of neuropsychological assessment he had been off of all treatment for 3 years and in complete cytogenetic remission. The patient had a Ph.D. in mathematics, was a professor at a university, and owned and operated a computer services business. He was diagnosed with CML one year after starting his business, which was rapidly expanding. His initial symptoms related to IFN-a therapy consisted of fatigue, difficulty concentrating, and difficulty retaining complex technical information. Memory loss extended to difficulty recalling names of associates and business contacts, necessitating the use of notes. IFNa treatment was discontinued after 5 years because the neuropsychiatric symptoms had become increasing more pronounced; extreme irritability, depression, paranoia, sleeplessness, forgetfulness, and distractibility. Just before treatment was discontinued he had expressed some suicidal thoughts.

78

C. A. Meyers

The patient reported that the more dramatic symptoms resolved after treatment was discontinued. However, he never fully recovered his ability to retain complex details without using notes and continues to have difficulty remembering the names of clients, appointments, and distinctions between various projects at work. The success of his company has suffered as a result. Formal neuropsychological testing performed 3 years after treatment cessation revealed low average to average performance on tests of information-processing speed, and severe impairments of verbal memory and fine motor skills. These findings are in marked contrast to his estimated pre-illness superior intellectual level. Personality assessment revealed mild anxiety and depression. The potential for irreversible neurologic toxicity from prolonged cytokine treatment may become more problematic as increasing numbers of patients are cured or have very long-term remissions from their disease.

4. PROSPECTIVE ASSESSMENT OF THE CNS EFFECTS OF IFN-a We are currently conducting a prospective study of CML patients before and during IFN-a treatment to better determine the incidence of neurotoxic side effects. So far we have assessed 27 patients. The demographic characteristics of these patients are listed in Table 1. It is expected that individuals will perform better on many of these tests the second time due to practice effects. In fact, the mean test performance declined on most tests during IFN-a therapy (Table 2). There was a marked decline on tests measuring graphomotor speed, verbal memory, executive function, and motor coordination, but no significant worsening on tests of simple attention span or verbal reasoning. On the Minnesota Multiphasic Personality Inventory there was a highly significant increase on scales measuring somatic concern, depression, hysteria, and impaired thinking ability. We have also assessed several patients with hepatitis before and during lower dose IFN-a treatment. At this time we have only 6 patients with pre-treatment and follow-up assessment. The dose for all of these individuals was 3 million units (MIU) three times per week and the follow-up interval was 4.2 months (range 3-5 months). The demographic characteristics of this sample is very similar to the CML patients (mean age = 42.8 years, mean education = 13.7 years). Although this data is very preliminary, we have found that these patients tended to perform better on the cognitive tests at follow-up, the expected pattern due to practice effects. Only one of the 6 patients experienced increased depression (as assessed by the Beck Depression Inventory). Table 3 summarizes the preliminary results of these prospective studies. Our current prospective study includes patients on high-dose therapy as adjuvant treatment for melanoma. These patients are treated with 20MIU of IFN-a/m^ daily for 4 weeks, and then placed on maintenance therapy (lOMIU/m^ three times a week for

Table 1. Demographic characteristics of CML patients in prospective study (N = 27) Age Education (years) Follow-up interval (months) Dose (MIU daily)*

Mean

SD

Range

49.5 15.4 4.7 7.0

12.5 2.4 3.0 2.8

18-74 12-19 2-16 1-11

''At follow-up, many patients had alreadv undergone dose reduction.

Mood and Cognitive Disorders in Cancer Patients Receiving Cytokine Therapy

79

Table 2. Preliminary test results of prospective study of CML patients on interferon

Attention" Graphomotor speed" Verbal reasoning" Verbal memory'' Executive function' Verbal fluency'' Motor coordination' MMPI-Hypochondriasis' MMPI-Depression' MMPI-Hysteria" MMPI-Schizophrenia'

Pre-IFN baseline

Follow-up

11.6 11.0 11.2 98.3 66.2 42.0 70.3 60.8 61.5 63.0 51.8

11.4 10.3* 11.8 92.8** 81.2* 39.0** 82.5** 69.7* 72.4* 70.6* 61.3*

"Age-corrected scaled score. ''Number of words recalled over 12 trials. 'Number of seconds to complete task. ''Number of words produced beginning with a specific letter in 3 minutes. 'T-score (mean = 50, standard deviation = 10), >70 is clinically significant. * Significant difference (p < .05, two-tailed t-test). ** Approaches significance (p < .1, two-tailed t-test).

48 weeks). In addition to formal psychiatric and neuropsychological evaluations before treatment and at 1 month and 3 months after, the patients will undergo functional magnetic resonance imaging (fMRI) to better define the changes in brain activity associated with IFN-a therapy. For those individuals who become clinically depressed, antidepressant therapy will be instituted. The neuropsychological, mood, and fMRI assessments will then be repeated on antidepressant therapy.

5. POTENTIAL MECHANISMS OF THE CNS EFFECTS OF CYTOKINES There are a myriad of mechanisms by which cytokines can effect nervous system function. These include induction of secondary cytokines that affect neuronal function (Licinio, Kling, & Hauser, 1998), release of stress hormones that are known to cause mood and cognitive alterations (Meyers & Valentine, 1995), the development of autoimmune thyroid disease (Jones, Wadler, & Hupart, 1998), alterations of neurotransmitter function that are important in frontal-subcortical neuronal circuitry (Ho, Lu, Huo, Fan, Meyers, Tansey, Payne, & Levin, 1994), and possibly effects on cerebral endothelium (Meyers & Valentine, 1995). Additionally, most recent studies utilize recombinant cytokines. There is a recent report on the use of natural interferon in asymptomatic HIV-positive individuals that reported fewer side effects than is generTable 3. Summary of preliminary prospective studies of IFN-a neurotoxicity • 19/27 (70%) CML patients declined on one or more cognitive tests • 17/20 (85%) CML patients experienced a significant increase in emotional distress • 1/6 (17%) hepatitis patients experienced cognitive decHne • 1/6 (17%) hepatitis patients experienced increased depression • Patients who experienced cognitive decline were not necessarily the same patients who reported increased mood disturbance

80

C. A. Meyers

ally observed (Mapou, Law, Wagner, Malone, & Skillman, 1996). This report utilized different methodologies and had limited patient follow-up. It is known that recombinant proteins expressed in host cells may have differences in their folding compared to the natural protein. Over-expression of recombinant proteins may also lead to altered post-translational modifications (e.g., acylation, glycosylation, phosphorylation) that may be important for biological activity. However, if there are any differences between natural and recombinant cytokines, they must be cryptic since the recombinant protein produces a biological response predictive of the natural protein.

6. T R E A T M E N T O F C Y T O K I N E - I N D U C E D M O O D DISTURBANCE It is not uncommon that patient report of symptoms related to cytokine treatment may under- or over-represent the symptoms that might have been predicted based upon quantitative assessment. Many patients are alarmed at the potential consequences of dose reduction or treatment cessation if they complain of side effects, and thus may not report symptoms unless encouraged to do so. Conversely, some individuals are highly intolerant of even slight reductions in their cognitive efficiency. Prior to initiating therapy, assessment for risk factors that might predispose the individual to develop depression should be performed, including assessment of the patient's past and family history of psychiatry illness and past history of neurologic disease. Potential side effects of cytokine treatment should be discussed with patients prior to initiating therapy. Patients should be regularly assessed for depression, including laboratory evaluations, neurologic functioning, and depression rating scales (Valentine et al., 1998). The recent explosion in the elucidation of the mechanisms of cytokine-induced CNS effects has encouraged innovative trials of pharmacologic interventions, including antidepressants (Goldman, 1994; Levenson & Fallon, 1993), opiate antagonists, (Valentine, Meyers, & Talpaz, 1995), stimulants (Valentine et al., 1998), and corticosteroids (Amato et al., 1995). The synthetic steroid megestrol acetate may be an effective treatment for the anorexia and cachexia caused by cytokine treatment (Plata-Salaman, 1998). In the future it may be possible to specifically inhibit certain induced cytokines responsible for adverse side effects but maintain the therapeutic efficacy of the given cytokine (Taylor & Grossberg, 1998). Finally, non-pharmacologic approaches, including exercise and time management strategies, may be of benefit (Dalakas, Mock, & Hawkins, 1998). From these early experiences it is becoming clear that appropriate interventions will allow patients to continue on effective doses of cytokines for long periods of time.

7. S U M M A R Y Chronic treatment with cytokines is associated with the development of mood and cognitive changes that suggests frontal^subcortical cerebral dysfunction. There is large individual variability in the type and severity of specific symptoms that are reported. The CNS effects of cytokines can be disassociated from the effects of chronic disease, other treatments and medications, and psychological responses to illness. The length of treatment and dose are both important factors in the development of mood

Mood and Cognitive Disorders in Cancer Patients Receiving Cytokine Therapy

81

disturbance. The finding that treatment chronicity is important makes long-term followup of patients on cytokine therapy all the more vital. Most adverse effects of cytokines improve with appropriate treatment of symptoms, although dose reduction or cessation of therapy may be necessary in individual cases. Future studies will be needed to better identify at-risk individuals, to compare the efficacy of various interventions, including antidepressants, stimulants, and opiate antagonists, and to assess the feasibility of treating at-risk individuals prophylactically.

REFERENCES Amato, R., Meyers, C , Ellerhorst, J., Finn, L., Kilbourn, R., Sella, A., & Logothetis, C. (1995). A phase I trial of intermittent high-dose a-interferon and dexamethasone in metastatic renal cell carcinoma. Annals of Oncology, 6, 911-914. Cummings, J. L. (1990). Subcortical Dementia. New York: Oxford University Press. Dalakas, M. C , Mock, V., & Hawkins, M. J. (1998). Fatigue: definitions, mechanisms, and paradigms for study. Seminars in Oncology, 25 Supplement I, 48-53. Goldman, L. S. (1994). Successful treatment of interferon alfa-induced mood disorder with nortriptyline (Letter to the editor). Psychosomatics, 35, 412^13. Ho, B. T., Lu, J. G., Huo, Y. Y., Fan, S. H., Meyers, C. A., Tansey, L. W., Payne, R., & Levin, V. A. The opioid mechanism of interferon-a action. Anti-Cancer Drugs, 5, 90-94. Jones, T H. J., Wadler, S., & Hupart, K. H. (1998). Endocrine-mediate mechanisms of fatigue during treatment with interferon-a. Seminars in Oncology, 25 Supplement 1, 54-63. Levenson, J. L. & Fallon, H. J. (1993). Fluoxetine treatment of depression caused by interferon-alpha. American Journal of Gastroenterology, 88, 760-761. Licinio, X, Kling, M. A., & Hauser, P. (1998). Cytokines and brain function: relevance to interferon-a-induced mood and cognitive changes. Seminars in Oncology, 25 Supplement!, 30-38. Mapou, R. L., Law, W. A., Wagner, K., Malone, J. L., & Skillman, D. R. (1996). Neuropsychological effects of interferon alfa-n3 treatment in asymptomatic human immunodeficiency virus-1-infected individuals. Journal of Neuropsychiatry and Clinical Neurosciences, 8, 74—81. Meyers, C. A. & Abbruzzese, J. L. (1992). Cognitive functioning in cancer patients: effect of previous treatment. Neurology, 42, 434-436. Meyers, C. A. & Valentine, A. D. (1995). Neurologic and psychiatric adverse effects of immunological therapy. CNS Drugs, 3, 56-68. Meyers, C. A., Valentine, A. D., Wong, F. C. L., & Leeds, N. E. (1994). Reversible neurotoxicity of interleukin2 and tumor necrosis factor: correlation of SPECT with neuropsychological testing. Journal of Neuropsychiatry and Clinical Neurosciences, 6, 285-288. Meyers, C. A. & Yung, W K. A. (1993). Delayed neurotoxicity of intraventricular interleukin-2: a case report. Journal of Neuro-Oncology, 15, 265-267. Meyers, C. A., Obbens, E. A. M. T, Scheibel, R. S., & Moser, R. P. (1991). Neurotoxicity of intraventricularly administered alpha interferon for leptomeningeal disease. Cancer, 68, 88-91. Meyers, C. A., Scheibel, R. S., & Forman, A. D. (1991). Persistent neurotoxicity of systemically administered interferon-alpha. Neurology, 41, dll-dld. Pavol, M. A., Meyers, C. A., Rexer, J. L., Valentine, A. D., Mattis, R J., & Talpaz, M. (1995). Pattern of neurobehavioral deficits associated with interferon-alfa therapy for leukemia. Neurology, 45, 947-950. Plata-Salaman, C. R. (1998). Cytokines and anorexia: a brief overview. Seminars in Oncology, 25 Supplement 1, 64-72. Smith, A., Tyrrell, D., Coyle, K., & Higgins, P. (1988). Effects of interferon alpha on performance in man: a preliminary study. Psychopharmacology, 96, 414-416 Strife, D., Valentine, A. D., & Meyers, C. A. (1997). Manic episodes in two patients treated with interferon alpha. Journal of Neuropsychiatry and Clinical Neurosciences, 9, 273-276. Taylor, J. L. & Grossberg, S. E. (1998). The effects of interferon-a on the production and action of other cytokines. Seminars in Oncology, 25 Supplement 1, 23-29. Valentine, A. D., Meyers, C. A., & Talpaz, M. (1995). Treatment of neurotoxic side effects of interferon-a with naltrexone. Cancer Investigation, 13, 561-566. Valentine, A. D., Meyers, C. A., Kling, M. A., Richelson, E., & Hauser, P. (1998). Mood and cognitive side effects of interferon-a therapy. Seminars in Oncology, 25 Supplement 1, 39-41.

MECHANISMS OF THE BEHAVIOURAL EFFECTS OF CYTOKINES Robert Dantzer,' Arnaud Aubert,' Rose-Marie Bluthe,' Gilles Gheusi,' Sandrine Cremona,' Sophie Laye,' Jan-Pieter Konsman,' Patricia Parnet,' and Keith W. Keliey'^ ' Neurobiologie integrative, INRA-INSERM U394 Rue Camille Saint-Saens, 33077 Bordeaux Cedex, France ^ Laboratory of Immunophysiology, Department of Animal Sciences University of lUinois, Urbana, Ilhnois 61801

1. I N T R O D U C T I O N Sickness behavior refers to the coordinated set of behavioral changes that develop in sick individuals during the course of an infection. A sick individual typically displays depressed locomotor activity and little or no interest in his physical and social environment. Body care activities are usually absent and ingestive behavior is profoundly depressed despite the increased metabolism that is necessary for the fever response. Since the initial demonstration in the late eighties that similar symptoms are induced in healthy subjects by peripheral and central injection of the proinflammatory cytokines that are released by activated monocytes and macrophages during the host response to infection, the mechanisms of cytokine-induced sickness behaviour have been the subject of intense research, carried out at several levels of investigation. The purpose of the present chapter is to review the results that have been obtained in this field during the last decade. At the behavioural level, there is now clear evidence showing that sickness behaviour is not the result of weakness and physical debilitation affecting the sick individual, but the expression of a central motivational state that reorganizes the organism's priorities to cope with pathogenic microorganisms. At the organ level, this motivational aspect of sickness behavior is important since it implies that the endogenous signals of sickness are likely to act on the brain to activate a set of neural structures that are at the origin of the subjective, behavioural and physiological components of sickness. The presence of cytokine receptors in the brain is in accordance with this view. These receptors are activated not by peripheral cytokines entering the brain, but by cytokines that are produced locally in the brain in response to peripheral Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

83

84

R. Dantzer et al.

cytokines. At the cellular level, there is still some uncertainty concerning the identification of the exact cell types that are the source of cytokines in the brain, and those that are the targets of brain released cytokines. At the molecular level, the cytokines that are behaviourally active and the receptor mechanisms that mediate their effects are not yet fully understood. Because sickness behaviour dissipates gradually during recovery, it has been proposed that the production and effects of cytokines are opposed by endogenous regulatory factors. These factors have been termed cryogens, by analogy with pyrogens, and they include a wide variety of bioactive molecules, from glucocorticoids to anti-inflammatory cytokines. However, their exact role has not yet been fully elucidated.

2. THE CYTOKINE NETWORK The main proinflammatory cytokines that are responsible for the initial host response to infection are interleukin-1 (IL-1), interleukin-6 (IL-6), and tumor necrosis factor-a (TNFa) (Nicola, 1994). At the periphery, these cytokines are synthesized and released by activated monocytes and macrophages, and play a pivotal role in inflammation. A convenient way to induce the expression of these cytokines in healthy subjects is to inject lipopolysaccharide (LPS), the active fragment of endotoxin from Gram negative bacteria. LPS activates macrophages and monocytes by binding to the CD14 receptor. Although the biological actions of proinflammatory cytokines widely overlap, each cytokine has its own characteristic properties. IL-1 exists in two molecular forms, IL-la and IL-ip, which are encoded by two different genes (Dinarello, 1996). Another member of the IL-1 family is the IL-1 receptor antagonist (IL-lra). This last cytokine behaves as a pure endogenous antagonist of IL-1 receptors and blocks most biological effects of IL-la and IL-1 (3 in vitro and in vivo. IL-6 is mainly responsible for the synthesis of acute phase proteins by hepatocytes. TNFa plays a major role in the pathogenesis of the septic shock syndrome. IL-1 binds to two types of receptors belonging to the immunoglobuhn superfamily and representing two separate gene products. In the immune system, the type I IL-1 receptor is predominantly expressed on T cells and fibroblasts whereas the type IIIL1 receptor is found on B cells and macrophages. The type I IL-1 receptor transduces the signal to the cell nucleus via the activation of the nuclear transcription factor N F K B . In contrast, the type II IL-1 receptor does not transduce the signal and certainly functions as a "decoy" receptor that binds the excess of IL-1 to down regulate its actions. The ability of the type I IL-1 receptor to transduce a signal is dependent on the formation of a complex with an IL-1 receptor accessory protein (IL-IR AcP). IL-lra behaves as an antagonist of IL-1 receptors by preventing the formation of this complex after binding to the type I IL-1 receptor.

3. PROINFLAMMATORY CYTOKINES INDUCE SICKNESS BEHAVIOR Peripheral and central administration of IL-la, IL-ip, and TNF-a to healthy laboratory animals induce fever (Kluger, 1991), activation of the hypothalamic-pituitaryadrenal axis (Besedovsky, del Rey, Sorkin, & Dinarello, 1986) and behavioural

Mechanisms of the Behavioural Effects of Cytokines

85

symptoms of sickness (Kent, Bluthe, Kelley, & Dantzer, 1992b). Depressed activity and reactivity are commonly observed, together with decreases in food intake and the adoption of a curled posture (Dantzer & Kelley, 1989). The two behavioural end points that have been mostly used to investigate the mechanisms of cytokine-induced sickness behaviour are social exploration and food intake. In the first case, the effect of cytokines are assessed by measuring changes in the duration of exploration of a conspecific juvenile that is temporarily introduced into the home cage of the animal under test. This social stimulus normally induces a full sequence of close following and olfactory investigation, that is profoundly depressed in a time- and dose-dependent manner in mice and rats injected with LPS and recombinant proinflammatory cytokines (Kent et al., 1992b). Changes in ingestive behaviour that develop in sick animals can be assessed based on the disruption of food intake in animals provided with free food, or the decrease in operant responding in animals trained to press a lever or to poke their nose in a hole for a food reward (Kent, Bret-Dibat, Kelley, & Dantzer, 1996; Plata-Salaman, 1995, 1997). In both cases, LPS and recombinant proinflammatory cytokines have profound depressive effects on the behaviour under investigation. 4. SICKNESS B E H A V I O U R IS T H E E X P R E S S I O N O F A C E N T R A L M O T I V A T I O N A L STATE Most of the symptoms of sickness behavior look like the expression of a physical debihtation or a general weakness of the organism. During an infection episode, the sick individual remains prostrated and engages in little physicial activity. However, this interpretation is not tenable (Fig. 1). Hart (1988) has convincingly argued that sickness behaviour is not simply the consequence of a reduced efficiency of bodily functions but that it supports the metabolic and physiological changes that occur in the infected organism and increases their efficiency. This implies that the reduced behavioral activities that occur in ill individuals are not the result of the weakness and physical debilitation that accompany an infectious illness, but the expression of a highly organized strategy that is critical to the survival of the organism. If this is the case, then sick individuals should be able to reorganize their behaviour depending on its consequences and the internal and external constraints they are exposed to. This flexibility is charac-

General weakness

X Immune activation

^ Behavioral ^ ^ ^ ^ degradation

O •

^^

V

Lethargy, hypophagia, social disinterest,...

Motivational _ Behavioral changes ^ ^ ^ ^ reorganization Figure 1. The behavioral symptoms of sickness that develop during an infectious episode can be the result of physical debilitation and weakness (upper pathway) or the expression of a reorganization of the organism's priorities in face of the pathogen challenge (lower pathway). Much evidence has been collected during the recent years to support this last interpretation.

86

R. Dantzer et al.

teristic of what psychologists call a motivation. A motivation can be defined as a central state which reorganizes perception and action. A typical motivational state is fear. In order to escape a potential threat, a fearful individual focuses his attention on all potential sources of danger, and at the same time, gets ready to engage in the most appropriate defensive behavioural pattern that is available in his behavioural repertoire. In other terms, a motivational state does not trigger an unflexible behavioural pattern in response to an internal or external stimulus. Instead of this, it enables to uncouple perception from action and therefore to select the appropriate strategy in relation to the eliciting situation (BoUes, 1970). The first evidence that sickness behaviour is the expression of a motivational state rather than the consequence of weakness was provided by Miller in an invited review paper (Miller, 1964). While he was still in search of endogenous signals of the thirst motivational system. Miller was told that endotoxin has strong suppressive effects on drinking. He therefore injected rats with endotoxin and observed that this treatment made rats stop bar pressing for water. Apparently their motivation for drinking was attenuated but not suppressed since, when given water, endotoxin-treated rats drank it although to a lesser extent than normally. However, this effect was not restricted to thirst since the same endotoxin treatment also reduced bar pressing for food and even blocked responding in rats trained to press a bar for the rewarding effects of electrical stimulation in the lateral hypothalamus (brain self-stimulation). Interestingly enough, when rats were trained to turn off an aversive electrical stimulation in this brain area, endotoxin also reduced the rate of responding, but to a lesser extent than bar pressing for a rewarding brain stimulation. More to the point, however, was the observation that rats which were placed in a rotating drum that they could stop for brief periods by pressing a lever increased their response rate in response to endotoxin treatment. The mere fact that endotoxin treatment could decrease or increase behavioural output depending on its consequences was for Miller a strong indication in favour of a motivational effect of endotoxin. Since the behavioural effects he observed in endotoxin-treated rats took time to develop, in contrast to the effect of psychoactive drugs, he went as far as proposing that endotoxin triggers the release of endogenous signals for sickness, and that animals injected with endotoxin actually behave in a sick way. Despite their premonitory nature, these results remained buried in the literature for nearly two decades before the conditions were met for a revival of interest in the motivational properties of sickness. An important characteristic of a motivational state is that it competes with other motivational states for behavioural output. As a typical example, it is difficult if not impossible to search for food, and at the same time court a sexual partner, since the behavioural patterns of foraging and courtship are not compatible with each other. The normal expression of behaviour therefore requires a hierarchical structure of motivational states which is continuously updated according to the external and internal. When an infection occurs, the sick individual is at a life or death juncture and his physiology and behaviour must be altered so as to overcome the disease. However, this is a relatively long term process which needs to make room for more urgent needs when necessary. It is easy to imagine that if a sick person lying in his bed hears a fire alarm ringing in his house and sees flames and smoke coming out of the basement, he will be able to momentarily overcome his sickness behavior to escape danger. Translated into motivational terms, this means that fear competes with sickness, and fear motivated behaviour takes precedence over sickness behaviour. An example of this competition between fear and sickness is provided by the observation that the depressive effects of

Mechanisms of the Behavioural Effects of Cytokines

87

IL-ip on behaviour of mice are more pronounced when experimental animals are tested in the safe surroundings of their home cage than when they are placed into a new environment. To show that sickness does not interfere with the subject's ability to adjust his defensive behavioural strategies with regard to his needs and capacities, Aubert and colleagues (Aubert, Kelley, & Dantzer, 1997b) compared the effects of LPS on food hoarding and food consumption in rats receiving or not a food supplement in addition to the amount of food they obtained in the situation. Briefly, rats were trained to get food for 30min in an apparatus consisting of a cage connected to an alley with free food at its end. In this apparatus, rats normally bring back to their home cage the food that is available at the end of the alley, and the amount of food they hoard is lower when they receive a food supplement than when they have no supplement. In response to LPS, food intake was decreased to the same extent whether rats received a food supplement or not. However, food hoarding was less affected in rats that did not receive a food supplement compared to the rats provided with the food supplement. These results are important because they indicate that the internal state of sickness induced by LPS is more effective in suppressing the immediate response to food than the anticipatory response to future needs. Another example of the motivational aspects of sickness behaviour is provided by a study of the effects of cytokines on maternal behaviour. In rats and mice, neonates are poikilotherm and need the protection of a nest. Maternal behaviour is therefore critical for the survival of the progeny. In accordance with the motivational priority of maternal behaviour, lactating mice that were made sick by an appropriate dose of LPS were still able to retrieve their pups that had been removed from the nest, but did not engage in nest building when tested at 22°C. However, when the dams and their litters were placed at 6°C to increase the survival value of nest building, this last behaviour was no longer impaired (Fig. 2) (Aubert, Goodall, Dantzer, & Gheusi, 1997a).

5. SICKNESS BEHAVIOUR IS MEDIATED BY THE EXPRESSION OF BRAIN CYTOKINES IN RESPONSE TO PERIPHERAL CYTOKINES The fact that sickness is a central motivational state implies that the endogenous signals of sickness act on the brain. However, cytokines are relatively large protein molecules, e.g., about 15 kDa for IL-1, and their hydrophilic nature does not allow them to cross the blood-brain barrier. It has therefore been proposed that cytokines, like many other centrally acting peptides, enter the brain at the sites in which the bloodbrain barrier is deficient. These privileged sites for communication with the internal milieu are known as the circumventricular organs, because of their anatomical location around the brain ventricles. The organum vasculosum of the lamina terminalis (OVLT) has long been considered as the site of action of peripherally released cytokines, based on results of lesion and knife cut experiments (Blatteis, 1990). Blood-borne cytokines were assumed to diffuse into the brain side of the blood-brain barrier through fenestrated capillaries, and act on parenchymal astrocytes to induce the synthesis and release of prostaglandins of the E2 series. Prostaglandins would then freely diffuse to nearby target brain areas, such as the median preoptic area of the hypothalamus in the case of fever. However, there are at least two problems with this proposed mechanism. Tlie first one is that in contrast to hormones, most cytokines have a very limited range of

R. Dantzer et al.

88 Standard ambient tentperarure (24°C)

Coldamhient temperature (6°C)

30

30 r

S 25 « 7 20 E 15

25 -

10 5 0

to 4

9 o S a

r

O)

c •5

2

C= O

-K 1 V

n

. -

t2

1

A •

I•

t2

ii

t24

3 -

•• 1 •• ^ I I•

to

to

124

t

t2

1 -

124

10

t2

t24

Figure 2. Interactions between LPS-induced sickness behavior and maternal behavior in lactating mice. Lactating mice were injected with LPS (400 |xg/kg ip) (black columns) or apyrogenic saline (grey columns) at time 0 (tO).Two and 24h later, their nest was removed and they were given surgical cotton to build another nest. Then their pups were removed and aligned along the wall on the opposite side from the nest. Dependent variables on each test were the time needed to retrieve all the pups and the quahty of the nest built by the dam (from 1 = no nest to 4 = fully enclosed nest). *p < 0.05, ***p < 0.001 compared with saline (From Aubert et al., 1997a).

diffusion, and are tuned to act locally in an autocrine and paracrine manner. The second problem is that fully functional cytokine receptors are present in the brain (see below) where they mediate the effects of peripherally released or exogenously injected cytokines. For example, blockade of brain IL-1 receptors has been found to abrogate the effects of peripheral immune stimuli on behaviour (Kent, Bluthe, Dantzer, Hardwick, Kelley, Rothwell, & Vannice, 1992a), corticotropin releasing hormone gene expression in the hypothalamus (Kakucska, Qi, Clark, & Lechan, 1993), and body temperature (Klir, McClellan, & Kluger, 1994). The simplest way to interpret these last findings is to assume that cytokines are actively transported into the brain, as proposed for other peptides. For example, insulin has been shown to be able to enter the brain by a receptor-mediated, saturable transport process across brain capillary endothelial cells (Baura, Foster, Porte, Kahn, Bergman, Cobelli, & Schwartz, 1993; Wu, Yang, & Pardridge, 1997). Pharmacokinetic data with radiolabeled cytokines have revealed the existence of specific saturable transport systems for a number of proinflammatory cytokines (Banks, Ortiz, Plotkin, & Kastin, 1991). In the case of insulin, the transporter is nothing else than the insulin receptor. However, the molecular identity of cytokine transporters has not yet been identified. In any case, it is important to point out that the existence of a transport system is not sufficient to explain how peripheral cytokines act on the brain. There is still the problem that in order to be transported, cytokines need to be present in the general blood circulation, or synthesized and released close

Mechanisms of the Behavioural Effects of Cytoldnes

89

to their transporter. Furthermore, the intervention of a transport process does not appear to be necessary for peripheral cytokines to enter the brain, since independent experiments have provided evidence for the existence of a brain compartment of cytokines that is inducible by peripheral cytokines (Fig. 3) (Gatti & Bartfai, 1993, Laye, Parnet, Goujon, & Dantzer, 1994). Brain cells cannot produce cytokines in a coordinated manner in response to peripheral cytokines without a communication pathway for transmission of immune information from the periphery to the brain (Dantzer, 1994). As proposed by Kent et al. (1992b), neural afferents are good candidates for this role, since they already transmit the two sensory components of inflammation, calor and dolor (heat and pain). When LPS or cytokines are injected into the abdominal cavity, an inflammatory response develops locally. The main afferent pathway from the abdominal cavity to the brain is the vagus nerve. The role of the vagus nerve in the transmission of the immune information from the periphery to the brain has first been demonstrated by c-fos mapping experiments. The immediate early gene c-fos is differentially expressed in many regions of the central nervous system following various physiological challenges and can be used as a marker of neuronal activation. Intraperitoneal administration of LPS to rats induced the expression of Fos, the protein product of the c-fos gene, in the primary and secondary projections areas of the vagus nerve. This labeling was completely abrogated by subdiaphragmatic section of the vagus nerve (Wan, Wetmore, Sorensen, Greenberg, & Nance, 1994). Vagotomy also blocked the depressing effects of LPS and IL-ip on behaviour in rats and mice (Fig. 4) (Bluthe, Walter, Parnet, Laye, Lestage, Verrier, Poole, Stenning, Kelley, & Dantzer, 1994; Bret-Dibat, Bluthe, Kent, Kelley, & Dantzer, 1995).This protecting effect of vagotomy was not due to an impaired peripheral immune response since the increase in the levels of IL-ip that was induced in the peritoneal macrophages and plasma in response to an intraperitoneal injection of LPS was not impaired by vagotomy (Bluthe et al., 1994; Laye, Bluthe, Kent, Combe, Medina, Parnet, Kelley, & Dantzer, 1995). TTie involvement of the vagus nerve in the behavioural effects of cytokines is specific to cytokines that are released in the

RT-PCR

ELISA I L - l p

n Hypcliidansfl RasTB Figure 3. Systemic administration of LPS induces the expression of IL-ip and IL-lra in the brain of mice. Mice were injected intraperitoneally with LPS (lOng/mouse) and killed just before or at different times after injection. Total RNA was extracted from the hypothalamus and submitted to RT-PCR to determine transcripts for IL-ip and IL-lra, using beta2-microglobulin (P2mgl) as an internal standard. Levels of IL-ip were measured in the hypothalamus (pg/mg protein) and plasma (pg/ml) using a validated ELISA (mean of 5 different experiments).

R. Dantzer et al.

90

Sham - Solvent Sham - IL-1

VGX - Solvent VGX - IL-1

Q

Time (hr) Figure 4. Vagotomy attenuates the depressing effects of systemic IL-l(i on social exploration of a juvenile conspecific introduced into the home cage of the test animal. The experiment was carried out on rats that had been submitted to vagotomy (VGX) or sham surgery (sham) 4 weeks before the test. Solvent or recombinant rat IL-ip (15 jig/rat) was injected ip immediately after the first behavioral session that took place at time 0 (baseline) and the same animals were tested again with different juveniles 2, 4, and 6h later (n = 5 for each experimental group). Duration of exploration was expressed as percentage of baseline (81 to 100sec for 4min of test) (From Bluthe et al., 1996b).

abdominal cavity and sensed by afferent vagal terminals since vagotomy did not block sickness behaviour induced by subcutaneous and intravenous injections of IL-lp (Bluthe, Michaud, Kelley, & Dantzer, 1996a), and had no effect on sickness behaviour induced by central injection of IL-ip (Fig. 5) (Bluthe, Michaud, Kelley, & Dantzer, 1996b). According to the hypothesis of a neural transmission pathway of immune signals from the periphery to the brain, section of vagal afferents should abrogate the induction of brain cytokines expression in response to intraperitoneal administration of LPS and IL-1 p. In accordance with this prediction, vagotomy was demonstrated to abrogate LPS- and IL-l^-induced increases in IL-ip mRNA in the hypothalamus and

160

Sham Solvent Sham IL-1

VGX Solvent VGX IL-1

140 120 100 80 60 40 20 0

IL-ip. 45ng, ICV Time (hr)

Figure 5. Vagotomy does not block the depressing effect of icv IL-ip on social exploration. Solvent or recombinant rat IL-ip (45ng/rat) was injected icv immediately after the first behavioral session that took place at time 0 (baseline). Same conventions as in Fig. 4 (n = 5 for each experimental group). Baseline values varied from 96 to 127 sec for 4min of test). From Bluthd et al., 1996b).

Mechanisms of the Behavioural Effects of Cytokines

91

hippocampus but not in the pituitary (Laye et al., 1995; Hansen, Taishi, Chen, & Krueger, 1998). The brain actions of peripheral cytokines that are mediated by neural afferents also include fever and activation of the HPA axis since vagotomy attenuates LPSinduced fever and increase in plasma ACTH levels (Watkins, Maier, & Goehler, 1995). All of these data converge to support the hypothesis that vagal afferents transmit the immune message from the abdominal cavity to the brain. Whether other neural afferents are involved in the transmission of the immune message originating from other bodily sites to the brain and whether the circumventricular organs are also involved in the relatively rare case in which cytokine levels dramatically increase in the general circulation remain to be determined.

6. CELLULAR ORGANIZATION OF THE CYTOKINE NETWORK IN THE BRAIN The concept that peripheral immune activation induces the expression of cytokines at the mRNA and protein levels in the brain has initially been based on the observation that intraperitoneal administration of the cytokine inducer, LPS, results in the induction of IL-ip synthesis in the brain (Hopkins & Rothwell, 1995). The brain cells which are at the origin of this synthesis are meningeal and perivascular macrophages, and ramified microglial cells (Van Dam, Brouns, Louisse, & Berkenbosch 1992; Buttini & Boddeke, 1995). However, IL-1 is not the sole cytokine to be induced in the brain by peripheral immune stimuli. TNFa is also expressed at the mRNA and protein levels (Gatti & Bartfai, 1993; Breder, Hazuka, Ghayur, Klug, Huginin, Yasuda, Teno, & Saper, 1994), and its cellular sources are represented by perivascular cells and neurons. ITiere is still some controversy as to whether IL-6 is synthesized in the brain in response to peripheral LPS or imported from the periphery (Gatti & Bartfai, 1993; Laye et al., 1994; Schobitz, de Kloet, & Holsboer, 1994). At the periphery, LPS induces first the synthesis of TNFa, that is responsible for the synthesis of IL-1. IL-1 itself triggers the synthesis and release of IL-6. RT-PCR studies of brain transcripts for these different cytokines at various intervals following intraperitoneal administration of LPS revealed that the same sequence of events occurs in the brain.TNFa and IL-1 appeared first and were followed by IL-lra and IL-6 (Laye et al., 1994). The importance of IL-1 in this cascade was apparent from the observation that intracerebroventricular (ICV) administration of IL-lra to LPS-treated mice blocked the induction of expression of IL-1 (3, IL-6, and TNFa in the hypothalamus and attenuated it in the hippocampus (Laye, Parnet, & Dantzer, unpublished observations). Based on in situ hybridization studies, IL-lra appears to be mainly expressed in the meninges, choroid plexus and perivascular cells (Wong, Bongiorno, Rettori, McCann, & Licinio. 1997). The exact cellular targets of proinflammatory cytokines in the brain are still elusive. IL-1 receptors have been identified in brain tissues, using radioligand-binding studies, immunocytochemistry, or expression of receptor mRNA. Equilibrium binding of '"T-IL-la by brain membrane homogenates and brain slices was the first technique that was used, and allowed to demonstrate the presence of binding sites for IL-1 in the mouse brain (Haour, Ban, Milon, Baran, & Pillion, 1990; Takao, Tracey, Mitchell, & De Souza, 1991; Parnet, Amindari. Wu, Brunke-Rees, Goujon, Weyhenmeyer, Dantzer, & Kelley. 1994). These binding sites are located almost exclusively in the dentate gyrus of

92

R. Dantzer et al.

the hippocampus and choroid plexus. However, these findings are inconsistent with proposed sites of action of IL-1 receptor agonists in such areas as the hypothalamus and brain stem. In situ hybridization identified the type I IL-1 receptor mRNA in several regions of the rat brain, including the anterior olfactory nucleus, medial thalamic nucleus, posterior thalamic nucleus, basolateral amygdaloid nucleus, ventromedial hypothalamus nucleus, arcuate nucleus, median eminence, mesencephalic trigeminal nucleus, motor trigeminal nucleus, facial nucleus, and Purkinje cells of the cerebellum (Ericsson, Liu, Hart, & Sawchenko, 1995). Despite their importance for the understanding of the way IL-1 affects brain functions, the mechanisms of IL-1-mediated cell-to-cell communication in the nervous system have not yet been elucidated. IL-IRI is present on neurons, astrocytes, epithelial cells of the choroid plexus and ventricles, and endothelial cells. Using a rat antimouse type I IL-1 receptor monoclonal antibody, neurons and oligodendrocytes have been found to express type I IL-1 receptor protein (Parnet et al., 1994). In addition, transcripts for the type I, but not the type II, IL-1 receptors are expressed in murine neuronal cell lines whereas transcripts for both types of receptors are present in mouse brain (Fig. 6). Northern blot analysis indicates that IL-IR Ac? is constitutively expressed at high levels in mouse brain (Greenfeder, Nunes, Kwee, Labow, Chizzonite, & Ju, 1995), but its cellular localization and coexpression with IL-IRI are still unknown. Little information is available on the factors that regulate IL-1 receptors in the nervous system. IL-1 receptor expression on the cell surface is negatively regulated by IL-1 in monocytes and fibroblasts. Indirect evidence in favor of a similar mechanism in the brain comes from experiments showing a decrease in the number of IL-1 binding sites in the hippocampus of mice administered an intraperitoneal injection of LPS (Haour et al., 1990). The same phenomenon occurs in astrocytes and could be due to internalization of the ligand-receptor complex. Currently, relatively little information is available as to whether IL-1 receptors in the nervous system are the same as those identified in other tissues. Bristulf and colleagues (Bristulf, Gatti, Malinowsky, Bjork, Sundgren, & Bartfai, 1994) PCR-cloned a type II IL-1 receptor cDNA from an excitable rat insulinoma cell line. This receptor has structural features that are not found in the previously identified type II IL-1 receptor, suggesting that different subtypes of type II IL-1 receptors could be expressed in the CNS.

7. MECHANISMS OF ACTION OF CYTOKINES ON THEIR BRAIN TARGETS The exact way by which cytokines act on their brain targets is still elusive. There are several aspects to this question. The first one concerns the brain circuitry that is involved in the response to proinflammatory cytokines. The second one concerns the respective role of brain cytokines and intermediate molecules, such as prostanoids and nitric oxide (NO), in the activation of this circuitry. The functional circuits that underlie the brain response to immune stimuli have mainly been studied in the context of the activation of the hypothalamic-pituitaryadrenal axis, the fever response and the alterations in food intake that develop in animals injected with LPS and IL-1 p. Ericsson and colleagues (Ericsson, Kovacs, & Sawchenko, 1994) have examined the pathways that mediate activation of the hypothalamic-pituitary-adrenal axis in response to an intravenous injection of IL-1 p. Using

Mechanisms of the Behavioural Effects of Cytoicines

93

Hypothalamus D TL-iRi

RT-PCR 10

m IL-IRIT

3

sllllsllll

p

B CO. i f « -

Hippocampus

_ - _ ^ j

a, OP

u

0

t

6 Hrs

LPS Figure 6. RT-PCR demonstration of the expression of type I and type II IL-1 receptors in the pituitary, hippocampus and hypothalamus of mice and its regulation by systemic administration of LPS. Mice were killed either before or at different times after ip injection of LPS (lOng/mouse) and total RNA was extracted from various tissues and brain structures for RT-PCR analysis. Yeast total RNA served as a negative control. The left portion of the figure represents a typical acrylamide gel showing transcripts for the type I IL-1 receptor (702 bp) and the type II IL-1 receptor (297 bp). The right portion of the figure represents variations of IL-1 receptor mRNA in the hypothalamus and hippocampus (expressed as percent of beta2-microglobulin mRNA, mean of 3 experiments) in response to ip LPS.

Fos immunohistochemistry, they found that Fos was induced in CI noradrenergic neurons in the ventrolateral medulla. These neurons project to the CRH containing neurons of the paraventricular nucleus of the hypothalamus (PVN). Interruption of the input from the CI cells to the PVN by knife cuts prevented the CRH response to ILip. In the case of fever, Saper and colleagues (Elmquist, Scammel, & Saper, 1997) observed that intravenously injected LPS activated Fos in the ventromedial preoptic area of the hypothalamus (VMPO). Using retrograde tracing, they found that the VMPO has both direct and indirect (via the anterior perifornical area and parastrial nucleus) projections to the autonomic parvicellular divisions of the PVN. They proposed that activation of VMPO efferents following LPS disinhibit PVN neurons that are involved in thermogenesis by inhibition of warm sensitive neurons that are located in the anterior perifornical area, therefore raising the thermostatic set-point for thermoregulation (Elmquist et al., 1997). In the case of food intake, Plata-Salaman and colleagues (Plata-Salaman, Oomura, & Kai, 1988) showed that IL-ip suppressed the

94

R. Dantzer et al.

neuronal activity of the glucose-sensitive neurons in the lateral hypothalamic area, while activating glucose-sensitive neurons in the hypothalamic ventromedial nucleus. However, the evidence for a direct involvement of these hypothalamic neural structures in the anorexia induced by immune stimuli is still lacking. Whether cytokines directly act on the above described neural pathways or activate them via intermediates such as prostaglandins and NO is still a matter of controversy. There is no doubt that cytokines are powerful inducers of the expression of cyclooxygenase-2 (COX-2) (Breder & Saper, 1996; Cao, Matsumura, Yamagata, & Watanabe, 1996) in the brain. COX-2 is the rate limiting enzyme in the conversion of arachidonic acid to prostanoids. This enzyme is induced in brain vascular- and perivascular-associated cells in response to systemic administration of LPS and IL-ip, but not IL-6 (Lacroix and Rivest, 1998). Cytokines also induce NO synthase in the brain (Wong, Rettori, Al-Shekhlee, Bongiorno, Canteros, McCann, Gold, & Licinio, 1996). The inducible form of NO synthase (iNOS), also called type II NOS, is synthesised by macrophages in response to inflammatory stimuli. Using in situ hybridization, Wong et al. (1996) reported that there is no detectable expression of iNOS in the brain in basal conditions, but that this enzyme is rapidly induced in vascular, glial, and neuronal structures of the rat brain in response to IP LPS. The induction of this enzyme in the hypothalamus was confirmed by Satta and colleagues (Satta, Jacobs, Kaltsas, & Grossman, 1998), using RT-PCR. Since many of the functional neuroanatomical studies on the brain effects of immune stimuli were carried out using the intravenous route, and endothelial cells of the brain microvasculature respond to intraluminar cytokines by producing prostaglandins, Elmquist et al. (1997) proposed that these intermediates mediate the CRH and fever responses to immune stimuli. These actions of prostaglandins would take place in different regions of the brain, the CI noradrenergic neurons for the CRH reponse, and the VMPO neurons in the case of fever. Stimulation of COX-2 in the brain microvasculature of the MPOA results in an increased local production of prostaglandins. These intermediates would bind to specific receptors expressed on nearby neurons to promote the fever in response to peripheral immune stimuli (Elmquist et al., 1997; Lacroix & Rivest, 1998). In the case of a local inflammation, prostaglandins released at the site of inflammation would activate local sensory fibers, resulting in the transmission of the peripheral immune message to the brain. The way activation of vagal sensory fibers and brain production of IL-ip combine with each other to allow the development of the host response to infection has been examined by Konsman and colleagues (Konsman, Kelley, & Dantzer, 1998), using immunohistochemistry for Fos and IL-ip, and in situ hybridization for the expression of iNOS mRNA, taken as a marker of IL-1(5 bioactivity. In this study, rats were killed at different times after IP LPS. IL-lp-positive cells were found 2h after LPS in circumventricular organs and the choroid plexus. The cells that expressed IL-1(3 were isolectinpositive cells and their shape allowed to identify them as perivascular phagocytic cells. Fos expression was restricted at that time to brain parenchymal structures such as the nucleus tractus solitarius, medial preoptic area, paraventicular nucleus, supraoptic nucleus, and central amygdala. Eight hours after LPS, Fos became apparent in circumventricular organs. This was associated with a shift of expression of IL-ip from circumventricular organs to adjacent brain structures such as the nucleus tractus solitarius and medial preoptic area. Since this was accompanied by an expression of iNOS at the interface between circumventricular organs and adjacent brain nuclei, these results were interpreted as suggesting that IL-ip functions as a volume transmision

Mechanisms of the Behavioural Effects of Cytokines

95

signal, which originates from perivascular cells at the blood side of the circumventricular organs, is passively conveyed through the interstitium of circumventricular organs by pulsations of nearby arterioles, and acts on microglial cells that are located on the neural side of the circumventricular organs. From there, the IL-ip message would propagate itself throughout the brain parenchyma by recruiting adjacent microglial cells (Konsman et al., 1998) (Fig. 7). It is important to note that a similar picture was obtained by Quan and colleagues (Quan, Whiteside, Kim, & Herkenham, 1997), based on the induction of iKBa expression in the brain in response to LPS. IKB is a cytoplasmic protein that binds to the nuclear transcription factor N F K B , preventing its translocation to the nucleus. Upon cellular activation by immune signals, IKB is

Figure 7. Schematic drawing of the way IL-ip is synthesized in the brain in response to ip LPS and its mechanisms of action. The upper portion of the figure represents a sagittal section of the rat brain whereas the lower left hand portion of the figure represents a coronal section of the brain at the level of the area postrema (AP). The lower right hand of the figure represents the liver. Intraperitoneally administered LPS induces the release of IL-ipby Kupffer cellsin the liver, resulting in the activation of vagal afferents. This neural message is transmitted to the NTS, and from there to the parabrachial nuclei (PB), the ventrolateral medulla (VLM). and the forebrain nuclei that are implicated in the metabolic, neuroendocrine and behavioral components of the systemic host response to infection (Medial Preoptic area, MPO; Paraventricular Nucleus, PVN; Supraoptic Nucleus of the hypothalamus, SON; Central nucleus of the Amygdala, CeA; Bed nucleus of the Stria Terminalis, BST). Neural activation of these structures is apparent from Fos expression. IL-ip, represented by asteriks, is first synthesized and released in the circumventricular organs (Organum Vasculosum of the Lamina Terminalis, OVLT; Area Postrema, AP; Subfornical Organ, SFO; Median Eminence, ME) and the choroid plexus (ChP). In the AP, IL-1(3 is synthesized and released by perivascular phagocytic cells and then diffuses throughout the interstitium to act on IL-1 type I receptors that are present on AP neurons and project to the Nucleus Tractus Sohtarius (NTS). This results in a second wave of FOS expression in the NTS and its projection structures. The NTS plays therefore a pivotal role in the integration of peripheral and central IL-1 (3. A second wave of IL-ip expression occurs on the neural side of the blood-brain barrier and corresponds to the sequential activation of ramified microglial cells. Not represented in this figure is the fact that the actions of IL-1|3 on many of its target structures are mediated by intermediates such as prostaglandins and NO (From Konsman et al., 1999).

%

R. Dantzer et al.

phosphorylated and dissociates itself from N F K B . N F K B is then able to enter the nucleus and stimulate transcription of a wide array of genes, including those responsible for COX-2 and iNOS. Since activation of N F K B is followed by induction of IKB, the detection of IKB induction in the brain reveals the extent and cellular location of brain-derived immune molecules in response to peripheral immune stimuli. In response to IP LPS, IKB mRNA was first expressed in cells lining the blood side of the bloodbrain barrier and progressed to glial cells inside the brain (Quan et al., 1997).

8. SPECIFICITY OF THE INVOLVEMENT OF PERIPHERAL AND CENTRAL CYTOKINES IN DIFFERENT COMPONENTS OF SICKNESS BEHAVIOUR Although most of the work on the behavioural effects of proinflammatory cytokine has concentrated on IL-1, the different components of the behavioral syndrome that develops in sick individual do not have an identical molecular basis. Moreover, the role of a given cytokine in sickness depends on the compartment, peripheral or central, in which it exerts its effects. The evidence that supports this interpretation is based on time course studies of the effect of IL-ip on social exploration and food-motivated behaviour, on experiments examining the behavioral effects of LPS in animals pretreated with various cytokine and cytokine receptor antagonists, and on investigation of the effects of immune stimulation in mice in which the gene coding for a specific cytokine or its receptor has been deleted by the technique of homologous recombination. Time course studies of the behavioural effects of IL-ip show that the changes in social exploration gradually develop within 2 hours following peripheral administration of this cytokine, whereas the changes in food-motivated behavior reach a maximum by 1 hour following treatment (Kent et al., 1996).The behavioural alterations that are induced by central administration of IL-ip usually develop more slowly than those induced by peripheral administration of this cytokine, which certainly reflects the time that is necessary for the progression of the cytokine message from the lateral ventricle of the brain to its neural targets. The identification and cloning of IL-lra provide a powerful pharmacological tool to attempt to resolve the role of IL-1 in sickness behavior. This cytokine blocks most of the inflammatory and behavioral effects of IL-1 when co-administered at a 100- to 1000-fold excess dose with IL-1, either peripherally or centrally. To determine the importance of brain IL-1 in the depressing effects of peripherally administered IL-1 on social activities and food-motivated behavior, rats were pretreated with an ICV injection of IL-lra, followed by an IP administration of IL-1 p. The depression of social exploration of a conspecific juvenile that normally occurs within 2-6 hours following IL-ip injection was totally blocked in IL-lra pretreated rats. However, in rats trained to press a bar for a food reward in a Skinner box, the same treatment only partially attenuated the decrease in bar pressing that normally occurs within 1-8 hours following IL-lp injection (Kent et al., 1992a). This did not appear to be a dose-dependent effect since increasing the dose of IL-lra by nearly nine-fold did not augment this partial blockade. These results indicate that among the different cytokines that are induced in the brain in response to peripheral IL-1, IL-1 is certainly the predominant molecular signal for the depression in social behavior but not for the decrease in foodmotivated behavior.

Mechanisms of the Behavioural Effects of Cytokines

97

Similar results were obtained when IL-lra was administered at the periphery in rats injected with IP LPS. This treatment abrogated the depressing effects of LPS on social behavior (Bluthe, Dantzer, & Kelley, 1992) but had no effect at all on the depressing effects of LPS on food-motivated behavior (Kent, Kelley, & Dantzer, 1992c), indicating that cytokines other than IL-1 mediate the disruption in food intake that occurs in LPS-treated individuals. This specificity of the actions of cytokines with respect to the end point under consideration is not restricted to sickness behavior. In the case of fever for example, it has been shown that antagonism of TNFa potentiates, instead of attenuating LPSinduced fever, suggesting that TNFa is a cryogenic cytokine (Kluger, 1991). The previously described experiments show that IL-lp is important both at the periphery and in the brain for social activities but not for food-motivated behavior. In this last case, IL-ip appears to be more important in the brain than at the periphery. To test further this possibility, the depressing effects of LPS were examined on the food intake of mice pretreated with IP or ICV IL-lra. As expected, IP pretreatment with IL-lra had no effect. However, ICV administration of IL-lra attenuated the decrease in cumulative food intake that was induced by LPS (Gheusi, unpublished observations). The involvement of brain IL-ip in the depressing effects of LPS on food intake was confirmed in an experiment using mice which are deficient in the IL-1 (3 converting enzyme (ICE).This enzyme is very important for the maturation of IL-1[3. IL-ip is synthesized as an inactive 31 kD precursor molecule, that needs to be cut at two specific aspartic acid residues to give the mature form of 17 kD which is biologically active. This cleavage is carried out by ICE. Mice in which the gene coding for ICE has been deleted (ICE-/-) do not produce active IL-ip. Both wild-type and I C E - / - mice responded to IP and ICV IL-lp injection by a decrease in food intake, indicating that the effector mechanisms of this cytokine were not affected by the mutation. However, I C E - / - mice were less sensitive than wild type mice to the depressing effects of ICV LPS on food intake and food-motivared behavior, whereas they did not differ in their response to IP LPS (Burgess, Gheusi, Yao, Johnson, Dantzer, & Kelley, 1998). The importance of brain IL-lp in the depression of social behavior that occurs in sick individuals needs to be qualified. As described earlier, brain IL-ip is involved in the depression of social exploration that occurs in response to peripheral administration of IL-ip, and this has been observed both in rats and in mice (Bluthe et al., 1997; Kent et al., 1992a). This effect appears to be mediated by the type I IL-1 receptor since ICV administration of a neutralizing antibody directed against this receptor blocked the decrease in social behavior that was induced in mice by IP IL-1 (Cremona, Goujon, Kelley, Dantzer, & Parnet, 1998). However, central blockade of IL-1 receptors by ICV IL-lra did not abrogate the effects of IP LPS on social behavior (Bluthe et al., 1992), indicating that other cytokines can compensate for the lack of action of IL-ip in the brain of LPS-treated animals. This possibility has been confirmed by the demonstration that mice which are deficient for the type I IL-1 receptor and no longer respond to ICV IL-ip, were still sensitive to the depressing effects of ICV LPS on social behavior (Bluthe, Parnet, & Dantzer, unpublished observations). In general, the presence of a network of cytokines acting in a cascade fashion does not make easy the task of assessing the relative importance of individual molecules in the induction of the behavioral symptoms that are characteristic of sick individuals. Since cytokines act in the context of other cytokines, which have agonist and antagonist activities, the conclusions that can be drawn from a particular combination of cytokines induced by a specific immune stimulus does not necessarily apply

98

R. Dantzer et al.

to a different combination of the same and possibly other cytokines induced by a different immune stimulus. Even apparently straightforward pharmacological experiments, in which the concentration of a given cytokine is artificially modified can lead to erroneous conclusions. For example, ICV and IP administration of recombinant rat IL-6 to rats failed to cause any alteration in social behavior, whereas the same treatments reliably induced fever and activation of the pituitary-adrenal axis. These results could be interpreted to suggest that IL-6 plays no role in sickness behavior. However, ICV IL-6 was found to potentiate the effects of a subthreshold dose of ICV IL-lp on social behavior in rats. In addition, mice which are deficient in IL-6 were found to be less sensitive to the depressing effects of both IL-1 and LPS on social behavior when these molecules were administered either IP or ICV. Moreover, their sensitivity to these treatments was entirely due to IL-1 since pretreatment with ICV IL-lra abrogated the depressing effects of IP LPS on social behavior in IL-6 knock-out mice (Bluthe & Dantzer, unpublished observations). These results indicate that IL-6 is behaviorally active only in the context of other proinflammatory cytokines. The problem of identifying a major molecular signal for sickness behavior mirrors the difficulties experienced by pathologists in the selection of an appropriate molecular target for developping potential therapies for the treatment of septic shock and other pathological conditions in which cytokines play a critical role. This does not imply that proinflammatory cytokines are not important for the condition under examination, it just indicates that the redundant nature of the cytokine network imposes the choice of alternative strategies, based on the investigation of the mechanisms that are responsible for the commonality of action of proinflammatory cytokines, e.g., postreceptor mechanisms, and processes that regulate cytokine expression and actions.

9. ROLE OF VASOPRESSIN AND GLUCOCORTICOIDS IN THE REGULATION OF CYTOKINE ACTIONS IN THE BRAIN Proinflammatory cytokines are often viewed as a two-edge sword. At low concentrations, they have potent biological effects that play a key role in the development of the host response to infection. However, at high concentrations, they can lead to cell death by necrosis or apoptosis. It is therefore important for the activity of proinflammatory cytokines to be tightly regulated. The molecular factors that oppose cytokine actions in the brain have been called cryogens (Kluger, 1991). The neuropeptide vasopressin (VP) is one of these substances. This key factor in the regulation of water metabolism is produced by magnocellular hypothalamic neurons and accumulates in the terminals of these neurons in the posterior pituitary. It is released in the general circulation and acts as an antidiuretic hormone in conditions of water deprivation and in response to fever. The release of VP is enhanced during the febrile response, and this enhanced release can be of such a degree that miction does not occur before defervescence. VP also acts as a neurotransmitter in the brain. It is particularly present in neurons whose cell bodies are located in the bed nucleus of the stria terminalis (BNST), and whose terminals project to the lateral septum. This vasopressinergic pathway is highly sensitive to circulating androgens in rodents. Castration leads to a dramatic reduction in the content of VP mRNA in the BNST neuronal ceU bodies and to a reduction in immunoreactive VP in the terminal areas of the septum. This androgendependent pathway is also activated during fever. Central administration of VP has been shown to attenuate the depressing effects

Mechanisms of the Behavioural Effects of Cytokines

99

of centrally injected IL-ip on social exploration. Conversely, central injection of an antagonist of vasopressin receptors, which has no biological activity on its own but prevents endogenously released vasopressin to reach its receptors, sensitized rats to the behavioural effects of IL-1. These last results are important since they suggest that endogenous VP plays a physiological modulatory role in the behavioural effects of ILl.To determine whether this phenomenon is mediated by an androgen-dependent or independent pathway, castrated male rats were compared to intact male rats in their sensitivity to the modulatory role of the VP receptor antagonist. Castration by itself potentiated the depressing effects of IL-1 P on social exploration. Central administration of VP was more effective in attenuating the behavioural effects of IL-1 in castrated than in intact male rats and, conversely, ICV administration of the VP receptor antagonist was no longer active in potentiating the behavioural effects of IL-1 in castrated male rats lacking vasopressinergic innervation of the lateral septum (Dantzer, Bluthe, & Kelley, 1991). The mechanisms by which the brain vasopressinergic system is activated by cytokines and the way vasopressin interacts with the effect of cytokines on their target cells remain to be elucidated. The potent activating effects of proinflammatory cytokines on pituitary-adrenal activity have already been mentioned. Glucocorticoids represent another class of key molecules in the regulation of sickness behaviour. In immune cells, proinflammatory cytokine gene expression is mainly regulated by glucocorticoids. Glucocorticoids downregulate the synthesis and secretion of IL-1, IL-6, and TNFa by activated monocytes and macrophages (Lee, Tsou, Chan, Thomas, Petrie, Eugui, & Allison, 1988). LPSinduced increases in plasma levels of TNFa and IL-6 are much higher in adrenalectomized (ADX) than in sham-operated animals (Zuckerman, Shellhaas, & Butler, 1989). The enhanced release of these cytokines is inhibited by administration of glucocorticoids. At the molecular level, glucocorticoids inhibit transcriptional and post-transcriptional expression of the IL-ip gene and decrease stability of IL-ip mRNA. Regulation of the synthesis and release of proinflammatory cytokines by glucocorticoids has marked functional consequences since adrenalectomy sensitizes experimental animals to the septic shock syndrome whereas administration of glucocorticoids has the opposite effect (Bertini, Bianci, & Ghezzi, 1988; Rachamandra, Sehon, & Berczi, 1992). To assess the possible influence of endogenous glucocorticoids on cytokine expression in the brain, ADX mice and sham-operated mice were injected with saline or LPS, and the levels of transcripts for IL-1 a, IL-lp, IL-lra, IL-6, and TNFa were subsequently determined in the spleen, pituitary, hypothalamus, hippocampus, and striatum, using comparative reverse transcription polymerase chain reaction (RT-PCR) (Goujon, Parnet, Laye, Combe, & Dantzer, 1996). Levels of IL-1 (3 were also measured by a specific ELISA in the plasma and tissues of experimental animals. In accordance with the results of previous investigations, LPS induced the expression of proinflammatory cytokines at the mRNA level in most tissues under investigation. This effect was potentiated by adrenalectomy in the plasma and other tissues, including several brain structures (Fig. 8). However, the magnitude of this enhancement differed according to the tissue under investigation. LPS increased plasma and tissue levels of IL-1 (3, as determined by ELISA, and this effect was potentiated by adrenalectomy, in the plasma and tissues other than the spleen. Since IL-ip and other proinflammatory cytokines are mainly synthesized in the brain by microglial cells and meningeal and perivascular macrophages, and since glucocorticoids have been shown to inhibit in vitro the synthesis and secretion of IL-1, IL-6, and TNFa by activated monocytes and macrophages.

R. Dantzer et al.

100

DShamlADX

E

Saline

LPS

IL-lp mRNA Figure 8. Adrenalectomy increases expression of IL-ip mRNA in mice. Adrenalectomized (ADX) and sham operated mice were injected subcutaneously with saline or LPS (lOng/mouse) and killed 2h later.Total RNA was extracted from the hypothalamus and IL-ip mRNA, expressed as percentage of beta2-microglobulin mRNA was measured by comparative RT-PCR. The figure represents the mean results of 3 different experiments. Note that LPS increased hypothalamic IL-ip mRNA and adrenalectomy increased the expression of this cytokine in saline- and LPS-treated mice (From Goujon et al., 1996).

it is tempting to propose that the enhancing effects of adrenalectomy on cytokine gene expression in peripheral and central tissues is the result of the lack of glucocorticoids. This interpretation is strengthened by the observation that the effects of stress are contrary to those of adrenalectomy on the LPS-induced expression of cytokines in the brain (Goujon, Parnet, Laye, Combe, Kelley, & Dantzer, 1995c). In this experiment, mice injected with LPS or saline, were submitted to a 15-min restraint stress which resulted in a significant increase in plasma corticosterone levels. They were subsequently sacrificed to assess the effects of this stressor on the induction of proinflammatory cytokines in the spleen, pituitary, hypothalamus, hippocampus and striatum. LPS-induced cytokine gene expression, as determined by comparative RT-PCR, was lower in restrained than in nonrestrained mice. LPS increased plasma and tissue IL-ip levels, as determined by ELISA. This effect of LPS was also less marked in restrained than in nonrestrained mice. It is important to note that stress has also been reported to induce an increase in IL-ip mRNA levels in the hypothalamus of rats (Minami, Kuraishi, Yamaguchi, Nakai, Hirai, & Satoh, 1991) and an enhancement of plasma levels of IL-6 (Le May, Vander, & Kluger, 1990; Zhou, Kusnecov, Shurin, DePaoli, & Rabin, 1993). The reason for these contradictory results is not known. As already mentioned, IL-lp is synthesized as an inactive precursor which needs to be cleaved to be secreted in an active form. The enzyme which is responsible for this cleavage is a protease, called interleukin-ip converting enzyme (ICE).To determine whether those factors which regulate the expression of IL-1 in immune and nonimmune tissues are also able to regulate the expression of ICE, mice were injected with LPS and the levels of ICE mRNA in the spleen, pituitary, and brain of experimental animals were measured by comparative RT-PCR. ICE mRNAs were more abundant in the spleen and hippocampus than in the pituitary and hypothalamus, but they were not significantly altered by LPS treatment. In another experiment, mice were submitted to adrenalectomy or a 15-min restraint stress and injected with saline

Mechanisms of the Behavioural Efl'ects of Cytoitines

101

or LPS. ADX mice had significantly higher ICE mRNA levels whereas stressed mice had significantly lower ICE mRNA levels than their respective controls. These results can be interpreted to suggest that endogenous glucocorticoids regulate the expression of ICE in peripheral and brain tissues, in a manner which is commensurate with their effects on IL-ip gene expression (Laye, Goujon, Combe, VanHoy, Kelley, Parnet, & Dantzer, 1996). In addition to their action on the synthesis of proinflammatory cytokines in the periphery and the brain, glucocorticoids are also able to modulate the expression of IL-1 receptors. Although glucocorticoids induce the expression of IL-1 receptors in various immune and nonimmune tissues, mixed results have been reported in the brain (Ban, Marquette, Sarrieau, Fitzpatrick, Pillion, Millon, Rostene, & Haour, 1993; Betancur, Lledo, Borrell, & Guaza, 1994). The problem with the radioligand binding techniques which have been used for these studies is that they do not allow an accurate description of what is going on at the receptor level, since the same effect, e.g., decreased binding, can be seen as a result of very different processes, e.g., down regulation of receptors, or activation of receptors with internalization of receptor-ligand complexes. Studies at the mRNA level revealed that ADX was accompanied by a decrease in the levels of IL-IRI and IL-IRII transcripts in the mouse pituitary and brain, whereas stress had the opposite effects (Goujon, Laye, Parnet, & Dantzer, unpublished observations). When ADX mice were implanted with a corticosterone pellet to induce plasma levels of corticosterone intermediate between baseline and stress levels, IL-IRI and IL-IRII mRNAs were increased above baseline levels. In all cases, the effects of corticosterone supplementation and stress were more marked for IL-IRII than for IL-IRI mRNA, which would fit with the down-regulatory effects of glucocorticoids on IL-1 biological activity. The down regulatory effects of glucocorticoids on the proinflammatory cytokine network at the periphery and in the brain should lead to an increased sensitivity of ADX animals to the brain actions of these cytokines. TTiis appears to be the case since adrenalectomy increased the febrile response to LPS whereas administration of glucocorticoids had the opposite effect (Coehlo, Souza, & Pela, 1982; Morrow, McClellan, Conn, & Kluger, 1993). In the same manner, adrenalectomy enhanced the depression of social exploration induced by peripheral injection of IL-lp or LPS (Goujon, Parnet, Aubert, Goodall, & Dantzer, 1995a). This effect was mimicked by administration of the glucocorticoid type II receptor antagonist, RU 38486. Chronic replacement with a 15-mg corticosterone pellet which yielded plasma corticosterone levels intermediate between baseline and stress levels, abrogated the enhanced susceptibility of ADX mice to the lower dose of IL-1 (3 but had only partial protective effects on the response to a higher dose of IL-ip and to LPS. Taken together, these findings can be interpreted to suggest that the phasic response of the pituitary-adrenal axis to cytokines has an important regulatory role on the neural effects of cytokines. However, they do not reveal at what level (peripheral or central) glucocorticoids are acting. Central sites of action appear to be involved in the regulatory effects of glucocorticoids on the neural effects of cytokines, since central injection of RU 38486 potentiates fevers induced by a peripheral injection of LPS to rats (McClellan, Klir, Morrow, & Kluger, 1994). In the same manner, adrenalectomy sensitized mice to the depressing effects of intracerebroventricular administration of IL-1 on social exploration, and this effect was attenuated by corticosterone compensation (Goujon, Parnet, Cremona, & Dantzer, 1995b). Central administration of RU 38486 mimicked the effect of adrenalectomy.

102

R. Dantzer et al.

10. C O N C L U S I O N Although the scope of this review has been restricted mainly to the involvement of members of the interleukin-1 family in sickness behaviour, sufficient evidence is now available to indicate that the proinflammatory cytokines, which are released at the periphery by activated monocytes and macrophates, activate peripheral afferent nerves to induce the synthesis and release of proinflammatory cytokines in the brain. These centrally released cytokines act probably by volume transmission on those neural structures that are involved in the control of thermoregulation, metabolism and behaviour, resulting in the development of the brain components of the host response to infection. Glucocorticoids that are released by the adrenal cortex in response to the hypothalamic effects of proinflammatory cytokines regulate the expression and actions of cytokines not only in the periphery but also in the brain.

ACKNOWLEDGMENTS Supported by INRA, INSERM, DRET, Commission of the European Communities (BIOMED 1 PL93-1450, BIOMED 2 CT97-2492, andTMR CT97-0149), and the National Institute of Health (MH-51569 and DK-49311 to KWK).

REFERENCES Aubert, A., Goodall, G., Dantzer, R., & Gheusi, G. (1997a). Differential effects of lipopolysaccharide on pup retrieving and nest building in lactating mice. Brain Behavior & Immunity, 11,107-118. Aubert, A., Kelley, K. W., & Dantzer, R. (1997b). Differential effect of lipopolysaccharide on food hoarding behavior and food consumption in rats. Brain, Behavior, & Immunity, 11, 229-238. Ban, E., Marquette, C., Sarrieau, A., Fitzpatrick, F., Pillion, G., Millon, G., Rostene, W., & Haour, F. (1993). Regulation of interleukin-1 receptor expression in mouse brain and pituitary by lipopolysaccharide and glucocorticoids. Neuroendocrinology, 58, 581-587. Banks, W. A., Ortiz, L., Plotkin, S. R., & Kastin, A. J. (1991). Human interleukin (IL) 1 alpha, murine IL-1 alpha, and murine IL-1 beta are transported from blood to brain in the mouse by a shared saturable mechanism. Journal of Pharmacology and Experimental Therapeutics, 259, 988-996. Baura, G. D., Foster, D. M., Porte, D. Jr., Kahn, S. E., Bergman, R. N., CobelH, C , & Schwartz, M. W. (1993). Saturable transport of insulin from plasma into the central nervous system of dogs in vivo. A mechanism for regulated insulin delivery to the brain. Journal of Clinical Investigation, 92, 1824-1830. Bertini, R., Bianchi, M., & Ghezzi, P. (1988) Adrenalectomy sensitizes mice to lethal effects of interleukin-1 and tumor necrosis (actor. Journal of Experimental Medicine, 167, 1708-1712. Besedovsky, H. O., del Rey, A., Sorkin, E., & Dinarello, C. A. (1986). Immunoregulatory feedback between interleukin-1 and glucocorticoid hormones. Science, 233, 652-654. Betancur, C , Lledo, A., Borrell, J., & Guaza, C. (1994). Corticosteroid regulation of IL-1 receptors in the mouse hippocampus: effects of glucocorticoid treatment, stress, and adrenalectomy. Neuroendocrinology, 59,120-128. Blatteis, C. M. (1990). Neuromodulative actions of cytokines. Yale Journal of Biological Medicine, 63,133-142. Bluthe, R. M., Dantzer, R., & Kelley, K. W. (1992). Effects of interleukin-1 receptor antagonist on the behavioral effects of lipopolysaccharide in rat. Brain Research, 573, 318-320. Bluthe, R. M., Walter, V., Parnet, P, Laye, S., Lestage, J., Verrier, D., Poole, S., Stenning, B. E., Kelley, K. W., & Dantzer, R. (1994) Lipopolysaccharide induces sickness behaviour in rats by a vagal mediated mechanism. Comptes Rendus de VAcademic des Science (Paris), 317, 499-503. Bluthe, R. M., Michaud, B., Kelley, K. W., & Dantzer, R. (1996a). Vagotomy blocks behavioural effects of interleukin-1 injected via the intraperitoneal route but not via other systemic routes. NeuroReport, 7, 2823-2827.

Mechanisms of the Behavioural Effects of Cytokines

103

Bluthe, R. M., Michaud, B., Kelley, K. W., & Dantzer, R. (1996b). Vagotomy attenuates behavioral effects of interleukin-1 injected peripherally but not centrally. NeuroReport, 7,1485-1488. Bluthe, R. M., Michaud, B., Dantzer, R., & Kelley, K. W. (1997). Central mediation of the effects of interleukin-1 on social exploration and body weight in mice. Psychoneuroendocrinology, 22, 1-11. Bolles, R. C. (1970). Species-specific defense reactions and avoidance learning. Psychological Review, 77, 32^8. Breder, C. D. & Saper, C. B. (1996). Expression of inducible cyclooxygenase mRNA in the mouse brain after systemic administration of bacterial lipopolysaccharide. Brain Research, 713, 64-69. Breder, C. D., Hazuka, C , Ghayur, T., Klug, C , Huginin, M.. Yasuda, K., Teno, M., & Saper, C. B. (1994). Regional induction of tumor necrosis factor a expression in the mouse brain after systemic lipopolysaccharide administration. Proceedings of the National Academy of Sciences USA, 91, 11393-11397. Bret-Dibat, J. L., Bluthe, R. M., Kent, S., Kelley, K. W., & Dantzer, R. (1995). Lipopolysaccharide and interleukin-1 depress food-motivated behavior in mice by a vagal-mediated mechanism. Brain, Behavior, & Immunity, 9, 242-246. Bristulf, J., Gatti, S., Malinowsky, D , Bjork, L., Sundgren A. K., & Bartfai,T. (1994). Interleukin-1 stimulates the expression of type I and type II interleukin-1 receptors in the rat insulinoma cell line RinmSF: sequencing a rat type II interleukin-1 receptor cDNA. European Cytokine Network, 5, 319-330. Burgess, W., Gheusi, G., Yao, J., Johnson, R. W., Dantzer, R., & Kelley, K. W. (1998). Interleukin-1 P converting enzyme (ICE)-deficient mice are resistant to central but not systemic lipopolysaccharide-induced aphagia. American Journal of Physiology, 274, R1829-R1833. Buttini, M. & Boddeke, H. (1995). Peripheral lipopolysaccharide stimulation induces interleukin-1 p messenger RNA in rat brain microglial cells. Neuroscience, 65, 523-530. Cao, C. Y., Matsumura, K., Yamagata, K., & Watanabe, Y. (1996). Endothelial cells of the rat-brain vasculature express cyclooxygenase-2 messenger-RNA in response to systemic interleukin-1: a possible site of prostaglandin synthesis responsible for fever. Brain Research, 733,263-272. Coehlo, M. M., Souza, G. E. P., & Pela, I. R. (1982). Endotoxin-induced fever is modulated by endogenous glucocorticoids in rats. American Journal of Physiology, 263, R423-R427. Cremona S., Goujon, E., Kelley, K. W., Dantzer, R., & Parnet, P (1998). Brain type I but not type II interleukin-1 (IL-1) receptors mediate the effects of IL-lp on behavior in mice. American Journal of Physiology, 274, R735-R740. Dantzer, R. (1994). How do cytokines say hello to the brain? Neural versus humoral mediation. European Cytokine Network, 5, 271-273. Dantzer, R. & Kelley, K. W. (1989). Stress and immunity: An integrated view of relationships between the brain and the immune system. Life Sciences, 44,1995-2008. Dantzer R., Bluthe, R. M., & Kelley, K. W. (1991). Androgen-dependent vasopressinergic neurotransmission attenuates interleukin-1-induced sickness behavior. Brain Research, 557,115-120. Dinarello, C. A. (1996). Biologic basis for interleukin-1 in disease. Blood, 87, 2095-2147. Elmquist, J. K., Scammell, T. E., & Saper, C. B. (1997). Mechanisms of CNS response to systemic immune challenge: the febrile response. Trends in Neuroscience, 20, 565-570. Ericsson, A., Kovacs, K. J., & Sawchenko, P. E. (1994). A functional anatomical analysis of central pathways subserving the effects of interleukin-1 on stress-related neuroendocrine neurons. Journal of Neuroscience, 14, 897-913. Ericsson, A., Liu, C , Hart, R. P., & Sawchenko, P. E. (1995). Type 1 interleukin-1 receptor in the rat brain: distribution, regulation, and relationship to sites of IL-1 induced cellular activation. Journal of Comparative Neurology, 361, 681-698. Gatti, S. & Bartfai,T. (1993). Induction of tumor necrosis factor a mRNA in the brain after peripheral endotoxin treatment: comparison with interleukin-1 family and interleukin-6. Brain Research, 624,291-295. Goujon, E., Parnet, P., Aubert, A., Goodall, G., & Dantzer, R. (1995a). Corticosterone regulates behavioral effects of lipopolysaccharide and interleukin-1 P in mice. American Journal of Physiology, 269, R154-R159. Goujon, E., Parnet, P., Cremona, S., & Dantzer, R. (1995b). Endogenous glucocorticoids downregulate central effects of interleukin-1 p on body temperature and behavior in mice. Brain Research, 702,173-180. Goujon, E., Parnet, P, Laye, S., Combe, C , Kelley, K. W., & Dantzer, R. (1995c). Stress downregulates lipopolysaccharide-induced expression of proinflammatory cytokines in the spleen, pituitary and brain of mice. Brain, Behavior, & Immunity, 9, 292-303. Goujon, E., Parnet, P., Laye, S., Combe, C , & Dantzer, R. (1996). Adrenalectomy enhances proinflammatory cytokine gene expression in the spleen, pituitary and brain of mice in response to lipopolysaccharide. Molecular Brain Research, 36, 53-62.

104

R. Dantzer et al.

Greenfeder, S. A., Nunes, P., Kwee, L., Labow, M., Chizzonite, R. A., & Ju, G. (1995). Molecular cloning and characterization of a second subunit of the interleukin-1 receptor complex. Journal of Biological Chemistry, 270,13757-13765. Hansen, M. K., Taishi, P., Chen, Z., & Krueger, J. M. (1998). Vagotomy blocks the induction of interleukinip (IL-ip) mRNA in the brain of rats in response to systemic IL-lp. Journal of Neuroscience, 18, 2247-2253. Haour, F. G., Ban, E. M., Milon, G. M., Baran, D., & Pillion G. M. (1990). Brain interleukin-1 receptors. Characterization and modulation after lipopolysaccharide injection. Progress in NeuroEndocrinoImmunology, 3,196-204. Hart, B. L. (1988). Biological basis of the behavior of sick animals. Neuroscience & Biobehavioral Reviews, 12,123-137. Hopkins, S. J. & Rothwell, N. J. (1995). Cytokines and the nervous system. I: Expression and recognition. Trends in Neurosciences, 18, 83-87. Kakucska, I., Qi, Y., Clark, B. D., & Lechan, R. M. (1993). Endotoxin-induced corticotropin-releasing hormone gene expression in the hypothalamic paraventricular nucleus is mediated centrally by interleukin-1. Endocrinology, 133, 815-821. Kent S., Bluthe, R. M., Dantzer, R., Hardwick, A. J., Kelley, K. W., Rothwell, N. J., & Vannice, J. L. (1992a). Different receptor mechanisms mediate the pyrogenic and behavioral effects of interleukin-1. Proceedings of the National Academy of Sciences USA, 89, 9117-9120. Kent, S., Bluthe, R. M., Kelley, K. W., & Dantzer, R. (1992b). Sickness behavior as a new target for drug development. Trends in Pharmacological Sciences, 13, 24-28. Kent, S., Kelley, K. W., & Dantzer, R. (1992c). Effects of lipopolysaccharide on food-motivated behavior are not blocked by an interleukin-1 receptor antagonist. Neuroscience Letters, 145, 83-86. Kent, S., Bret-Dibat, J. L., Kelley, K. W., & Dantzer, R. (1996). Mechanisms of sickness-induced decreases in food-motivated behavior. Neuroscience & Biobehavioral Reviews, 20, 171-175. Klir, J. J., McClellan, J. L., & Kluger, M. J. (1994). Interleukin-lp causes the increase in anterior hypothalamic interleukin-6 during LPS-induced fever in rats. American Journal of Physiology, 266, R1845-R1848. Kluger, M. J. (1991). Fever: Role of pyrogens and cryogens. Physiological Reviews, 71, 93-127. Konsman, J. P., Kelley, K. W., & Dantzer, R. (1998). Temporal and spatial relationships between lipopolysaccharide-induced expression of Fos, interleukin-1 P and inducible NO synthase in rat brain. Neuroscience, L89, 535-548. Lacroix, S. & Rivest, S. (1998). Effect of acute systemic inflammatory response and cytokines on the transcription of the genes encoding cycloxygenase enzymes (COX-1 and COX-2) in the rat brain. Journal of Neurochemistry, 70, 452^66. Laye, S., Parnet, P., Goujon, E., & Dantzer, R. (1994). Peripheral administration of lipopolysaccharide induces the expression of cytokine transcripts in the brain and pituitary of mice. Molecular Brain Research, 27,157-162. Laye S., Bluthe, R. M., Kent, S., Combe, C , Medina, C , Parnet, P, Kelley, K. W., & Dantzer, R. (1995). Subdiaphragmatic vagotomy blocks induction of IL-lp mRNA in mice brain in response to peripheral LPS. American Journal of Physiology, 268, R1327-R1331. Laye, S., Goujon, E., Combe, C , VanHoy, R., Kelley K. W., Parnet, P, & Dantzer, R. (1996). Effects of lipopolysaccharide and glucocorticoids on expression of interleukin-1 P converting enzyme in the pituitary and brain of mice. Journal of Neuroimmunology, 68, 61-66. Lee, S. W., Tsou, A. P, Chan, H., Thomas, J., Petrie, K., Eugui, E. M., & Allison, C. A. (1988). Glucocorticoids selectively inhibit the transcription of the interleukin-lp gene and decrease the stabihty of interleukini p mRNA. Proceedings of the National Academy of Sciences USA, 85,1204-1208. Le May, L. G., Vander, A. J., & Kluger, M. J. (1990). The effect of psychological stress on plasma interleukin6 activity in rats. Physiology & Behavior, 47, 957-961. McClellan, J. L., Klir, J. J., Morrow, L. E., & Kluger, M. J. (1994). Central effects of glucocorticoid receptor antagonist RU-38486 on lipopolysacharide and stress-induced fever. American Journal of Physiology, 267, R705-R711. Miller, N. E. (1964). Some psychophysiological studies of motivation and of the behavioral effects of illness. Bulletin of the British Psychology Society, 17,1-20. Minami, M., Kuraishi, Y., Yamaguchi,T., Nakai, S., Hirai, Y., & Satoh, M. (1991). Immobilization stress induces interleukin-1 beta mRNA in the hypothalamus. Neuroscience Letters, 123, 254-256. Morrow, L. E., McClellan, J. L., Conn, C. A., & Kluger M. J. (1993). Glucocorticoids alter fever and IL-6 responses to psychological stress and to lipopolysaccharide. American Journal of Physiology, 264, R1010-R1016.

Mechanisms of the Behavioural Effects of Cytokines

105

Nicola, N. A. (ed) (1994). Guidebook to cytokines and their receptors. Oxford University Press, Oxford. Parnet P., Amindari, S., Wu, C , Brunke-Reese, D., Goujon, E., Weyhenmeyer, J. A., Dantzer, R., & Kelley, K. W. (1994). Expression of type I and type II interleukin-1 receptors in mouse brain. Molecular Brain Research, 27, 63-70. Plata-Salaman, C. R. (1995). Cytokines and feeding suppression: An integrative view from neurological to molecular levels. Nutrition, 11,674-677. Plata-Salaman, C. R. (1997). Anorexia during acute and chronic disease. Relevance of neurotransmitterpeptide-cytokine interactions. Nutrition, 13,159-160. Plata-Salaman, C. R., Oomura, Y., & Kai, Y. (1988). Tumor necrosis factor and interleukin-1 P suppression of food intake by direct action in the central nervous system. Brain Research, 448,106-114. Quan, N., Whiteside, M., Kim, L., & Herkenham, M. (1997). Induction of inhibitory factor kappaBalpha mRNA in the central nervous system after peripheral lipopolysaccharide administration: an in situ hybridization histochemistry study in the rat. Proceedings of the National Acdemy of Sciences USA, 94,10985-10990. Rachamandra, R. N., Sehon A. H., & Berczi, I. (1992). Neuro-hormonal host defence in endotoxin shock. Brain, Behavior & Immunity, 6,157-169. Satta, M. A., Jacobs, R. A., Kaltsas, G. A., & Grossman, A. B. (1998). Endotoxin induces interleukin-1 P and nitric oxide synthase mRNA in rat hypothalamus and pituitary. Neuroendocrinology, 67,109-116. Schobitz, B., de Kloet, E. R., & Holsboer, E (1994). Gene expression and function of interleukin 1, interleukin 6 and tumor necrosis factor in the brain. Progress in Neurobiology, 44,397-432. Takao, T., Tracey, D. E., Mitchell, W. M., & De Souza, E. B. (1991). Interleukin-1 receptors in mouse brain: characterization and neuronal localization. Endocrinology, 127, 3070-3078. Van Dam A. M., Brouns, M., Louisse, S., & Berkenbosch, E (1992). Appearance of interleukin-1 in macrophages and in ramified microglia in the brain of endotoxin-treated rats: a pathway for the induction of non-specific symptoms of sickness. Brain Research, 588, 291-296. Wan, W., Wetmore, W., Sorensen, C. M., Greenberg, A. H., & Nance, D. M. (1994). Neural and biochemical mediators of endotoxin and stress-induced c-fos expression in the rat brain. Brain Research Bulletin, 34,1-U. Watkins, L. R., Maier, S. E, & Goehler, L. E. (1995). Cytokine-to-brain communication: A review & analysis of alternative mechanisms. Life Sciences, 57,1011-1026. Wong, M. L., Bongiorno, P B., Rettori, V., McCann, S. M., & Licinio, J. (1997). Interleukin (IL) ip, IL-1 receptor antagonist, IL-10, and IL-13 gene expression in the central nervous system and anterior pituitary during systemic inflammation: pathophysiological implications. Proceedings of the National Academy of Sciences USA, 94, 227-232. Wong, M. L., Rettori, v., Al-Shekhlee, A., Bongiorno, P B., Canteros, G., McCann, S. M., Gold, P W., & Licinio, J. (1996). Inducible nitric oxide synthase gene expression in the brain during systemic inflammation. Nature Medicine, 2, 581-584. Wu, D., Yang, J., & Pardridge, W. M. (1997). Drug targeting of a peptide radiopharmaceutical through the primate blood-brain barrier in vivo with a monoclonal antibody to the human insulin receptor. Journal of Clinical Investigation, 100, 1804-1812. Zhou, D., Kusnecov, A. W, Shurin, M. R., DePaoli, M., & Rabin, B. S. (1993). Exposure to physical and psychological stressors elevates plasma interleukin-6: relationship to the activation of hypothalamicpituitary-adrenal axis. Endocrinology, 133, 2523-2530. Zuckerman, S. H., Shellhaas, J., & Butler, L. D. (1989). Differential regulation of lipopolysacharide-induced interleukin-1 and tumor necrosis factor synthesis: effects of endogenous glucocorticoids and the role of pituitary-adrenal axis. European Journal of Immunology, 19, 301-305.

EFFECTS OF CYTOKINES ON GLUCOCORTICOID RECEPTOR EXPRESSION AND FUNCTION Glucocorticoid Resistance and Relevance to Depression

Andrew H. Miller,* Carmine M. Pariante, and Bradley D. Pearce Department of Psychiatry and Behavioral Sciences Emory University School of Medicine 1639 Pierce Dr., Suite 4000, Atlanta, Georgia 30322

1. INTRODUCTION A large body of data has been amassed which convincingly demonstrates that the immune system and the nervous system extensively interact (Ader, Cohen, & Felten, 1991; Miller, In Press). Immune cells and tissues express receptors for a wide range of transmitters associated with and regulated by the nervous system including neurotransmitters, peptides and hormones, and nervous system innervation of lymphoid tissues has been well characterized. Moreover, the presence of soluble immune products (cytokines) and their receptors have been found in multiple nervous system and endocrine tissues (Besedovsky & del Rey, 1996). Alterations of nervous system function by exposure to a variety of stressors has been shown to result in dramatic changes in immune system function, and exposure to cytokines or various types of immune activation has been shown to significantly alter CNS function (Miller, In Press; McEwen, Biron, Brunson, Bulloch, Chambers, Dhabhar, Goldfarb, Kitson, Miller, Spencer, & Weiss, 1997; Kent, Bluthe, Kelley, & Dantzer, 1992). Given the vast potential for mutual regulation involving the immune system and the brain, there has been considerable interest in the possibility that immune system processes may be involved in the pathophysiology of psychiatric syndromes including major depression. One possible * 404-727-8260 (phone), 404-727-3233 (FAX), [email protected] (email) Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

107

108

A. H. Miller et al.

mechanism whereby the immune system might contribute to the development of pathology which manifests as depressive symptomatology involves the capacity of cytokines to inhibit functioning of the receptors for glucocorticoids, thus inducing a state of glucocorticoid resistance. Glucocorticoid resistance is reliably found in patients with major depression and may contribute to reduced negative feedback by glucocorticoids on key pathways which may mediate behavioral toxicity in the depressive disorders including excessive release of corticotropin releasing hormone (CRH) and/or proinflammatory cytokines, such as interleukin (IL)-l or tumor necrosis factor (TNF)alpha. This chapter will review the role of cytokines in glucocorticoid resistance and examine the potential relevance of glucocorticoid resistance in expression of behavioral toxicity and morbidity/mortality. Finally, the capacity of antidepressants to facilitate glucocorticoid receptor function and thereby reverse glucocorticoid resistance and its related pathologies will be discussed.

2. GLUCOCORTICOID RESISTANCE IN DEPRESSION Hyperactivity of the hypothalamic pituitary adrenal (HPA) axis in patients with major depression is one of the most consistent findings in biological psychiatry. Patients with major depression exhibit elevated Cortisol concentrations in plasma, urine, and cerebrospinal fluid (csf); an enlargement of both the pituitary and adrenal glands, and an exaggerated Cortisol response to ACTH (Owens & Nemeroff, 1993; Holsboer & Barden, 1996). The etiology of these HPA axis alterations is believed to be a function of hypersecretion of CRH. CRH is a key regulatory peptide in HPA axis regulation and has a multitude of behavioral effects in animals which are similar to those seen in patients suffering from depression including alterations in activity, appetite, and sleep (Owens & Nemeroff, 1993). Moreover, patients with depression have been found to exhibit increased concentrations of CRH in the csf, increased mRNA in the paraventricular nucleus of the hypothalamus, and a blunted ACTH response to CRH challenge (likely reflecting downregulation of pituitary CRH receptors) (Owens & Nemeroff, 1993; Holsboer & Barden, 1996). Finally, downregulation of receptors for CRH in the frontal cortex of victims of suicide (presumably secondary to hypersecretion of CRH) has been described (Owens & Nemeroff, 1993). Although much attention has focused on the potential neurotransmitter pathways that may be involved in excessive CRH activity, one mechanism by which CRH hypersecretion may occur is through altered feedback inhibition of CRH release by endogenous glucocorticoids. Endogenous glucocorticoids serve as potent negative regulators of the synthesis and release of CRH through binding to glucocorticoid receptors (GR) in relevant HPA axis tissues including the hypothalamus and the hippocampus. Data supporting the notion that glucocorticoid mediated feedback inhibition is impaired in major depression comes from a number of studies demonstrating nonsuppression of Cortisol secretion following administration of the synthetic glucocorticoid, dexamethasone (DEX) (Riberio, Tandon, Grunhaus, & Greden, 1993). This resistance to the inhibitory effects of glucocorticoid-mediated feedback inhibition (glucocorticoid resistance) is also apparent in studies showing lack of inhibition of ACTH responses to CRH following DEX pretreatment (Holsboer & Barden, 1996). Altered responsiveness in the DEX-CRH test recently has been demonstrated in the first degree relatives of patients with major depression, suggesting a genetic liability to glucocorticoid resistance in at risk probands of depressed patients (Holsboer & Barden, 1996).

Glucocorticoid Resistance and Relevance to Depression

109

Table 1. Review of literature: glucocorticoid receptors in depression Glucocorticoid receptor number — 11 studies identified — 7 found no difference in receptor number between depressives and controls — 4 found a significant decrease in receptor number in depressed patients — 3 of 4 studies which found a significant decrease in receptor number in depressed patients used a cytosolic radioligand binding assay — all negative studies used the whole cell binding technique Glucocorticoid receptor function — 6 studies identified — all 6 found evidence of glucocorticoid resistance as determined by an in vitro (3) or an in vivo (3) hormone challenge — 1 study on recovered depressives found no difference in steroid sensitivity compared to controls

As noted, glucocorticoid-mediated feedback inhibition is mediated via GR, and therefore the impaired inhibitory response to DEX in patients with major depression has been hypothesized to reflect alterations in the GR. Accordingly, a number of studies have examined GR expression and/or function in depression (Pariante, Nemeroff, & Miller, 1995). The majority of these studies have utilized readily accessible peripheral blood mononuclear cells which express relatively high amounts of GR and have been shown in animal studies to downregulate following chronic glucocorticoid exposure (Spencer, Miller, Stein, & McEwen, 1991). While some investigations have demonstrated decreased GR binding in depressed patients (mostly those using cytosolic radioligand binding assays), the majority have not (mostly those using whole cell assays) (Pariante et al., 1995). There are several important technical differences between these assays which are relevant regarding the interpretation of this data, including the relative compartmentalization of the receptor and relative receptor affinity. Nevertheless, a much more consistent finding has been the decreased sensitivity of peripheral blood mononuclear cells from depressed patients to glucocorticoid effects on functional endpoints measured in vitro (typically assessments of immunologic function) compared to cells from healthy controls (Pariante et al., 1995). This glucocorticoid resistance is consistent with data showing nonsuppression of HPA axis function following dexamethasone, and further support the notion that alterations in glucocorticoid inhibition of HPA axis activity underlies the pathologic hypersecretion of CRH found in this disorder.

3. CYTOKINES AND GLUCOCORTICOID RESISTANCE Given the potential role of glucocorticoid resistance in the pathophysiology of major depression, attention has focused on what factors might contribute to glucocorticoid resistance in depression. One interesting possibihty involves the immune system. At least two disorders recently have been associated with glucocorticoid resistance in the context of immune activation, including steroid resistant asthma and infection with the human immunodeficiency virus (Miller, Pearce, Ruzek, & Biron, In Press). In both of these cases, cells isolated from relevant patient populations have demonstrated decreased sensitivity to glucocorticoids in vitro, and in both cases, the findings either relate to or can be reproduced by specific cytokines. Work in our own laboratories have also demonstrated that viral infection is associated with changes in GR function (Miller, Spencer, Tanapat, Leung, Dhabhar, McEwen, & Biron, 1997).

110

A. H. Miller et al.

Complementary to the notion that immune activation may be involved in the glucocorticoid resistance found in patients with depression is the finding that certain patients with depression exhibit evidence of an activated immune response. Studies have reported elevated serum concentrations of the proinflammatory cytokine, IL-6, as well as increased acute phase proteins including haptoglobin, C-reactive protein and alpha-1-acid glycoprotein in patients with major depression (Maes, 1993; Maes, Meltzer, Bosmans, Bergmans, Vandoolaeghe, Ranjan, & Desnyder, 1995; Sluzewska, Rybakowski, Bosmans, Sobieska, Berghmans, Maes, & Wiktorowicz, 1996; Musselman, Pisell, Lewison, Pearce, Knight, Ninan, Nemeroff, & Miller, 1997). In addition, cellular markers of immune activation have been described (Maes, 1993). The source of the immune activation in major depression is unknown, although studies have shown that stress and CRH are capable of inducing proinflammatory cytokines in the absence of a formal immune challenge (Labeur, Arzt, Wiegers, Holsboer, & Reul, 1995; Schulte, Bamberger, Elsen, Herrmann, Bamberger, & Barth, 1994; Zhou, Kusnecov, Shurin, DePaoli, & Rabin 1993). Elevations in cytokines (especially proinflammatory cytokines like IL-6) have drawn much interest, because cytokines have been shown to have potent stimulatory effects on the HPA axis, in large part through activation of CRH pathways (Besedovsky & del Rey, 1996; Ericsson, Kovacs, & Sawchenko, 1994; Rivier, 1995). Moreover, proinflammatory cytokines have been shown to alter monoamine neurotransmission in multiple brain regions including the hypothalamus (Dunn & Wang, 1995). Finally, administration of cytokines or cytokine inducers have been shown to cause a syndrome of "sickness behavior" in humans and laboratory animals which shares many features in common with major depression including anhedonia, listlessness, altered sleep patterns, reduced appetite and social withdrawal (Kent et. al., 1992, Yirmiya, 1996). These findings support a role for cytokines in many of the pathophysiological features of depression. Nevertheless, taken together, the evidence of immune activation in depression in conjunction with the association between immune activation and glucocorticoid resistance, it is logical to consider that cytokines might also directly influence GR function leading to glucocorticoid resistance in major depression and thereby participate in the pathophysiology of the disorder.

3,1. Cytokine Effects on GR Expression and Function There is a large body of data on the impact of cytokines on the GR with over 20 studies documenting significant effects on both GR number and function (for review Miller, et al.. In Press). Interestingly, in terms of receptor number, the studies are polarized into those finding increased GR number following cytokine administration and those finding decreased GR number. Of note is that significant technical issues may be confounding the results, since 8 out of 12 studies using whole cell assay binding techniques found GR number to be increased (4 found no change) following treatment with a host of cytokines including IL-1, IL-2, and IL-4, whereas the majority of studies (6 out of 9) using a cytosolic radioligand binding assay found GR to be decreased (2 found no change) following treatment with the same group of cytokines. Based on findings of increased GR following cytokine treatment, some investigators have posited that relevant cells would be more sensitive to glucocorticoids, however, most studies examining the effects of cytokines (especially, pro-inflammatory cytokines and cytokines that mediate lymphocyte growth and differentiation, e.g., IL-2 and IL-4) on GR function have found inhibition of GR function. In fact, the inhibitory impact of

Glucocorticoid Resistance and Relevance to Depression

111

Table 2. Review of literature: cytokine effects on glucocorticoid receptors Glucocorticoid receptor number and affinity — 23 studies identified Receptor number — 10 studies found an increase in receptor number following treatment with a host of cytokines and cytokine inducers including TNF, IL-1, IL-6, IL-2, IL-4, IFN-alpha, LPS, and sepsis — 6 studies found a decrease in receptor number following treatment with a similar group of cytokines and cytokine inducers including TNF, IL-1, IL-6, IFN-beta, and endotoxin — 9 out of 10 studies that reported an increase in receptor number used a whole cell assay and tended to use longer incubation times — All 6 studies that reported a decrease in receptor number used a cytosohc radioligand binding assay and tended to use shorter incubation times Receptor affinity — 8 studies found a decrease in receptor affinity (increased Kd) following treatment with a host of cytokines and cytokine inducers including TNF, IL-1, IL-6, IL-2, IL-4, IFN-alpha, LPS/endotoxin, and sepsis Glucocorticoid receptor function — 8 studies identified — 5 out of eight demonstrated induction of relative glucocorticoid resistance by cytokines including TNF, ILl alpha and beta, IL-2, IL-4, and Transforming Growth Factor-beta — 2 studies reported increased glucocorticoid receptor mediated cellular activity following TNF or IL-1 beta

IL-2 and IL-4 on GR function is hypothesized to play a central role in steroid resistant asthma (Miller et al., In Press). To examine potential mechanisms of cytokine effects on GR function, work in our laboratory has focused on the impact of proinflammatory cytokines on GR translocation from cytoplasm to nucleus and hormone-induced, GR-mediated gene transcription. According to the "nucleocytoplasmic traffic" model, the GR in its unactivated form resides primarily in the cytoplasm, and after being bound by steroid undergoes a conformational change ("activation"), dissociates from a multimeric complex including several heat shock proteins (hsp), and translocates from the cytoplasm to the nucleus, where it binds to glucocorticoid response elements (GREs) and interacts with other transcription factors (Guiochon-Mantel, Delabre, Lescop, & Milgrom, 1996). We have evaluated GR translocation in L929 cells (mouse fibroblasts), treated with the proinflammatroy cytokine, IL-1 alpha, for 24 h in the presence or absence of the synthetic steroid, DEX (Pariante, Pearce, Pisell, & Miller, 1996). Our initial characterizations of cytokine-GR interactions were with IL-1 alpha, because much of the focus on neuroendocrine-immune interactions has been based on the neuroendocrine effects of IL-1, and recent work in our experimental murine viral infections has shown that IL-1 alpha in particular induces IL-6 to stimulate a surge in endogenous glucocorticoids during infection with murine cytomegalovirus (MCMV) (Ruzek, Miller, Opal, Pearce, & Biron, 1997). GR translocation was investigated using GR immunostaining and cytosolic radioligand receptor binding. In addition, GR-mediated gene transcription was measured by means of L929 cells stably transfected with the MMTV-chloramphenicol acetyltransferase (MMTV-CAT) reporter gene which is under hormonal control by virtue of several GREs residing within the MMTV-LTR region (upstream of the CAT reporter gene). IL-lalpha was found to both increase receptor protein expression in the cytoplasm, and block GR translocation from the cytoplasm to the nucleus in cells that were pretreated with IL-1 alpha. In addition, in cells that were either

112

A. H. Miller et al.

pretreated with IL-1 alpha and then incubated with DEX or coincubated with IL-1 alpha plus DEX for 24 hours, there was a significant reduction in DEX-induced GR-mediated gene transcription. These results suggest that proinflammatory cytokines like IL-1 may have direct effects on GR function which lead to glucocorticoid resistance. These results also might explain the mechanism underlying glucocorticoid resistance (dexamethasone nonsuppression) found in animals chronically treated with LPS (Weidenfeld & Yirmiya, 1996) and in patients with HIV infection (Norbiato, Bevilacqua, & Vago, 1997).

4. CONSEQUENCES OF GLUCOCORTICOID RESISTANCE In order to establish the potential relevance of glucocorticoid resistance to disease expression, it is important to establish the role of endogenous glucocorticoids in regulating cytokine and cellular responses during immune activation. To accomplish this task, we have recently examined the consequences of viral infection with MCMV in mice rendered glucocorticoid deficient by adrenalectomy. MCMV infection was chosen because this virus is associated with a surge in glucocorticoids which have been shown to peak at 36 hours after infection and reach serum values of up to 200ng/ml (Ruzek et al., 1997). Mice with cytokine-deficiencies or neutralized cytokine function have demonstrated that IL-6 is the pivotal mediator of the glucocorticoid response, with IL-1 (alpha) contributing to IL-6 production. Of note, the IL-6 requirement appears to be specific for virus-type stimuli as polyI:C, a non replicating immune stimulus which is to viral infection as LPS is to bacterial infection, also induced IL-6-dependent glucocorticoid release (Ruzek et al., 1997). Treatment with the bacterial product, LPS, and a physical restraint stressor elicited IL-6-independent responses. Mice were sham-operated or adrenalectomized (ADX) and treated with various doses of virus and were sacrificed at varying times after infection (Ruzek, Miller, Pearce, Spencer, & Biron, 1997). Removal of corticosterone resulted in sustained increases in serum levels of IFN-gamma, IL-12, IL-6, and TNF-alpha. In addition, removing endogenous corticosterone resulted in an increased susceptibility of infected mice to MCMV-induced disease such that the majority (80%) of mice died through a TNF-dependent mechanism within 150 hours of infection. Interestingly, the log viral titer in the liver of ADX mice was significantly lower than sham-operated, infected, control mice, indicating that increases in relevant cytokines were mediating stronger antiviral immunity. Cytokine-mediated death was reversed by administering "surge" levels of corticosterone in the drinking water. These results demonstrate that endogenous glucocorticoids regulate the magnitude of systemic cytokine responses to MCMV infection and protect the host from acute, virus-induced cytokine disease. Given the importance of endogenous glucocorticoids in protection against the toxic and ultimately lethal effects of cytokines during viral infection, it is logical to consider that the presence of changes in relative glucocorticoid sensitivity (such as in the case of glucocorticoid resistance) might in turn influence the relative ability of glucocorticoids to mediate negative feedback on immune activation and immunemediated toxicity. Therefore, in the context of the the previously noted, well-characterized glucocorticoid resistance in certain patients with depression (as reflected in nonsuppression on the dexamethasone suppression test and reduced sensitivity to in vitro glucocorticoid exposure) as well as relevant first degree relatives, it is possible that

Glucocorticoid Resistance and Relevance to Depression

113

increased levels of proinflammatory cytokines as induced by stress, CRH or some other source of inflammation might contribute to or in some cases cause this glucocorticoid resistance in depressed patients, thus leading to HPA axis hyperactivity in conjunction with cytokine stimulation of CRH pathways. Taken together, it is possible then that a feedforward cascade may occur in depression whereby increased CRH induces increased proinflammatory cytokines which in turn both further stimulate CRH release and induce glucocorticoid resistance (thus limiting feedback inhibition on further CRH and cytokine release). This inflammatory glucocorticoid cascade would lead to an accentuation of behavioral and other toxicities which would include sickness behavior.

5. TREATMENT CONSIDERATIONS Accompanying the idea that impairment in GR number and/or function as a consequence of genetic or environmental factors (e.g. immune processes) contributes to the pathophysiology of major depression, recent studies have suggested that a possible mechanism by which antidepressants treat depression is through effects on the GR. A number of animal studies have shown that in vivo treatment with a range of antidepressant agents including both tricyclic antidepressants and selective serotonin reuptake inhibitors is capable of enhancing glucocorticoid feedback inhibition and increasing GR protein and/or mRNA in key brain regions including the hippocampus (which has been shown to mediate an inhibitory influence on CRH in the paraventricular nucleus) (Holsboer & Barden, 1996). Antidepressants have also been found to facilitate glucocorticoid-mediated feedback inhibition and increase GR in animal models of HPA axis dysregulation including a transgenic model of impaired glucocorticoid receptor function (Pepin, Pothier, & Barden, 1992) and a rat model of early developmental stress (maternal deprivation) which leads to hypersecretion of CRH in response to stress in adulthood (Plotsky, P.M., personal communication). Moreover, antidepressant medications have been associated with the resolution of HPA axis alterations in patients with depression (Holsboer & Barden, 1996). Of note is that in a study by Rossby, Nalepa, Huang, Perrin, Burt, Schmidt, Gillespie, and Sulser (1995), the in vivo effects of the tricyclic antidepressant, desipramine (DMI), on GR were independent of noradrenergic reuptake blockade, suggesting that the effects of this antidepressant on GR may represent a direct action of this class of drugs. Consistent with this data are in vitro studies which have shown that DMI is capable of directly enhancing GR function as measured by increased activity of a reporter gene whose regulation is dependent on upstream glucocorticoid response elements (GRE) which bind the activated, GR (Pepin, Govindan, & Barden, 1992). Taken together, the data suggest that impaired GR function may contribute to the pathophysiology of major depression, and a potentially important mechanism of action of antidepressants may be to directly enhance GR function. In order to examine the potential mechanism of the effects of DMI on the GR, in conjunction with our studies on the impact of cytokines on GR function, we have also examined the influence of DMI on GR using a range of assessments of receptor protein including cytosolic receptor binding, immunocytochemistry (ICC), western blot analysis of the immunoprecipitated GR, and GR-mediated gene transcription as measured in stably transfected cell lines containing a MMTV-CAT reporter plasmid (Pariante, Pearce, Pisell, Owens, & Miller, 1997). All studies were performed on the mouse fibroblast cell line, L929, which has been used extensively to investigate GR

A. H. Miller et al.

114

function. Interestingly, we have found evidence that the mechanism by which DMI enhances GR function is through facihtation of the translocation of the GR from the cytoplasm to the nucleus (Pariante et al., 1997). 24 hour treatment with DMI (1-10|a,M) was found to translocate the GR from cytoplasm to nucleus in both the presence and absence of steroid as evidenced by decreased cytosolic receptor binding, and increased immunoreactivity of GR in the nucleus and decreased immunoreactive GR in the cytoplasm as determined by both ICC and western blots of immunoprecipated GR protein in cytosolic and nuclear fractions. Moreover, incubation of cells with a threshold dose of dexamethasone (10 nM) plus DMI led to an enhancement of dexamethasoneinduced GR translocation and a 2-3 fold increase in GR-mediated CAT activity. DMI alone had no effect on GR-mediated gene transcription in the absence of hormone (charcoal stripped serum). No evidence was found for upregulation of GR protein in incubations for up to 96 hours. These results indicate that DMI is capable of translocating the GR in the absence of ligand and in the presence of ligand facilitates GR effects on gene expression. The finding that antidepressants can facilitate GR function is intriguing and suggests a possible mechanism whereby pretreatment of animals with antidepressants is capable of reducing the relative toxicity of endotoxin administration (Yirmiya, 1996). We would suggest that antidepressants facilitate feedback inhibition pathways and thereby enhance negative feedback of endogenous glucocorticoids on cytokine responses.

Infection, Inflammation Cytokines

—ii

Glucocorticoidsri^Q f^

t

Antidepressants

I

*

n •

Inhibitory Stimulatory

IL-1

1^ Neurotoxicity (Sickness Behavior)

I

MAJOR DEPRESSION

CRF

^ Anxiety, Behavioral Despair

It

Stressors Figure 1. Role of Glucocorticoids and the Glucocorticoid Receptor in Mediating Feedback Inhibition on IL-1 and CRH: Potential Influence of Cytokines and Antidepressants. Stressors and immune activation secondary to inflammation and/or infection serve as potent stimuli to corticotropin releasing factor (CRH) and interleukin (IL)-l respectively. Overexpression of these factors in turn can lead to behavioral disorders whose symptoms overlap and fall under the rubric of major depression. Glucocorticoids via their receptors mediate feedback inhibition on CRH and IL-1 release, thus providing a brake on excessively exuberant responses. The relative functioning of the glucocorticoid receptor (GR), therefore serves as a critical determinant of sensitivity to feedback signals. Proinflammatory cytokines have been shown to inhibit GR function, and antidepressants have been shown to enhance GR function. Thus both of these factors serve to change relevant setpoints regarding glucocorticoid sensitivity as well as sensitivity to major depression via the pathways of stress and immune activation.

Glucocorticoid Resistance and Relevance to Depression

115

6. SUMMARY AND CONCLUSIONS Our data indicate that the proinflammatory cytokine, IL-lalpha inhibits GR translocation and hormone-induced GR-mediated gene transcription, and, in conjunction with previous in vivo and in vitro studies, can be interpreted to suggest that cytokines have the capacity to contribute to glucocorticoid resistance and thus the pathophysiology of depression. In addition, data from our mouse viral studies in glucocorticoid deficient animals demonstrate that endogenous glucocorticoids modulate a delicate balance between viral defense and cytokine toxicity. Finally, the antidepressant, DMI, has been found to enhance GR translocation and GR-mediated gene transcription and thus may provide a useful strategy for adjusting neuroendocrine setpoints in vivo. Taken together, these findings suggest that factors which modulate glucocorticoid action (e.g. cytokines and antidepressants) will be relevant contributors to disease expression including behavioral toxicity and sickness behavior.

REFERENCES Ader, R., Cohen, N., & Felten, D. (1991). Psychoneuroimmunology II. New York, Academic Press. Besedovsky, H. O. & del Rey, A. (1996). Immune-neuro-endocrine interactions: Facts and hypotheses. Endocrine Reviews, 17, 64-102. Dunn, A.J. & Wang. J. (1995). Cytokine effects on CNS biogenic amines. Neuruimmunomodulation, 2, 319-328. Ericsson, A., Kovacs, J.,& Sawchenko, P. E. (1994). A functional anatomical analysis of central pathways subserving the effects of interleukin-1 on stress-related neuroendocrine neurons. Journal of Neuroscience. I4,H97-9\3. Guiochon-Mantel, A., Delabre, K., Lescop, P., & Miigrom, E. (1996). Intracellular traffic of steroid hormone receptors. Journal of Steroid Biochemistry and Molecular Biology. 56, 3-9. Holsboer, F. & Barden, N. (1996). Antidepressants and hypothalamic-pituitary-adrenocortical regulation. Endocrine Reviews, 17,187-205. Kent. S., Bluthe, R. M., Kelley, K. W., & Dantzer, R. (1992). Sickness behavior as a new target for drug development. Trends in Pharmacological Sciences. 13, 24-28. Labeur, M. S., Arzt, E., Wiegers, G. J.. Holsboer, F, & Reul, J. M. H. M. (1995). Long-term intracerebroventricular corticotropin-reieasing hormone administration induces distinct changes in rat splenocyte activation and cytokine expression. Endocrinology. 136, 2678-2688. Maes. M. (1993). A review on the acute phase response in major depression. Reviews in the Neurosciences, 4,407-416. Maes. M.. Meltzer, H. Y., Bosmans, E.. Bergmans, R., Vandooiaeghe, E.. Ranjan, R., & Desnyder, R. (1995). Increased plasma concentrations of interleukin-6, soluble interleukin-6, soluble interleukin-2, and transferrin receptor in major depression. Journal of Affective Disorders. 34, 301-309. McEwen, B. S., Biron, C. A., Brunson, K. W., Bulloch, K., Chambers, W. H., Dhabhar, F S., Goldfarb. R. H., Kitson. R. P., Miller, A. H., Spencer, R. L., & Weiss, J. M. (1997). The role of adrenocorticoids as modulators of immune function in health and disease: neural, endocrine, and immune interactions. Brain Research Reviews, 23,19--\?>?>. Miller. A. H., Pearce. B. D.. Ruzek, M. C . & Biron. C. A. (In Press). Interactions between the hypothalamicpituitary-adrenal axis and immune system during viral infection: pathways for environmental effects on disease expression. In B. S. McEwen (Ed.). Handbook of Physiology. Miller. A. H. (1998). Neuroendocrine and immune system interactions in stress and depression. Psychiatric Clinics of North America. Philadelphia. W B. Saunders, 21. 443^63. Miller. A. H., Spencer, R. L.. Tanapat. P, Leung, J. J.. Dhabhar, F S.. McEwen. B. S.. & Biron. C. A. (1997). Effects of viral infection on corticosterone secretion and glucocorticoid receptor binding in immune tissues. Psychoneuroendocrinology, 22. 455-474. Musselman, D. L.. Pisell.T. L.. Lewison. B., Pearce, B. D., Knight, B.T., Ninan, PT., Nemeroff, C. B.. & Miller, A. H. (1997). lnterleukin-6 plasma concentrations in cancer patients with depression. Society for Neuroscience Abstract, 23, 996.

116

A. H. Miller et al.

Norbiato, G., Bevilacqua, M., & Vago, T. (1997). Glucocorticoids and the immune system in AIDS. Psychoneuroendocrinology, 22, S19-S25. Owens, M. J. & Nemeroff, C. B. (1993). The role of CRF in the pathophysiology of affective disorders: laboratory and clinical studies, Ciba Foundation Symposium on Corticotropin-Releasing Factor 172 (W. W. Vale, Ed.), New York, John Wiley & Sons, 293-316. Pariante, C. M., Pearce, B. D., Pisell, T. L., & Miller, A. H. (1996). Interleukin-1 alpha inhibits neucleocytoplasmic traffic of the glucocorticoid receptor. International Congress of the International Society for Neuroimmunomodulation Abstract. Pariante, C. M., Nemeroff, C. B., & Miller, A. H. (1995). Glucocorticoid receptors in depression. Israel Journal of Medical Science, 31, 705-712. Pariante, C. M., Pearce, B. D., Pisell, T. L., Owens, M. J., & Miller, A. H. (1997). Steroid-Independent translocation of the glucocorticoid receptor by the antidepressant desipramine. Molecular Pharmacology, 52, 571-581. Pepin, M.-C, Pothier, E, & Barden, N. (1992). Antidepressant drug action in a transgenic mouse model of the endocrine changes seen in depression. Molecular Pharmacology, 42, 991-995. Pepin, M.-C, Govindan, M. V., & Barden, N. (1992). Increased glucocorticoid receptor gene promoter activity after antidepressant treatment. Molecular Pharmacology, 41, 1016-1022. Ribeiro, S. C. M., Tandon, R., Grunhaus, L., & Greden, J. E (1993). The DST as a predictor of outcome in depression: a meta-analysis. American Journal of Psychiatry, 150, 1618-1629. Rivier, C. (1995). Influence of immune signals on the hypothalamic-pituitary axis of the rodent. Frontiers in Neuroendocrinology, 16, 151-182. Rossby, S. P, Nalepa, I., Huang, M., Perrin, C , Burt, A. M., Schmidt, D. E., Gillespie, D. D., & Sulser E (1995). Norepinephrine-independent regulation of GRII mRNA in vivo by a tricyclic antidepressant. Brain Research, 687, 79-82. Ruzek, M., Miller, A. H., Pearce, B. D., Spencer, R. L., & Biron, C. A. (1997). Evidence for endogenous glucocorticoid protection against MCMV-induced lethality. Society for Neuroscience Abstract, 23, 715. Ruzek, M. C , Miller, A. H., Opal, S. M., Pearce, B. D., & Biron, C. A. (1997). Characterization of early cytokine responses and an IL-6-dependent pathway of endogenous glucocorticoid induction during murine cytomegalovirus infection. Journal of Experimental Medicine, 185, 1185-1192. Schulte, H. M., Bamberger, C. M., Elsen, H., Herrmann, G., Bamberger, ••, & Barth, J. (1994). Systemic interleukin-a alpha and interleukin-2 secretion in response to acute stress and to corticotropinreleasing hormomone in humans. European Journal of Clinical Investigation, 24,111>-111. Sluzewska, A., Rybakowski, J., Bosmans, E., Sobieska, M., Berghmans, R., Maes, M., & Wiktorowicz, K. (1996). Indicators of immune activation in major depression. Psychiatry Research, 64, 161-167. Spencer, R. L., Miller, A. H., Stein, M., & McEwen, B. S. (1991). Corticosterone regulation of type I and type II adrenal steroid receptors in brain, pituitary, and immune tissue. Brain Research, 549, 236-246. Weidenfeld, J. & Yirmiya, R. (1996). Effects of bacterial endotoxin on the glucocorticoid feedback regulation of adrenocortical response to stress. Neuroimmunomodulation, 3, 352-357. Yirmiya, R. (1996). Endotoxin produces a depressive-like episode in rats. Brain Research, 711,163-174. Zhou, D., Kusnecov, A. W., Shurin, M. R., DePaoli, M., & Rabin, B, S. (1993). Exposure to physical and psychological stressors elevates plasma interleukin 6: Relationship to the activation of the hypothalamicpituitary-adrenal axis. Endocrinology, 133, 2523-2530.

8

EFFECTS OF CYTOKINES ON CEREBRAL NEUROTRANSMISSION Comparison with the Effects of Stress

Adrian J. Dunn,' Jianping Wang,^ and Tetsuya Ando^ ' Department of Pharmacology and Therapeutics, Louisiana State University Medical Center, Shreveport, Louisiana ^Present address: Inflammatory Joint Diseases Section, NIAMS/NIH Building 10, Room 9N228, Bethesda, Maryland 20892-1820 ^ Present address: Division of Psychosomatic Research National Institute of Mental Health, National Center of Neurology and Psychiatry 1-7-3 Konodai, Ichikawa, 272 Japan

1. INTRODUCTION 1.1. Responses Associated with Stress Stress is normally associated with coactivation of the sympathoadrenal system (sympathetic nervous system plus the adrenal medulla) and the hypothalamo-pituitaryadrenocortical (HPA) axis. However, extensive work in the past 30 years has indicated that responses also occur in the central nervous system. The major response occurs in noradrenergic (NA) neurons. Most studies have also noted responses in dopaminergic (DA) and serotonergic (5-HT) systems (Dunn & Kramarcy, 1984; Stone, 1975). Whether or not adrenergic (adrenaline-containing) neurons respond is not resolved, although there is evidence that adrenergic neurons (along with NA and 5-HT neurons) are involved in the regulation of hypothalamic corticotropin-releasing factor (CRF) secretion which initiates HPA activation (Plotsky, Cunningham, & Widmaier, 1989). The NA response is widespread and appears to affect to similar extents both the locus coeruleus (A6) system innervating the dorsal structures (cortex, hippocampus, cerebellum, etc.), and the nucleus tractus solitarius (A1/A2) system innervating the ventral structures (e.g., the hypothalamus). The DA response is also widespread with all the major neuronal systems (nigrostriatal, mesolimbic, mesocortical) showing responses, but the magnitude of the response is particularly large in the mesocortical system (i.e. in the Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

117

118

A. J. Dunn et al

prefrontal and cingulate cortices). The 5-HT response is not markedly regionally specific (although some (e.g., Kirby, Allen, & Lucki, 1995) have reported regional differences). There is also a robust elevation of concentrations of tryptophan (the natural precursor of 5-HT) in all regions of the brain. This is quite uniform in magnitude, and does not appear to be related in any obvious way to the extent of the serotonergic innervation of a region (Curzon, Joseph, & Knott, 1972; Dunn, 1988a). There may be changes in the metabolism and secretion of many other neurotransmitters (e.g., acetylcholine, and y-aminobutyric acid (GABA), see Dunn & Kramarcy, 1984), but the evidence that these are a specific or universal response in stress is less compelling than for the catecholamines and 5-HT. There are also responses in certain peptidergic systems. CRF secretion is critical in the activation of the HPA axis (Vale, Spiess, Rivier, & Rivier, 1981), and some extrahypothalamic CRF neurons are activated too (Merlo Pich, Lorang, Yeganeh, Rodriguez De Fonseca, Raber, Koob, & Weiss, 1995). A number of other peptides are also affected in stress.

1.2. Neurochemical Responses to Infection Before discussing the responses to cytokines, it is useful first to compare the neurochemical and physiological responses to infections and to endotoxin (LPS) with those described above for more commonly studied stressors. When we studied the effects of influenza virus infection in mice, we observed a progressive activation of the HPA axis (Dunn, Powell, Meitin, & Small, 1989). This observation was consistent with scattered reports in the literature, and with our own previous observations of animals that became sick during experiments from a variety of causes (frequently unknown). It is important that this HPA activation was continuous and not transient as it is to stressors such as electric shock. Secondly, we observed an increased metabolism of NA (i.e. production of 3-methoxy,4-hydroxyphenylethyleneglycol, MHPG) in all brain regions studied. But, unlike the response to footshock or restraint, the magnitude of the response was significantly greater in the hypothalamus (Dunn et al., 1989), and appeared to be associated with a preferential activation of the ventral noradrenergic system which originates largely from nucleus tractus solitarius, although it receives a contribution from the locus coeruleus. DA metabolism was not significantly affected. Tryptophan concentrations and 5-HT metabolism (i.e. production of 5-hydroxyindoleacetic acid, 5-HIAA) were elevated as observed following footshock or restraint. The earliest time at which these changes appear is around 24 h after infection. Subsequent studies have shown similar changes associated with infection with other viruses and bacteria (Beisel, 1981; Ben Hur, Rosenthal, Itzik, & Weidenfeld, 1996; Guo, Qian, Peters, & Liu, 1993; Kass & Finland, 1958; Miller, Spencer, Pearce, Pisell, Tanapat, Leung, Dhabhar, McEwen, & Biron, 1997; Weidenfeld, Wohlman, & Gallily, 1995). We conclude that there is a marked similarity between the neurochemical and physiological responses to physical and psychological stressors, and infections or other illnesses as depicted in Table l.The responses are not identical; the major differences being that infections and illness are associated with larger NA responses in the hypothalamus than other brain regions, and a lack of DA responses.

1.3. Neurochemical Responses to Endotoxin (Lipopolysaccharide, LPS) It has long been known that administration of LPS activates the HPA axis (Bliss, Migeon, Eik-Nes, Sandberg, & Samuels, 1954). In our studies, intraperitoneal (ip) injec-

Effects of Cytokines on Cerebral Neurotransmission

119

Table 1. Physiological responses to physical stressors (shock or restraint) and viral infection HPA axis Sympathetic nervous system Adrenal medulla Brain NA Brain DA Brain 5-HT Brain tryptophan

Shock or restraint

Viral infection

Tt TT TT TT T T TT

TT T T TT 0

T TT

T Indicates an increase in activity; TT indicates a larger increase; 0 indicates no change.

tion of mice with low doses of LPS (1 |Lig per mouse) induces a rapid elevation of plasma ACTH and corticosterone reaching a peak at around 2h. NA metabolism is also activated (Pohorecky, Wurtman, Taam, & Fine, 1972), and like the response to influenza virus, the response was greatest in the hypothalamus (Dunn, 1992a), suggesting a relatively greater activation of the ventral than the dorsal NA system. In contrast to the effects of influenza virus, there were small effects on DA activity, as indicated by production of DOPAC.TTie peak responses of both catecholamines (DA and NA) occurred around 2h. In contrast to the responses to electric shock and restraint, the increases in DOPAC were not particularly regionally selective. The increases in the prefrontal cortex were less marked and were similar in magnitude to those found in the hypothalamus and other brain regions (Dunn, 1992a). Tryptophan and 5-HIAA were also increased in a regionally nonspecific manner. However, these latter changes reached a peak much later at around 8h (Dunn, 1992a). Very similar changes were observed following intracerebroventricular (icv) LPS and the effective doses were similar (Dunn, 1992a). Changes in NA and 5-HT consistent with these have been described by a number of other authors (Delrue, Deleplanque, Rouge-Pont, Vitiello, & Neveu, 1994; Linthorst, Flackskamm, Holsboer, & Reul, 1996; Molina-Holgado & Guaza, 1996). Thus administration of LPS induces a pattern of neuroendocrine and neurochemical responses rather similar to those described above for infection with influenza virus, albeit with a more rapid onset.

2. NEUROCHEMICAL RESPONSES TO CYTOKINES 2.1. Neurochemical Responses to Interleukin-l (IL-1) 2.1.1. Catecholamines and Indoleamines. With the above as background, we studied the responses to peripheral administration of IL-1, because it had been shown that this cytokine was a potent stimulator of the HPA axis (Besedovsky, del Rey, Sorkin, & Dinarello, 1986). IL-1 administration to mice markedly increased brain NA metabolism, closely parallelling in dose and time the HPA activation, indicated by elevation of plasma corticosterone (Dunn, 1988b). The pattern of activation resembled that we had observed previously with influenza virus infection and LPS; a greater response in the hypothalamus than in the cortex and other brain regions, and there was no response in DA metabolism. Furthermore, IL-1 induced a regionally nonspecific increase in tryp-

120

A. J. Dunn et al

tophan concentrations, as well as increases in 5-HIAA (Dunn, 1988b). Thus IL-1 mimicked quite closely the response to influenza virus infection, except that the IL-1induced responses dissipated within a few hours. These results were subsequently replicated in rats (Kabiersch, del Rey, Honegger, & Besedovsky, 1988). It was also shown in rats that decreases in hypothalamic NA occurred after ip IL-1 (Fleshner, Goehler, Hermann, Relton, Maier, & Watkins, 1995). Depletions of NA are more readily obtained in rats than in mice (Stone, 1975 and Dunn, unpublished observations). Zalcman et al. also observed IL-1-induced increases in MHPGrNA and 5-HIAA:5HT ratios in the prefrontal cortex and hippocampus, but also found increases in mouse prefrontal cortex DOPAC (Zalcman, Green-Johnson, Murray, Nance, Dyck, Anisman, & Greenberg, 1994). In more than 100 experiments with ip IL-1 administration to mice, we have observed statistically significant increases of DOPAC in only a very few experiments. Subsequent work has shown that IL-1 elicits a modest sympathoadrenal activation (Berkenbosch, De Goeij, del Rey, & Besedovsky, 1989). We have obtained evidence that the increases in tryptophan (in response to footshock, restraint, IL-1, and LPS) may be dependent upon peripheral sympathetic activity, because they can be blocked by pretreatment with the ganglionic blocker chlorisondamine, and largely prevented by P-adrenergic, but not by muscarinic, receptor antagonists (Dunn & Welch, 1991). Thus, the increases in brain tryptophan may reflect sympathetic activation. The increased tryptophan can drive and may be essential for the observed increases in 5HIAA. Administration of tryptophan (50mg/kg ip), results in small increases in brain tryptophan and 5-HIAA (Dunn, unpublished observations), although it is not clear that the increased 5-HIAA reflects increased synaptic release. Also, the chlorisondamine and p-adrenoreceptor antagonist treatments that prevent the footshock-, restraint, IL1-, and LPS-induced increases in tryptophan, also prevent the increases in 5-HIAA (Dunn & Welch, 1991). All these effects of IL-1 are characteristic of various forms of IL-1 administered by various routes. We have observed rather similar responses to different samples of recombinant human or mouse IL-1 a and human or mouse IL-ip from different sources, including natural IL-1 purified from our own animals (Dunn, 1988b; Dunn, 1992a). The neurochemical responses are observed in response to ip or subcutaneous injections, and to icv administration at considerably lower doses (Dunn, 1992a). Intravenous administration has lesser neurochemical effects (Dunn & Chuluyan, 1992). These effects are not due to contamination with endotoxin, (which, as described above, has very similar effects to IL-1) because when the IL-1 is heated to 100°C for 10 minutes, all the activities are lost. Subsequent studies have indicated that the effect of IL-1 on NA most likely reflects increased synaptic release, because in vivo microdialysis studies have indicated that IL-1 elicits increased extracellular concentrations of NA in the medial hypothalamus in the region of the paraventricular nucleus (Ishizuka, Ishida, Kunitake, Kato, Hanamori, Mitsuyama, & Kannan, 1997; MohanKumar & Quadri, 1993; Smagin, Swiergiel, & Dunn, 1996). It has been reported that icv hIL-ip administration increases tyrosine hydroxylase activity in the median eminence of rats (Abreu, Llorente, Hernandez, & Gonzalez, 1994). The NA and tryptophan/5-HT responses appear to be independent. The NA response to ip IL-1 and LPS peaks at around 2 hours along with the increases in plasma ACTH and corticosterone; whereas the peak responses in tryptophan and 5-HIAA appear later (IL-1—4 hours, LPS—8 hours). Stronger evidence for the dissociation was derived from studies with endotoxin-resistant (C3H/HeJ) mice which exhibit

Effects of Cytokines on Cerebral Neurotransmission

121

very small HPA responses to LPS and no changes in MHPG, while the increases in tryptophan and 5-HIAA were similar to those in a control (C3H/HeN) strain (Dunn & Chuluyan, 1994). The HPA and neurochemical responses to IL-1 were similar in both endotoxin-resistant and normal strains. Furthermore, pretreatment with nitric oxide synthase (NOS) inhibitors prevents the tryptophan and 5-HIAA response, with no detectable effects on the HPA or NA responses (Dunn, 1993). Thus the NA and indoleamine responses can be independently manipulated. Recent experiments with selective NOS inhibitors suggest that the inducible form of NOS (iNOS) is involved in this response. 2.7.7. Responses of Other Neurotransmitters. There are a few reports of effects of IL-1 administration on other neurotransmitters. Ip administration of hIL-ip decreased the secretion of acetylcholine from the hippocampus, as measured by microdialysis in freely moving rats (Rada, Mark, Vitek, Mangano, Blume, Beer, & Hoebel, 1991), but only at high doses (20 or 50)xg/kg), 7.5|ig/kg was ineffective. Also, it was found that ip injection of hIL-ip (20fig/kg, but not 10|ig/kg) decreased hippocampal concentrations of glutamate, glutamine and GABA one hour later (Bianchi, Ferrario, Zonta, & Panerai, 1995). Most authors have found 3-5|ig/kg IL-1 to induce maximal HPA activation in the rat, so the physiological significance of these effects at substantially higher doses is not clear.

2.2. Neurochemical Responses to Interleukm-2 (IL-2) Zalcman et al. (1994) reported that IL-2 (200 ng) increased MHPG concentrations and MHPG: NE ratios in the hypothalamus of BALB/c mice 2 hours after ip injection. DOPAC:DA ratios were also increased in prefrontal cortex, but plasma corticosterone was not affected. Bianchi et al. (1995) reported small increases in glutamine in the cortex and hippocampus of mice one hour after ip injection of 5|ig/kg hIL-2.

2.3. Neurochemical Responses to Interleukin-6 (IL-6) Many researchers have reported that IL-6 administration activates the HPA axis as indicated by elevations of plasma ACTH and corticosterone. In a recent systematic study, we tested recombinant mouse IL-6 injected iv and ip into mice. Injection of mIL-6 by either route elicited modest increases in plasma ACTH and corticosterone; maximal concentrations were much lower than observed with IL-1 (Wang & Dunn, 1998). The corticosterone elevations were maximal at 30-60 minutes and were no longer evident by 2 hours. In earlier studies, we did not observe effects of IL-6 on cerebral biogenic amines (Dunn, 1992b), consistent with other reports (Terao, Oikawa, & Saito, 1993). However, Zalcman et al. (1994) reported that hIL-6 injected into mice (200 ng, ip) elevated 5-HIAA in hippocampus, and 5-HIAA and DOPAC in prefrontal cortex. In our recent study, we found mIL-6 (iv or ip) consistently elevated tryptophan in all brain regions at around 2 hours, and 5-HIAA in the brain stem at the same time, but had no effect on MHPG or DOPAC in any brain region (Wang & Dunn, 1998). lev mIL-6 had similar endocrine and neurochemical effects. Because IL-1 and LPS administration both stimulate the synthesis and secretion of IL-6, it is possible that IL-6 is the mediator of the indoleamine responses to IL-1 and LPS.

122

A. J. Dunn et al.

2.4. Neurochemical Responses to Tumour Necrosis Factor a (TNFa) Human tumour necrosis factor a (hTNFa) activates the HPA axis in man (Michie, Spriggs, Manogue, Sherman, Revhaug, O'Dwyer, Arthur, Dinarello, Cerami, Wolff, Kufe, & Wilmore, 1988) and rodents (Besedovsky, del Rey, Kinsman, Furukawa, Monge Arditi, & Kabiersch, 1991; Perlstein, Whitnall, Abrams, Mougey, & Neta, 1993; Sharp & Matta, 1993). TNFa was significantly less potent than IL-1 in rats (Besedovsky et al., 1991; del Rey & Besedovsky, 1992) and mice (Dunn, 1992b), although it has also been reported that hTNFa was almost equipotent with hIL-l|3 in rats (Sharp, Matta, Peterson, Newton, Chao, & McAllen, 1989). We found human TNFa to be ineffective in elevating plasma corticosterone in mice, except when it was injected iv (Dunn, 1992b). However, 0.1 and 1 }ig mouse TNFa ip elicited modest increases in plasma corticosterone. In a recent study, we examined the response of mice to iv and ip mTNFa. Injection of mTNFa by either route caused a weak activation of the HPA axis; 1 |ig was necessary to produce an elevation of plasma corticosterone, and the response was shortlived, like that to IL-6 (Ando & Dunn, 1998). TNFa increased brain MHPG and tryptophan, but only at the higher doses (Ifxg or more) so this result is not inconsistent with our earlier studies (Dunn, 1992b) or that of Terao et al. (1993). In contrast, Elenkov, Kovacs, Duda, Stark, & Vizi, (1992) reported that TNFa inhibited NE release from the median eminence, and Hurst & Collins (1994) reported that TNFa inhibited NE release from the rat myenteric plexus.

2.5. Neurochemical Responses to Interferons We did not observe significant response in plasma corticosterone or cerebral biogenic amines to human interferon-a (hlFNa: 1000 or 10000 Units ip, Dunn, 1992b). In a recent collaborative study, we tested the responses to mouse IFNa administration at doses and times that were behaviourally active (Crnic & Segall, 1992). We observed no significant changes in plasma corticosterone, or catecholamine or serotonin metabolism in any of five brain regions 2 hours following ip administration of 400,800 or 1600 Units of mIFNa (Leenhouwers, Crnic, & Dunn, unpublished observations).

3. DISCUSSION 3.1. The Role of Cytokines in Responses to LPS and Infection Because the responses to IL-1 resemble so closely those to influenza virus infection and LPS, an obvious question is the extent to which this cytokine accounts for the responses. Surprisingly, studies with antibodies to IL-1 and the endogenously synthesised IL-1-receptor antagonist (IL-lra) suggest that IL-1 is not the only factor involved. In the case of LPS, it is important that whereas LPS is a potent stimulator of IL-1 production, little appears in less than one hour and the peak response in the plasma is around 2 hours. Therefore, if IL-1 were the mediator, the endocrine and neurochemical responses to LPS should be delayed more than is actually observed (Dunn, 1992a). Studies with antibodies to IL-1 and IL-lra failed to show blockade of either the increases in plasma ACTH and corticosterone, or those in MHPG or tryptophan (Dunn, 1992b). More recent studies using higher doses of IL-lra produced similar results, although there was a trend towards an attenuation of the HPA activation at 3 ^

Effects of Cytokines on Cerebral Neurotransmission

123

hours (Dunn & Brown, 1996), as would be expected from the time considerations indicated above. These results are consistent with observations on the responses to LPS in mice that lack the gene for IL-ip (Fantuzzi, Zheng, Faggioni, Benigni, Ghezzi, Sipe, Shaw, & Dinarello, 1996). Treatment with a neutralising antibody to TNFa also failed to prevent the HPA and neurochemical responses to LPS, even when supplemented with IL-lra (Dunn, 1992b). Studies with antibodies to IL-6 have suggested a role for IL-6 in the HPA responses to both the administration of LPS (Perlstein et al., 1993; Wang & Dunn, 1996) and IL-1 (Wang & Dunn, 1996). In our experiments, these effects were confined to the later phases of the HPA response, consistent with the delay necessary for the production and secretion of IL-6. However, pretreatment with a neutralising monoclonal antibody to mouse IL-6 also attenuated the increases in tryptophan and 5-HIAA following LPS administration to mice, but not that to IL-1, suggesting that IL-6 only contributes to the indoleamine response to LPS (Wang & Dunn, 1996).

3.2. The Functional Significance of the Neurochemical Responses to Cytokines An important question is whether any of the neurochemical responses observed following cytokine administration can be linked to a functional response. The most useful functional responses in this respect are the HPA responses, and various behavioural responses. 3.2.1. Involvement in the HPA Responses. Both NA and 5-HT have been implicated in HPA activation (Plotsky et al., 1989), and thus are obvious candidates. As described above, the NA response and that of plasma ACTH and corticosterone are closely linked in time, and in our experiments, have been highly correlated over a large number of experiments involving a large number of different manipulations. A microdialysis study in freely moving rats also indicated a very close temporal relationship between extracellular concentrations of NA in the hypothalamus and plasma concentrations of corticosterone following both iv and ip IL-ip (Smagin et al., 1996). When the neurotoxin 6-hydroxydopamine (6-OHDA) was injected into the VNAB or the PVN of rats, resulting in depletions of PVN NA of 75% or more, the plasma corticosterone response to ip IL-1 was largely prevented (Chuluyan, Saphier, Rohn, & Dunn, 1992). A similar result was obtained using icv IL-1 in 6-OHDA-treated rats (Weidenfeld, Abramsky, & Ovadia, 1989). Surprisingly however, when whole brain NA was depleted by around 98% using 6-OHDA in mice, no impairment of the plasma corticosterone response to IL-1 (3 was observed (Dunn, Swiergiel, & Stone, 1996, and other unpublished observations). 3.2.2. Involvement in the Behavioural Responses. The other functional correlate we have studied is IL-1-induced hypophagia. Many authors have demonstrated IL-1induced reductions of food intake, including ourselves (Swiergiel, Smagin, & Dunn, 1997). This cytokine-induced hypophagia is sensitive to cyclo-oxygenase inhibitors. Because, the peak of this hypophagia occurs around 2h, we tested the ability of adrenergic receptor antagonists to reverse the hypophagia. However, neither a p , a2-, nor nonspecific a- or (3-adrenergic receptor antagonists, even in combination induced significant reductions in IL-1-induced hypophagia in mice (Dunn & Swiergiel, 1997). Likewise, pretreatment with 6-OHDA or DSP-4 to deplete cerebral NA failed to alter

124

A. J. Dunn et al.

the hypophagic response to IL-1 (Dunn & Swiergiel, 1997). Surprisingly, we have also failed to find any effects of cerebral 5-HT depletion (with 5,7-dihydroxytryptamine) or pretreatment with a variety of 5-HT-receptor antagonists. However, studies of other cytokine-induced behaviours has produced evidence imphcating noradrenergic mechanisms. For example, Ovadia, Abramsky, and Weidenfeld (1989) found that pretreatment with 6-OHDA, the p-adrenoreceptor blocker, propranolol, or the tta-adrenoreceptor blocker, yohimbine, prevented IL-1-induced fever. Also, Bianchi and Panerai (1995) reported that 6-OHDA pretreatment or prazosin could prevent the antinociceptive effect of icv hIL-la determined in the hot-plate test.

4. SUMMARY Table 2 summarizes the reported responses of the HPA axis, as well as catecholamines and indoleamines to the cytokines discussed above. Cytokine administration to animals can elicit a number of effects on the brain, including neuroendocrine and behavioural effects, and also alters the metabolism of neurotransmitters. The most well documented effect is the activation by interleukin1 (IL-1) of the hypothalamo-pituitary-adrenocortical (HPA) axis, which is accompanied by a stimulation of cerebral noradrenaline (NA) metabolism, probably reflecting increased NA secretion. IL-1 also stimulates indoleamine metabolism, most prominently increasing tryptophan concentrations, and increasing the metabolism of serotonin (5-hydroxytryptamine, 5-HT). IL-6 induces a short-lived activation of the HPA axis, and has effects on tryptophan and 5-HT similar to those of IL-1. Tumour necrosis factor a (TNFa) has effects on the HPA axis similar to those of IL-6, but affects NA and tryptophan only at high doses. Interferon a had no effects on the parameters studied. The effects of IL-1 are remarkably similar to those observed following administration of endotoxin (lipopolysaccharide, LPS), and infections, such as influenza virus. They also resemble quite closely the responses that are observed to stressors commonly studied in laboratory animals, such as electric shock or restraint. The major differences are: that the NA response to shock or restraint is very uniform throughout the brain, whereas that to IL-1, LPS or infection is significantly greater in the hypothalamus; and, responses in dopaminergic (DA) systems are normally observed to shock or restraint, with especially prominent responses in the limbic cortex, whereas DA responses are rarely observed in response to IL-1 and immune stimuli, and when they do occur, the mesocortical system is not selectively affected.

Table 2. HPA and brain neurochemical responses to infection, LPS, and cytokines Stimulus Influenza virus LPS IL-la/IL-ip IL-2 IL-6 TNFa IFNa

Corticosterone

NE

DA

Tryptophan

5-HT

+ + + 0 + + 0

+ + + + 0

0 + 0 + 0 0 0

+ + + ND +

+ + + 0 + 0 0

(+) 0

(+) 0

+ indicates an increase; ++ indicates a larger increase; 0 indicates no change; (+) indicates increases only at the highest dose of TNFa (1 ng or more), and therefore may not be physiologically relevant; ND not determined.

Effects of Cytokines on Cerebral Neurotransmission

125

The neurochemical responses to cytokines may underlie some of the endocrine and behavioural responses. The NA response to IL-1 is apparently related to the HPA activation, but not the hypophagia. The significance of the indoleaminergic responses is not known.

ACKNOWLEDGMENTS The author's research described in this review was supported by the Office of Naval Research (N0001-4-85K-0300), the U.S. National Institute of Mental Health (MH46261) and the U.S. National Institute of Neurological Diseases and Stroke (NS25370). We are grateful for the excellent technical assistance of Bunney Powell, Sandra Vickers, Michael Adamkiewicz, and Lynn Pittman-Cooley, and the clerical assistance of Rhonda Hiers, Sharon Farrar, and Karen Scott.

REFERENCES Abreu, P., Llorente, E., Hernandez, M. M., & Gonzalez, M. C. (1994). Interleukin-ip stimulates tyrosine hydroxylase activity in the median eminence. NeuroReport, 5,1356-1358. Ando,T. & Dunn, A. J. (1998). Mouse tumor necrosis factor a increases brain tryptophan concentrations and norepinephrine metabolism while activating the HPA axis in mice. Neuroimmunomodulation, 5, in press. Beisel, W. R. (1981). Alterations in hormone production and utilization during infection. In M. C. Powanda & P. G. Canonico (Eds.), Infections: The Physiologic and Metabolic Responses of the Host (pp. 147-172). Elsevier/North-Holland Biomedical Press. Ben Hur, T., Rosenthal, J., Itzik, A., & Weidenfeld, J. (1996). Adrenocortical activation by herpes virus: involvement of IL-ip and central noradrenergic system. NeuroReport, 7, 927-931. Berkenbosch, F., De Goeij, D. E. C , del Rey, A., & Besedovsky, H. O. (1989). Neuroendocrine, sympathetic and metabolic responses induced by interleukin-l. Neuroendocrinology, 50, 570-576. Besedovsky, H. O., del Rey, A., Klusman, I., Furukawa, H., Monge Arditi, G., & Kabiersch, A. (1991). Cytokines as modulators of the hypothalamus-pituitary-adrenal axis. Journal of Steroid Biochemistry and Molecular Biology, 40, 613-618. Besedovsky, H. O., del Rey, A., Sorkin, E., & Dinarello, C. A. (1986). Immunoregulatory feedback between interleukin-l and glucocorticoid hormones. Science, 233, 652-654. Bianchi, M., Ferrario, P., Zonta, N., & Panerai, A. E. (1995). Effects of interleukin-lP and interleukin-2 on amino acids levels in mouse cortex and hippocampus. NeuroReport, 6, 1689-1692. Bianchi, M. & Panerai, A. E. (1995). CRH and the noradrenergic system mediate the antinociceptive effects of central interleukin-l a in the rat. Brain Research Bulletin, 36,113-117. Bliss, E. L., Migeon, C. J., Eik-Nes, K., Sandberg, A. A., & Samuels, L. T. (1954). The effects of insulin, histamine, bacterial pyrogen, and the antabuse-alcohol reaction upon the levels of 17-hydroxycorticosteroids in the peripheral blood of man. Metabolism, 3, 493-501. Chuluyan, H., Saphier, D., Rohn, W. M., & Dunn, A. J. (1992). Noradrenergic innervation of the hypothalamus participates in the adrenocortical responses to interleukin-l. Neuroendocrinology, 56,106-111. Crnic, L. S. & Segall, M. A. (1992). Behavioral effects of mouse interferon-a and interferon-y and human interferon-a in mice. Brain Research, 590, 277-284. Curzon, G., Joseph, M. H., & Knott, P. J. (1972). Effects of immobilization and food deprivation on rat brain tryptophan metabolism. Journal of Neurochemistry, 19,1967-1974. del Rey, A. & Besedovsky, H. O. (1992). Metabolic and neuroendocrine effects of pro-inflammatory cytokines. European Journal of Clinical Investigation, 22, 10-15. Delrue, C , Deleplanque, B., Rouge-Pont, F., Vitiello, S., & Neveu, P. J. (1994). Brain monoaminergic, neuroendocrine, and immune responses to an immune challenge in relation to brain and behavioral lateralization. Brain Behavior and Immunity, 8,137-152. Dunn, A. J. (1988a). Changes in plasma and brain tryptophan and brain serotonin and 5-hydroxyindoleacetic acid after footshock stress. Life Sciences, 42,1847-1853.

126

A. J. Dunn et al.

Dunn, A. J. (1988b). Systemic interleukin-1 administration stimulates hypothalamic norepinephrine metabohsm parallelling the increased plasma corticosterone. Life Sciences, 43,429-435. Dunn, A. J. (1992a). Endotoxin-induced activation of cerebral catecholamine and serotonin metabolism: comparison with interleukin-1. Journal of Pharmacology and Experimental Therapeutics, 261, 964-969. Dunn, A. J. (1992b). The role of interleukin-1 and tumor necrosis factor a in the neurochemical and neuroendocrine responses to endotoxin. Brain Research Bulletin, 29, 807-812. Dunn, A. J. (1993). Nitric oxide synthase inhibitors prevent the cerebral tryptophan and serotonergic responses to endotoxin and interleukin-1. Neuroscience Research Communications, 13, 149-156. Dunn, A. J. & Brown, R. F. (1996). The interleukin-1-receptor antagonist fails to alter the neurochemical and HPA responses to LPS. Society for Neuroscience Abstracts, 22, 1456. Dunn, A. J. & Chuluyan, H. (1992).The role of cyclo-oxygenase and Hpoxygenase in the interleukin-1-induced activation of the HPA axis: dependence on the route of injection. Life Sciences, 51, 219-225. Dunn, A. J. & Chuluyan, H. E. (1994). Endotoxin elicits normal tryptophan and indolamine responses but impaired catecholamine and pituitary-adrenal responses in endotoxin-resistant mice. Life Sciences, 54, 847-853. Dunn, A. J. & Kramarcy, N. R. (1984). Neurochemical responses in stress: relationships between the hypothalamic-pituitary-adrenal and catecholamine systems. In L. L. Iversen, S. D. Iversen, & S. H. Snyder (Eds.), Handbook of Psychopharmacology (pp. 455-515). New York: Plenum Press. Dunn, A. J., Powell, M. L., Meitin, C , & Small, P. A. (1989). Virus infection as a stressor: influenza virus elevates plasma concentrations of corticosterone, and brain concentrations of MHPG and tryptophan. Physiology and Behavior, 45, 591-594. Dunn, A. J. & Swiergiel, A. H. (1997). Cytokines and behavior: The role of biogenic amines in hypophagia. Neuroimmunomodulation, 4, 213. Dunn, A. J., Swiergiel, A. H., & Stone, E. A. (1996). The role of cerebral noradrenergic systems in the Fos response to interleukin-1. Brain Research Bulletin, 41, 61-64. Dunn, A. J. & Welch, J. (1991). Stress- and endotoxin-induced increases in brain tryptophan and serotonin metabolism depend on sympathetic nervous system activation. Journal of Neurochemistry, 57, 1615-1622. Elenkov, I. J., Kovacs, K., Duda, E., Stark, E., & Vizi, E. S. (1992). Presynaptic inhibitory effect of TNF-a on the release of noradrenaline in isolated median eminence.. Journal of Neuroimmunology, 413,117-120. Fantuzzi, G., Zheng, H., Faggioni, R., Benigni, F, Ghezzi, P, Sipe, J. D., Shaw, A. R., & Dinarello, C. A. (1996). Effect of endotoxin in IL-ip-deficient mice. Journal of Immunology, 157, 291-296. Fleshner, M., Goehler, L. E., Hermann, J., Relton, J. K., Maier, S. F, & Watkins, L. R. (1995). Interleukin-1 p induced corticosterone elevation and hypothalamic NE depletion is vagally mediated. Brain Research Bulletin, 37, 605-610. Guo, Z. M., Qian, C. G., Peters, C. J., & Liu, C.T. (1993). Changes in platelet-activating factor, catecholamine, and serotonin concentrations in brain, cerebrospinal fluid, and plasma of Pichinde virus-infected guinea pigs. Laboratory Animal Science, 43, 569-574. Hurst, S. M. & Collins, S. M. (1994). Mechanism underlying tumor necrosis factor-a suppression of norepinephrine release from rat myenteric plexus. American Journal of Physiology, 266, Gl 123-1129. Ishizuka, Y., Ishida, Y., Kunitake, T, Kato, K., Hanamori, T, Mitsuyama, Y., & Kannan, H. (1997). Effects of area postrema lesion and vagotomy on interleukin-1 P-induced norepinephrine release in the hypothalamic paraventricular nucleus region in the rat. Neuroscience Letters, 223, 57-60. Kabiersch, A., del Rey, A., Honegger, C. G., & Besedovsky, H. O. (1988). Interleukin-1 induces changes in norepinephrine metabolism in the rat brain. Brain Behavior and Immunity, 2, 267-274. Kass, E. H. & Finland, M. (1958). Corticosteroids and infections. 9, 45-80. Kirby, L. G., Allen, A. R.,& Lucki, I. (1995). Regional differences in the effects of forced swimming on extracellular levels of 5-hydroxytryptamine and 5-hydroxyindoleacetic acid. Brain Research, 682,189-196. Linthorst, A. C. E., Flackskamm, C , Holsboer, F , & Reul, J. M. H. M. (1996). Activation of serotonergic and noradrenergic neurotransmission in the rat hippocampus after peripheral administration of bacterial endotoxin: involvement of the cyclo-oxygenase pathway. Neuroscience, 72, 989-997. Merlo Pich, E., Lorang, M., Yeganeh, M., Rodriguez De Fonseca, F, Raber, J., Koob, G. F, & Weiss, F (1995). Increase of extracellular corticotropin-releasing factor-like immunoreactivity levels in the amygdala of awake rats during restraint stress and ethanol withdrawal as measured by microdialysis. Journal of Neuroscience, 15, 5439-5447. Michie, H. R., Spriggs, D. R., Manogue, K. R., Sherman, M. L., Revhaug, A., O'Dwyer, S. T, Arthur, K., Dinarello, C. A., Cerami, A., Wolff, S. M., Kufe, D. W, & Wilmore, D. W. (1988). Tumor necrosis factor and endotoxin induce similar metabolic responses in human beings. Surgery, 104, 280-286. Miller, A. H., Spencer, R. L., Pearce, B. D., Pisell, T. L., Tanapat, P, Leung, J. J., Dhabhar, F. S., McEwen,

Effects of Cytokines on Cerebral Neurotransmission

127

B. S., & Biron, C. A. (1997). Effects of viral infection on corticosterone secretion and glucocorticoid receptor binding in immune tissues. Psychoneuroendocrinology, 22, 455-474. MohanKumar, P. S. & Quadri, S. K. (1993). Systemic administration of interleukin-1 stimulates norepinephrine release in the paraventricular nucleus. Life Sciences, 52,1961-1967. Molina-Holgado, F. & Guaza, C. (1996). Endotoxin administration induced differential neurochemical activation of the rat brain stem nuclei. Brain Research Bulletin, 40, 151-156. Ovadia, H., Abramsky, O., & Weidenfeld, J. (1989). Evidence for the involvement of the central adrenergic system in the febrile response induced by interleukin-1 in rats. Journal of Neuroimmunology, 25, 109-116. Perlstein, R. S., Whitnall, M. H., Abrams, J. S., Mougey, E. H., & Neta, R. (1993). Synergistic roles of interleukin-6, interleukin-1, and tumor necrosis factor in the adrenocorticotropin response to bacterial lipopolysaccharide in vivo. Endocrinology, 132, 946-952. Plotsky, P. M., Cunningham, E. T., & Widmaier, E. P. (1989). Catecholaminergic modulation of corticotropinreleasing factor and adrenocorticotropin secretion. Endocrine Reviews, 10, 437^58. Pohorecky, L. A., Wurtman, R. J., Taam, D., & Fine, J. (1972). Effects of endotoxin on monoamine metabolism in the rat. Proceedings of the Society for Experimental Biology and Medicine, 140,139-1 Ab. Rada, P, Mark, G. P, Vitek, M. R, Mangano, R. M., Blume, A. J.. Beer, B., & Hoebel, B. G. (1991). Interleukin1(3 decreases acetylcholine measured by microdialysis in the hippocampus of freely moving rats. Brain Research, 550, 287-290. Sharp, B. M. & Malta, S. G. (1993). Prostaglandins mediate the adrenocorticotropin response to tumor necrosis factor in rats. Endocrinology, 132, 269-274. Sharp, B. M., Malta, S. G., Peterson, P K., Newton, R., Chao, C , & McAllen, K. (1989). Tumor necrosis factora is a potent ACTH secretagogue: comparison to interleukin-1 (3. Endocrinology, 124. 3131-3133. Smagin, G. N., Swiergiel, A. H., & Dunn, A. J. (1996). Peripheral administration of interleukin-1 increases extracellular concentrations of norepinephrine in rat hypothalamus: comparison with plasma corticosterone. Psychoneuroendocrinology, 21, 83-93. Stone, E. A. (1975). Stress and catecholamines. In A. J. Friedhoff (Ed.), Catecholamines and Behavior. Vol. 2 Neuropsychopharmacology (pp. 31-72). New York: Plenum Press. Swiergiel, A. H., Smagin, G. N., & Dunn, A. J. (1997). Influenza virus infection of mice induces anorexia: comparison with endotoxin and interleukin-1 and the effects of indomethacin. Pharmacology Biochemistry and Behavior, 57. 389-396. Terao, A., Oikawa, M., & Saito, M. (1993). Cytokine-induced change in hypothalamic norepinephrine turnover: involvement of corticotropin-releasing hormone and prostaglandins. Brain Research, 622, 257-261. Vale, W.. Spiess, J., Rivier, C , & Rivier, J. (1981). Characterization of a 41-residue ovine hypothalamic peptide that stimulates secretion of corticotropin and p-endorphin. Science, 213, 1394-1397. Wang, J. P. & Dunn, A. J. (1996). The role of interleukin-6 in the activation of the hypothalamo-pituitaryadrenocortical axis induced by endotoxin and IL-1. Society for Neuroscience Abstracts, 22, 1354. Wang, J. P. & Dunn, A. J. (1998). Mouse interleukin-6 stimulates the HPA axis and increases brain tryptophan and serotonin metabolism. Neurochemistry International, 33, 143-154. Weidenfeld, J., Abramsky, O., & Ovadia, H. (1989). Evidence for the involvement of the central adrenergic system in interleukin 1-induced adrenocortical response. Neuropharmacology, 28, 1411-1414. Weidenfeld, J., Wohlman, A., & Gallily, R. (1995). Mycoplasma fermentans activates the hypothalamopituitary adrenal axis in the rat. NeuroReport. 6. 910-912. Zalcman, S., Green-Johnson, J. M., Murray, L., Nance, D. M., Dyck, D., Anisman, H., & Greenberg, A. H. (1994). Cytokine-specific central monoamine alterations induced by interleukin-1, -2, and -6. Brain Research. 643, 40-49.

INFLAMMATION AND BRAIN FUNCTION UNDER BASAL CONDITIONS AND DURING LONG-TERM ELEVATION OF BRAIN CORTICOTROPIN-RELEASING HORMONE LEVELS Astrid C. E. Linthorst* and Johannes M. H. M. Reul Max Planck Institute of Psychiatry, Section Neuropsychopharmacology Kraepelinstrasse 2-10, D-80804 Munich, Germany

1. OVERVIEW It is now well accepted that immune challenges elicit a variety of physiological processes, such as activation of the hypothalamic-pituitary-adrenocortical (HPA) axis, fever, and sickness behaviour, to promote survival and return of homeostasis. The underlying neurotransmitter responses in the brain, however, still need to be further characterized. Using an in vivo microdialysis method in rats, we have shown that peripheral inflammation (induced by intraperitoneal (i.p.) injection of endotoxin (lipopolysaccharide; LPS)) results in a highly differentiated serotonergic and noradrenergic neurotransmitter response in the brain. LPS caused a profound increase in extracellular levels of serotonin (5-HT) in the hippocampus, but not in the preoptic area. In contrast, this treatment induced a dramatic rise in extracellular levels of noradrenaline (NA) in the preoptic area, but had only moderate effects in the hippocampus. Based on studies using biotelemetry and observation of behavioural activity simultaneously with microdialysis, we have proposed that the increase in preoptic NA is involved in fever and/or HPA axis activation during inflammation. The rise in hippocampal 5-HT seems to be associated with the development of sickness behaviour. Disturbances in the functioning of the immune system as described in major

* Phone: ++49 89 30622 589; Fax: ++49 89 30622 371 or ++49 89 30622 569; E-mail: [email protected] Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

129

130

A. C. E. Linthorst and J. M. H. M. Reul

preoptic NA is involved in fever and/or HPA axis activation during inflammation. Tiie rise in hippocampal 5-HT seems to be associated with the development of sickness behaviour. Disturbances in the functioning of the immune system as described in major depression, triggered our interest to study neuroimmunoendocrine interactions during pathophysiological conditions. Because one of the characteristics of major depression is an elevated central drive of corticotropin-releasing hormone (CRH), and this neuropeptide is highly involved in physiological responses to stress, we developed an animal model of CRH hyperactivity, i.e. the long-term intracerebroventricularly (i.e.v.) CRH-infused rat. On day seven of the i.e.v. treatment, rats were i.p. injected with LPS. Whereas LPS caused a profound rise in body temperature in control animals, only a blunted fever response developed in long-term CRH-treated rats. Moreover, LPS produced an attenuated hippocampal 5-HT response and a delayed onset of HPA axis activation and behavioural inhibition in CRH-treated rats. These attenuated brainmediated responses were not caused by suppression of the release of plasma cytokines, because CRH-infused animals showed enhanced LPS-induced plasma bioactivities of interleukin (IL)-l and IL-6 (but not of tumor necrosis factor (TNF)). These data indicate that inflammation induces highly distinct neurotransmitter responses in the brain that might be responsible for the initiation of specific physiological responses to cope with this challenge. In addition, a chronically elevated CRH drive results in aberrant brain-mediated responses to an acute inflammatory challenge; a situation which, in the long run, may represent a putatively detrimental threat for health.

2. INTRODUCTION During the past decade it has become increasingly clear that multi-level interactions exist between the immune system and the central nervous system. The brain is no longer considered to be an immune-privileged organ. Moreover, evidence has accumulated showing that the immune system is not an autonomously operating entity, but that its function is regulated by influences from the central nervous system (Felten & Felten, 1991). When confronted with an infectious agent or inflammatory stimulus, the organism responds with a wide variety of defence mechanisms. As a first step, the immune system is activated, resulting in, among other things, the secretion of cytokines into the blood circulation. Stimulation of the immune system subsequently leads to a rise in the levels of circulating glucocorticoids, which accelerate the course of the immune response and prevent it from overshooting (Munck, Guyre, & Holbrook, 1984; Wiegers & Reul, 1998; Wiegers, Labeur, Stec, Klinkert, Holsboer, & Reul, 1995). Moderate to severe systemic infectious and inflammatory states are accompanied by physiological and behavioural changes such as fever, loss of appetite, fatigue, and social disinterest. The behavioural changes observed during activation of the immune system are now collectively termed "sickness behaviour" (Dantzer, Bluthe, Kent, & Kelley, 1991; Hart, 1988), which is thought to be an important behavioural mechanism to help overcoming the infection or inflammation. Although the understanding of brain-mediated physiological, endocrine, and behavioural responses to immune challenges has increased notably during the last

Inflammation and Brain Function

131

years, comprehensive research on the neurochemical mechanisms underlying these phenomena is still needed. Our research group devotes part of its work to the characterization of neurotransmitter systems responsive to stimulation of the immune system. Moreover, we are attempting to elucidate functional implications of the activation of specific neurotransmitter pathways during immune system challenges. Experiments in rats were first conducted under basal conditions. Immune system changes observed in major depression (Irwin, Lacher, & Caldwell, 1992; Maes, 1995; Sluzewska, Rybakowski, Bosmans, Sobieska, Berghmans, Maes, & Wiktorowicz, 1996), however, triggered our interest to study neuro-immune interactions also during pathophysiological states. One of the characteristic features of major depression is an increased central drive of CRH (for references see Holsboer & Barden, 1996). Because CRH plays a key role in the coordination of neuroendocrine, autonomic and behavioural responses to stress (for references see Vale, Vaughan, & Perrin, 1997) and is implicated in the regulation of immune system functioning (Irwin, Vale, & Britton, 1987), we developed a rat model with long-term elevated levels of CRH in the brain. In this review, we describe the effects of peripheral inflammation on brain neurotransmitter systems with the emphasis on 5-HT and NA.The role of specific cytokines in inflammation-induced neurotransmitter changes will be discussed, on the basis of the results of our own experiments as well as data provided by other research groups. Moreover, the consequences of long-term elevation of CRH for immune system functioning, brain-mediated physiological and behavioural responses, and underlying neurochemical changes are described. 3. P E R I P H E R A L I N F L A M M A T I O N C A U S E S H I G H L Y DIFFERENTIATED N E U R O T R A N S M I T T E R RESPONSES IN THE BRAIN Although conclusive evidence has accumulated showing bidirectional communication between the immune system and the brain, the underlying neurochemical mechanisms have been hardly characterized. Until now, most studies have assessed concentrations of neurotransmitters and their metabolites in brain tissue samples as an index of turnover. Unfortunately, such studies provide only limited information on the extracellular (effective) concentrations of neurotransmitters and the time resolution of changes. Therefore, we have used an in vivo microdialysis technique in rats, which allows simultaneous assessment of various neurotransmitters and metabolites, as well as continuous monitoring of neurotransmitters over extended time periods (hours and days). In addition, we have developed a method to measure corticosterone levels in dialysates, providing an index of HPA axis activity. Because the extracellular space of the brain is devoid of corticosterone binding globulin, dialysate concentrations of corticosterone represent the free (biological active) fraction of this glucocorticoid. To elucidate the functional significance of inflammation-induced neurotransmission changes, in vivo microdialysis was combined with simultaneous measurement of body temperature and locomotor activity using a biotelemetry method. Moreover, behavioural activity (locomotion, rearing, grooming, eating, drinking, resting) was monitored by visual observation of the animals, and HPA axis activity was determined by the, already mentioned, measurement of corticosterone in dialysates (Figure 1). Because peripheral inflammation induces, apart from immune, neuroendocrine,

A. C. E. Linthorst and J. M. H. M. Reul

132

in vivo mlcrodlalysia comMnad with ptiyvioroglcal and behavioural moasuroments HPLC

540" 5-HIAA HA MHPG

M

ehfamMnmw

RIA CortkMtwona

Sehavkiural actlvFty scoring saminumplirKi: 1: &- 2 min activi S: 2-15 mln active 3; > 15 min activ*

Blotalamatry C

j

' boitr tBinp«tiir» >h«wtr«t» 'lacofTititlon

locanwUon, graonilns, Htlnti, drinking, M c

Figure 1. Schematic overview of in vivo microdialysis combined with the measurement of other physiological and behavioural parameters. Microdialysates are collected for the determination of extracellular levels of 5-HT and NA, and their metabolites by a HPLC method, and free corticosterone by a radioimmunoassay (RIA). Body temperature, heart rate, and locomotion can be assessed by means of biotelemetry (using an i.p. implanted transmitter and a receiver underneath the cage). Visual observation of the animals is used to determine the effects of various treatments on specific aspects of behaviour.

and autonomic responses, also changes in behaviour (i.e., vigilance state and affect), we were interested in putative neurotransmitter alterations in higher limbic brain structures. Hence, experiments were focused on the hippocampus, which plays a key role in behavioural responses to stress. To be able to demonstrate regionally distinct patterns of neurotransmitter activation in the brain during inflammation, we compared responses in the hippocampus with those in the preoptic area, a brain structure known to be involved in autonomic (body temperature) and endocrine regulation. The results of our studies on the effects of i.p. administration of bacterial endotoxin on 5-HT and NA, two major neurotransmitter systems in both the hippocampus and the preoptic area, are described in the following sections and compared with data available from the literature. Moreover, the possible role of brain cytokines in these effects will be discussed.

3.1. Serotonin 3.1.1. Serotonin and the Hippocampus 3.1.1.1. Peripheral immune stimulation and hippocampal serotonergic neurotransmission. Stimulation of the immune system by i.p. administration of LPS (30-300 |xg/kg body weight) induced a dose-dependent increase in hippocampal extracellular levels of 5-HT (maximum between 200-250% of baseline) and its metabolite 5hydroxyindoleacetic acid (5-HIAA; maximum about 150% of baseline) (Linthorst, Flachskamm, Miiller-Preuss, Holsboer, & Reul, 1995b; Linthorst, Flachskamm, Holsboer, & Reul, 1996) (Figure 2A,B). The doses of LPS used in this and our subsequent studies (100|ig/kg body weight) induced fever and a clear reduction in behavioural

133

Inflammation and Brain Function

-

OE-il

0£ZI

-^ ^ O

-*o . 0E'9l

-<

OG'Sl

-Ko«

OC't't

-m -•

OE El

I \ -A '0

, OE'Zl

I

/

-•<

0

\ -<

«

1/

OE'Ol

«

(M

1-

.

OEZl

•«-o

< <

&-

5 o ^

08 U

X

cr- .^•

0£'60

91

i

•. o 0)

OE'80 I

.

1

.

£ 'c ^ < -^ £ i: rt o

. oeoi

u

- ^ o< V « 1

oe El -Z i:,

. >

t\Z\\

.1 i .i_ _ 1

- ^ 00 - a

•\ •< ? 0

i

1-

1

OEfrt

i

\ 1 < -0

\/ -%

o a. o a a.

0£'9L

\ o

/

It •o

to

•S -S a

-^ -o

^

•n

oe 91

-o

1

1

I

.

I

OG'80

_1

,

in

. OC-60

L

.

oc •

I V

T-

— 73 XI 1^ O „

( 3 U ! | 9 s e q j o =/,) V V | H ' 9

(aunaseq jo %) I H - 5

• a v^ ^ o

—

JJ C3\

I- G. r-l

^

O

g a

^ 'u «

2 6^ i3

• " OS

o .y ^ ^—- T3

OEiL

•%

OG'SL

-p

-<

-^\

•<

OE^I?!-

«

orci ;:

'0

*< ^

o o

OGfrl

o

oeei

J OE'U

X

0E"Ol

/

CO

oe'60

OE'Dl

o. £

1

.

1

,

r

.

J

OEEO

.

1 ^

(dunasBq i o %) VVIH-S

a s e q ^o %) I H - S

CQ

x_-

t^S

o c s?

:= _ o ^

I c^ i : o- g< .S E

¥ m

o Q.

DE'SO

. OE H

w 0< M -^

F

-a lu

2 « 2

oesi E

\\

CO D.

C

t/5 "O

•X

^n

o

&0 - ^

o

O/

\

E

00 *^

i

-•<

\ -A < <

^ ^ OGfil

O

-•<

0£-3L

> o oe9i

-O

V\

oesi.

o

^ -S ^

. OE i l

E

O c/5

oe'80

Z <

W

(IN

03

OH

-J

134

A. C. E. Linthorst and J. M. H. M, Reul

activity of the animals together with other signs of sickness behaviour, such as a curledup body posture and piloerection. Moreover, the activity of the HPA axis was stimulated as indicated by a marked and prolonged rise in free corticosterone concentrations (Linthorst, Flachskamm, Holsboer, & Reul, 1995a; Linthorst et al, 1995b). Extracellular levels of 5-HIAA represent an index of the amount of newly synthesized 5-HT, and do not reflect metabolism of released 5-HT (Auerbach, Minzenberg, & Wilkinson, 1989; Crespi, Garratt, Sleight, & Marsden, 1990; Kalen, Strecker, Rosengren, & Bjorklund, 1988; Linthorst, Flachskamm, Holsboer, & Reul, 1994). Hence, the rise in 5-HIAA concentrations shows that inflammation results in an enhanced synthesis of this indoleamine. This is of particular interest, because preliminary results from 5-HTIA receptor autoradiography experiments have shown that i.p. administration of LPS caused a long-term increase in the concentration of these receptors in the hippocampus (Reul et al., unpublished observations). Taken together, these results indicate that the rise in extracellular concentrations of the parent amine 5-HT together with the stimulation of its synthesis and the upregulation of 5-HTIA receptors may ensure a profound and extended activation of 5-HT-regulated postsynaptic events in the hippocampus. Recent microdialysis studies carried out by Merali and colleagues (Merali, Lacosta, & Anisman, 1997) have shown that i.p. injection of IL-1 also increased extracellular levels of 5-HT and 5-HIAA in the hippocampus. Earlier studies on brain (tissue) neurotransmitter levels in rats and mice have demonstrated that systemic application of IL-1 (Kabiersch, Del Rey, Honegger, & Besedovsky, 1988; Zalcman, GreenJohnson, Murray, Nance, Dyck, Anisman, & Greenberg, 1994) and IL-6 (Zalcman et al., 1994) increased the turnover of 5-HT (as indicated by a rise in the ratio of 5HIAA/5-HT) and/or increase 5-HIAA levels. These findings indicate that systemically produced IL-1 can affect 5-HT neurotransmission; a mechanism which may be involved in the effects of LPS on hippocampal serotonergic neurotransmission as found in our experiments. Stimulatory effects of inflammation on higher brain structures are not confined to the hippocampus. Studies on brain neurotransmitter (tissue) levels have shown that peripheral injection of endotoxin, IL-1, and IL-6 enhanced 5-HT turnover in the frontal cortex (Dunn, 1992; Dunn & Welch, 1991; Zalcman et al., 1994). Moreover, Lavicky and Dunn (1995) reported increased levels of 5-HIAA in dialysates from the frontal cortex after i.p. administration of LPS. Unfortunately, no information on the parent amine was provided in this study. It is now well accepted that arachidonic acid metabolites play an important role in the actions of cytokines released after an immune challenge (Gottschall, Komaki, & Arimura, 1992; Roth well, 1991). Systemic administration of LPS induced the release of prostaglandin E2 in the hypothalamus (Sirko, Bishai, & Coceani, 1989; Smith, Hewson, Quarrie, Leonard, & Cuzner, 1994). Moreover, IL-lp increased levels of prostaglandin E2 in rat astrocyte cultures and rat brain (Katsuura, Gottschall, Dahl, & Arimura, 1989; Komaki, Arimura, & Koves, 1992). The cyclo-oxygenase inhibitor indomethacin suppressed the production of prostaglandins and inhibited the IL-1-induced release of ACTH (Katsuura, Gottschall, Dahl, & Arimura, 1988). Prostaglandins also play an important role in neurotransmission responses during inflammation. Pretreatment of rats with indomethacin (lOmg/kg body weight i.p.) largely but not completely blocked the LPS-induced increase in hippocampal extracellular levels of 5-HT and 5-HIAA (Linthorst et al., 1996). The indomethacin pretreatment, however, was able to prevent the fever, tachycardia and sickness behaviour responses entirely. The incomplete

Inflammation and Brain Function

135

inhibition of hippocampal 5-HT by indomethacin suggests the involvement of other mediators. Nitric oxide might represent a candidate because results of various studies point to a role of nitric oxide in brain responses to inflammation (Karanth, Lyson, & McCann, 1993; Raber & Bloom, 1994; Rivier & Shen, 1994). 3.1.1.2. Central cytokines and hippocampal serotonergic neurotransmission. As described above, LPS causes profound rises in extracellular concentrations of 5-HT and 5-HIAA, and in the levels of 5 - H T ] A receptors in various regions of the hippocampus. Although LPS induces the synthesis and release of many (proinflammatory) cytokines, which cytokines are involved in the LPS-induced brain-mediated responses needs to be clarified. Therefore, a series of studies was designed to delineate the role of cytokines in hippocampal serotonergic neurotransmission. First, it was examined to what extent singly administered cytokines would be able to mimic the effects of i.p. injected LPS. Two proinflammatory cytokines were tested, i.e. TNF-a and IL-ip, and compared with the actions of the non-inflammatory cytokine IL-2. Biotelemetry measurements showed that i.e.v. administration of these cytokines increased body temperature and free corticosterone levels (Kluger, 1991; Linthorst et al., 1995b; Pauli, Linthorst, & Reul, 1998). However, whereas IL-lp and IL-2 induced sickness behaviour and profound rises in hippocampal levels of 5-HT and 5-HIAA, TNF-a had no effects on these parameters (using doses effective in inducing fever and HPA axis activation; Figure 3A-C) (Linthorst et al., 1995b; Pauli et al., 1998). A recent study on brain neurotransmitter (tissue) levels showed that i.e.v. administration of IL-ip increased hippocampal 5-HIAA levels in rats, whereas IL-2, IL-6, and TNF-a were ineffective (Connor, Song, Leonard, Merali, & Anisman, 1998). The discrepancy between the latter report and our observation that i.e.v. IL-2 stimulates hippocampal 5-HT and 5-HIAA, may not only be explained by methodological differences (brain tissue levels vs. microdialysis), but also by the early time point of killing of the rats (45min after administration of IL-2) in the study of Connor et al. Hence, it may be concluded that both central IL-ip and IL-2, but not TNF-a, are able to stimulate hippocampal serotonergic neurotransmission. IL-2 is an important cytokine within the immune system where it acts as the main T-cell growth factor. IL-2 stimulates the release of various cytokines from immune cells, among which IL-1 (Janssen, Mulder, The, & Deleij, 1994). Therefore, we were interested to determine whether IL-2 exerts its effects on the level of the brain also via the induction of IL-1. For this purpose, rats were pretreated i.c.v. with the IL-1 receptor antagonist (IL-lra). Pretreatment with IL-lra completely blocked the i.c.v. IL-2induced increases in hippocampal 5-HT and 5-HIAA (Figure 3C), but only attenuated the rise in free corticosterone (Pauli et al., 1998).These results clearly demonstrate that the effects of IL-2 on hippocampal serotonin are caused via the release of IL-1, whereas IL-1 plays only a minor role in the IL-2-induced activation of the HPA axis. The finding that central administration of IL-ip has marked effects on hippocampal 5-HT has fed the notion that IL-1 may be involved in the LPS-induced changes in the levels of this neurotransmitter. Indeed, i.c.v. pretreatment of rats with IL-lra (lOjig) resulted in a 50% reduction of the LPS-induced activation of hippocampal 5-HT release (Figure 3D), pointing to a role of brain IL-1 in the effects of LPS on 5-HT in the hippocampus (Linthorst et al, 1995b). Moreover, we have demonstrated that local infusion of IL-ip into the hippocampus also resulted in elevations of extracellular 5-HT (Linthorst et al., 1994). Based on the observations that on the one hand i.c.v. and intrahippocampal IL-ip mimic the effects of LPS on hippocampal

4(ID

A

i.c.v. IL-1[3

* 1

^250 It * 200 a a -150 ^100

O i.c.v. vehicle 1 O

O

O

A i,c.v. IL-1B

O

O

D

O

O

O

O

rtoioi-nrivui^rK

time {hr)

B

, j.c.v. TNF-alpha ^2S0 c

1 •

©aoo ^150

jl^ioo w

50

O

i.C.v. vehicle

A

i.c.v. TNF

J n> «

rt »

A

rt D

i

en -

T

^^ i

4

TO n

4

M U

w ]

i

a

rt r

<

rt -

tliTifl (hr)

300

. i,C.v. IL-2 *

c •Saoo D

- I S O

"^100

r ^

O J.cv, vehicle A i,c.v. JL-2 + i.c.v. IL-Ira + IL-2

* 1

50 0

time (hr) 300

. i.cv. IL-1 ra + .p. LPS *1

£ »aoo
A - ISO

o

^100 X "* 50

: Snf^^fP a l.p- LPS

•

1

1

1

t

• 1

'

time (hr)

i.c.v, IL-1ra + i.p. LPS 1

1

1

1

Figure 3. Effects of i.c.v. administration of vehicle, IL-lp (100ng),TNF-a (100ng), or IL2 (500ng) (A-C) on hippocampal extracellular levels of 5-HT. Intracerebroventricular pretreatment of the rats with the interleukin1 receptor antagonist (IL-lra; 10|j.g) inhibits the i.c.v. IL-2- and attenuates the i.p. LPSinduced increase in hippocampal 5-HT (C, D). The time point of injection(s) is indicated by the arrow. *indicates the time period during which data from i.c.v. vehicle- and cytokinetreated (A, C) and from i.p. LPS- and i.c.v. IL-lra/i.p.LPS-treated animals (D) differ significantly (MANOVA with repeated measures). Data taken from Linthorst et al., 1995b and Pauli et al., 1998 by permission of the Journal of Neuroscience and the European Neuroscience Association, respectively.

Inflammation and Brain Function

137

serotonergic neurotransmission, and on the other hand i.c.v. IL-lra attenuates the LPS-induced effects on hippocampal 5-HT, it may be concluded that IL-lp released/produced locally in the hippocampus is involved in the activation of 5-HT in this brain structure during peripheral inflammation, 3.1.2. Serotonin and the Preoptic Area 3.1.2.1. Peripheral immune stimulation and preoptic serotonergic neurotransmission. To compare the effects of LPS on serotonergic neurotransmission in the hippocampus with those in another brain structure, we have performed similar studies in the preoptic area. The preoptic area (often studied as an entity with the anterior hypothalamus) plays a key role in the regulation of autonomic and endocrine processes, such as body temperature, food intake, sleep, and sexual behaviour (Blatteis, 1990; Rivest & Rivier, 1993; Swanson, 1987). In contrast to the increase in hippocampal 5-HT, i.p. administration of LPS had no effects on preoptic extracellular levels of 5-HT (Figure 2C). However, about two hours after the injection of LPS a marginal rise in extracellular concentrations of 5-HIAA rise was observed (maximum about 120% of baseline; Figure 2D) (Linthorst et al., 1995a). Of interest in this respect is the finding of Wilkinson et al. (Wilkinson, Auerbach, & Jacobs, 1991), who have showed, using in vivo microdialysis in cats, that systemic administration of the pyrogen muramyl dipeptide had no effect on preoptic 5-HT. Other studies have focused on serotonergic neurotransmission in various other parts of the hypothalamus. The general picture arising from these studies is that systemic immune stimulation and administration of IL-1 increase the 5-HIAA/5-HT ratio (Dunn, 1992; Mefford & Heyes, 1990) and dialysate levels of 5-HIAA (Lavicky & Dunn, 1995). 3.1.2.2. Central cytokines and preoptic serotonergic neurotransmission. Information in the literature on the effects of central administration of cytokines on preoptic serotonergic neurotransmission is too limited to deduct definite conclusions. Data from two reports, however, suggest that IL-1 may play a role in the effect of LPS on 5HIAA as observed in our experiments. Intracerebroventricular and local application of IL-1 increased preoptic 5-HIAA levels as assessed by in vivo voltammetry and microdialysis, respectively (Gemma, Ghezzi, & De Simoni, 1991; Shintani, Kanba, Nakaki, Nibuya, Kinoshita, Suzuki, Yagi, Kato, & Asai, 1993). Two reports have assessed the effects of central administration of cytokines on the hypothalamus. However, whereas one group, using an in vivo push-pull method, demonstrated that local application of IL-1 enhanced 5-HIAA levels (Mohankumar, Thyargarajan, & Quadri, 1993), i.c.v. injection of IL-1 and TNF-a was recently reported to decrease tissue levels of 5-HIAA in the hypothalamus (Connor et al., 1998).

3.2. Noradrenaline 3.2.1. Noradrenaline and the Hippocampus. Intraperitoneal injection of LPS (100 |ig/kg) moderately increased hippocampal extracellular levels of NA (maximum about 150% of baseline; Figure 4A) and its major metabolite MHPG (maximum about 170% of baseline; Figure 4B) (Linthorst et al., 1996). Pretreatment of the animals with indomethacin almost completely blocked the LPS-induced increases in NA and MHPG (Linthorst et al., 1996), pointing to an important role of prostaglandins. The incomplete inhibition, however, shows that, apart from prostaglandins, other mediators released during inflammation are involved in the stimulation of hippocampal noradrenergic

138

A. C. E. Linthorst and J. M. H. M. Reul

neurotransmission (see 3.1.1.1.). Studies of Zalcman et al. suggest that IL-1, released peripherally during inflammation, may play a key role in the LPS-induced effects on hippocampal NA. Intraperitoneal administration of IL-1 (3, but not of IL-2 or IL-6, increased the ratio MHPG/NA in hippocampal tissue of mice (Zalcman et al., 1994). Unfortunately, there is hardly any information available on the effects of centrally administered cytokines on noradrenergic neurotransmission in the hippocampus. Recently, it has been described that i.e.v. injection of IL-1, IL-2, IL-6, and TNF-a exerts no effects on NA in this brain structure (Connor et al., 1998). Interestingly, changes in noradrenergic neurotransmission during inflammation have also been found in the prefrontal cortex. Using brain tissue samples of mice and in vivo microdialysis in rats, Dunn and colleagues showed that peripheral administration of LPS enhanced the ratio of MHPG/NA and extracellular levels of NA and MHPG in the prefrontal cortex, respectively (Dunn, 1992; Lavicky & Dunn, 1995). An involvement of IL-1 released into the circulation during inflammation is still controversial, since increases in the ratio MHPG/NA (Dunn, 1992) as well as no effects (Zalcman et a l , 1994) after i.p. injection of IL-1 have been reported in the literature. 3.2.2. Noradrenaline and the Preoptic Area. Whereas peripherally administered LPS has a moderate effect on hippocampal NA, this treatment resulted in dramatic increases in extracellular NA (about 500% of basehne; Figure 4C) and MHPG (about 400% of baseline; Figure 4D) in the preoptic area (Linthorst et al., 1995a). NA and MHPG concentrations remained elevated for at least 6 hours. Prostaglandins serve as important mediators in these effects, because pretreatment with indomethacin largely prevented LPS-induced rises in preoptic extracellular concentrations of NA and MHPG. Similar as found for NA and 5-HT in the hippocampus, the incomplete inhibition of preoptic NA by indomethacin suggests an involvement of additional mediators. Unfortunately, only one other research group has studied the effects of peripheral inflammation on NA in the preoptic area. Terao and colleagues showed that i.p. administration of IL-1 increased the turnover of NA in this brain area (Terao, Oikawa, & Saito, 1993). Shintani and colleagues (1993) have shown, using a microdialysis method, that local perfusion of the preoptic area with IL-1 resulted in marked and sustained elevations of extracellular NA and MHPG. Because NA plays a key role in the regulation of HPA axis activity during stress, several research groups have extensively studied the effects of immune stimulation on noradrenergic neurotransmission in the hypothalamus and paraventricular nucleus (PVN). The general picture arising from these studies is that peripheral inflammation results in elevations in the turnover and extracellular levels of hypothalamic NA (Dunn, 1992; Lavicky & Dunn, 1995; Mefford & Heyes, 1990). Peripheral induction of IL-1 may be responsible for the effects of LPS on NA in the hypothalamus and PVN, because i.p. injection of IL-1, but not of IL-6, increased the ratio of MHPG/NA in the hypothalamus of mice (Dunn, 1988; Mefford & Heyes, 1990; Zalcman et al., 1994). Moreover, systemic administration of IL-1 increased the extracellular levels of NA in the PVN (Mohankumar & Quadri, 1993; Smagin, Swiergiel, & Dunn, 1996). The role of central cytokines is not conclusive yet. Intracerebroventricular administration of IL1 enhanced the turnover of NA in the hypothalamus (Terao et al., 1993), although one study showed no effect of IL-1 in this brain structure (Connor et al., 1998). No effects of i.c.v. injection of TNF-a, IL-6, and IL-2 were reported (Connor et al., 1998; Terao et al., 1993). In the PVN, i.c.v. treatment with IL-1 enhanced the levels of extracellular NA (Terrazzino, Perego, & De Simoni, 1995).

BOO

, hippocampal NA

_500 •I •

M

m ^300 o

_*_

< z

'O

100 1

1

1

! •

-o-- 0 ^ _ -o^o

I

time (hr)

B

soo

hifjpocampal

MHPG

I 400 5 300 ^ 200 [ a. S 100

•-a^4_^.

lime (hr)

GOO

preoptic NA

lA'Vi-i.1

^500 Z

*-'*Kj

MO

41

n

f 300 o £-200

<

I

100 0

t i m e (hr)

D

preoptic MHPG ^400

Figure 4. Effects of i.p. administration of saline or LPS (lOO^g/kg body weight) on extracellular levels of NA and MHPG in the hippocampus (A, B) and the preoptic area (C, D) of the rat. The time point of injection is indicated by the arrow. *indicates the time period during which data from i.p. saline- and LPS-treated animals differ significantly (MANOVA with repeated measures). Data taken from Linthorst et al., 1995a and 1996 by permission of Oxford University Press and IBRO, respectively.

ID

2 300

/

Q ^200 CD a. ^ 100

/ *==*^«-*-»-^-4^o^ o - o - o ^ o,^o-o-o- O - o O

D

O

time ( h r )

O Lp. saline

A

i.p. L P S

<9

O

140

A. C. E. Linthorst and J. M. H. M. Reul

3.3. Distinct Responsiveness of Brain Serotonin and Noradrenaline to Inflammation: A Hypothesis on the Underlying Mechanisms From our experiments and the available literature data described under 3.1. and 3.2., it is evident that the CNS is able to generate a highly differentiated neurotransmission response during inflammation. Whereas we observed that extracellular NA levels in the preoptic area increased dramatically after peripheral injection of bacterial endotoxin, only a relatively moderate effect on this neurotransmitter was found in the hippocampus. On the other hand, systemic inflammation resulted in marked elevations in extracellular 5-HT in the hippocampus, but was ineffective at the level of the preoptic area. These data have led to a working hypothesis taking into account 1) the distinct neuroanatomical innervation of the preoptic area and the hippocampus, and 2) the divergent responsiveness to inflammatory challenges of the immediate early gene c-fos in the noradrenergic and serotonergic cell body-containing brain regions. While the preoptic area (and the PVN) receives its noradrenergic input mainly from the A1/A2 region in the brainstem, the hippocampus is almost exclusively innervated by noradrenergic neurons located in the locus coeruleus (Pacak, Palkovits, Kopin, & Goldstein, 1995; Swanson, 1987; Swanson, Kohler, & Bjorklund, 1987). Studies using c-fos mRNA as a marker for neuronal activity have shown that the A1/A2 region and the locus coeruleus respond differently to the peripheral administration of LPS and IL1. Whereas both the systemic injection of LPS and IL-1 induce c-fos expression in the nucleus of the solitary tract (containing the A2 region), the locus coeruleus is only activated by LPS (Brady, Lynn, Herkenham, & Gottesfeld, 1994; Elmquist, Ackermann, Register, Rimler, Ross, & Jacobson, 1993; Ericsson, Kovacs, & Sawchenko, 1994; Sagar, Price, Kasting, & Sharp, 1995; Wan, Janz, Vriend, Sorensen, Greenberg, & Nance, 1993). Taken together, these neuroanatomically distinct responses of brain regions containing noradrenergic cell bodies may underlie the different impact of i.p. administration of LPS on extracellular levels of NA in the preoptic area and the hippocampus. However, at present it cannot be excluded that processes at the level of the noradrenergic terminals are (also) involved in the differences between NA responses in the preoptic area and the hippocampus. Until now several research groups have characterized c-fos expression in the brain after peripheral administration of endotoxin or IL-1 (Brady et al., 1994; Elmquist et al, 1993; Ericsson et al., 1994; Sagar et al., 1995; Wan et al., 1993; Bilang-Bleuel, Linthorst, & Reul, unpublished observations). To our knowledge, no study has so far described c-fos expression in the raphe nuclei after stimulation of the immune system. Two other observations also suggest that the raphe nuclei are not responsive to immune system challenges. Molina-Holgado and Guaza (1996) have recently reported that i.p. administration of LPS had no effect on NA and 5-HT in the raphe nuclei, but that this treatment enhanced brain tissue levels of both neurotransmitters in the NTS. Moreover, preliminary studies of our group have shown that i.p. administration of LPS increased the level of 5 - H T I A receptors in various regions of the hippocampus, but had no effect on the autoreceptors in both the dorsal and median raphe nucleus (Reul et al., unpublished observations). However, although LPS at a dose of 100|ig/kg body weight had no effect on extracellular levels of 5-HT in the preoptic area, this dose caused a marked increase in the hippocampus. Hence, it may be speculated that mechanisms at the terminal level in the hippocampus are involved in the LPS-induced increases in 5-HT in this brain structure. As described above, we have provided evidence that a local action of IL-1 may play a role in the inflammation-induced rises in hippocampal 5-HT (Linthorst et al., 1994,1995b).

Inflammation and Brain Function

141

3.4. Functional Implications of Endotoxin-Induced Changes in Serotonergic and Noradrenergic Neurotransmission The characterization of neurotransmitter systems in the brain that are responsive to activation of the immune system is of utmost importance. However, such characterization should be followed by delineation of the functional significance of the specific neurotransmitter responses. Combining the in vivo microdialysis technique with biotelemetry gives the opportunity to monitor neurotransmitter changes, dialysate free corticosterone levels, body temperature, locomotion, and behavioural activity simultaneously (Figure 1). We have performed combined microdialysis/biotelemetry to study the temporal correlates of the rises in NA in the preoptic area during peripheral inflammation. I.p. injection of LPS (100fj.g/kg body weight) induced, apart from elevations in free corticosterone and sickness behaviour, a biphasic fever, and tachycardia. Comparison of the time courses of the LPS-induced increase in NA and the induction of fever, tachycardia, and HPA axis activation indicates that preoptic NA elevations may be responsible for the first phase of the fever response and/or the increase in free corticosterone (Linthorst et al., 1995a). This interpretation is based on the fact that the rise in NA runs in parallel with the first increase in body temperature and the increase in corticosterone levels. Moreover, maximum NA levels are reached about 60min earlier than the peak in body temperature and corticosterone concentrations. A possible role of NA changes in the PVN in HPA axis stimulation during inflammation has been suggested by Dunn and his coworkers. Their hypothesis is based on studies which show, using brain tissue samples of mice and microdialysis in rats, that LPS-induced activation of noradrenergic neurotransmission in the hypothalamus and PVN, and increases in plasma corticosterone levels follow parallel time courses (Dunn, 1988; Smagin et al., 1996). However, as recently suggested by these authors, divergences in the time courses indicate that other mechanisms are also involved (see Smagin et al., 1996). We also aimed at elucidating the functional implication of the activation of the hippocampal serotonergic system in response to peripheral administration of LPS and postulated that the increased levels of 5-HT in this brain structure might be instrumental in (some aspects of) sickness behaviour. Serotonin is known to regulate stressrelated behaviour and to be implicated in the processing of sensory information. Several studies in the cat and rat have shown that there is a positive relationship between behavioural activity of the animal on the one hand, and the firing rate of serotonergic neurons and extracellular levels of 5-HT in the terminal areas on the other hand (Jacobs & Azmitia, 1992; Kalen, Rosegren, Lindvall, & Bjorklund, 1989; Linthorst et al., 1994). We have observed the strongest correlation between extracellular 5-HT and behavioural activity in the hippocampus, although such a relationship was also evident in the preoptic area and the neocortex (Linthorst et al., 1995a, 1995b). Various stressors have been demonstrated to increase hippocampal extracellular levels of 5-HT in parallel with behavioural activation (Jacobs & Azmitia, 1992; Kalen et al., 1989; Rueter & Jacobs, 1996). However, the situation during immune stress is unique, in the sense that the rise in hippocampal 5-HT levels is accompanied by a decrease in behavioural activity (sickness behaviour). As already pointed out above (3.3.), this effect seems to be generated solely at the level of the hippocampus; i.e. without a concomitant increase in the firing rate of the serotonergic cell bodies. We have proposed that, in this way, a situation is realized in which sensory information processing in the hippocampus is modulated by 5-HT, but a parallel increase in behavioural activity by facilitation of motor output via 5-HT at the level of the brain stem is prevented (Reul, Labeiir. Wiegers. & Linthorst. 19981 An involvement of hioDOcamoal 5-HT in sickness

142

A. C. E. Linthorst and J. M. H. M. Reul

behaviour is underscored by the observation that behavioural lethargy during immune stimulation is always accompanied by enhanced hippocampal levels of 5-HT (for instance in the case of LPS, IL-lp, IL-2). TNF-a at doses that induce HPA axis activation and fever, however, neither generate a reduction in behavioural activity nor increase hippocampal 5-HT concentrations (Pauli et al., 1998). Another argument supporting this hypothesis is derived from experiments on long-term i.c.v. CRHinfused rats. These rats show an attenuated rise in hippocampal 5-HT levels during peripheral inflammation together with a delayed development of sickness behaviour (see below). The above-described notions on the functional implications of LPS-induced neurotransmitter changes are still speculative and based on correlative evidence. Experiments applying specific antagonists, antisense oligodeoxynucleotide treatments, lesion paradigms, and knockout animals are needed to further elucidate the role of brain NA and 5-HT in inflammation and infection.

4. NEUROIMMUNOENDOCRINE ACTIONS DURING LONG-TERM ELEVATION OF BRAIN CORTICOTROPIN-RELEASING HORMONE CRH, a neuropeptide widely distributed in the brain, is the key mediator of the overall responses of the organism to stress. CRH not only functions as the principal neuropeptide regulating ACTH release, but is also responsible for autonomic and behavioural responses associated with stress (for references see Vale et al., 1997). Moreover, central administration of CRH reduces natural killer cell activity, pointing to a role of CRH in the regulation of the activity of the immune system (Irwin et al., 1987). Disturbances of the brain CRH system are thought to be implicated in several stress-related disorders such as major depression and anorexia/bulimia, but also in Alzheimer's disease. There is evidence that major depression is associated with an increased central CRH drive. Patients suffering from major depression have elevated levels of CRH in the cerebrospinal fluid (Nemeroff, Widerlov, Bissette, Walleus, Karlsson, Eklund, Kilts, Loosen, & Vale, 1984), increased numbers of CRH expressing neurons and CRH mRNA levels in their PVN (Raadsheer, Hoogendijk, Stam, Tilders, & Swaab, 1994; Raadsheer, Van Heerikhuize, Lucassen, Hoogendijk, Tilders, & Swaab, 1995) and reduced CRH receptor density in the frontal cortex (Owens & Nemeroff, 1993). In addition, these patients present many symptoms reminiscent of the effects of centrally administered CRH in experimental animals, such as an activated HPA axis, increased sympathetic outflow, anxiety, increased emotionality, and loss of appetite and sexual interest (Owens & Nemeroff, 1991; Holsboer & Barden, 1996). In contrast, a role for a hypoactive CRH system is suggested in the pathophysiology of Alzheimer's disease, as indicated by the decreased CRH content and the upregulation of CRH receptors in postmortem cerebral cortex of patients (De Souza, Whitehouse, Kuhar, Price, & Vale, 1986). Moreover, cognitive impairment in patients with Alzheimer's disease seems to be associated with a reduced amount of CRH in the cerebrospinal fluid (Pomara, Singh, Deptula, LeWitt, Bissette, Stanley, & Nemeroff, 1989). From the above-described functions of CRH and the putative involvement of disturbed CRH functioning in brain diseases, it is clear that a well-balanced CRH

Inflammation and Brain Function

143 The long-term i>c.v. CRH-lnfused rat

in Vivo Treatment • Lorg-ternn central CRH-infusion ( 0 . 1 - 1 ng/h, 7 days)

Figure 5. Schematic drawing of the long-term i.c.v. CRHinfused rat. CRH is pumped into the ventricular system during 7 days via an i.c.v. cannula connected to a miniosmotic pump (Alza Corporation, Palo Alto, U.S.A.). In addition, some endocrine and physical data are depicted.

Neuroendi^drlne and Physical Effects • Plasma ACTH and Corticosterone • Anterior pituitary POMC mRNA • Adrenal weight • Spleen and Thymus weight • Body weight gain

system is essential for health. Although CRH system disturbances have been suggested in major depression and Alzheimer's disease, until now, studies in humans could not identify which physiological and behavioural abnormalities, associated with these illnesses, are caused by hyper- or hyposecretion of CRH. Experimental paradigms for the investigation of central CRH dysfunction in humans are rather limited. Despite the extensive molecular and neuroanatomical characterization studies on the CRH system and the study of physiological and behavioural effects of acutely administered CRH, hardly any investigations have focussed on the neurobiological consequences of long-term central CRH dysfunction. The latter should be of prime concern given that chronic disturbances in this system are most likely to occur during and to be involved in the development of stress-related disorders. We have, therefore, developed an animal model of central CRH hyperactivity, i.e. the long-term i.c.v. CRH-infused rat (Figure 5). Rats are i.c.v. infused with CRH for seven days via a miniosmotic pump, positioned subcutaneously in the dorsal region, connected to an i.c.v. cannula via polyethylene tubing (Alza Corporation, Palo Alto, U.S.A.). This animal model enables the study of the consequences of a disturbed CRH system for physiological, immune, neurochemical, and behavioural processes under baseline conditions and after stressful challenges. Because 1) CRH plays an important role in neuroimmunoendocrine interactions (Irwin et al., 1987; Rivier & Rivest, 1993), and 2) immune system alterations have been described in major depression (Maes, 1995; Sluzewska et al., 1996), we decided to assess the effects of peripheral inflammation in the long-term i.c.v. CRH-treated rat.

4.1. Characteristics of Long-Term i.e.v. CRH-lnfused Rats under Basal Conditions and during a Peripheral Inflammatory Challenge 4.1.1. The Immune System. In a first series of experiments the effects of longterm i.c.v. CRH infusion on the immune system were investigated (Labeur, Arzt, Wiegers, Holsboer, & Reul, 1995). CRH-infused rats showed a marked involution of

144

A. C. E. Linthorst and J. M. H. M. Reul

the spleen and the thymus. In vitro T-cell prohferation was tested in primary splenocytes collected at day seven of CRH treatment. As expected, splenocytes of long-term CRH-treated animals had reduced proliferative responses to the mitogen concanavalin A. Surprisingly, however, it was found that IL-2 levels were elevated in the supernatant of splenocyte cultures of CRH-infused rats, despite the high level of circulating glucocorticoid in this animal model (see below). This is most likely caused by a reduced uptake/internalization of IL-2 resulting from a diminished gene expression of the IL2 receptor a chain as observed in splenic T cells of CRH-treated animals. The effects of long-term elevation of brain CRH on T cell proliferation and IL-2 receptors may be evoked by substances released from the adrenal gland, because adrenalectomy prior to CRH infusion prevents these changes in immune function (Labeur et al., 1995). Next, effects of the CRH treatment on the level of the cytokines were investigated. It was observed that seven days of i.e.v. CRH treatment elevated the basal levels of IL-ip mRNA expression in splenocytes. Moreover, in vitro stimulation of the splenocytes with LPS induced a higher increase in IL-lp mRNA in splenocytes of long-term i.c.v. CRH-infused rats than in those of control animals (Labeur et al, 1995). These results point to a possible hyperresponsiveness of IL-1 in the immune system of longterm i.c.v. CRH-infused animals, which is underscored by experiments on the in vivo effects of LPS on cytokines. On day seven of the CRH treatment, rats were i.p. injected with LPS or saline. After administration of saline, plasma levels of bioactive IL-1,TNF, and IL-6 were below the detection limit of their respective assays. However, administration of LPS caused a much more pronounced elevation in circulating levels of IL1 and IL-6, but not of TNF, at three hours after the injection in long-term CRH-infused rats than in control rats (Linthorst, Flachskamm, Hopkins, Hoadley, Labeur, Holsboer, & Reul, 1997). Hence, it may be concluded that the immune system of long-term CRHinfused rats responds with a much higher release of IL-1 and IL-6 into the circulation during inflammatory challenges. This may represent a compensatory mechanism for the elevated levels of glucocorticoids which are well known to negatively regulate cytokines and cytokine actions (Munck & Guyre, 1991). On the other hand, this situation may be potentially dangerous, given the involvement of cytokines in septic shock and tissue degeneration (Dinarello, 1991). 4.1.2. The HPA Axis. Long-term infusion of CRH enhanced plasma levels of ACTH and corticosterone (Labeur et al., 1995). In addition, microdialysis experiments on day seven of treatment showed that free corticosterone levels were elevated in CRH-infused rats, resulting in a disappearance of the normal diurnal rhythm. Such a condition is also observed in chronic stress and major depression (Owens & Nemeroff, 1991; Holsboer & Barden, 1996). Other observations pointing to a sustained hyperactive HPA axis are hyperplasia of the adrenals, involution of the thymus and elevated expression of proopiomelanocortin (POMC) mRNA in the anterior pituitary of CRHtreated rats (Labeur et al., 1995; Linthorst et al., 1997). Intraperitoneal injection of LPS elevated free corticosterone levels in both i.c.v. CRH-infused and control rats (Figure 6A). Whereas the maximum levels reached in the two treatment groups were similar, the rise in free corticosterone was significantly delayed in long-term CRH-infused animals (Linthorst et al., 1997). Hence, these data indicate that the HPA axis in long-term CRH-infused rats is still able to respond to an inflammatory challenge, despite the higher level of circulating glucocorticoids and the possible desensitization of CRH receptors (see 4.1.3.). However, a pathophysiological consequence of such delayed response in corticosterone may be that glucocorticoid-

Inflammation and Brain Function

145

facilitated and glucocorticoid-regulated processes are executed (too) late (Wiegers & Reul, 1998; Wiegers et al., 1995). 4.1.3. Body Temperature and Locomotion. We have used a biotelemetry method to assess the effects of long-term i.c.v. infusion of CRH on body temperature and locomotor activity. Continuous i.c.v. infusion of CRH increased body temperature during the light as well as the dark period of the diurnal cycle, resulting in a flattened circadian rhythm especially during the first 3-4 days of treatment. Moreover, the i.c.v. treatment enhanced locomotor activity during the light and the dark period of the first 2-3 days (Linthorst et al., 1997). The fading-out of the CRH-induced increases in body temperature and locomotion over the treatment period points to a desensitization of central CRH effector mechanisms. Such a desensitization may take place at the level of the CRH receptors, because various situations which elevate CRH levels, including chronic stress and adrenalectomy, decrease CRH receptor density and CRH mRNA concentrations in the anterior pituitary, the hypothalamus and/or the frontal cortex (Anderson, Kant, & De Souza, 1993; De Souza, Insel, Perrin, Rivier, Vale, & Kuhar, 1985; Hauger, Millan, Lorang, Harwood, & Aguilera, 1988; Luo, Kiss, Rabadandiehl, & Aguilera, 1995; Makino, Schulkin, Smith, Pacak, Palkovits, & Gold, 1995; Tizabi & Aguilera, 1992). Intraperitoneal administration of LPS (100|i.g/kg body weight) caused a marked increase in body temperature in control animals, whereas this fever response was highly attenuated in long-term i.c.v. CRH treated rats (Figure 6B). Long-term peripherally CRH-infused rats displayed normal LPS-induced fever responses. Peripherally and i.c.v. CRH-infused animals showed similar elevations in free corticosterone levels under baseline conditions. Thus, the unchanged fever response in peripherally CRH-treated rats suggests that elevated levels of corticosterone are not responsible for the attenuated fever that was observed in long-term i.c.v. CRH-infused subjects (Linthorst et al., 1997). 4.1.4. Hippocampal Serotonergic Neurotransmission and Behavioural Activity. Long-term i.c.v. infusion of CRH had no consequences on basal extracellular levels of 5HT and 5-HIAA in the hippocampus as assessed on day 7 of treatment using in vivo microdialysis (Linthorst et al., 1997). This is a striking observation, because we have shown that acute i.c.v. injection of CRH caused a profound increase in both 5-HT and 5HIAA in the hippocampus (Reul & Linthorst, 1997). Hence, it may be hypothesized that long-term i.c.v. CRH treatment results in pertinent regulatory changes in 5-HT neurotransmission; an effect possibly related to desensitization of CRH receptors (see 4.1.3.). Intraperitoneal injection of LPS is known to increase hippocampal extracellular levels of 5-HT and 5-HIAA (see 3.1.1.1.). The LPS-induced rise in 5-HT was, however, significantly reduced in long-term i.c.v. CRH-infused rats (Figure 6C). Moreover, a delayed elevation of 5-HIAA after i.p. administration of LPS was found in i.c.v. CRHtreated animals (Linthorst et al., 1997). These observations clearly show that long-term elevation of brain CRH resulted in an attenuated responsiveness of the hippocampal 5-HT system to an acute stressful challenge. Whether CRH receptor desensitization or putative alterations in brain cytokines (IL-1?) are underlying the reduced responsiveness in 5-HT needs to be further elucidated. LPS (100|ig/kg body weight i.p.) induced sickness behaviour, as indicated by a reduced behavioural activity of the animals, in both i.c.v. treatment groups (CRH and vehicle). The induction of sickness behaviour was, however, delayed by about 1.5-2 hr in long-term i.c.v. CRH-treated rats (Figure 6D) (Linthorst et al., 1997). Because sick-

A. C. E. Linthorst and J. M. H. M. Reul

146

.70

free corticosterone

,60

^ 0 SO .40

O

«> o u

30

tp 0. 20 u

loh 00

k

*—•—•—A—. -a—*£ i

fli—f-

O

^-

»-

«

«N

n

time {hr)

B o

-1.00

time (hr) Figure 6. Effects of i.p. administration of LPS on free corticosterone (A), body temperature (B), hippocampal extracellular levels of 5-HT (C), and behavioural activity (D) in long-term i.e.v. vehicle- and i.e.v. CRH-infused rats. In Figure A, * and + indicate the first time point of statistical significance (as compared to i.p. saline-injected animals; not shown here) for i.e.v. vehicle- and CRH-infused rats, respectively (Student's t test). In Figures B and C, *indicates the time period during which data from long-term i.e.v. vehicle- and CRH-treated animals differ significantly (MANOVA with repeated measures). In Figure D, *indicates a significant difference between vehicle- and CRH-infused animals (Student's t test). The time point of injection is indicated by the arrow. The shaded area in Figure B represents the dark period of the diurnal cycle. Data taken from Linthorst et al., 1997 with permission of the Journal of Neuroscience.

Inflammation and Brain Function Z2S

147

hippocampal 5-HT

200 I

I

-;-k! /i-H'-l-S'^H-

50 -1

o

o

o

^

1

I

I

i

i

o

o

o

o

o

'

o

I

o

L

o

o

o

o

tlm« (hr)

D

3.00

o

o

o

o

o

o

o

o

o

o

o

o

o

O

t4

4

lii iri e

«

n

o

o n

o

o O

a

O

n

time (hr)

long-term i.e.v. CRH

long-term i-c.v. ^ vehicle

Figured. {Continued}

ness behaviour is as an important, adaptive response supporting the control of infection and inflammation, a delayed initiation of this behaviour potentially represents a detrimental situation for an organism.

4.2. Implications of Long-Term Hyperactivity of Brain CRH We have demonstrated here that long-term central infusion of CRH induces profound changes in various physiological systems. These changes show a striking resem-

148

A. C. E. Linthorst and J. M. H. M. Reul

blance with some aspects of major depression. Long-term i.c.v. CRH-infused animals display elevated levels of corticosterone without an apparent diurnal rhythm. Moreover, our study is the first to demonstrate a possible link between disturbances in central CRH and serotonergic neurotransmission, a situation which may also develop in stress-related disorders such as major depression. The general immunosuppression in combination with the hyperresponsiveness of IL-1 and IL-6 during peripheral inflammation, suggests that long-term elevation of brain CRH exerts distinct influences on the functioning of the immune system. The mechanisms underlying the changes caused by long-term i.c.v. infusion of CRH still need to be resolved. However, desensitization of CRH receptors, chronic HPA hormone hypersecretion, and elevated sympathetic outflow may play important roles.

5. CONCLUDING REMARKS The immune system, the endocrine system and the CNS form a complex network, and interact with each other via bidirectional communication pathways. We have provided evidence that during peripheral inflammation, a highly distinct activation of neurotransmitter pathways in the brain is generated. Moreover, the cytokine IL-1 seems to play an important role in inflammation-induced neurotransmitter responses in the brain. Whereas under baseline conditions the neuroimmunoendocrine network may be perfectly capable to react adequately to disturbances of homeostasis, we have demonstrated that chronic elevation of brain CRH, a situation which may also occur during chronic stress and stress-related disorders, evolves in aberrant immune, neuroendocrine, neurochemical, autonomic, and behavioural responses to inflammation. Chronically elevated levels of brain CRH, may, therefore, compromise biological defence mechanisms to cope with stressful challenges. Such defective responses may, ultimately, have deleterious consequences for homeostasis, promote disease, and limit the chance of survival of the organism.

ACKNOWLEDGMENTS The authors thank Ms. B. Burkart-Lauer for excellent artwork. The work of the authors described in this chapter was subsidized by the Volkswagen Foundation (grantnos. 1/68 430 and 1/70 543) and by the BIOMED Concerted Action Program "Cytokines in the brain."

REFERENCES Anderson, S. M., Kant, G. J., & De Souza, E. B, (1993). Effects of chronic stress on anterior pituitary and brain corticotropin-releasing factor receptors. Pharmacology, Biochemistry, & Behavior, 44, 755-761. Auerbach, S. B., Minzenberg, M. J., & Wilkinson, L. O. (1989). Extracellular serotonin and 5-hydroxyindoleacetic acid in hypothalamus of the unanesthetized rat measured by in vivo dialysis coupled to high-performance liquid chromatography with electrochemical detection: dialysate serotonin reflects neuronal release. Brain Research, 499, 281-290. Blatteis, C. M. (1990). Neuromodulative actions of cytokines. Yale Journal of Biology & Medicine, 63,133-146. Brady, L. S., Lynn, A. B., Herkenham, M., & Gottesfeld, Z. (1994). Systemic interleukin-1 induces early and late patterns of c-fos mRNA expression in brain. Journal of Neuroscience, 14, 4951. Connor, T. J., Song, C , Leonard, B. E., MeraU, Z., & Anisman, H. (1998). An assessement of the effects of

Inflammation and Brain Function

149

central interleukin-1-beta, -2, -6, and tumor necrosis factor-alpha administration on some behavioural, neurochemical, endocrine, and immune parameters in the rat. Neuroscience, 84, 923-933. Crespi, F, Garratt, J. C , Sleight, A. J., & Marsden, C. A. (1990). In vivo evidence that 5-hydroxytryptamine (5-HT) neuronal firing and release are not necessarily correlated with 5-HT metabolism. Neuroscience, 35,139-144. Dantzer, R., Bluthe, R. M., Kent, S., & Kelley, K. W. (1991). Behavioural effects of cytokines. In N. Rothwell & R. Dantzer (Eds.), Interleukin-1 in the brain (pp. 135-150). Oxford: Pergamon Press. De Souza, E. B., Insel, T. R., Perrin, M. H., Rivier, J., Vale, W. W., & Kuhar, M. J. (1985). Differential regulation of corticotropin-releasing factor receptors in anterior and intermediate lobes of pituitary and in brain following adrenalectomy in rats. Neuroscience Letters, 56, 121-128. De Souza, E. B., Whitehouse, P J., Kuhar, M. J., Price, D. L., & Vale, W. W. (1986). Reciprocal changes in corticotropin-releasing factor (CRF)-like immunoreactivity and CRF receptors in cerebral cortex of Alzheimer's disease. Nature, 319, 593-595. Dinarello, C. A. (1991). Interleukin-1 and interleukin-1 antagonism. Blood, 77, 1627-1652. Dunn, A. J. (1988). Systemic interleukin-1 administration stimulates hypothalamic norepinephrine metabolism parallelling the increased plasma corticosterone. Life Sciences, 43,429-435. Dunn, A. J. (1992). Endotoxin-induced activation of cerebral catecholamine and serotonin metabolism: comparison with interleukin-1. yowwfl/ of Pharmacology and Experimental Therapeutics, 26L 964-969. Dunn, A. J. & Welch, J. (1991). Stress- and endotoxin-induced increases in brain tryptophan and serotonin metabolism depend on sympathetic nervous system acWsiiy. Journal of Neurochemistry,57,1615-1622. Elmquist, J. K., Ackermann, M. R., Register, K. B., Rimler, R. B., Ross, L. R., & Jacobson, C. D. (1993). Induction of fos-like immunoreactivity in the rat brain following pasteurella-multocida endotoxin administration. Endocrinology, 133, 3054-3057. Ericsson, A., Kovacs, K. J., & Sawchenko, P. E. (1994). A functional anatomical analysis of central pathways subserving the effects of interleukin-1 on stress-related neuroendocrine neurons. Journal of Neuroscience, 14, 897-913. Felten, S. Y. & Felten, D. L. (1991). Innervation of lymphoid tissue. In R. Ader, D. L. Felten, & N. Cohen (Eds.), Psychoneuroimmunology (pp. 27-69). San Diego: Academic Press. Gemma, C , Ghezzi, P., & De Simoni, M. G. (1991). Activation of the hypothalamic serotoninergic system by central interleukin-1. European Journal of Pharmacology, 209, 139-140. Gottschall, P. E., Komaki, G., & Arimura, A. (1992). Interleukin-1 p activation of the central nervous system. In N. Rothwell & R. Dantzer (Eds.), lnterleukin-1 in the brain (pp. 2 7 ^ 9 ) . Oxford: Pergamon Press. Hart, B. L. (1988). Biological basis of the behavior of sick animals. Neuroscience & Biobehavioral Reviews, 12, 123-137. Hauger, R. L., Millan, M. A., Lorang, M., Harwood, J. P., & Aguilera, G. (1988). Corticotropin-releasing factor receptors and pituitary adrenal responses during immobilization stress. Endocrinology, 123, 396-405. Holsboer, F & Barden, N. (1996). Antidepressants and hypothalamic pituitary adrenocortical regulation. Endocrine Reviews, 17, 187-205. Irwin, M., Lacher, U, & Caldwell, C. (1992). Depression and reduced natural killer cytotoxicity: a longitudinal study of depressed patients and control subjects. Psychological Medicine, 22, 1045-1050. Irwin, M. R., Vale,W., & Britton, K.T. (1987). Central corticotropin-releasing factor suppresses natural killer cytotoxicity. Brain, Behavior, & Immunity, 1, 81-87. Jacobs, B. L. & Azmitia, E. C. (1992). Structure and function of the brain serotonin system. Physiological Reviews, 72, 165-229. Janssen, R. A. J., Mulder, N. H., The, T. H., & Deleij, L. (1994). The immunobiological effects of interleukin2 in vivo. Cancer Immunology Immunotherapy, 39, 207-216. Kabiersch, A., Del Rey, A., Honegger, C. G., & Besedovsky, H. O. (1988). Interleukin-1 induces changes in norepinephrine metabolism in the rat brain. Brain Behavior and Immunity, 2, 261-21 A. Kalen, P., Strecker, R. E., Rosengren, E., & Bjorklund, A. (1988). Endogenous release of neuronal serotonin and 5-hydroxyindoleacetic acid in the caudate-putamen of the rat as revealed by intracerebral dialysis coupled to high-performance liquid chromatography with fluorimetric detection. Journal of Neurochemistry, 51, 1422-1435. Kalen, P., Rosegren, E., Lindvall, O., & Bjorklund, A. (1989). Hippocampal noradrenaline and serotonin release over 24 hours as measured by the dialysis technique in freely moving rats: correlation to behavioural activity state, effect of handling and tail-pinch. European Journal of Neuroscience, 1, 181-188. Karanth, S., Lyson, K., & McCann, S. M. (1993). Role of nitric oxide in interleukin 2-induced corticotropinreleasing factor release from incubated hypothalami. Proceedings of the National Academy of Sciences of the United States of America, 90, 3383-3387. Katsuura, G, Gottschall, P. E., Dahl, R. R., & Arimura, A. (1988). Adrenocorticotropin release induced by

150

A. C. E. Linthorst and J. M. H. M. Reul

intracerebroventricular injection of recombinant human interleukin-1 in rats: possible involvement of prostaglandin. Endocrinology, 122,1773-1779. Katsuura, G., Gottschall, P. E., Dahl, R. R., & Arimura, A. (1989). Interleukin-1 beta increases prostaglandin E2 in rat astrocyte cultures: modulatory effect of neuropeptides. Endocrinology, 124, 3125-3127. Kluger, M. J. (1991). Fever: role of pyrogens and cryogens. Physiological Reviews, 71, 93-127. Komaki, G., Arimura, A., & Koves, K. (1992). Effect of intravenous injection of IL-lbeta on PGE2 levels in several brain areas as determined by microdialysis. American Journal of Physiology, 262, E246-E251. Labeur, M. S., Arzt, E., Wiegers, G. J., Holsboer, E, & Reul, J. M. H. M. (1995). Long-term intracerebroventricular corticotropin-releasing hormone administration induces distinct changes in rat splenocyte activation and cytokine expression. Endocrinology, 136, 2678-2688. Lavicky, J. & Dunn, A. J. (1995). Endotoxin administration stimulates cerebral catecholamine release in freely moving rats as assessed by microdialysis. Journal of Neuroscience Research, 40, 407^13. Linthorst, A. C. E., Flachskamm, C., Holsboer, F, & Reul, J. M. H. M. (1994). Local administration of recombinant human interleukin-1 beta in the rat hippocampus increases serotonergic neurotransmission, hypothalamic-pituitary-adrenocortical axis activity, and body temperature. Endocrinology, 135, 520-532. Linthorst, A. C. E., Flachskamm, C , Holsboer, F , & Reul, J. M. H. M. (1995a). Intraperitoneal administration of bacterial endotoxin enhances noradrenergic neurotransmission in the rat preoptic area: relationship with body temperature and hypothalamic-pituitary- adrenocortical axis activity. European Journal of Neuroscience, 7, 2418-2430. Linthorst, A. C. E., Flachskamm, C , Muller-Preuss, R, Holsboer, F, & Reul, J. M. H. M. (1995b). Effect of bacterial endotoxin and interleukin-1 beta on hippocampal serotonergic neurotransmission, behavioral activity, and free corticosterone levels: an in vivo microdialysis study. Journal of Neuroscience, 15, 2920-2934. Linthorst, A. C. E., Flachskamm, C , Holsboer, F, & Reul, J. M. H. M. (1996). Activation of serotonergic and noradrenergic neurotransmission in the rat hippocampus after peripheral administration of bacterial endotoxin: involvement of the cyclo-oxygenase pathway. Neuroscience, 72, 989-997. Linthorst, A. C. E., Flachskamm, C , Hopkins, S. J., Hoadley, M. E., Labeur, M. S., Holsboer, F, & Reul, J. M. H. M. (1997). Long-term intracerebroventricular infusion of corticotropin-releasing hormone alters neuroendocrine, neurochemical, autonomic, behavioral, and cytokine responses to a systemic inflammatory challenge. Journal of Neuroscience, 17, 4448-4460. Luo, X., Kiss, A., Rabadandiehl, C., & Aguilera, G. (1995). Regulation of hypothalamic and pituitary corticotropin-releasing hormone receptor messenger ribonucleic acid by adrenalectomy and glucocorticoids. Endocrinology, 136, 3877-3883. Maes, M. (1995). Evidence for an immune response in major depression: a review and hypothesis. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 19, 11-38. Makino, S., Schulkin, J., Smith, M. A., Pacak, K., Palkovits, M., & Gold, P W. (1995). Regulation of corticotropin-releasing hormone receptor messenger ribonucleic acid in the rat brain and pituitary by glucocorticoids and stress. Endocrinology, 136, 4517-4525. Mefford, I. N. & Heyes, M. P. (1990). Increased biogenic amine release in mouse hypothalamus following immunological challenge: antagonism by indomethacin. Journal of Neuroimmunology, 27, 55-61. Merali, Z., Lacosta, S., & Anisman, H. (1997). Effects of interleukin-1-beta and mild stress on alterations of norepinephrine, dopamine, and serotonin neurotransmission—a regional microdialysis study. Brain Research, 761,225-235. Mohankumar, P. S.,Thyagarajan, S., & Quadri, S. K. (1993). Interleukin-1 beta increases 5-hydroxyindoleacetic acid release in the hypothalamus in vivo. Brain Research Bulletin, 31, 745-748. Mohankumar, P. S. & Quadri, S. K. (1993). Systemic administration of interleukin-1 stimulates norepinephrine release in the paraventricular nucleus. Life Sciences, 52, 1961-1967. Molina-Holgado, F & Guaza, C. (1996). Endotoxin administration induced differential neurochemical activation of the rat brain stem nuclei. Brain Research Bulletin, 40, 151-156. Munck. A., Guyre, P. M., & Holbrook, N. J. (1984). Physiological functions of glucocorticoids in stress and their relation to pharmacological actions. Endocrine Reviews, 5, 25-44. Munck, A. & Guyre, P. M. (1991). Glucocorticoids and immune function. In R. Ader, D. S. Felten, & N. Cohen (Eds.), Psychoneuroimmunology (pp. 447-474). San Diego: Academic Press. Nemeroff, C. B., Widerlov, E., Bissette, G., Walleus, H., Karlsson, I., Eklund, K., Kilts, C. D., Loosen, P T., & Vale, W. (1984). Elevated concentrations of CSF corticotropin-releasing factor-like immunoreactivity in depressed patients. Science. 226, 1342-1344. Owens, M. J. & Nemeroff, C. B. (1991). Physiology and pharmacology of corticotropin-releasing factor. Pharmacological Reviews, 43. 425-473.

Inflammation and Brain Function

151

Owens, M. J. & Nemeroff, C. B. (1993). The role of corticotropin-releasing factor in the pathophysiology of affective and anxiety disorders: laboratory and clinical studies. In D. J. Chadwick, J. Marsh, & K. Ackrill (Eds.), Corticotropin-Releasing Factor, Ciha Foundation Symposium 172 (pp. 296-308). Chichester: John Wiley & Sons. Pacak, K., Palkovits, M., Kopin, 1. J., & Goldstein, D. S. (1995). Stress-induced norepinephrine release in the hypothalamic paraventricular nucleus and pituitary-adrenocortical and sympathoadrenal activity: In vivo microdialysis studies. Frontiers in Neuroendocrinology, 16, 89-150. Pauli, S., Linthorst, A. C. E., & Reul, J. M. H. M. (1998). Tumour necrosis factor-alpha and interleukin-2 differentially affect hippocampal serotonergic neurotransmission, behavioural activity, body temperature and hypothalamic-pituitary-adrenocortical axis activity in the rat. European Journal of Neuroscience, 10, 868-878. Pomara, N., Singh, R. R., Deptula, D., LeWitt, P A., Bissette, G., Stanley. M., & Nemeroff, C. B. (1989). CSF corticotropin-releasing factor (CRF) in Alzheimer's disease: its relationship to severity of dementia and monoamine metabolites. Biological Psychiatry, 26, 500-504. Raadsheer, E C , Hoogendijk, W. J. G., Stam, F. C.Tilders, E J. H., & Swaab, D. E (1994). Increased numbers of corticotropin-releasing hormone expressing neurons in the hypothalamic paraventricular nucleus of depressed patients. Neuroendocrinology, 60, 436-444. Raadsheer, E C , Van Heerikhuize, J. J., Lucassen, P. J., Hoogendijk, W. J. G., Tilders, E J. H., & Swaab, D. E (1995). Corticotropin-releasing hormone mRNA levels in the paraventricular nucleus of patients with Alzheimer's disease and depression. American Journal of Psychiatry, 152, 1372-1376. Raber, J. & Bloom, E E. (1994). IL-2 induces vasopressin release from the hypothalamus and the amygdala: Role of nitric oxide-mediated signaling. Journal of Neuroscience, 14, 6187-6195. Reul, J. M. H. M.. Labeur, M. S., Wiegers, G. J., & Linthorst, A. C. E. (1998). Altered neuroimmunoendocrine communication during a condition of chronically increased brain corticotropin-releasing hormone drive. Annals of the New York Academy of Sciences, 840, 444-455. Reul, J. M. H. M. & Linthorst, A. C. E. (1997). Central administration of corticotropin-releasing hormone profoundly stimulates rat hippocampal serotonergic neurotransmission and behavioral activity: no effect of adrenalectomy. Proceedings of the 26th Annual Meeting of the Society for Neuroscience, New Orleans, LA, 56.7. Rivest, S. & Rivier, C. (1993). Central mechanisms and sites of action involved in the inhibitory effects of CRF and cytokines on LHRH neuronal activity. Annals of the New York Academy of Sciences, 697, 117-141. Rivier, C. & Rivest, S. (1993). Mechanisms mediating the effects of cytokines on neuroendocrine functions in the rat. In D. J. Chadwick, J. Marsh, & K. Ackrill (Eds.), Corticotropin-Releasing Factor, Ciha Foundation Symposium 172 (pp. 204-225). Chichester: John Wiley & Sons. Rivier, C. & Shen, G. H. (1994). In the rat, endogenous nitric oxide modulates the response of the hypothalamic-pituitary-adrenal axis to interleukin-1 beta, vasopressin, and oxytocin. Journal of Neuroscience, 14, 1985-1993. Rothwell, N. J. (1991). Functions and mechanisms of interleukin 1 in the brain. Trends in Pharmacological Sciences, 12, 430-436. Rueter, L. E. & Jacobs, B. L. (1996). A microdialysis examination of serotonin release in the rat forebrain induced by behavioral environmental manipulations. Brain Research, 739, 57-69. Sagar, S. M., Price, K. J., Kasting, N. W., & Sharp, E R. (1995). Anatomic patterns of EOS immunostaining in rat brain following systemic endotoxin administration. Brain Research Bulletin, 36, 381-392. Shintani, E, Kanba, S., Nakaki,T., Nibuya, M., Kinoshita, N., Suzuki, E., Yagi, G., Kato, R., & Asai, M. (1993). lnterleukin-1 beta augments release of norepinephrine, dopamine, and serotonin in the rat anterior hypothalamus. yow/-Ajfl/ of Neuroscience, 13, 3574-3581. Sirko, S., Bishai, 1., & Coceani, E (1989). Prostaglandin formation in the hypothalamus in vivo: effect of pyrogens. American Journal of Physiology, 256, R616-24. Sluzewska, A., Rybakowski, J., Bosmans, E., Sobieska, M., Berghmans, R., Maes, M., & Wiktorowicz, K. (1996). Indicators of immune activation in major depression. Psychiatry Research, 64, 161-167. Smagin, G. N., Swiergiel, A. H., & Dunn, A. J. (1996). Peripheral administration of interleukin-1 increases extracellular concentrations of norepinephrine in rat hypothalamus: Comparison with plasma corticosterone. Psychoneuroendocrinology, 21, 83-93. Smith,T., Hewson, A. K., Quarrie, L., Leonard, J. P,& Cuzner, M. L. (1994). Hypothalamic PGE(2) and cAMP production and adrenocortical activation following intraperitoneal endotoxin injection—In vivo microdialysis studies in Lewis and Fischer rats. Neuroendocrinology, 59, 396-405. Swanson, L. W. (1987). The hypothalamus. In A. Bjorklund, T. Hokfelt, & L. W Swanson (Eds.), Handbook of Chemical Neuroanatomy. Vol. 5: Integrated systems of the CNS, Part I. (pp. 1-124). Amsterdam: Elsevier Science.

152

A. C. E, Linthorst and J. M. H. M. Reul

Swanson, L. W., Kohler, C , & Bjorklund, A. (1987). The limbic region. I: The septohippocampal system. In A. Bjorklund, T. Hokfelt, & L. W. Swanson (Eds.), Handbook of Chemical Neuroanatomy. Vol. 5: Integrated Systems of the CNS, Part I. (pp. 125-277). Amsterdam: Elsevier Science. Terao, A., Oikawa, M., & Saito, M. (1993). Cytokine-induced change in hypothalamic norepinephrine turnover: involvement of corticotropin-releasing hormone and prostaglandins. Brain Research, 622, 257-261. Terrazzino, S., Perego, C , & De Simoni, M. G. (1995). Noradrenaline release in hypothalamus and ACTH secretion induced by central interleukin-1 beta. Neuroreport, 6,2465-2468. Tizabi, Y. & Aguilera, G. (1992). Desensitization of the hypothalamic-pituitary-adrenal axis following prolonged administration of corticotropin-releasing hormone or vasopressin. Neuroendocrinology, 56, 611-618. Vale, W., Vaughan, J., & Perrin, M. (1997). Corticotropin-releasing factor (CRF) family of ligands and their receptors. Endocrinologist, 7, S3-S9. Wan, W., Janz, L., Vriend, C. Y., Sorensen, C. M., Greenberg, A. H., & Nance, D. M. (1993). Differential induction of c-Fos immunoreactivity in hypothalamus and brain stem nuclei following central and peripheral administration of endotoxin. Brain Research Bulletin, 32, 581-587. Wiegers, G. J., Labeur, M. S., Stec, I. E. M., Klinkert, W. E. F., Holsboer, F , & Reul, J. M. H. M. (1995). Glucocorticoids accelerate anti-T cell receptor-induced T cell growth. Journal of Immunology, 155, 1893-1902. Wiegers, G. J. & Reul, J. M. H. M. (1998). Induction of cytokine receptors by glucocorticoids: functional and pathological significance. Trends in Pharmacological Sciences, 19,317-321. Wilkinson, L. O., Auerbach, S. B., & Jacobs, B. L. (1991). Extracellular serotonin levels change with behavioral state but not with pyrogen-induced hyperthermia. Journal of Neuroscience, 11,2732-2741. Zalcman, S., Green-Johnson, J. M., Murray, L., Nance, D. M., Dyck, D., Anisman, H., & Greenberg, A. H. (1994). Cytokine-specific central monoamine alterations induced by interleukin-1, -2, and -6. Brain Research, 643, 40-49.

10

DYNAMIC REGULATION OF PROINFLAMMATORY CYTOKINES Linda R. Watkins,^ Kien T. Nguyen,' Jacqueline E. Lee,^ and Steven F. Maier' ' Department of Psychology ^Department of Molecular, Cellular, and Developmental Biology University of Colorado at Boulder Boulder, Colorado 80309

1. OVERVIEW Proinflammatory cytokines, including interleukin-lbeta (IL-ip), IL-6, and tumor necrosis factor-alpha (TNF-a), are proteins created and released by a wide variety of immune cells in response to infection, inflammation, and tissue damage. These proteins have long been recognized as critical communication factors for orchestrating early responses of immune cells to such immune challenges. More recently, it has become accepted that these same proinflammatory cytokines are key mediators of immune-tobrain communication as well. In this role, proinflammatory cytokines communicate to the brain, thereby orchestrating sickness responses including fever, decreased food, and water intake, increased sleep, increased pain responsivity, and so forth (Kent, Bluthe, Kelley, & Dantzer, 1992; Maier & Watkins, 1998). Thus, it is clear that proinflammatory cytokines are importantly involved in the response of both the immune system and the central nervous system in these situations. Despite the importance of proinflammatory cytokines for regulating immune and central nervous system responses, we are only beginning to understand how proinflammatory cytokines are regulated. One aspect of cytokine regulation that has attracted increasing interest in recent years is the impact of environmental stress. Even at this early stage, it is clear that acute and chronic stressors can dramatically impact proinflammatory cytokine production. The purpose of the first part of this chapter is to summarize this literature. What will be disturbingly obvious is that, at present, there is no straightforward conclusion that can be drawn from this stress literature; predictable patterns to the results will be hard to find. What will be clear is that stress cannot simply be stated to increase or decrease cytokine production. It will be argued that the seeming chaos is actually the logical result. One of the Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

153

154

L. R. Watkins et al.

basic difficulties inherent in this literature is the extreme diversity of experimental procedures all lumped together under the single heading of "stress". Stressors that have been studied in the animal and/or human literature with regards to their effects on proinflammatory cytokine regulation include psychological stress (humans watching a disturbing film or animals introduced into a novel environment), mild voluntary exertion, severe forced exertion, physical exertion in elite athletes who have had years of daily conditioning prior to testing, physical exertion in untrained subjects, surgery (with its inherent problems with anesthetic effects, tissue damage, introduction of endotoxin, etc.), severe immobilization, mild restraint, exposure to loud noise, cold water immersions, and so forth. These "stressors" are obviously extremely diverse with regards to their metabolic demands, endocrine profiles, neurochemical mediators, duration of effect, associated trauma, and so forth. Beyond even these procedural differences, another basic difficulty inherent in this literature has to do with whether the cytokines are measured in the presence or absence of superimposed immune challenge (i.e., lipopolysaccharide [LPS] or other immune stimulant); measured in vivo, in vitro, or ex vivo; when they are measured relative to stress; whether physiological or pharmacological drug doses are tested; and so forth. If proinflammatory cytokine regulation were simple and straightforward, perhaps simple, straightforward answers could still be found despite such diverse experimental procedures. However, cytokine regulation is not merely a matter of increasing or decreasing some single rate-limiting step in production. The last section of this chapter is focused on the "nuts-and-bolts" of proinflammatory cytokine regulation, using interleukin-1 (3 (IL-ip) as the example. The point of this section will be to clarify the multiple levels at which cytokine production and function is dynamically regulated, and to define what is presently known about how and when such regulation occurs. From this, it will become evident that to understand the impact of stress on cytokine regulation, one must first have a thorough appreciation for the underlying control mechanisms involved.

2. IMPACT OF STRESS ON PROINFLAMMATORY CYTOKINES 2.1. Impact of "Stress-Hormones" on Peripheral Cytokine Production 2.1.1. Evidence from Animal Studies. If stress is to influence cytokine production in the periphery, then substances released in response to stress must be able to influence cytokine production by peripheral immune cells. Substance P, for example, is released from sympathetic terminals into the peritoneal fluid in response to stressors. By this means, it would be accessible for binding by peritoneal macrophages. Indeed, stress enhances the ability of substance P to bind to peritoneal macrophages, by as yet unknown mechanisms (Chancellor-Freeland, Zhu, Kage, Beller, Leeman, & Black, 1995). Once bound, substance P activates phagocytic and chemotactic capacity in macrophages, as well as pro-inflammatory cytokine production (Chancellor-Freeland et al., 1995). In vivo, stress (immobilization) increases TNF-a levels in mice, an effect blocked by substance P antagonists (Arck, Merali, Stanisz, Stead, Chaouat, Manuel, & Clark, 1995). In vitro, preincubating (priming) macrophages with substance P prior to LPS enhances proinflammatory cytokine release (Berman, Chancellor-Freeland, Zhu, & Black, 1996) and addition of substance P to cultures of macrophages from stressed animals greatly enhances LPS induced TNF-a production

Dynamic Regulation of Proinflammatory Cytokines

155

relative to its effect on macrophages from non-stressed animals (Chancellor-Freeland et al., 1995). Sympathetic activation releases catecholamines, as well as substance P. Alphaadrenergics appear to mediate rises in IL-lp in supernatants from murine bone marrow cells produced by prolonged (lOhr) immobilization, since this effect was blocked by alpha (but not beta) adrenergic antagonists injected at the time of stress (Khlusov, Dygai, & Goldberg, 1993). Epinephrine, derived almost exclusively from the adrenal medulla, also increases plasma IL-6 in rats, an effect blocked by the beta adrenergic antagonist propranolol and mimicked by beta adrenoreceptor agonists (DeRijk, Boelen,Tilders, & Berkenbosch, 1994; vanGool, vanVugt, Helle, & Aarden, 1990). Epinephrine also increases LPS-induced IL-6, but decreases LPS-induced TNF-a, from liver via a beta-adrenergic mechanism (Liao, Keiser, Scales, Kunkel, & Kluger, 1995b). In contrast, central rather than peripheral, beta adrenergics are implicated in increased plasma IL-6 produced by open field stress since centrally (but not peripherally) acting beta adrenergic antagonists can block this stress induced rise in IL-6 (LeMay, Vander, & Kluger, 1990; Soszynski, Kozak, Conn, Rudolph, & Kluger, 1996). Lastly, stressors release glucocorticoids from the adrenal cortex. Effects of glucocorticoids on proinflammatory cytokine production have been mixed, with one possible general conclusion being that very low glucocorticoid doses enhance cytokine production whereas higher doses are suppressive (Liao, Keiser, Scales, Kunkel, & Kluger, 1995a). Infusion of corticosterone at both nonstressed and stressed levels enhance basal TNF-a and IL-6 release from isolated rat liver (Liao et al., 1995a). In contrast, when corticosteroid is combined with LPS, enhancement and suppression of cytokine release are found with nonstressed and stressed corticosterone levels, respectively (Liao et al., 1995a). High glucocorticoid doses have also been reported to suppress LPS-induced TNF-a, IL-ip, and IL-6 from perfused rat liver (Tran-Thi,Weinhold, Weinstock, Hoffmann, Schulze-Specking, Northoff, & Decker, 1993). Inhibition of glucocorticoid release (by cyanoketone or antibodies against corticotropin releasing hormone [CRH]) enhances LPS-induced serum TNF-a in mice (Fantuzzi, Di Santo, Sacco, Genigni, & Ghezzi, 1995). When blood from stressed mice or CRH-treated mice is treated ex vivo with LPS, less TNF-a is produced than in blood from control animals. Importantly, normal TNF-a production is restored ex vivo by addition of the glucocorticoid receptor antagonist RU-486, implicating endogenous glucocorticoids in the observed TNF-a suppression (Fantuzzi et al., 1995). While glucocorticoids can suppress IL-6 production as well (Dobbs, Feng, Beck, & Sheridan, 1996; Waage, Slupphaug, & Shalaby, 1990), it is perhaps more striking that several studies report enhancement of IL-6 due to glucocorticoids. Glucocorticoids are implicated in mediating elevated plasma IL-6 in rats produced by conditioned aversive stimuli, since elevated IL-6 in response to this stressor is blocked by adrenalectomy, but not by the beta-receptor antagonist propranolol (Zhou, Kusnecov, Shurin, DePaoli, & Rabin, 1993). Glucocorticoids are likewise implicated in altered IL-6 responses produced by chronic restraint stress in virally infected mice. Restraint stress in virally infected mice enhances IL-6 release from regional lymph nodes; on the other hand, restraint inhibits splenic IL-6 release (Dobbs et al., 1996). Both the enhancing and inhibiting effects of restraint were prevented by RU-486, thus implicating glucocorticoids (Dobbs et al., 1996). Thus, to date, there appears to be evidence for substance P induced increases in at least TNF-a, alpha-adrenergic induced increases of at least IL-ip, and betaadrenergic induced increase of at least IL-6. Whether this last finding reflects a central or peripheral action is unclear. Regarding glucocorticoids, low doses may increase

156

L. R. Watkins et ah

cytokines, while higher doses are generally suppressive. The only exception to this last conclusion may be IL-6 which tends across studies to be increased rather than decreased by adrenal steroids. 2.1.2. Evidence from Humans. While a potential role for substance P has not been examined in humans, studies have examined the effect of both catecholamines and glucocorticoids on cytokine production. With regard to catecholamines, norepinephrine inhibits LPS-induced TNF-a and IL-6 production in human whole blood. This effect is blocked by beta 1 antagonists but not by alpha antagonists (VanDerPool, Jansen, Endert, Sauerwein, & VanDeventer, 1994). In accordance with these findings, beta adrenergic agonists mimic the effect of norepinephrine, whereas alpha adrenergic agonists are without effect (VanDerPool et al., 1994). Such results cannot be explained on the basis of internalization of LPS receptors in response to norepinephrine, since beta adrenergic agonists have no effect on the expression of the LPS-receptor, CD14 (VanDerPool et al., 1994). In contrast to the cytokine-suppressive effects of catecholamines cited above, elevations in plasma epinephrine and norepinephrine elicited by strenuous treadmill exercise directly correlate with elevations in plasma IL-6 (Papanicolaou, Petrides, Tsigos, Bina, Kalogeras, Wilder, Gold, Deuster, & Chrousos, 1996). Furthermore, IL-lp secretion by monocytes is increased by physiological concentrations of epinephrine in vitro (Cannon, Evans, Hughes, Meredith, & Dinarello, 1986). Glucocorticoids have generally been found to have no effect on, or to suppress, cytokine production. LPS-induced IL-ip, TNF-a, and IL-6 are all suppressed following administration of pharmacological doses of hydrocortisone, whereas a physiological dose of hydrocortisone only suppresses TNF-a production (DeRijk, Michelson, Karp, Petrides, Galliven, Deuster, Paciotti, Gold, & Sternberg, 1997). While higher concentrations of hydrocortisone in the physiological range are without effect on IL-ip release, combined administration of epinephrine and hydrocortisone suppress this measure (Cannon et al., 1986). Stress induced levels of glucocorticoids elicited by strenuous exercise suppress LPS-induced IL-lp and TNF-a production but have no effect on IL-6 (DeRijk et al., 1997). Dexamethasone and hydrocortisone have also been found to depress strenuous treadmill exercise induced IL-6 levels, an effect not accounted for by dexamethasone induced alterations in plasma catecholamines (Papanicolaou et al., 1996). In contrast to the cytokine-suppressive effects of glucocorticoids cited above, i.v. CRH (which increases serum glucocorticoid levels) increases serum IL-la levels (Schulte, Bamberger, Elsen, Herrmann, Bamberger, & Barth, 1994) and low concentrations of hydrocortisone augment IL-lp release in vitro (Cannon et al., 1986). Thus, to date, beta adrenergics appear to inhibit LPS-induced cytokine production. Basal (i.e., not induced by LPS) levels of cytokines, in contrast, may be increased by physiological levels of catecholamines. Glucocorticoids can either suppress or have no effect on cytokine production. It is notable that these data do not readily agree with the results from laboratory animals.

2.2. Impact of Acute Stress on Peripheral Cytokine Production 2.2.1. Evidence from Animal Studies. A variety of acute environmental stressors have been found to alter proinflammatory cytokine production. Regarding IL-1 effects, IL-la is increased in plasma by rotation stress (Korneva, Rybakina, Orlov, Shamova,

Dynamic Regulation of Proinflammatory Cytokines

157

Shanin, & Kokryakov, 1997) and by combined stress (2hr of 4-5°C followed by room temperature immobilization for 20hr) (Korneva et al., 1997). In contrast, plasma IL-la is not increased by acute cold water stress (Korneva et al., 1997). IL-ip is also affected by stress. IL-lp is increased in plasma by rotation stress (Korneva, Rybakina, Fomicheva, Kozinets, & Shkhinek, 1992) and an IL-1-like factor is increased in splenic adherent cells by social isolation of 14 day old chicks (Cunnick, Kojic, & Hughes, 1994). LPS-induced IL-ip plasma levels are increased by a prior metabolic (2-deoxyglucose) stress (Miller, Koebel, & Sonnenfeld, 1993b), yet LPS-induced IL-ip in plasma is decreased by brief (15min) restraint stress (Goujon, Parnet, Laye, Combe, Kelley, & Dantzer, 1995). Whereas TNF-a is increased in serum by ultra sonic sound stress by a rodent repellant device (Arck, Troutt, & Clark, 1997), heat stress of rats I h prior to LPS decreased LPS-induced serum TNF-a (Ribeiro, Villar, Downey, Edelson, & Slutsky, 1996). This latter procedure also produces decreased TNF-a release from cultured alveolar macrophages (Ribeiro et al., 1996). Forced exhaustive treadmill running up to 6hr prior to LPS also transiently blunted LPS-induced increases in plasma TNF-a (Bagby, Sawaya, Crouch, & Shepherd, 1994). IL-6 is increased in plasma by a variety of stressors, including open field stress (LeMay et al., 1990; Soszynski et al., 1996), restraint and immobilization stress (Kitamura, Konno, Morimatsu, Jung, Kimura, & Saito, 1997; Kusnecov, Shurin, Armfield, Litz, Wood, Zhou, & Rabin, 1995;Takaki, Huang, Somogyvari-Vigh, & Arimura, 1994; Zhou et al., 1993), footshock (Kusnecov et al., 1995; Zhou et al., 1993), and conditioned aversive stimuli (Kusnecov et al., 1995; Zhou et al., 1993). Metabolic (2-deoxyglucose) stress also enhances subsequent LPS-induced IL-6 plasma levels (Miller et al., 1993b). Furthermore, while both immobilization and LPS increase mRNA for IL-6 in liver (Kitamura et al., 1997), these changes occur in parenchymal cells after immobilization stress, versus in sinusoidal mononuclear cells after LPS, suggesting that stress and immune stimuli increase IL-6 in these tissues by different mechanisms (Kitamura et al., 1997). In ex vivo studies, heat stress suppresses LPS-induced rise in plasma TNF-a (Kluger, Rudolph, Soszynski, Conn, Leon, Kozak, Wallen, & Moseley, 1997). In contrast, LPS-induced IL-ip and TNF-a (but not IL-6) release is enhanced following footshock stress (Persoons, Schornagel, Breve, Berkenbosch, & Kraal, 1995). Furthermore, restraint stress, footshock, and prolonged (lOhr) immobilization each increase ex vivo IL-ip production (Khlusov et al., 1993; Persoons et al., 1995), effects not correlated with either ACTH or glucocorticoid levels (Persoons et al., 1995). Intriguingly, these effects dissipated by 3 days post-stress but reappeared l-2wks later, suggesting longterm effects of stress on cytokine regulation that are largely overlooked and unanticipated by most researchers (Persoons et al., 1995). In summary, then, IL-la, IL-ip, TNF-a may or may not be affected by various stressors. There appears to be no apparent predictability as to which stressors will or won't alter the levels of these cytokines, or even whether "altered levels" imply increases or decreases in production. IL-6, in contrast, appears to be consistently elevated by stress. 2.2.2. Evidence from Humans. Regarding stress-induced changes in cytokine production, by far the most frequently studied stress in humans is exercise. Cytokine responses to strenuous exercise has been investigated by multiple investigators with highly divergent results (for review, see (Northoff, Weinstock, & Berg, 1994)). Exercise increases plasma IL-ip, correlated with exercise intensity (Bury, Louis, Radermecker,

158

L. R. Watkins et al.

& Pirnay, 1996). An incremental exercise procedure fails to alter monocyte production of either TNF-a or IL-6; however, spontaneous release of both cytokines is inhibited by epinephrine added in vitro (Rivier, Pene, Chanez, Anselme, Caillaud, Prefaut, Godard, & Bousquet, 1994). One hr of bicycle exercise either increases (Haahr, Pedersen, Fomsgaard, Tvede, Diamant, Klarlund, Halkjaer-Kristensen, & Bendtzen, 1991; Ullum, Haahr, Diamant, Palmo, Halkjaer-Kristensen, Pedersen, & Palm, 1994) or has no effect on (Smith, Telford, Baker, Hapel, & Weidemann, 1992) plasma IL-6; neither TNF-a, IL-la, nor IL-ip is altered by this procedure (Haahr et al., 1991; Smith et al., 1992; Ullum et al., 1994). mRNA for IL-la, IL-1 p, IL-6, and TNF-a are also not affected by this procedure (Ullum et al., 1994). Running a 5 Km race elevates serum TNF-a levels 2hrs, but not 24 hrs, later (Espersen, Elbaek, Ernst, Toft, Kaalund, Jersild, & Grunnet, 1990). A 2.5 hr running test increases serum TNF-a only I h r after the run (Dufaux & Order, 1989). Serum IL-ip is elevated immediately, but not 2hr, after maximal bicycling exercise in highly trained cyclists (Lewicki,Tchorzewski, Majewska, Nowak,& Baj, 1988). Running a marathon (42 Km; 2.75-4.75 hrs) elevates plasma TNFa and IL-6 but not IL-la (Camus, Poortmans, Nys, Deby-Dupont, Duchateau, Deby, & Lamy, 1997; Castell, Poortmans, Leclercq, Brasseur, Duchateau, & Newsholme, 1997). A 6hr endurance run (approx. 65 Km) in trained athletes elevates plasma IL-1 receptor antagonist (IL-lra) and IL-6, but not IL-1 (3 or TNF-a (Drenth, VanUum, VanDeuren, & Pesman, 1995). Ex vivo LPS challenge of blood from these athletes reveals suppressed IL-ip and TNF-a production (Drenth et al., 1995). Following a 250 Km road race, plasma levels of TNF-a and IL-6 were transiently elevated (Gannon, Rhind, Suzui, Shek, & Shephard, 1997). A maximal treadmill stress test decreases peripheral blood mRNA for TNF-a, but no change in mRNA levels of IL-la, IL-1 [3, or IL-6 (Natelson, Zhou, Ottenweller, Bergen, Sisto, Drastal, Tapp, & Gause, 1996). A competition swim by elite swimmers fails to produce changes in plasma IL-1 [3, IL-6, or TNF-a (Espersen, Elbaek, Schmidt-Olsen, Ejlersen, Varming, & Grunnet, 1996) whereas one hour of bicycle exercise elevates IL-1 activity in plasma several hrs later (Cannon et al., 1986). Psychological/emotional stressors have also been examined in a few studies. Emotional stress induced by parachute jumping does not alter plasma levels of IL-lp,TNFa, or IL-6 (Aloe, Bracci-Laudiero, AUeva, Lambiase, micera, & Tirassa, 1994; Dugue, Leppanen, Teppo, Fyhrquist, & Grasbeck, 1993). Similarly, the psychological stress of viewing and recalling a "gruesome surgery film" fails to alter circulating IL-1 [3 (Zakowski, McAllister, Deal, & Baum, 1992). In contrast, academic examination in medical students elevates IL-lb immediately after the exam (Dobbin, Harth, McCain, Martin, & Cousin, 1991), while interferon levels produced by leukocytes were decreased by a similar stressor (Glaser, Rice, Speicher, Stout, & Kiecolt-Glaser, 1986). In summary, then, exercise has very inconsistent effects on cytokine production, likely because the duration, intensity, subject pool (untrained subjects versus professional athletes), etc. vary so dramatically across studies. What is most striking about this literature is that very small, transient effects are seen at most. Cytokines are also typically not affected by psychological stressors.

2.3. Impact of Chronic Stress on Peripheral Cytokine Production 2.3.1. Evidence from Animal Studies. In terms of chronic stress, by far the most well-studied paradigm is antiorthostatic (head down) suspension, which is believed to simulate space flight. Antiorthostatic suspension for 11 days does not alter macrophage

Dynamic Regulation of Proinflammatory Cytokines

159

secretion of TNF-a or IL-1(3, despite elevated serum corticosterone (Kopydlowski, McVey, Woods, landolo, & Chapes, 1992). Chronic restraint, whether orthostatic or antiorthostatic or equivalent duration of just harness restraint, reduces IL-1 (3 responses to mitogen (ConA) stimulation (Berry, Murphy, Smith, Taylor, & Sonnenfeld, 1991). In contrast, chronic antiorthostatic or orthostatic (head up) suspension results in a fivefold increase in serum basal IL-ip, in the absence of elevated ACTH or corticosterone (Mekaouche, Givalois, Barbanel, Siaud, Maurel, Malaval, Bristow, Boissin, Assenmacher, & Ixart, 1994). In addition, both groups of suspended rats show enhanced serum IL-ip in response to LPS. This effect does not correlate with altered endocrine response to LPS, since antiorthostatic suspended rats show decreased ACTH and enhanced corticosterone, and orthostatic suspended controls show no difference in ACTH and corticosterone, relative to non-suspended LPS-injected controls (Mekaouche et al., 1994). The effects of chronic cold water swims have also been examined by several groups. Twice daily 5min 10°C cold water swims for 4 days leads to increased spontaneously released IL-ip, increased cell-associated IL-1 (3, and enhanced LPS-induced IL6 and TNF-a release from peritoneal macrophages; changes which are positively correlated with elevated substance P levels in the peritoneal cavity (ChancellorFreeland et al., 1995; Cheng, Morrow-Tesch, Beller, Levy, & Black, 1990; Zhu, Chancellor-Freeland, Berman, Kage, Leeman, Beller, & Black, 1996). Capsaicin pretreatment, which depletes substance P, reduces the effect of cold water stress on macrophage cytokine production (Chancellor-Freeland et al., 1995). Elevated substance P appears to be a key mediator of this effect, as the exaggerated IL-6 response is prevented by a substance P antagonist (Zhu et al., 1996) In contrast, twice daily 3-8 min 10-15°C cold water swims for 14 days has been reported to suppresses macrophage TNF-a release (Aarstad, Kolset, & Seljelid, 1991). Four other chronic stressors have been examined in a few studies, with regards to their effect on proinflammatory cytokines. Restraint stress in virally infected mice inhibits splenic IL-6 release but enhances IL-6 release from regional lymph nodes (Dobbs et al., 1996). Noise stress (24hr of 80dB "rock" music) in rats significantly suppresses IL-1 (3 release from neutrophils and macrophages, compared to controls (McCarthy, Ouimet, & Daun, 1992). Chronic repeated tail shock suppresses IL-2 production from mitogen stimulated lymphocytes. Lastly, diet independent increases in plasma IL-1 (3 were found the first 2 days of weaning in 3wk old pigs (McCracken, Gaskins, Ruwe-Kaiser, Klasing, & Jewell, 1995). Thus, to date, chronic restraint and chronic cold water swims tend to increase cytokine response. 2.3.2. Evidence from Humans. By far the most frequently studied chronic stress in humans is, again, exercise. Chronic exercise has divergent effects across studies. Twelve weeks of intense, progressive resistance strength training does not change plasma IL-ip, IL-6, or TNF-a (Rail, Roubenoff, Cannon, Abad, Dinarello, & Meydani, 1996). A 5-7 day military training course involving continuous physical exercise, calorie deficiency, and deprivation also produces no change in serum IL-ip, but did reduce serum IL-6 (Boyum, Wiik, Gustavsson, Veiby, Reseland, Haugen, Opstad, & Byum, 1996). On the other hand, the effects of physical exercise on TNF-a production in LPSstimulated blood samples vary according to exercise history. Blood samples drawn preexercise from athletes who train once or twice every day show the lowest LPS-induced TNF-a production. Samples from the group that trains the least (4 or 5 times total per

160

L. R. Watkins et al

month) show the highest LPS-induced TNF-a levels, pre-exercise. Post-exercise, blood samples from all groups show markedly suppressed TNF-a production to LPS (Kvernmo, Olsen, & Osterud, 1992). Endurance training (3 times per wk for 12wks) elevates TNF-a plasma levels in women, but not men, independent of plasma Cortisol levels (Home, Bell, Fisher, Warren, & Janowska-Wieczorek, 1997). Chronic psychological stressors of various types have also been studied, and found to produce divergent effects. Peripheral blood leucocytes from Alzheimer caregivers produce less IL-ip mRNA in response to LPS (Kiecolt-Glaser, Marucha, Malarkey, Mercado, & Glaser, 1995). When Alzheimer caregivers are given influenza virus vaccination, their cells also produce less virus-specific IL-ip in vitro (Kiecolt-Glaser, Glaser, Gravenstein, Malarkey, & Sheridan, 1996). In contrast, combat-related posttraumatic stress disorder (PTSD) subjects show elevated serum IL-ip levels which correlate with duration of PTSD symptoms, but not with either Cortisol levels, severity of PTSD, anxiety, or depressive symptoms (Spivak, Shohat, Mester, Avraham, Gil-Ad, Bleich, Valevski, & Weizman, 1997). Chronic subclinical anxiety in college freshmen is correlated with lower serum IL-ip, compared either to controls or to subjects with negative attributional styles (Zorrilla, Redei, & DeRubeis, 1994). Lastly, chronic cold water (18 immersions across 6wks in 14°C for 60min) produces only a trend toward increased plasma IL-6 and no change in plasma IL-ip (Jansky, Pospisilova, Honzova, Ulicny, Sramek, Zeman, & Kaminkova, 1996). Thus, to date, chronic exercise gives mixed results, likely for the same problems with subject selection, exertion level, and so forth. No consistent findings have emerged from other chronic stress paradigms, either.

2.4. Impact of Acute Stress on Central Cytokine Production 2.4.7. Evidence from Animal Studies. Given the obvious difficulties involved, no studies in humans have examined the effect of acute or chronic stressors on central cytokine production. There also exist no laboratory animal studies on the effect of chronic stressors on such measures, and only a very few studies that have examined the effect of acute stressors. Regarding effects on IL-ip, restraint stress in mice (15min) suppresses LPSinduced IL-ip protein in plasma and discrete brain regions (i.e., hippocampus and striatum, but not in hypothalamus) (Goujon et al., 1995). In contrast, inescapable tailshock (but not restraint) increases hypothalamic IL-ip protein levels (Nguyen, Deak, Owens, Kohno, Fleshner, Watkins, & Maier, 1998). Since more widespread and dramatic stressinduced changes are seen in brain IL-ip protein levels in adrenalectomized rats than in sham controls, these data suggest that stress-induced increases in circulating glucocorticoids suppress stress-induced increases in brain IL-ip protein (Nguyen et al., 1998) (see Maier et al. chapter in this volume for a comprehensive discussion of our work in this area). Immobilization stress for 60-120 min also increases IL-lp bioassayable activity in rat hypothalamus (Shintani, Nakaki, Kanba, Sato, Yagi, Shiozawa, Also, Kato, & Asai, 1995). This same study found that IL-1 receptor antagonist delivered through the microdialysis probe to the hypothalamus inhibits immobilization induced increases in both plasma ACTH and in hypothalamic norepinephrine, dopamine, serotonin, and their metabolites, suggesting that immobilization releases endogenous IL-ip (Shintani et al., 1995). Immobilization stress also produces peak increases in IL-ip mRNA and IL-lra mRNA within I h r and 2 ^ h r , respectively, after initiation of immobilization (Minami, Kuraishi, Yamaguchi, Nakai, Hirai, & Satoh, 1991; Suzuki, Shintani, Kanba,

Dynamic Regulation of Proinflammatory Cytokines

161

Asai, & Nakaki, 1997). Lastly, interleukin-1 converting enzyme (ICE), which cleaves pro-IL-ip into its active form (see below) is also affected by stress. Specifically, restraint stress decreases ICE levels in pituitary, hippocampus, and hypothalamus, an effect proposed to be due to stress-increased elevations in glucocorticoids (Laye, Goujon, Combe, VanHoy, Kelley, Parnet, & Dantzer, 1996). Regarding effects of stress on other proinflammatory cytokines, no data are yet available for TNF-a, and only one study to date has examined IL-6. This work reported that immobilization for 4hrs elevates IL-6 mRNA in midbrain, but has no effect on IL6 mRNA in hypothalamus (Shizuya, Komori, Fujiwara, Miyahara, Ohmori, & Nomura, 1997). In summary, the small number of studies to date are reasonably consistent in finding increases in brain IL-lp levels after stress. ICE can be depressed by stress, whereas IL-6 can be site-specifically increased by stress. No information is yet published regarding the effect of stress on central TNF-a.

3. REGULATION OF PROINFLAMMATORY CYTOKINE PRODUCTION: IL-lp As noted above, highly discrepant results have been found for the effects of various stressors on proinflammatory cytokine production. No clear pattern emerges from the stress literature, even when the effects of stress on a single cytokine are considered. Such seemingly chaotic findings may indeed be predictable from the fact that proinflammatory cytokine production and function are complexly regulated at multiple levels. To illustrate this point, the sections that follow will use IL-ip as the "model" proinflammatory cytokine for illustrating how production and function of cytokines are regulated.

3.1. Regulation of IL-lp Transcription (mRNA Formation) 3.1.1. Extracellular Signals. Increase in the rate of IL-lp gene transcription can be achieved by a variety of extracellular signals, including cross-linking of ICAM-1 adhesion molecules (Koyama,Tanaka, Saito, Abe, Nakatsuka, Morimoto, Auron, & Eto, 1996), granulocyte-macrophage colony stimulating factor (GM-CSF) (Fernandez, Walters, & Marucha, 1996; Oster, Brach, Gruss, Mertelsmann, & Herrmann, 1992), macrophage-CSF (M-CSF) (Oster et al., 1992), transforming growth factor beta 1 (TGFbetal) (Chantry, Turner, Abney, & Feldmann, 1989), IL-la (Slack, Nemunaitis, Andrews, Singer, & Andrews, 1990), IL-lp (Fenton, Vermeulen, Clark, Webb, & Auron, 1988), TNF-a (Slack et al., 1990), IL-2 (Fenton, 1992), prostaglandin E2 (PGE2) (Ohmori, Strassman, & Hamilton, 1990). Bradykinin (Pan, Zuraw, Lung, Prossnitz, Browning, & Ye, 1996), and complement factor 5a (C5a) (Geiger, Rordorf, Galakatos, Seligmann, Henn, Lazdins, & Vosbeck, 1992). In addition, interferon gamma (IFNy) enhances IL-ip mRNA transcription in LPS-stimulated cells (Fenton, 1992) as well as in both resident and thioglycoUate-elicited peritoneal macrophages (Collart, Belin, Vassalli, de Kossodo, & Vassalli, 1986). Synergy in inducing IL-ip transcription is also observed between IFNy, GM-CSF and IL-3 (Fenton, 1992); and between histamine and IL-la (Vannier & Dinarello, 1993). Extracellular substances that suppress IL-ip gene activation include antiinflammatory cytokines and steroids. That is, IL-4, IL-6, and IL-10 can each suppress

162

L. R. Watkins et al.

LPS-induced IL-ip expression (Fenton, 1992). Glucocorticoids and dexamethasone also inhibit IL-ip transcription, effects blocked by RU-486 (Busso, Collart, Vassalli, & Belin, 1987; Lee, Tsou, Chan, Thomas, Petrie, Eugui, & Allison, 1988). Since these steroids act via intracellular receptors, the mechanisms involved will be discussed below. 3.1.2. Cytoplasmic Signals and Transcription Factors. A variety of intracellular signals increase IL-lp gene transcription. cAMP, for example, generally increases ILi p gene transcription (Sung & Walters, 1991), although it decreases IL-ip transcription in astrocytes (Willis & Nisen, 1995). PGE2 and TNF-a increase IL-lp transcription, in part, via increases in cAMP (Lorenz, Furdon, Taylor, Verghese, Chandra, Kost, Haneline, Roner, & Gray, 1995; Ohmori et al., 1990). In addition, cAMP increases IL-lp gene activation via activation of protein kinase C (PKC) (Serkkola & Hurme, 1993). PKC is a serine-threonine kinase activated by phorbol ester to produce IL-ip mRNA and protein. PKC is also involved in LPS effects since LPS-induced IL-ip mRNA is downregulated by a PKC inhibitor (Serkkola, 1995). A variety of protein kinases, in addition to PKC, lead to IL-ip gene activation. One way that such protein kinases (ERK, p38, JNK, etc) are activated is by formation of reactive oxygen intermediates. These rapidly form upon stimulation by IL-la, ILip, TNF-a, ultraviolet light, or growth factors (Rosette & Karin, 1996). The activated protein kinases, in turn, phosphorylate, and thus activate, transcription factors imphcated in IL-ip transcription (see below) (Koj, 1996; Raingeaud, Gupta, Rogers, Dickens, Han, & Ulevitch, 1995). Arachidonic acid (but not its metabolites) also activates protein kinases such as JNK which, in turn, activate transcription factors (Rizzo & Carlo-Stalla, 1996). Dexamethasone, amongst its other actions, inhibits activation of JNK (Swantek, Cobb, & Geppert, 1997). A variety of classical transcription factors can either enhance or suppress IL-lp gene transcription. Transcription factors which increase IL-ip gene activation include: (a) AP-1 (activator protein-1; a heterodimer of c-Jun and c-Fos) created by crosslinking ICAM-1 adhesion molecules and also by IL-ip, cAMP, and PKC (Abe,Tanaka, Saito, Shirakawa, Koyama, Goto, & Eto, 1997; Fenton, 1992; Koyama et al., 1996; Serkkola & Hurme, 1993; Turner, Chantry, Buchan, Barrett, & Feldmann, 1989); (b) a tyrosine-phosphorylated protein induced by both LPS and IL-ip, which is similar to Signal Transducers and Activators of Transcription [STAT] proteins used by other cytokines (Tsukada, Waterman, Koyama, Webb, & Auron, 1996); (c) a heterodimer of nuclear factor (NF)-IL-6 and cAMP response element binding (CREB) protein induced by LPS (Tsukada, Saito, Waterman, Webb, & Auron, 1994); (d) a heterodimer of NF-IL-6 and a non-CREB protein induced by LPS (Tsukada et al., 1994); (e) NFIL-6 induced by lipoarabinomannan (Mycobacterium tuberculosis cell wall component), LPS, and TNF-a (Zhang & Rom, 1993); (f) CREB proteins and other cAMP responsive CREB/activating transcription factor (ATF) family members induced by LPS, phorbol ester, and TNF-a (Chandra, Cogswell, Miller, Godlevski, Stinnett, Noel, Kadwell, Kost, & Gray, 1995); (g) NF-betaA and immediate-early 1 (lEl) induced by cytomegalovirus (Crump, Geist, Auron, Webb, Stinski, & Hunninghake, 1992; Hunninghake. Monks, Geist, Monick, Monroy, Stinski, Webb, Dayer, Auton, & Fenton, 1992), and (h) NF-kappaB induced by multiple stimuli, including GM-CSF, M-CSF, ILlp, bradykinin, TNF-a, PKC activators, increases in cAMP, and Sendai virus induced myeloid extracts (Abe et al., 1997; Cogswell, Godlevski, Wisely, Clay, Leesnitzer, Ways,

Dynamic Regulation of Proinflammatory Cytoicines

163

& J.G, 1994; Hiscott, Marois, Garoufalis, D'Addario, Roulston, Kwan, Pepin, Lacoste, Nguyen, Bensi, & Fenton, 1993; Newton, Adcock, & Barnes, 1996; Oster et al., 1992; Pan, Zuraw, Lung, Prossniz, Browning, & Ye, 1996; Strulovici, Daniel-Issakani, Oto, Jr., Chan, & Tsou, 1989). In terms of negative transcriptional signals, both tonic and phasic factors are involved. Evidence exists for tonic suppression of IL-ip transcription by a short-lived (and as yet unidentified) transcriptional repressor protein, since IL-ip transcription is rapidly induced by cycloheximide, a compound which disrupts protein synthesis (Collart et al., 1986) (cf (Turner et al., 1989)). This, or similar, tonic transcriptional repressor protein is known to physically prevent the binding of the transcription factor NF-kappaB to the IL-ip gene (Lebedeva & Singh, 1997). Transcription factors that physically suppress IL-ip gene activation include those activated by extracellular substances and/or cellular stressors. IL-ip gene transcription is suppressed by heat shock factor 1 (Cahill, Waterman, Xie, Auron, & Calderwood, 1996) and by a 150kDa complex formed by activation of intracellular glucocorticoid receptors. This glucocorticoid-receptor complex translocates to the nucleus where it binds to a negative glucocorticoid response element (nGRE) in the IL-lp gene (Zhang, Zhang, & Duff, 1997). There is evidence that this binding of activated glucocorticoid receptors to nOREs results in a physical disruption of the ability of AP-1, CREB, NF-IL-6, and the p65, p50, and c-Rel subunits of NF-kappa B to bind, thus disrupting gene activation (Adcock, Brown, Gelder, Shirasaki, Peters, & Barnes, 1995; Ray & Prefontaine, 1994; Scheinman, Gualberto, Jewell, Cidlowski, & Baldwin, 1995).

3.2. Regulation of Steady State Levels of IL-lp mRNA The experimental measure reported in many studies is total IL-ip mRNA. Changes in total mRNA could be caused by changes in frequency of transcription initiation, changes in mRNA stability, etc. Regardless of underlying mechanism, the end result is an alteration in the total amount of mRNA potentially available for translation. The work summarized below reflects literature which reports levels of IL-ip mRNA, rather than a specific description of where the alteration in gene expression occurred. A variety of extracellular and intracellular signals increase total IL-ip mRNA levels, including LPS (Knudsen, Dinarello, & Strom, 1986), IL-la (Elias, Reynolds, Kotloff, & Kern, 1989), IL-ip (Elias et al., 1989),TNF-a (Kaushansky, Broudy, Harlan, & Adamson, 1988),TNF-p (Kaushansky et al., 1988), epidermal growth factor (Schluns, Cook, & Le, 1997), IL-2 (Adachi, Inoue, Arinaga, Li, Ueo, Mori, & Akiyoshi, 1997), and activation of PKC (Smith, Kueppers, & Lee, 1991). CRH upregulates IL-lp mRNA, an effect mediated by cAMP and blocked by a CRH antagonist (Paez, Sauer, Perez, Finkielman, Stalla, Holsboer, & Arzt, 1995). Beta-adrenergic agonists increase IL-1(3 mRNA levels in microglia, yet decrease IL-113 release; in contrast, alpha adrenergic agonists have no effect (Hetier, Ayala, Bousseau, & Prochiantz, 1991). Synergy is observed between TNF-a and either IL-la or IL-ip (Elias, Reynolds, Kotloff, & Kern, 1989). IFNy has been reported to have no effect on IL-1 (3 mRNA levels yet enhance release of mature IL-13, in response to LPS (Sihvola & Hurme, 1989). In contrast, in response to silica, IFNy has also been reported to reduce steady state levels of IL-1 (3 mRNA concomitant with reduced release of mature IL-lp (Sihvola & Hurme, 1989). IL-ip mRNA induction is again blocked by steroids and anti-inflammatory cytokines. Increases of LPS-induced IL-ip mRNA are blocked by dexamethasone

164

L. R. Watkins et al

(Knudsen et al., 1986), and IL-4 inhibits IL-lp mRNA expression (Brantschen, Gauchat, De Week, & Stadler, 1989).

3.3. Stabilization of IL-lp mRNA The stabiUty of an mRNA determines its half-Ufe. Stabilization versus destabiHzation can be imparted by a variety of factors inherent in the structure of the specific mRNA under study. The mRNA of proinflammatory cytokines, including IL-ip, contain structures that impart inherent instability. This leads to an IL-1|3 mRNA half-life of only about 4-5 hrs in monocytes and macrophages (Herzyk, Allen, Marsh, & Wewers, 1992) However, the half-life of IL-lp mRNA can be influenced by multiple factors. GMCSF, M-CSF, and IFNy each increase IL-ip mRNA stability (Fenton, 1992; Fernandez et al., 1996; Oster et al., 1992). TNF-a also increases IL-lp mRNA half-life (Gorospe, Kumar, & Baglioni, 1993; Kang, Hammerbert, & Cooper, 1996). This effect appears to be mediated by PKC, since PKC stimulators increase IL-ip mRNA stability and PKC inhibitors prevent accumulation of IL-ip mRNA by TNF-a and accelerate decay of ILl p mRNA in cells pretreated with TNF-a (Gorospe et al., 1993; Yamato, el-Hajjaoui, & Koeffler, 1989). On the other hand, glucocorticoids, histamine, calmodulin antagonists, and calcium antagonists each decrease IL-lp mRNA half-life (Amano, Lee, & Allison, 1993; Kimberlin, Willis, Jr., & Nisen, 1995; Lee et al., 1988; Ohmori & Hamilton, 1992; Vannier & Dinarello, 1993). The anti-inflammatory cytokine IL-4 also decreases IL-ip mRNA half-life, with this effect produced via induction of a protein that directly destabilizes the mRNA (Donnelly, Fenton, Kaufman, & Gerrard, 1991; Ohmori & Hamilton, 1992). 3.4. Translation of I L - l p m R N A into Inactive P r o - I L - l p Protein Although it may seem that translation (protein production) should naturally occur as long as the mRNA is present, specific structures within the mRNA of proinflammatory cytokines decrease translation efficiency, thereby basally reducing the efficacy with which protein is formed (Kruys, Marinx, Shaw, Deschamps, & Huez, 1989). Indeed, transcription can be an isolated event from translation. A variety of stimuli (LPS, TNF-a,TGFbetal, cAMP, etc.) can increase IL-ip transcription without translation (Chantry et al., 1989; Fenton, 1992; Knudsen et al., 1986; Sung & Walters, 1991). Although, as noted above, LPS has been reported to induce transcription without translation, other studies have reported that LPS can in fact stabilize IL-ip mRNA, allowing enhanced translation (Caput, Beutler, Hartog, Thayer, Brown-Shimer, & Cerami, 1986; Dinarello, 1992). Dexamethasone inhibits IL-ip translation, likely via increases in cAMP, a second messenger known to block IL-lp translation (Knudsen et al., 1986). In contrast, translation of IL-lp is increased by epidermal growth factor (Schluns, Cook, & Le, 1997), CRH (Paez et al., 1995) as well as by cross-linking ICAM-1 adhesion molecules (Koyama et al., 1996). Lastly, GM-CSF treated monocytes produce greater pro-ILip in response to LPS (Laliberte, Perregaux, McNiff, & Gabel, 1997). 3.5. Post-Translational Processing of P r o - I L - l p Protein IL-lp is first synthesized as an inactive 31kDa protein. This pro-IL-ip must be proteolytically cleaved to become active. Chloride anions are suggested as a kev

Dynamic Regulation of Proinflammatory Cytoicines

165

element in the post-translational processing of IL-l(3.This is a result of selective anion transport leading to a volume-induced change in the intracellular ionic environment (Perregaux, Laliberte, & Gabel, 1996a). Most cells that produce pro-IL-ip also produce the IL-1 converting enzyme (ICE) that cleaves the Asp(116)-Ala(117) bond to generate 17.5 kDa mature IL-lp (Miller, Calaycay, Chapman, Howard, Kostura, Molineaux, & Thronberry, 1993a). This cleavage of pro-IL-ip into mature IL-ip is thought to be closely associated with secretion of IL-ip (Miller et al., 1993a). Some cells (e.g., fibroblasts, keratinocytes) do not synthesize ICE, but rather release pro-IL-ip into the extracellular fluid where it is cleaved by various proteolytic enzymes present at inflammation sites (Hazuda, Strickler, Kueppers, Simon, & Young, 1990; Mizutani, Schechter, Lazarus, Black, & Kupper, 1991). ICE, like IL-lp, is first synthesized as an inactive precursor. Active ICE is formed by generating a heterodimer of a 20 kDa and a 10 kDa polypeptide cleavage fragment (Keane, Giegel, Lipinski, Callahan, & Shivers, 1995; Singer, Scott, Chin, Bayne, Limjuco, Weidner, Miller, Chapman, & Kostura, 1995). Like pro-IL-ip, proICE and mature ICE are found free within the cytoplasm (Singer et al., 1995). ICE activity must therefore be regulated, since both ICE and pro-IL-ip co-exist without the mature IL-ip being formed (Laliberte, Perregaux, Svensson, Pazoles, & Gabel, 1994). However, there is evidence that active ICE can be associated with the outer cell surface of human monocytes, supporting the idea that mature IL-ip is generated via cleavage by ICE following association of pro-IL-ip with the plasma membrane, during secretion (Singer et al., 1995). Little is yet known regarding regulation of ICE gene expression and activation. Increased ICE mRNA is observed in brain (pituitary, hypothalamus, hippocampus) within 4hrs of LPS in rat (Keane et al., 1995;Tingsborg, Zetterstrom, Alheim, Hasanvan, Schultzberg, & Bartfai, 1996) but not after LPS in mouse (Laye et al., 1996). LPS also increases ICE mRNA in various peripheral tissues, including spleen, testis, and adrenal (Keane, 1995;Tingsborg et al., 1996). On the other hand, glucocorticoids appear to suppress ICE transcription, since ICE mRNA increases in mouse brain following adrenalectomy and decreases upon stress exposure (Laye et al., 1996).

3.6. Release of Mature IL-lp Protein Studies of IL-ip release into the extracellular fluid do not typically assess mechanism of action. Thus, changes in protein product release could result from effects anywhere along the molecular biological cascade from transcription to post-translational processing and release. Processing of proIL-ip into its mature form does not require protein synthesis. Low levels of stimulation of human monocytes cause these cells to accumulate significant levels of intracellular pro-IL-lp (Chin & Kostura, 1993). Treatment of LPSpre-stimulated monocytes with cycloheximide was used to prevent new protein synthesis. Since cycloheximide treatment does not inhibit LPS induced release of mature IL-lp, this supports the conclusion that pre-existing pools of pro-IL-ip occurs in these cells (Chin & Kostura, 1993). Enhanced release of mature IL-ip has been reported following leukotriene inhibitors (Sirko, Schindler, Doyle, Weisman, & Dinarello, 1991), P2Z purinergic receptor agonists such as extracellular ATP (Ferrari, Chiozzi, Falzoni, Dal Susino, Melchiorri, Baricordi, & Di Virgiho, 1997; Griffiths, Stam, Downs, & Otterness, 1995; Perregaux, Svensson, & Gabel, 1996b), cAMP (Okamoto, Oh, & Nakano, 1990), substance P fMartin. Charles. Sanderson. & Merrill. 1992; Viae. Gueniche. Doutremeouich.

166

L. R. Watkins et al.

Reichert, Claudy, & Schmitt, 1996), calcium ionophores (Martin et al., 1992; Suttles, Giri, & Mizel, 1990), and IFNy (Sihvola & Hurme, 1989). In contrast, release of IL-lp has been reported to be inhibited by dexamethasone (Kern, Lamb, Reed, Daniele, & Nowell, 1988), IL-4 (Miossec, Briolay, Dechanet, Wijdenes, Martinez-Valdez, & Banchereau, 1992), cAMP (Iwaz, Kouassi, Lafont, & Revillard, 1990; Viherluoto, Palkama, Silvennoinen, & Hurme, 1991; Yoshimura, Kurita, Nagao, Usami, Nakao, Watanabe, Kobayashi, Yamazaki,Tanaka, Inagaki, & Nagai, 1997) PKC inhibitors (Bakouche, Moreau, & Lachman, 1992; Krakauer, 1996; Shapira, Takashiba, Champagne, Amar, & VanDyke, 1994), beta-adrenergic agonists (Hetier et al., 1991; Yoshimura et al., 1997), IFNy (Chujor, Klein, & Lam, 1996; Sihvola & Hurme, 1989), and intracellular calcium buffers (Martin et al., 1992).

4. REGULATION OF IL-ip BIOAVAILABILITY In addition to regulation of IL-1|3 production, regulation of IL-ip function also occurs. This clearly adds a new level of complexity to the understanding of how proinflammatory cytokines are dynamically regulated in vivo. As will be reviewed below, the function of IL-lp can be regulated by binding to decoy receptors, by competition with IL-lra for binding with IL-1 receptors, by disruption of IL-ip induced intracellular signaling by intracellular IL-lra, and by alterations in subunits of the IL1 receptor complex.

4.1. Increased IL-1 Type II Decoy Receptors Soluble type II IL-1 receptors have been referred to as decoy receptors since binding does not elicit physiological response. These IL-1 decoy receptors modulate IL-ip activity by interacting both with extracellular pro-IL-l[5 (from fibroblasts, keratinocytes, etc.) as well as by binding to extracellular mature IL-ip (Symons, Young, & Duff, 1995). Binding of pro-IL-1 (3 to decoy receptors prevents its processing to a mature form. Binding of mature IL-1 (3 block its ability to bind to the physiologically functional type I IL-1 receptors expressed by nearby cells (Symons et al., 1995). Anti-inflammatory cytokines (e.g., IL-4, IL-13) induce the expression and release of IL-1 decoy receptors, and glucocorticoids increase the expression of these anti-inflammatory cytokines (Colotta, Re, Muzio, Bertini, Polentarutti, Sironi, Giri, Dower, Sims, & Mantovani, 1993; Colotta, Re, Muzio, Polentarutti, Minty, Caput, Ferrara, & Mantovani, 1994). In addition, glucocorticoids directly increase transcription and half-life of mRNA for soluble IL-1 decoy receptors, as well as increasing the release of the IL-1 decoy receptor (Colotta et al., 1994; Re, Muzio, DeRossi, Polentarutti, Giri, Mantovani, & Colotta, 1994). Increased release of IL-1 decoy receptors also occurs in response to reactive oxygen intermediates (Sambo, Fadlon, Sironi, Matteucci, Introna, Mantovani, & Colotta, 1996). 4.2. Increased IL-1 Receptor Antagonist Interleukin-1 receptor antagonist (IL-lra) competitively inhibits binding of ILi p to its receptor.Three forms of IL-lra exist: soluble IL-lra (sIL-lra) which is released extracellularly, and two forms of intracellular IL-lra (icIL-lra, types I and II) (Levine, Wu, & Shelhamer, 1997; Mantovani & Colotta, 1995). The affinity of IL-lra for the type

Dynamic Regulation of Proinflammatory Cytokines

167

I IL-1 receptor is approximately equal to that of IL-ip. However, due to the marked "spare receptor" effect for the IL-lp system, whereby very few receptors need be bound for a physiological response to occur, IL-lra is a weak inhibitor of physiological responses to IL-ip. IL-lra has only weak affinity for the IL-1 decoy receptor, in keeping with the role of the decoy receptor in suppressing IL-1 (3 actions (Evans & Robbins, 1994). Like all other aspects of IL-1 (3 regulation, IL-lra formation can be modulated as well. Increases in sIL-lra mRNA and/or protein have been reported in response to ILl a (Vannier & Dinarello, 1993), IL-lp (Gabay, Smith, Eidlen, & Arend, 1997), IL-6 (Gabay et al., 1997), anti-inflammatory cytokines (e.g., IL-4, IL-13) (Muzio, Re, Sironi, Polentarutti, Minty, Caput, Ferrara, Mantovani, & Colotta, 1994a; Re, Mengozzi, Muzio, Dinarello, Mantovani, & Colotta, 1993), IFNy (Levine et al., 1997), and CRH (Paez et al., 1995). The effect of CRH appears to be via elevated cAMP (Paez et al., 1995), whereas IL-lp and IL-6 actions are mediated via activation of NF-kappaB and C/EBP transcription factors (Gabay et al., 1997). sIL-lra half-life is enhanced by TGFbetal (Muzio, Sironi, Polentarutti, Mantovani, & Colotta, 1994b). In contrast, sIL-lra levels are reported to not be affected by histamine (Vannier & Dinarello, 1993). icIL-lra is constitutively expressed in epithelial cells and can also be induced in other cell types (e.g., macrophages) after stimulation by LPS or phorbol ester (Jenkins, Drong, Shuck, Bienkowski, Slightom, Arend, & Smith, 1997). The function of icIL-lra has been unknown until very recently. icIL-lra has been reported to alter IL-1 P induced gene expression, not by blocking IL-ip binding to extracellular receptors, but rather by intracellular actions. That is, icIL-lra disrupts IL-ip signaUng by disrupting the effects that IL-ip induces within the cell it to which it is bound (Watson, Lofquist, Rinehart, Olsen, Makarov, Kaufman, & Haskill, 1995). Indeed, icIL-lra does not even block mRNA induction by bound IL-ip. Rather, icIL-lra disrupts the response to bound ILip by altering the stability/degradation of the mRNAs induced by bound IL-ip (Watson et al, 1995). Like sIL-lra, icIL-lra formation is regulated. icIL-lra mRNA and icILlra protein levels have been reported to not be affected by either TNF-a or TGFbetal (Kang et al., 1996). On the other hand, icIL-lra mRNA level and half-life have also been reported to be enhanced by TGFbetal (Muzio et al., 1994b). icIL-lra mRNA and protein are also increased in response to either IL-4 or IL-13 (Levine et al., 1997; Muzio et al., 1994a; Re et al., 1993), corticosteroids (Levine, Benfield, & Shelhamer, 1996; Levine et al., 1997), and IFNy (Levine et al., 1997).

4.3. Alterations in IL-1 Receptor Accessory- or IL-1 Receptor-Related Proteins IL-ip exerts its physiological effects via binding to IL-1 type I receptors (IL-IR type I). In many tissues, including brain, IL-IR type I is physically associated with a subunit termed the IL-1 receptor accessory protein (IL-lRAcP) (Gabellec, JafarianTehrani, Griffais, & Haour, 1996; Greenfeder, Nunes, Kwee, Labow, Chizzonite, & Ju, 1995; Liu, Chalmers, Maki, & DeSouza, 1996). IL-lRAcP forms a complex with IL-IR type I and either IL-la or IL-ip, but not IL-lra (Greenfeder et al., 1995). IL-lRAcP serves to increase binding affinity between the IL-IR type I and either IL-la or IL-ip (Greenfeder et al., 1995). Indeed, cell lines that express IL-IR type I without IL-lRAcP do not respond to IL-ip (Korherr, Hofmeister, Wesche, & Falk, 1997; Wesche, Neumann, Resch, & Martin, 1996). Binding of IL-1 p induces the formation of a complex

168

L. R. Watkins et al

containing both IL-IR type I and IL-lRAcP. IL-lRAcP is required for both internalization of the activated IL-1 receptor complex as well as for intracellular signaling to occur (Korherr et al., 1997). Intracellular signaling ensues following activation of IL-1 receptor associated kinase (IRAK), a serine-threonine protein kinase attached to the receptor complex via its linkage to IL-lRAcP (Huang, Gao, Li, & Cao, 1997; Volpe, Clatworthy, Kaptein, Maschera, Griffin, & Ray, 1997). Very little is known to date regarding regulation of IL-lRAcP. Chronic intracerebroventricular IL-ip increases both soluble IL-lRAcP mRNA and IL-IR type I mRNA in brain (Gayle, Ilyin, & Plata-Salaman, 1997). In contrast, LPS does not alter IL-IRAcP mRNA in brain, despite increasing IL-IR type I mRNA (Gabellec et al., 1996). In liver, LPS decreases IL-IR type I and IL-lRAcP mRNAs in parallel (Liu et al., 1996). In contrast to IL-lRAcP, IL-1 receptor-related proteins (IL-lRrp) are thought to be receptors despite the fact that this group of proteins fails to bind any known IL-1 ligands (Lovenberg, Crowe, Liu, Chalmers, Liu, Liaw, Clevenger, Oltersdorf, DeSouza, & Maki, 1996; Parnet, Garka, Bonnert, Dower, & Sims, 1996). Fit-1 proteins are also thought to be receptors distantly related to IL-IR, but bind IL-1 only with very low affinity (Reikerstorfer, Holz, Stunnenberg, & Busslinger, 1995). To date, no information is available regarding potential regulation of these IL-1 related receptor groups, but simply that they exist further substantiates the growing diversity in possibilities for IL1 regulation.

5. CONCLUSIONS As can be appreciated from the discussion above, IL-ip is regulated at the level of pre-transcriptional processes (second messengers, creation of transcription factors, etc), transcription, mRNA stability, translation of mRNA into an inactive pro-IL-ip protein, post-translational processing of pro-IL-ip into its active form, regulation of the enzyme required for post-translational processing, and release of the mature protein. The function of IL-lp is regulated as well, both by release of "decoy" receptors that bind IL-lp without creating receptor-mediated effects, by receptor accessory proteins that modulate IL-lp signaling, and by IL-lra which modulates IL-ip actions extracellularly and, apparently, intracellularly as well. What becomes apparent from this body of work is that there are many substances that influence IL-ip production and function. Some of these substances exert apparently opposing effects at different levels of gene regulation (increasing mRNA but decreasing protein release, for example); some of these substances increase IL-lp production by themselves yet decrease IL-lp production in the presence of other modulators; some exert no effect on their own yet modulate IL-lp production induced by immune stimulators such as endotoxin; some do not regulate IL-ip directly but rather cause other cells to release IL-1 decoy receptors or alter soluble or intracellular IL-lra. From this, it becomes obvious that one should naturally expect that different laboratories using different stressors will report different effects on proinflammatory cytokines, such as IL-lp. This is because different stressors with different parameters will produce different mixes of products that are capable of influencing IL-ip in differing ways and in opposing directions. The temporal dynamics of these factors will also differ for different stress paradigms, making the outcomes difficult to predict. Clearly, diverse experimental paradigms for inducing and assessing stress effects will inherently

Dynamic Regulation of Proinflammatory Cytokines

169

vary, at minimum, with regard to (a) the constellation and temporal pattern of hormonal and neural mediators released, (b) the ongoing pattern of cytokines and other immune modulators present during the stress procedure, (c) whether the immune cells are tested in the presence of this natural mix of stress-induced substances or whether the isolated cells are placed in culture, and (d) which step of gene regulation or function is being examined. Perhaps the only solution will be systematic in-depth mechanistic studies with a fixed stress paradigm. If this is done for a number of stressors, it may be that general patterns will emerge.

REFERENCES Aarstad, H. J., Kolset, S. O., & Seljelid, R. (1991). The effect of stress in vivo on the function of mouse macrophages in vitro. Scandinavian Journal of Immunology, 33(6), 673-681. Abe, M., Tanaka, Y., Saito, K., Shirakawa, F, Koyama, Y., Goto, S., & Eto, S. (1997). Regulation of interleukin (IL)-l beta gene transcription induced by IL-1 beta in rheumatoid synovial fibroblast-hke cells, E l l , transformed with simian virus 40 large T antigen. Journal of Rheumatology, 24(3), 420-429. Adachi, M., Inoue, H., Arinaga, S., Li, J., Ueo, H., Mori, M., & Akiyoshi, T. (1997). Quantitative analysis of cytokine mRNA expression in peripheral blood mononuclear cells following treatment with interleukin-2. Cancer Immunology & Immunotherapy, 44, 329-334. Adcock, I. M., Brown, C. R., Gelder, C. M., Shirasaki, H., Peters, M. J., & Barnes, P. J. (1995). Effects of glucocorticoids on transcription factor activation in human peripheral blood mononuclear cells. American Journal of Physiology, 268(2 Pt I), C331-C338. Aloe, L., Bracci-Laudiero, L., Alleva, E., Lambiase, A., micera, A., & Tirassa, P. (1994). Emotional stress induced by parachute jumping enhances blood nerve growth factor levels and the distribution of nerve growth factor receptors in lymphocytes. Proceedings of the National Academy of Sciences USA, 91(22), 10440-10444. Amano, Y., Lee, S. W., & Allison, A. C. (1993). Inhibition by glucocorticoids of the formation of interleukin1 alpha, interleukin-1 beta, and interleukin-6: mediation by decreased mRNA stabihty. Molecular Pharmacology, 43(2), 176-182. Arck, R C , MeraH, F. S., Stanisz, A. M., Stead, R. H., Chaouat, G., Manuel, J., & Clark, D. A. (1995). Stressinduced murine abortion associated with substance P-dependent alteration in cytokines in maternal uterine decidua. Biology of Reproduction, 53(4), 814-819. Arck, R C , Troutt, A. B., & Clark, D. A. (1997). Soluble receptors neutralizing TNF-alpha and IL-1 block stress-triggered murine abortion. American Journal of Reproductive Immunology, 37(3), 262-266. Bagby, G. J., Sawaya, D. E., Crouch, L. D., & Shepherd, R. E. (1994). Prior exercise suppresses the plasma tumor necrosis factor response to bacterial hpopolysaccharide. yoarna/ of Applied Physiology, 77(3), 1542-1547. Bakouche, O., Moreau, J. L., & Lachman, L. B. (1992). Secretion of IL-1: role of protein kinase C. Journal of Immunology, 148(1), 84-91. Berman, A. S., Chancellor-Freeland, C , Zhu, G., & Black, P. H. (1996). Substance P primes murine peritoneal macrophages for an augmented proinflammatory cytokine response to Upopolysaccharide. Neuroimmunomodulation, 5(2-3), 141-149. Berry, W. D., Murphy, J. D., Smith, B. A., Taylor, G. R., & Sonnenfeld, G (1991). Effect of microgravity modeling on interferon and interleukin responses in the rat. Journal of Interferon Research, 11(4), 243-249. Boyum, A., Wiik, R, Gustavsson, E., Veiby, O. R, Reseland, J., Haugen, A. H., Opstad, R K., & Byum, A. (1996). The effect of strenuous exercise, calorie deficiency and sleep deprivation on white blood cells, plasma immunoglobulins and cytokines. Scanadavian Journal of Immunology, 43(2), 228-235. Brantschen, S., Gauchat, J. F, De Week, A. L., & Stadler, B. M. (1989). Regulatory effect of recombinant interleukin (IL)3 and IL4 on cytokine gene expression of bone marrow and peripheral blood mononuclear cells. European Journal of Immunology, 19(11), 2017-2023. Bury, T. B., Louis, R., Radermecker, M. F, & Pirnay, F (1996). Blood mononuclear cells mobilization and cytokines secretion during prolonged exercises. International Journal of Sports Medicine, 17(2), 156-160. Busso, N., Collart, M., Vassalli, J. D., & Behn, D. (1987). Antagonist effect of RU 486 on transcription of glucocorticoid-regulated genes. Experimental Cell Research, 173(2), 425-430.

170

L. R. Watkins et al.

Cahill, C. M., Waterman, W. R., Xie, Y., Auron, P. E., & Calderwood, S. K. (1996). Transcriptional repression of the prointerleukin ip gene by heat shock factor 1. Journal of Biological Chemistry, 271(40), 24874-24879. Camus, G., Poortmans, J., Nys, M., Deby-Dupont, G., Duchateau, J., Deby, C., & Lamy, M. (1997). Mild endotoxaemia and the inflammatory response induced by a marathon race. Clininical Science (Colch), 92(4), 415-422. Cannon, J. G., Evans, W. J., Hughes, V. A., Meredith, C. N., & Dinarello, C. A. (1986). Physiological mechanisms contributing to increased interleukin-1 secretion. Journal of Applied Physiology, 61(5), 1869-1874. Caput, D., Beutler, B., Hartog, K., Thayer, R., Brown-Shimer, S., & Cerami, A. (1986). Identification of a common nucleotide sequence in the 3'-untranslated region of mRNA molecules specifying inflammatory mediators. Proceedings of the National Academy of Sciences USA, 83, 1670-1674. Castell, L. M., Poortmans, J. R., Leclercq, R., Brasseur, M., Duchateau, J., & Newsholme, E. A. (1997). Some aspects of the acute phase response after a marathon race, and the effects of glutamine supplementation. European Journal of Applied Physiology, 75(1), 47-53. Chancellor-Freeland, C , Zhu, G. R, Kage, R., Beller, D. I., Leeman, S. E., & Black, R H. (1995). Substance P and stress-induced changes in macrophages. Annul of the New York Academy of Sciences, 77/, 472-484. Chandra, G., Cogswell, J. R, Miller, L. R., Godlevski, M. M., Stinnett, S. W., Noel, S. L., Kadwell, S. H., Kost, T. A., & Gray, J. G. (1995). Cyclic AMP signaling pathways are important in IL-1 beta transcriptional regulaiion. Journal of Immunology, 155(10), 4535-4543. Chantry, D., Turner, M., Abney, E., & Feldmann, M. (1989). Modulation of cytokine production by transforming growth iactOT-heta. Journal of Immunology, 142(12), 4295-4300. Cheng, G. J., Morrow-Tesch, J. L., Beller, D. I., Levy, E. M., & Black, P. H. (1990). Immunosuppression in mice induced by cold water stress. Brain Behavior and Immunity, 4(4), 278-291. Chin, J. & Kostura, M. J. (1993). Dissociation of IL-1|3 synthesis and secretion in human blood monocytes stimulated with bacterial cell wall products. Journal of Immunology, 151, 5574-5585. Chujor, C. S., Klein, L., & Lam, C. (1996). Selective inhibition of interleukin-1 beta gene expression in activated RAW 264.7 macrophages by interferon-gamma. European Journal of Immunology, 26(6), 1253-1259. Cogswell, J. R, Godlevski, M. M., Wisely, G. V., Clay, W. C . Leesnitzer, L. M., Ways, J. P, & J.G, G. (1994). NFkappa B regulates IL-1 beta transcription through a consensus NF-kappa B binding site and a nonconsensus CRE-like site. Journal of Immunology 153(2),1\2-123. Collart, M. A., Belin, D., Vassalli, J. D., de Kossodo, S., & Vassalh, P. (1986). Gamma interferon enhances macrophage transcription of the tumor necrosis factor/cachectin, interleukin-1, and urokinase genes, which are controlled by short-lived repressors. Journal of Experimental Medicine, 164(6), 2113-2118. Colotta, F, Re, F, Muzio, M., Bertini, R., Polentarutti, N., Sironi, M., Giri, J. G., Dower, S. K., Sims, J. E., & Mantovani, A. (1993). Interleukin-1 Type II Receptor: A Decoy Target for IL-1 That is Regulated by IL-4. Science, 261, 472-475. Colotta, F, Re, F, Muzio, M., Polentarutti, N., Minty, A., Caput, D., Ferrara, P., & Mantovani, A. (1994). Interleukin-13 induces expression and release of interleukin-1 decoy receptor in human polymorphonuclear cells. Journal of Biological Chemistry, 269(17), 12403-12406. Crump, J. W, Geist, L. J., Auron, P E., Webb, A. C , Stinski, M. F , & Hunninghake, G. W. (1992). The immediate early genes of human cytomegalovirus require only proximal promoter elements to upregulate expression of interleukin-1 beta. American Journal of Respiratory Cell Molecular Biology, 6(6), 674-677. Cunnick, J. E., Kojic, L. D., & Hughes, R. A. (1994). Stress-induced changes in immune function are associated with increased production of an interleukin-1-like factor in young domestic fowl. Brain Behavior and Immunity, 8(2), 123-136. DeRijk, R., Michelson, D., Karp, B., Petrides, J., Galliven, E., Deuster, P, Paciotti, G., Gold, P W, & Sternberg, E. M. (1997). Exercise and circadian rhythm-induced variations in plasma Cortisol differentially regulate interleukin-1 beta (IL-1 beta), IL-6, and tumor necrosis factor-alpha (TNF-alpha) production in humans: high sensitivity of TNF-alpha and resistance of IL-6. Journal of Clinical Endocrinology and Metabolism, 82(7), 2182-2191. DeRijk, R. H., Boelen, A., Tilders, F J., & Berkenbosch, F (1994). Induction of plasma interleukin-6 by circulating adrenaline in the rat. Psychoneuroendocrinology, 19(2), 155-163. Dinarello, C. A. (1992). Dissociation of transcription from translation of human IL-1 beta: induction of steady state mRNA by adherence or recombinant C5a in the absence of translation. Proceedings of the Society for Experimental Biology and Medicine, 200(2), 228-232. Dobbin, J. P, Harth, M., McCain. G. A., Martin. R. A., & Cousin, K. (1991). Cytokine production and lymphocyte transformation during stress. Brain Behavior and Immunity. 5(4), 339-348.

Dynamic Regulation of Proinflammatory Cytokines

171

Dobbs, C. M., Feng, N., Beck, F. M., & Sheridan, J. F. (1996). Neuroendocrine regulation of cytokine production during experimental influenza viral infection: effects of restraint stress-induced elevation in endogenous corticosterone. TioMma/ of Immunology, 157(5), 1870-1877. Donnelly, R. P., Fenton, M. J., Kaufman, J. D., & Gerrard,T. L. (1991). IL-1 expression in human monocytes is transcriptionally and posttranscriptionally regulated by IL-4. Journal of Immunology, 146(10), 3431-3436. Drenth, J. P., VanUum, S. H., VanDeuren, M., & Pesman, G. J. (1995). Endurance run increases circulating IL6 and IL-lra but downregulates ex vivoTNF-alpha and IL-1 beta pvoduclion. Journal of Applied Physiology, 79,1497-1503. Dufaux, B. & Order, U. (1989). Plasma elastase-alpha 1-antitrypsin, neopterin, tumor necrosis factor, and soluble interleukin-2 receptor after prolonged exercise. International Journal of Sports Medicine, 10(6), 434^38. Dugue, B., Leppanen, E. A.,Teppo, A. M., Fyhrquist, F., & Grasbeck, R. (1993). Effects of psychological stress on plasma interleukins-1 beta and 6, C-reactive protein, tumour necrosis factor-alpha, anti-diuretic hormone, and serum Cortisol. Scandanavian Journal of Clinical Laboratory Investigations, 53(6), 555-561. Elias, J. A., Reynolds, M. M., Kotloff, R. M., & Kern, J. A. (1989). Fibroblast interleukin-1 beta: synergistic stimulation by recombinant interleukin-1 and tumor necrosis factor and posttranscriptional regulation. Proceedings of the National Academy of Sciences USA, 86(16), 6171-6175. Espersen, G. T., Elbaek, A., Ernst, E.,Toft, E., Kaalund, S., Jersild, C , & Grunnet, N. (1990). Effect of physical exercise on cytokines and lymphocyte subpopulations in human peripheral blood. Acta Pathology Microbiology Immunology Scandanavia, 98(5), 395-400. Espersen, G. T., Elbaek, A., Schmidt-Olsen, S., Ejlersen, E., Varming, K., & Grunnet, N. (1996). Short-term changes in the immune system of elite swimmers under competition conditions. Different immunomodulation induced by various types of sport. Scandanavian Journal of Medical Science of Sports, 6(3), 156-163. Evans, C. H. & Robbins, P. D. (1994). The interleukin-1 receptor antagonist and its delivery by gene transfer. Receptor, 4(1), 9-15. Fantuzzi, G., Di Santo, E., Sacco, S., Genigni, F., & Ghezzi, P. (1995). Role of the hypothalamuspituitary-adrenal axis in the regulation of TNF production in mice. Effect of stress and inhibition of endogenous glucocorticoids. Journal of Immunology, 155(7), 3552-3555. Fenton, M. J. (1992). Review: transcriptional and post-transcriptional regulation of interleukin-1 gene expression. International Journal of Immunopharmacology, 14(3), 401-411. Fenton, M. J., Vermeulen, M. W., Clark, B. D., Webb, A. C , & Auron, P E. (1988). Human pro-IL-1 beta gene expression in monocytic cells is regulated by two distinct pathways. Journal of Immunology, 140(7), 2267-2273. Fernandez, M. C , Walters, J., & Marucha, P. (1996). Transcriptional and post-transcriptional regulation of GM-CSF-induced IL-1 beta gene expression in PMN. J Leukocyte Biology, 59(4), 598-603. Ferrari, D , Chiozzi, P, Falzoni, S., Dal Susino, M., Melchiorri, L., Baricordi, O. R., & Di Virgilio, F (1997). Extracellular ATP triggers IL-1 beta release by activating the purinergic P2Z receptor of human macrophages. Journal of Immunology, 159(3), 1451-1458. Gabay, C , Smith, M. F., Eidlen, D., & Arend, W. P. (1997). Interleukin 1 receptor antagonist (IL-lRa) is an acute-phase protein. Journal of Clinical Investigation, 99(12), 2930-2940. Gabellec, M. M., Jafarian-Tehrani, M., Griffais, R., & Haour, F. (1996). Interleukin-1 receptor accessory protein transcripts in the brain and spleen: kinetics after peripheral administration of bacterial lipopolysaccharide in mice. Neuroimmunomodulation, 3, 304-309. Gannon, G. A., Rhind, S. G , Suzui, M., Shek, P M., & Shephard, R. J. (1997). Circulating levels of peripheral blood leucocytes and cytokines following competitive cycling. Canadian Journal of Applied Physiology, 22(2), 133-147. Gayle, D,, Ilyin, S. E., & Plata-Salaman, C. R. (1997). Central nervous system IL-1 beta system and neuropeptide Y mRNAs during IL-1 beta-induced anorexia in rats. Brain Research Bulletin, 44,311-317. Geiger, T , Rordorf, C , Galakatos, N., Seligmann, B., Henn, R., Lazdins, J., & Vosbeck, K. (1992). Recombinant human C5a induces transcription but not translation of interleukin-1 beta mRNA in human monocytes. Research in Immunology, 143,117-123. Glaser, R., Rice, J., Speicher, C. E., Stout, J. C , & Kiecolt-Glaser, J. K. (1986). Stress depresses interferon production by leukocytes concomitant with a decrease in natural killer cell activity. Behavioral Neurosciences, 100, 675-678. Gorospe, M., Kumar, S., & Baglioni, C. (1993). Tumor necrosis factor increases stability of interleukin-1 mRNA by activating protein kinase C. Journal of Biological Chemistry, 268(9), 6214-6220. Goujon, E., Parnet, R, Laye, S., Combe, C , Kelley, K. W., & Dantzer, R. (1995). Stress downregulates

172

L. R. Watkins et al.

lipopolysaccharide-induced expression of proinflammatory cytokines in the spleen, pituitary, and brain of mice. Brain Behavior and Immunity, 9(4), 292-303. Greenfeder, S. A., Nunes, P., Kwee, L., Labow, M., Chizzonite, R. A., & Ju, G. (1995). Molecular cloning and characterization of a second subunit of the interleukin-1 receptor complex. Journal of Biological Chemistry, 270,13757-13765. Griffiths, R. J., Stam, E. J., Downs, J. T., & Otterness, I. G. (1995). ATP induces the release of IL-1 from LPSprimed cells in vivo. Journal of Immunology, 154(6), 2821-2828. Haahr, P. M., Pedersen, B. K., Fomsgaard, A., Tvede, N., Diamant, M., Klarlund, K., Halkjaer-Kristensen, J., & Bendtzen, K. (1991). Effect of physical exercise on in vitro production of interleukin-1, interleukin6, tumour necrosis factor-alpha, interleukin-2, and interferon-gamma. International Journal of Sports Medicine, 12(2), 223-227. Hazuda, D. J., Strickler, J., Kueppers, E, Simon, P. L., & Young, P. R. (1990). Processing of precursor interleukin-1 beta and inflammatory disease. Journal of Biological Chemistry, 265(11), 6318-6322. Herzyk, D. J., Allen, J. N., Marsh, C. B., & Wewers, M. D. (1992). Macrophage and monocyte ILl-beta regulation differs at multiple sites. Messenger RNA expression, translation, and post-translational processing. Journal of Immunology, 149, 3052-3058. Hetier, E., Ayala, J., Bousseau, A., & Prochiantz, A. (1991). Modulation of interleukin-1 and tumor necrosis factor expression by beta-adrenergic agonists in mouse ameboid microglial cells. Experimental Brain Research, 86(2), 407-413. Hiscott, J., Marois, J., Garoufalis, J., D'Addario, M., Roulston, A., Kwan, I., Pepin, N., Lacoste, J., Nguyen, H., Bensi, G., & Fenton, M. (1993). Characterization of a functional NF-kappa B site in the human interleukin-1 beta promoter: evidence for a positive autoregulatory loop. Molecular and Cellular Biology, 13(10), 6231-6240. Home, L., Bell, G., Fisher, B., Warren, S., & Janowska-Wieczorek, A. (1997). Interaction between Cortisol and tumour necrosis factor with concurrent resistance and endurance training. Clinical Journal of Sports Medicine, 7(4), 247-251. Huang, J., Gao, X., Li, S., & Cao, Z. (1997). Recruitment of IRAK to the interleukin-1 receptor complex requires interleukin-1 receptor accessory protein. Proceedings of the National Academy of Sciences, 94,12829-12832. Hunninghake, G. W., Monks, B. G , Geist, L. J., Monick, M. M., Monroy, M. A., Stinski, M. F , Webb, A. C , Dayer, J. M., Auton, P. E., & Fenton, M. J. (1992). The functional importance of a cap site-proximal region of the human prointerleukin-1 beta gene is defined by viral protein transactivation. Molecular and Cellular Biology, 12(8), 3439-3448. Iwaz, K., Kouassi, E., Lafont, S., & Revillard, J. P. (1990). Elevation of cyclic adenosine monophosphate levels independently down regulates IL-1, IL-2, and IL-2 receptor (CD25) synthesis. International Journal of Immunopharmacology, 12(6), 631-637. Jansky, L., Pospisilova, D., Honzova, S., Ulicny, B., Sramek, P., Zeman, V., & Kaminkova, J. (1996). Immune system of cold-exposed and cold-adapted humans. European Journal of Applied Physiology, 72(5-6), 445-450. Jenkins, J. K., Drong, R. F , Shuck, M. E., Bienkowski, M. J., Slightom, J. L., Arend, W P, & Smith, M. F J. (1997). Intracellular IL-1 receptor antagonist promoter: cell type-specific and inducible regulatory regions. Journal of Immunology, 158(2), 748-755. Kang, K., Hammerbert, C., & Cooper, K. D. (1996). Differential regulation of IL-1 and IL-1 receptor antagonist in HaCaT keratinocytes by tumor necrosis factor-alpha and transforming growth factor-beta 1. Experimental Dermatology, 5(4), 218-226. Kaushansky, K., Broudy, V. C , Harlan, J. M., & Adamson, J. W. (1988). Tumor necrosis factor-alpha and tumor necrosis factor-beta (lymphotoxin) stimulate the production of granulocyte-macrophage colonystimulating factor, macrophage colony-stimulating factor, and IL-1 in vivo. Journal of Immunology, 74/f/Oj, 3410-3415. Keane, K. M., Giegel, D. A., Lipinski, W. J., Callahan, M. G., & Shivers, B. D. (1995). Cloning, tissue expression, and regulation of rat interleukin-1 beta converting enzyme. Cytokine, 7(2), 105-110. Kent, S., Bluthe, R.-M., Kelley, K. W., & Dantzer, R. (1992). Sickness behavior as a new target for drug development. Trends in Pharmacological Sciences, 13, 24-28. Kern, J. A., Lamb, R. J., Reed, J. C , Daniele, R. P., & Nowell, P. C. (1988). Dexamethasone inhibition of interleukin 1 beta production by human monocytes. Posttranscriptional mechanisms. Journal of Clinical Investigation, 81(1), 237-244. Khlusov, I. A., Dygai, A. M., & Goldberg, E. D. (1993). The adrenergic regulation of interleukin production by bone marrow cells during immobilization stress. Bioligcal and Experimental Biology and Medicine, 116(12), 570-572.

Dynamic Regulation of Proinflammatory Cytoi(ines

173

Kiecolt-Glaser, J. K., Glaser, R., Gravenstein, S., Malarkey,W. B., & Sheridan, J. (1996). Chronic stress aUers the immune response to influenza virus vaccine in older adults. Proceedings of the National Academy of Sciences USA, 93(7), 3043-3047. Kiecolt-Glaser, J. K., Marucha, P. T, Malarkey, W. B., Mercado, A. M., & Glaser, R. (1995). Slowing of wound healing by psychological stress. Lancet, 546(8984), 1194-1196. KimberUn, D. W., Willis, S. A., Jr., M. G. H., & Nisen, P. D. (1995). Protein synthesis-dependent induction of interleukin-1 beta by lipopolysaccharide is inhibited by dexamethasone via mRNA destabilization in human astrogUal cells. Journal of Clinical Immunology, 15, 199-204. Kitamura, H., Konno, A., Morimatsu, M., Jung, B. D., Kimura, K., & Saito, M. (1997). Immobilization stress increases hepatic IL-6 expression in mice. Biochemistry and Biophysics Research Communications, 238(3), 707-111. Kluger, M. I , Rudolph, K., Soszynski, D., Conn, C. A., Leon, L. R., Kozak, W., Wallen, E. S., & Moseley, P L. (1997). Effect of heat stress on LPS-induced fever and tumor necrosis factor. American Journal of Physiology, 273, R858-R863. Knudsen, P. J., Dinarello, C. A., & Strom, T. B. (1986). Prostaglandins posttranscriptionally inhibit monocyte expression of interleukin 1 activity by increasing intracellular cyclic adenosine monophosphate. Journal of Immunology, 137(10), 3189-3194. Koj, A. (1996). Initiation of acute phase response and synthesis of cytokines. Biochimica. Biophysica Acta, 1317,84-94. Kopydlowski, K. M., McVey, D. S., Woods, K. M., landolo, J. J., & Chapes, S. K. (1992). Effects of antiorthostatic suspension and corticosterone on macrophage and spleen cell function. Journal of Leukocyte Biology, 52(2), 202-208. Korherr, C , Hofmeister, R., Wesche, H., & Falk, W. (1997). A critical role for interleukin-1 receptor accessory protein in interleukin-1 signaling. European Journal of Immunology, 27, 262-267. Korneva, E. A., Rybakina, E. G., Fomicheva, E. E., Kozinets, 1. A., & Shkhinek, E. K. (1992). Altered interleukin-1 production in mice exposed to rotation stress. International Journal of Tissue Reactivity, 14(5), 219-224. Korneva, E. A., Rybakina, E. G., Orlov, D. S., Shamova, O. V., Shanin, S. N., & Kokryakov, V. N. (1997). Interleukin-1 and defensins in thermoregulation, stress, and immunity. Annals of the New York Academy of Sciences, 813, 465-473. Koyama, Y., Tanaka, Y., Saito, K., Abe, M., Nakatsuka, K., Morimoto, I., Auron, P. E., & Eto, S. (1996). Crosshnking of intercellular adhesion molecule 1 (CD54) induces AP-1 activation and IL-lbeta transcription. Journal of Immunology, 157(11), 5097-5103. Krakauer, T. (1996). Evidence for protein kinase C pathway in the response of human peripheral blood mononuclear cells to cholera toxin. Cellular Immunology, 172(2), 224-228. Kruys, v., Marinx, O., Shaw, G., Deschamps, J., & Huez, G. (1989).Translational blockade imposed by cytokinederived UA-rich sequences. Science, 245, 852-855. Kusnecov, A. W, Shurin, M. R., Armfield, A., Litz, J., Wood, P, Zhou, D., & Rabin, B. S. (1995). Suppression of lymphocyte mitogenesis in different rat strains exposed to footshock during early diurnal and nocturnal time periods. Psychoneuroendocrinology, 20(8), 821-835. Kvernmo, H., Olsen, J. O., & Osterud, B. (1992). Changes in blood cell response following strenuous physical exercise. European Journal of Applied Physiology, 64(4), 318-322. Laliberte, R., Perregaux, D., Svensson, L., Pazoles, C. J., & Gabel, C. A. (1994). Tenidap modulates cytoplasmic pH and inhibits anion transport in vitro II. Inhibitation of IL-1 beta production from ATP-treated monocytes and macrophages. yoMrna/ of Immunology, 153(5), 2168-2179. Laliberte, R. E., Perregaux, D. G., McNiff, P, & Gabel, C. A. (1997). Human monocyte ATP-induced IL-1 beta posttranslational processing is a dynamic process dependent on in vitro growth conditions. Journal of Leukocyte Biology, 62(2), 227-239. Laye, S., Goujon, E., Combe, C , VanHoy, R., Kelley, K. W., Parnet, P, & Dantzer, R. (1996). Effects of lipopolysaccharide and glucocorticoids on expression of interleukin-1 beta converting enzyme in the pituitary and brain of mice. J Neuroimmunology, 68(1-2), 61-66. Lebedeva, T. V. & Singh, A. K. (1997). Constitutive activity of the murine IL-1 beta promoter is regulated by a transcriptional repressor. Biochimica Biophysica Acta, 1353(1), 32-38. Lee, S. W.,Tsou, A. R, Chan, H., Thomas, J., Petrie, K., Eugui, E. M., & Allison, A. C. (1988). Glucocorticoids selectively inhibit the transcription of the interleukin-1 beta gene and decrease the stability of interleukin-1 beta mRNA. Proceedings of the National Academy of Sciences USA, 85(4), 1204-1208. LeMay, L. G, Vander, A. J., & Kluger, M.J. (1990). The effects of psychological stress on plasma interleukin6 activity in rats. Physiology and Behavior, 47(5), 957-961. Levine, S. J., Benfield, T, & Shelhamer, J. H. (1996). Corticosteroids induce intracellular interleukin-1

174

L. R. Watkins et al.

receptor antagonist type I expression by a human airway epithelial cell line. American Journal of Respiratory and Cellular Molecular Biology, 15(2), 245-251. Levine, S. J., Wu, T, & Shelhamer, J. H. (1997). Extracellular release of the type 1 intracellular IL-1 receptor antagonist from human airway epithelial cells: differential effects of IL-4, IL-13, IFN-gamma, and corticosteroids. Journal of Immunology, 158(12), 5949-5957. Lewicki, R., Tchorzewski, H., Majewska, E., Nowak, Z., & Baj, Z. (1988). Effect of maximal physical exercise on T-lymphocyte subpopulations and on interleukin-1 (IL-1) and interleukin-2 (IL-2) production in vitro. International Journal of Sports Medicine, 9, 114-117. Liao, J., Keiser, J. A., Scales, W. E., Kunkel, S. L., & Kluger, M. J. (1995a). Role of corticosterone in TNF and IL-6 production in isolated perfused rat liver. American Journal of Physiology, 268, R699-R706. Liao, J., Keiser, J. A., Scales, W. E., Kunkel, S. L., & Kluger, M. J. (1995b). Role of epinephrine in TNF and IL-6 production from isolated perfused rat liver. American Journal of Physiology, 268, R896-R901. Liu, C , Chalmers, D., Maki, R., & DeSouza, E. B. (1996). Rat homolog of mouse interleukin-1 receptor accessory protein: cloning, localization, and modulation studies. Journal of Neuroimmunology, 66, 41-48. Lorenz, J. J., Furdon, P. J., Taylor, J. D., Verghese, M. W., Chandra, G., Kost, T A., Haneline, S. A., Roner, L. A., & Gray, J. G. (1995). A cyclic adenosine 3',5'-monophosphate signal is required for the induction of IL-1 beta by TNF-alpha in human monocytes. Journal of Immunology, 155(2), 836-844. Lovenberg, T W., Crowe, P. D., Liu, C , Chalmers, D. T, Liu, X. J., Liaw, C , Clevenger, W., Oltersdorf, T, DeSouza, E. B., & Maki, R. A. (1996). Cloning of a cDNA encoding a novel interleukin-1 receptor related protein (IL \R-Tp2). Journal of Neuroimmunology, 70, 113-122. Maier, S. F. & Watkins, L. R. (1998). Cytokines for psychologists: implications of bi-directional immune-tobrain communication for understanding behavior, mood, and cognition. Psychological Review, 105, 83-107. Mantovani, A. & Colotta, F. (1995). Cloning and characterization of a newisoform of the interleukin-1 receptor antagonist. Journal of Experimental Medicine, 182(2), 623-628. Martin, F. C , Charles, A. C , Sanderson, M. J., & Merrill, J. E. (1992). Substance P stimulates IL-1 production by astrocytes via intracellular calcium. Brain Research, 599(1), 13-18. McCarthy, D. O., Ouimet, M. E., & Daun, J. M. (1992). The effects of noise stress on leukocyte function in rats. Research in Nursing and Health. 15, 131-137. McCracken, B. A., Gaskins, H. R., Ruwe-Kaiser, P J., Klasing, K. C , & Jewell, D. E. (1995). Diet-dependent and diet-independent metabolic responses underlie growth stasis of pigs at v^eamng. Journal of Nutrition, 125(11), 2838-2845. Mekaouche, M., Givalois, L., Barbanel, G., Siaud, P., Maurel, D., Malaval, F., Bristow, A. F, Boissin, J., Assenmacher, I., & Ixart, G. (1994). Chronic restraint enhances interleukin-1-beta release in the basal state and after an endotoxin challenge, independently of adrenocorticotropin, and corticosterone release. Neuroimmunomodulation, 1,292-299. Miller, D. K., Calaycay, J. R., Chapman, K. T, Howard, A. D., Kostura, M. J., Molineaux, S. M., & Thronberry, N. A. (1993a). The IL-1 beta converting enzyme as a therapeutic target. Annals of the New York Academy of Sciences, 696, 133-148. Miller, E. S., Koebel, D. A., & Sonnenfeld, G. (1993b). The metabolic stressor 2-deoxy-D-glucose (2-DG) enhances LPS-stimulated cytokine production in mice. Brain Behavior and Immunity, 7(4), 317-325. Minami, M., Kuraishi, Y., Yamaguchi,T.,Nakai, S., Hirai, Y., & Satoh, M. (1991). Immobilization stress induces interleukin-1 beta mRNA in the rat hypothalamus. Neuroscience Letters, 123(2), 254-256. Miossec, P., Briolay, J., Dechanet, J., Wijdenes, J., Martinez-Valdez, H., & Banchereau, J. (1992). Inhibition of the production of proinflammatory cytokines and immunoglobulins by interleukin-4 in an ex vivo model of rheumatoid synovitis. Arthritis and Rheumatology. 35(8), 874-883. Mizutani, H., Schechter, N., Lazarus, G., Black, R. A., & Kupper,T. S. (1991). Rapid and specific conversion of precursor interleukin-1 beta (IL-1 beta) to an active IL-1 species by human mast cell chymase. Journal of Experimental Medicine, 174(4). 821-825. Muzio. M., Re, E, Sironi, M., Polentarutti, N., Minty, A., Caput. D., Ferrara, P., Mantovani, A., & Colotta, F (1994a). Interleukin-13 induces the production of interleukin-1 receptor antagonist (IL-lra) and the expression of the mRNA for the intracellular (keratinocytc) form of IL-lra in human myelomonocytic cells. Blood, 83(7), 1738-1743. Muzio, M., Sironi, M., Polentarutti, N , Mantovani, A., & Colotta, F (1994b). Induction by transforming growth factor-beta 1 of the interleukin-1 receptor antagonist and of its intracellular form in human polymorphonuclear cells. European Journal of Immunology, 24(12), 3194-3198. Natelson, B. H., Zhou, X., Ottenweller, J. E., Bergen, M. T, Sisto, S. A., Drastal. S.. Tapp, W. N., & Cause, W. L. (1996). Effect of acute exhausting exercise on cytokine gene expression in men. International Journal of Sports Medicine. 17(4). 299-302.

Dynamic Regulation of Proinflammatory Cytoi(ines

175

Newton, R.,Adcock, I. M.,& Barnes, P. J. (1996).Superinduction of NF-kappa B by actinomycin D andcycloheximide in epithelial cells. Biochemical Biophysical Research Communications, 218(2), 518-523. Nguyen, K. T., Deak, T., Owens, S. M., Kohno, T., Fleshner, M., Watkins, L. R., & Maier, S. F. (1998). Exposure to acute stress induces brain interleukin-1 beta protein in the rat. Journal of Neurosciences, 18, 2239-2246. Northoff, H., Weinstock, C , & Berg, A. (1994). The cytokine response to strenuous exercise. International Journal of Sports Medicine, /5(Suppl 3), S167-171. Ohmori, Y. & Hamilton,T. A. (1992). Ca2+ and calmodulin selectively regulate lipopolyysaccharide-inducible cytokine mRNA expression in murine peritoneal macrophages. Journal of Immunology, 148(2), 538-545. Ohmori, Y., Strassman, G., & Hamilton, T. A. (1990). cAMP differentially regulates expression of mRNA encoding IL-1 alpha and IL-] beta in murine peritoneal macrophages. Journal of Immunology, 145(10), 3333-3339. Okamoto, H., Oh, C , & Nakano, K. (1990). Possible involvement of adenosine 3':5'-cyclic monophosphate and extracellular calcium ions in histamine stimulation of interleukin-1 release from macrophage-like P388D1 cells. Immunology, 70(2), 186-190. Oster, W., Brach, M. A., Gruss, H. J., Mertelsmann, R., & Herrmann, F. (1992). Interleukin-1 beta (IL-1 beta) expression in human blood mononuclear phagocytes is differentially regulated by granulocytemacrophage colony-stimulating factor (GM-CSF), M-CSE and IL-3. Blood, 79(5), 1260-1265. Paez, R M., Sauer, J., Perez, C. C , Finkielman, S.. Stalla, G. K., Holsboer, F, & Arzt, E. (1995). Corticotropinreleasing hormone differentially modulates the interleukin-1 system according to the level of monocyte activation by endotoxin. Endocrinology, 136(12), 5504-5510. Pan, Z. K., Zuraw, B. L., Lung, C. C , Prossnitz, E. R., Browning, D. D., & Ye, R. D. (1996). Bradykinin stimulates NF-kappaB activation and interleukin-1 beta gene expression in cultured human fibroblasts. Journal of Clinical Investigation, 98(9), 2042-2049. Papanicolaou, D. A., Petrides, J. S., Tsigos, C , Bina. S., Kalogeras, K. T., Wilder, R., Gold, P. W., Deuster, P. A., & Chrousos, G. P. (1996). Exercise stimulates interleukin-6 secretion: inhibition by glucocorticoids and correlation with catecholamines. American Journal of Physiology, 271(3 Pt 1), E6()l-6()5. Parnet, P., Garka, K. E., Bonnert, T. P., Dower, S. K., & Sims. J. E. (1996). IL-lRrp is a novel receptor-like molecule similar to the type I interleukin-1 receptor and its homologues T1/ST2 and IL-IR AcP. Journal of Biological Chemistry, 271, 3967-3970. Perregaux, D. G., Laliberte, R. E., & Gabel, C. A. (1996a). Human monocyte interleukin-1 beta posttranslational processing. Evidence of a volume-regulated response. Journal of Biological Chemistry. 271 (147), 29830-29838. Perregaux, D. G., Svensson, L., & Gabel, C. A. (1996b).Tenidap and other anion transport inhibitors disrupt cytolytic T lymphocyte-mediated IL-1 beta post-translational processing. Journal of Immunology, 157(1), 57-64. Persoons, J. H., Schornagel, K., Breve, J., Berkenbosch, F, & Kraal, G. (1995). Acute stress affects cytokines and nitric oxide production by alveolar macrophages differently. American Journal of Respiratory and Critical Care Medicine, 152(2), 619-624. Raingeaud, J., Gupta, S., Rogers, J. S., Dickens, M., Han, J., & Ulevitch, R. J. (1995). Proinflammatory cytokines and environmental stress cause p38 mitogen-activated protein kinase activation by dual phosphorylation on tyrosine and threonine. Journal of Biological Chemistry, 270, 7420-7426. Rail, L. C, Roubenoff, R., Cannon. J. G., Abad, L. W., Dinarello, C. A., & Meydani. S. N. (1996). Effects of progressive resistance training on immune response in aging and chronic inflammation. Medical Science and Sports Exercise, 28(11), 1356-1365. Ray, A. & Prefontaine, K. E. (1994). Physical association and functional antagonism between the p65 subunit of transcription factor NF-kappa B and the glucocorticoid receptor. Proceedings of the National Academy of Sciences, 91, 752-756. Re, F, Mengozzi, M., Muzio, M., Dinarello, C. A., Mantovani, A., & Colotta, F (1993). Expression of interleukin-1 receptor antagonist (IL-lra) by human circulating polymorphonuclear cells. European Journal of Immunology, 23(2), 570-573. Re, F, Muzio, M., DeRossi, M., Polentarutti, N., Giri, J. G., Mantovani, A., & Colotta, F (1994). The type II "receptor" as a decoy target for interleukin-1 in polymorphonuclear leukocytes: characterization of induction by dexamethasone and ligand binding properties of the released decoy receptor. Journal of Experimental Medicine, 179(2). 739-743. Reikerstorfer, A., Holz, H., Stunnenberg, H. G., & Busslinger, M. (1995). Low affinity binding of interleukin1 beta and intracellular signaling via NF-kappa B identify Fit-1 as a distant member of the interleukin1 receptor family. Journal of Biological Chemistry, 270, 17645-17648.

176

L. R. Watkins et al

Ribeiro, S. P., Villar, J., Downey, G. P., Edelson, J. D., & Slutsky, A. S. (1996). Effects of the stress response in septic rats and LPS-stimulated alveolar macrophages: evidence for TNF-alpha posttranslational regulation. American Journal of Respiratory and Critical Care Medicine, 154(6 Pt 1), 1843-1850. Rivier, A., Pene, J., Chanez, P, Anselme, R, Caillaud, C , Prefaut, C , Godard, R, & Bousquet, J. (1994). Release of cytokines by blood monocytes during strenuous exercise. International Journal of Sports Medicine, 15(4), 192-198. Rizzo, M. T. & Carlo-Stalla, C. (1996). Arachidonic acid mediates interleukin-1 and tumor necrosis factoralpha-induced activation of the c-jun amino-terminal kinases in stromal cells. Blood, 88, 3792-3800. Rosette, C. & Karin, M. (1996). Ultraviolet Hght and osmotic stress: activation of the JNK cascade through multiple growth facto and cytokine receptors. Science, 274,1194-1197. Sambo, P., Fadlon, E. J., Sironi, M., Matteucci, C , Introna, M., Mantovani, A., & Colotta, F. (1996). Reactive oxygen intermediates cause rapid release of the interleukin-1 decoy receptor from human myelomonocytic cells. Blood, 87(5), 1682-1686. Scheinman, R. I., Gualberto, A., Jewell, C. M., Cidlowski, J. A., & Baldwin, A. S. J. (1995). Characterization of mechanisms involved in transrepression of NF-kappa B by activated glucocorticoid receptors. Molecular and Cellular Biology, 15(2), 943-953. Schluns, K. S., Cook, J. E., & Le, P. T. (1997). TGF-beta differentially modulates epidermal growth factormediated increases in leukemia-inhibitory factor, IL-6, IL-1 alpha, and IL-1 beta in human thymic epithelial cells. Journal of Immunology, 158(6), 2704-2712. Schulte, H. M., Bamberger, C. M., Elsen, H., Herrmann, G., Bamberger, A. M., & Barth, J. (1994). Systemic interleukin-1 alpha and interleukin-2 secretion in response to acute stress and to corticotropinreleasing hormone in humans. European Journal of Clinical Investigation, 24(11), 773-777. Serkkola, E. & Hurme, M. (1993). Synergism between protein-kinase C and cAMP-dependent pathways in the expression of the interleukin-1 beta gene is mediated via the activator-protein-1 (AP-1) enhancer activity. European Journal of Biochemistry, 213(1), 243-249. Serkkola, E. M. (1995). Regulation of interleukin-1 gene expression by protein kinase C and cychc AMP. Dissertation Abstracts International, 56(1), 230. Shapira, L.,Takashiba, S., Champagne, C , Amar, S., & VanDyke,T. E. (1994). Involvement of protein kinase C and protein tyrosine kinase in lipopolysaccharide-induced TNF-alpha and IL-1 beta production by human monocytes. ioMrna/ of Immunology, 153, 1818-1824. Shintani, F , Nakaki, T., Kanba, S., Sato, K., Yagi, G., Shiozawa, M., Aiso, S., Kato, R., & Asai, M. (1995). Involvement of interleukin-1 in immobilization stress-induced increase in plasma adrenocorticotropic hormone and in release of hypothalamic monoamines in the rat. Journal of Neuroscience, 15, 1961-1970. Shizuya, K., Komori,T., Fujiwara, R., Miyahara, S., Ohmori, M., & Nomura, J. (1997). The influence of restraint stress on the expression of mRNAs for IL-6 and the IL-6 receptor in the hypothalamus and midbrain of the rat. Life Sciences, 61(10), PL 135-140. Sihvola, M. & Hurme, M. (1989). The effect of gamma interferon on interleukin-1 release of human monocytes. Dependence on the activating agent and correlation to the mRNA levels. Scandanavian Journal of Immunology, 29(6), 689-698. Singer, 1.1., Scott, S., Chin, J., Bayne, E. K., Limjuco, G., Weidner, J., Miller, D. K., Chapman, K., & Kostura, M. J. (1995). The interleukin-1 beta-converting enzyme (ICE) is localized on the external cell surface membranes and in the cytoplasmic ground substance of human monocytes by immuno-electron microscopy. Journal of Experimental Medicine, 182(5), 1447-1459. Sirko, S. P., Schindler, R., Doyle, M. I , Weisman, S. M., & Dinarello, C. A. (1991). Transcription, translation and secretion of interleukin-1 and tumor necrosis factor: effects of tebufelone, a dual cyclooxygenase/5-lipoxygenase inhibitor. European Journal of Immunology, 21(2), 243-250. Slack, J. L., Nemunaitis, J., Andrews, D. F. d.. Singer, J. W., & Andrews, D. F (1990). Regulation of cytokine and growth factor gene expression in human bone marrow stromal cells transformed with simian virus 40. Blood, 75(12), 2319-2327. Smith, J. A., Telford, R. D., Baker, M. S., Hapel, A. J., & Weidemann, M. J. (1992). Cytokine immunoreactivity in plasma does not change after moderate endurance exercise. Journal of Applied Physiology, 73(4), 1396-1401. Smith, M. F. J., Kueppers, F , & Lee, J. C. (1991). Differential regulation of interleukin-1 alpha and interleukin1 beta mRNA expression in human monocytes: evidence for protein kinase C-dependent and -independent pathways. Lymphokine Cytokine Research, 10, 397-403. Soszynski, D , Kozak, W., Conn, C. A., Rudolph, K., & Kluger, M. J. (1996). Beta-adrenoceptor antagonists suppress elevation in body temperature and increase in plasma IL-6 in rats exposed to open field. Neuroendocrinology, 63, 459-467.

Dynamic Regulation of Proinflammatory Cytokines

177

Spivak, B., Shohat, B., Master, R., Avraham, S., Gil-Ad, I., Bleich, A., Valevski, A., & Weizman, A. (1997). Elevated levels of serum interleukin-1 beta in combat-related posttraumatic stress disorder. Biological Psychiatry, 42(5), 345-348. Strulovici, B., Daniel-Issakani, S., Oto, E., Jr., N. J., Chan, H., & Tsou, A. P. (1989). Activation of distinct protein kinase C isozymes by phorbol esters: correlation with induction of interleukin-l beta gene expression. Biochemistry, 28(8), 3569-3576. Sung, S. S. & Walters, J. A. (1991). Increased cychc AMP levels enhance IL-1 alpha and IL-1 beta mRNA expression and protein production in human myelomonocytic cell lines and monocytes. Journal of Clinical Investigation, 88(6), 1915-1923. Suttles, J., Giri, J. G., & Mizel, S. B. (1990). IL-1 secretion by macrophages. Enhancement of IL-1 secretion and processing by calcium ionophores./owr«a/ of Immunology, 144(1), 175-182. Suzuki, E., Shintani, E, Kanba, S., Asai, M., & Nakaki,T. (1997). Immobilization stress increases mRNA levels of interleukin-1 receptor antagonist in various rat brain regions. Cellular and Molecular Neurobiology, 17(5), 557-562. Swantek, J. L., Cobb, M. H., & Geppert, T. D. (1997). Jun N-terminal kinase/stress-activated protein kinase (JNK/SAPK) is required for lipopolysaccharide stimulation of tumor necrosis factor-alpha (TNFalpha) translation: glucocorticoids inhibit TNF-alpha translation by blocking JNK/SAPK. Molecular and Cellular Biology, 17, 6274-6282. Symons, J. A., Young, P R., & Duff, G. W. (1995). Soluble type II interleukin 1 (IL-1) receptor binds and blocks processing of IL-1 beta precursor and loses affinity for IL-1 receptor antagonist. Proceedings of the National Academdy of Sciences USA, 92(5), 1714-1718. Takaki, A., Huang, Q. H., Somogyvari-Vigh, A., & Arimura, A. (1994). Immobilization stress may increase plasma interleukin-6 via central and peripheral catecholamines. Neuroimmunomodulation, 1(6), 335-342. Tingsborg, S., Zetterstrom, M., Alheim, K., Hasanvan, H., Schultzberg, M., & Bartfai, T. (1996). Regionally specific induction of ICE mRNA and enzyme activity in the rat brain and adrenal gland by LPS. Brain Research, 712, 153-158. Tran-Thi,T.-A., Weinhold, L.,Weinstock, C , Hoffmann, R., Schulze-Specking, A., Northoff, H., & Decker, K. (1993). Production of tumor necrosis factor-alpha, interleukin-1, and interleukin-6 in the perfused rat liver. European Cytokine Network, 4, 363-370. Tsukada, J., Saito, K., Waterman, W R., Webb, A. C , & Auron, P E. (1994). Transcription factors NF-IL6 and CREB recognize a common essential site in the human prointerleukin-1 beta gene. Molecular and Cellular Biology, 14(11), 7285-7297. Tsukada, J., Waterman, W R., Koyama, Y., Webb, A. C , & Auron, P E. (1996). A novel STAT-like factor mediates lipopolysaccharide, interleukin-1 (IL-1), and IL-6 signaling and recognizes a gamma interferon activation site-like element in the IL-ip gene. Molecular and Cellular Biology, 16(5), 2183-2194. Turner, M., Chantry, D., Buchan, G., Barrett, K., & Feldmann, M. (1989). Regulation of expression of human IL-1 alpha and IL-1 beta genes. Journal of Immunology, 143(11), 3556-3561. Ullum, H., Haahr, P M., Diamant, M., Palmo, J., Halkjaer-Kristensen, J., Pedersen, B. K., & Palm, J. (1994). Bicycle exercise enhances plasma IL-6 but does not change IL-1 alpha, IL-1 beta, IL-6, or TNF-alpha pre-mRNA in BMNC. Journal of Applied Physiology, 77(1), 93-97. VanDerPool, T, Jansen, J., Endert, E., Sauerwein, H. P., & VanDeventer, S. J. (1994). Noradrenaline inhibits lipopolysaccharide-induced tumor necrosis factor and interleukin-6 production in human whole blood. Infection and Immunity, 62, 2046-2050. vanGool, J., vanVugt, H., Helle, M., & Aarden, L. A. (1990).The relation among stress, adrenalin, interleukin6, and acute phase proteins in the rat. Clin Immunol Immunopathology, 57(2), 200-210. Vannier, E. & Dinarello, C. A. (1993). Histamine enhances interleukin (IL)-l-induced IL-1 gene expression and protein synthesis via H2 receptors in peripheral blood mononuclear cells. Comparison with IL-1 receptor antagonist. Journal of Clinical Investigation, 92(1), 281-287. Viae, J., Gueniche, A., Doutremepuich, J. D., Reichert, U., Claudy, A., & Schmitt, D. (1996). Substance P and keratinocyte activation markers: an in vitro approach. Archives of Dermatology Research, 288(2), 85-90. Viherluoto, J., Palkama,T., Silvennoinen, O., & Hurme, M. (1991). Cyclic adenosine monophosphate decreases the secretion, but not the cell-associated levels, of interleukin-1 beta in lipopolysaccharide-activated human monocytes. Scandanavian Journal of Immunology, 34(1), 121-125. Volpe, F , Clatworthy, J., Kaptein, A., Maschera, B., Griffin, A. M., & Ray, K. (1997). The IL-1 receptor accessory protein is responsible for the recruitment of the interleukin-1 receptor associated kinase to the IL-l/IL-1 receptor I complex. Federation of Experimental Biological Ssciences Letters, 419,

41^4.

178

L. R. Watkins et al.

Waage, A., Slupphaug, G., & Shalaby, R. (1990). Glucocorticoids inhibit the production of IL-6 from monocytes, endotheHal cells, and fibroblasts. European Journal of Immunology, 20, 2439-2443. Watson, J. M., Lofquist, A. K., Rinehart, C. A., Olsen, J. C , Makarov, S. S., Kaufman, D. G., & Haskill, J. S. (1995). The intracellular IL-1 receptor antagonist alters IL-1-inducible gene expression without blocking exogenous signaling by IL-1 heia. Journal of Immunology, 155(9), 4467-4475. Wesche, H., Neumann, D., Resch, K., & Martin, M. U. (1996). Co-expression of mRNA for type I and type II interleukin-1 receptors and the IL-1 receptor accessory protein correlates to IL-1 responsiveness. Federation of Experimental Biological Ssciences Letters, 391, 104-108. Willis, S. A. & Nisen, P. D. (1995). Inhibition of lipopolysaccharide-induced IL-1 beta transcription by cyclic adenosine monophosphate in human astrocytic cells. Journal of Immunology, 154(3), 1399-1406. Yamato, K., el-Hajjaoui, Z., & Koeffler, H. R (1989). Regulation of levels of IL-1 mRNA in human fibroblasts. Journal of Cell Physiology, 139(3), 610-616. Yoshimura, T., Kurita, C , Nagao, T, Usami, E., Nakao, T, Watanabe, S., Kobayashi, J., Yamazaki, F., Tanaka, H., Inagaki, N., & Nagai, H. (1997). Inhibition of tumor necrosis factor-alpha and interleukin-1-beta production by beta-adrenoceptor agonists from lipopolysaccharide-stimulated human peripheral blood mononuclear cells. Pharmacology, 54(3), 144-152. Zakowski, S. G., McAllister, C. G., Deal, M., & Baum, A. (1992). Stress, reactivity, and immune function in healthy men. Health Psychology, 11(4), 223-232. Zhang, G , Zhang, L., & Duff, G. W. (1997). A negative regulatory region containing a glucocorticosteroid response element (nGRE) in the human interleukin-1 beta gene. DNA Cell Biology, 16(2), 145-152. Zhang, Y. & Rom, W. N. (1993). Regulation of the interleukin-1 beta (IL-1 beta) gene by mycobacterial components and lipopolysaccharide is mediated by two nuclear factor-IL-6 motifs. Molecular and Cellular Biology, 13(6), 3831-3837. Zhou, D., Kusnecov, A. W., Shurin, M. R., DePaoli, M., & Rabin, B. S. (1993). Exposure to physical and psychological stressors elevates plasma interleukin-6: relationship to the activation of hypothalamicpituitary-adrenal axis. Endocrinology, 133(6), 2523-2530. Zhu, G. F., Chancellor-Freeland, C , Berman, A. S., Kage, R., Leeman, S. E., Beller, D. I., & Black, R H. (1996). Endogenous substance P mediates cold water stress-induced increase in interleukin-6 secretion from peritoneal macrophages, yourna/ of Neuroscience, 16(11), 3745-3752. Zorrilla, E. P., Redei, E., & DeRubeis, R. J. (1994). Reduced cytokine levels and T-cell function in healthy males: relation to individual differences in subclinical anxiety. Brain Behavior and Immunity, 8(4), 293-312.

11

CROSS-SENSITIZATION BETWEEN IMMUNE AND NON-IMMUNE STRESSORS A Role in the Etiology of Depression?

F. J. H. Tilders and E. D. Schmidt Research, Institute Neurosciences Vrije Universiteit Facuhy of Medicine Department of Pharmacology van der Boechorststraat 7, 1081 BT Amsterdam, The Netherlands

1. INTRODUCTION Organisms have a tremendous capacity to adapt to environmental conditions. This requires not only immediate responses to acute alterations in the internal or external milieu but also mechanisms that allow experiences to shape the response to later stimuli. Together with its genetic background, it are the specific experiences and their long lasting consequences that model the individual. Aspects of this can be found in almost all cells and organs, and are most prominent in the central nervous system and the immune system. The immune system shows an amazing capacity to store information about pathogens to the extent that a first encounter can prevent the organism from becoming ill by a renewed contact with the pathogen for a lifetime. Information about a wide variety of experiences can be stored in the brain to affect the responses to later events. How this information is stored is still poorly understood, but synaptic plasticity is generally considered to play a crucial role. Not only during early development but also in adulthood single or sporadic exposure to particular environmental stimuli can affect the response to the same stimulus weeks to years later. Such a change in response as a result of a specific earlier experience is often denoted as the "priming" effect of the stimulus. In its most simple form, priming can result in a reduction of the response to a second encounter, a phenomenon called desensitization, habituation or tolerance. Alternatively, priming may lead to exaggerated responses upon renewed exposure, a phenomenon known as sensitization. Both habituation and sensitization of brain-mediated responses represent expressions of neuro-plasticity in the adult brain and plays a key role in physiological adaptations but also in the etiology of various psychiatric diseases Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

179

180

F. J. H. Tilders and E. D. Schmidt

in particular of affective disorders. Habituation and sensitization of a wide variety of responses have been described including behavioral, metabolic, autonomous, neural, endocrine, and immune responses, and reflect altered vulnerability to a particular stimulus. In this chapter we will briefly describe long-lasting sensitization and crosssensitization to various stimuli as seen in adult animals, focus attention on immune activation induced long-lasting cross-sensitization, and discuss its possible contribution to the etiology of affective disorders.

2. LONG-LASTING SENSITIZATION 2.1. Sensitization to Stressors In experimental animals and in humans, single or sporadic exposure to certain stimuli including stressors, immune activation, and drugs of abuse can result in long-lasting (weeks-lifetime) sensitization. Although the information is scarce, the mechanisms underlying long-lasting sensitization may be different from sensitization seen during chronic or repeated stimulation (Antelman & Caggiula, 1996). In most chronic or repeated stress protocols that lead to sensitization, the neurochemical, behavioral, metabolic, and neuroendocrine responses undergo progressive alterations during stress exposure, and tend to return to their pre-stress profiles after cessation of the stress load (Van Dijken, 1992). Therefore, we and others have looked upon such chronic stress induced alterations as an effect of cumulation of rather short term effects (De Goeij, Binnekade, & Tilders, 1992a). Long-lasting sensitization by a single event follows a different pattern. For instance, exposure of adult rats to a single and short (15min) session of electric foot shocks induces behavioral sensitization to novel stressors, which progressively develops to reach a maximum approximately two weeks after the challenge (Van Dijken, Van Der Heijden, Mos, & Tilders, 1992c), and can be maintained for two months as illustrated in Fig l.Thus, behavioral sensitization can develop in the absence of further stimuh. Because this pattern of behavioral sensitization is paralleled by sensitization of neuroendocrine and autonomic stress responses (Antelman, De Giovanni, & Kocan, 1989, Fig IB, IC), this form of long- lasting stress sensitization may reflect a state of increased stress vulnerability as seen in various psychiatric disorders (Akil & Ines Morano, 1995). In support of this, footshock-induced long-lasting behavioral sensitization was associated with increased anxiety related behavior, and anxiolytics were found to reverse these aberrant responses (Van Dijken, Mos, Van Der Heijden, & Tilders, 1992a; Van Dijken, Tilders, Olivier, & Mos, 1992b). Several other stressors including social defeat have been reported to induce alterations in behavioral and metabolic parameters that last for a week or more (Meerlo, Overkamp, & Koolhaas, 1997; Sherman & Petty 1982). Thus, although the effects of most stressors on subsequent responses dissipate within a few days, certain stressful events appear to provoke progressive and long-lasting alterations in stress responsive brain circuits that has precipitated in the "Once is enough" hypothesis (Van Dijken, 1992). There is a large body of literature on long-lasting effects of perinatal stress exposure on stress vulnerability, stress responses and changes in brain stress circuits. Collectively, these observations show that the developing organism is extremely

181

Cross-Sensitization between Immune and Non-Immune Stressors

200 ^160 a m

^ 120 80 40

B

400

1 g 300 X

O 200

<

m^-*—*--

g

a.

r-* 0

2

4

G

a

Post-shock intetval (weeks) ~ • ~ control

~*~ shocked

Figure 1. Long-lasting stress induced sensitization. Adult male rats were exposed to a single session of electric footshocks (shocked) or to the cage in which the shock was delivered without receiving electric current (control), and placed back in their home cage (Van Dijken et al., 1992a). At various time intervals thereafter (1 day-8 weeks) groups (n = 8) of shocked and control rats were exposed to a novel environment (noise test. Van Dijken et al., 1992a,b).Their behavior was recorded for 3min in the presence and absence of 65dB white noise, and animals were decapitated immediately thereafter (cf. Van Dijken et al., 1993). Panel A: behavioral immobility expressed as cumulative duration during the 3min noise off period. Panel B: plasma levels of ACTH at the end of the noise test. Panel C: numbers of fecal boluses produced during the noise test, which serves as an index of autonomic reactivity *p < 0.05; **p <4 0.01; ***p < 0.001 shocked vs time matched controls.

sensitive in the sense that stress shapes the individual's responses to stressors in later life (Meaney, Diorio, Francis, Widdowson, Laplante, Caldji, Sharma, Seckl, & Plotsky, 1996; Weinstock, 1997). As discussed above, such long lasting sensitization is not an exclusive property of the developing brain. Also in adulthood certain stressors can affect stress responses and the underlying brain circuits for prolonged periods of time.

182

F. J. H. Tilders and E. D. Schmidt

2.2. Sensitization to Drugs of Abuse Repeated exposure of experimental animals to drugs of abuse in particular to psychostimulants, can induce progressive and long-lasting (months) increases in behavioral responses induced by renewed drug exposure, and progressively increased motivational responses to obtain the drug. This "behavioral sensitization" to drugs of abuse is associated with lasting functional changes in certain brain circuits, i.e., mesolimbic dopamine systems (Kalivas & Steward, 1991; Robinson & Berridge, 1993). In humans, it is believed that the development of similar long-lasting neuroplastic changes determine the individual's vulnerability to drug addiction and relapse after detoxification. Studies in rats have demonstrated that sporadic or even single administration of a psychostimulant such as amphetamine or cocaine can lead to sensitization of the motivational, locomotor, neurochemical, and neuroendocrine i.e. hypothalamus-pituitary-adrenal (HPA) responses to the drug weeks later (Antelman, Eichler, Black, & Kocan, 1980; MacLennan & Maier, 1983; Robinson, 1984; Wilcox, Robinson, & Becker, 1986; Paulson, Camp, & Robinson, 1991; Steward & Badiani, 1993; Schmidt, Tilders, Janszen, Binnekade, De Vries, & Schoffelmeer, 1995b; Schmidt, Tilders, Binnekade, Schoffelmeer, & De Vries, in press; De Vries, Schof-felmeer, Binnekade, Mulder, & Vanderschuren, 1998). As for long-lasting sensitization to stressors, behavioral sensitization to drugs also appears to develop progressively in the absence of the stimulus. Several researchers demonstrated that neuroch-emical changes induced by chronic drug exposure show similarities with, but also differences from the long lasting changes seen after single or sporadic drug administration. This indicates that single and chronic drug administration can induce differential effects on the central nervous system.

2.3. Sensitization to Immune Stimuli Activation of the immune system by pathogenic organisms or antigens leads to pathogen-specific responses of the immune system. In addition, immune stimuli elicit a host-defense response which is largely not pathogen specific. This non-specific illness response also occurs during tissue damage and inflammation and part of it is mediated by the brain. Thus, neuroendocrine, neurochemical, metabolic, and autonomic responses, and changes in mood, motivational state, sleeping patterns, eating behavior, social, and locomotor activities etc. all represent brain-mediated responses to illness (Kluger, 1991; Kent, Bluthe, Kelly, & Dantzer, 1992;Tilders, De Rijk, Van Dam, Vincent, Schotanus, & Persoons, 1994; Rothwell & Hopkins, 1995; Besedovsky & del Rey, 1996). The interest in these brain-mediated responses was boosted by the finding that endotoxins such as bacterial lipopolysacharide (LPS), and cytokines such as tumor necrosis factor alpha (TNF), interleukin-1 (IL-1), and IL-6 mimic most if not all of the brainmediated symptoms of illness induced by pathogens and inflammatory stimuli. How these large molecular cytokines, that cannot readily pass the blood-brain-barrier, affect the brain is still not fully understood but may involve neural afferents (Watkins,Wiertelak, Goehler, Mooney-Heiberger, Martinez, Furness, Smith, & Maier, 1994; Gaykema, Dijkstra, & Tilders, 1995; Rivest, 1995; Sehic & Blatteis, 1996; Maier, Goehler, Fleshner, & Watkins, 1998) and humoral mechanisms (Van Dam, De Vries, Kuiper, Zijlstra, Tilders, & Berkenbosch, 1996; Tilders et al., 1994; Rivest, 1995; Ericsson, Liu, Hart, & Sawchenko, 1995; Ericsson, Arias, & Sawchenko, 1997). A prominent symptom of illness is activation of the hypothalamus-pituitary-

Cross-Sensitization between Immune and Non-Immune Stressors

183

adrenal (HPA) system, which results in elevated plasma concentrations of adrenal glucocorticoids (Cortisol, corticosterone). Because glucocorticoids exert potent anti-inflammatory activity and in general inhibit the production and effects of proinflamatory cytokines and other mediators, increased corticosterone levels are considered to play a crucial role in restraining the host-defense response. In its most typical form, immune activation by single administration of endotoxin, IL-1, IL-6 or TNF leads to transient activation of the HPA-system which, dependent on dose and route of administration, dissipates within 6-24 hours. After chronic administration of these cytokines, plasma ACTH, and corticosterone levels are initially elevated but progressively fall in time, which may point towards desensitization of the HPA system (Van der Meer, Sweep, Pesman, Tilders, & Hermus, 1996). In an acute model of experimental allergic encephalomyelitis, which is a T-cell and macrophage mediated autoimmune response corticosterone levels increase in parallel with the severity of the neurological symptoms and normalize after their disappearance (MacPhee, Antoni, & Mason, 1989). Experimental models of chronic inflammatory diseases such as adjuvant-induced arthritis show chronic activation of the HPA system (Sarlis, Chowdrey, Stephanou, & Lightman, 1992; Harbuz & Lightman, 1992) and reduced responses to stressors (Shanks, Harbuz, Jessop, Perks, Moore, & Lightman, 1998). Thus, chronic immune stimulation may lead to desensitization of ongoing HPA responses. In contrast, short immune activation by single administration of IL-1 to adult rats can increase the HPA response to a second stimulus one to two weeks later, showing that a single exposure to this immune stimulus induces long-lasting sensitization of the HPA response (Schmidt, Janszen, Wouterlood, & Tilders, 1995a). Similar long-lasting (1-12 weeks) sensitization of the HPA response to IL-2 has been reported in humans (Denicoff, Durkin, Lotze, Quinlan, Davis, Listwak, Rosenberg, & Rubinkow, 1989). Whether or not long lasting IL-1 induced sensitization of the HPA axis is associated with sensitization of other brain mediated host-defense responses is not clear (Anisman, Kokkinidis, Borowski, & Merali, 1998). Preliminary findings do not show lL-1 induced long-lasting sensitization of the febrile response. In fact, this is not surprising because glucocorticoids are known to suppress the febrile response to IL-1 and other immune stimuli (Watanabe, Makisuma, Tan, Nakamori, Hakaura, & Murakami, 1995).Therefore, exaggerated corticosterone responses may mask possible sensitization of brain mechanisms driving febrile or other illness responses. Alternatively, it is not inconceivable that immune activation induces long-lasting sensitization of some but not of other host-defense responses.

3. LONG-LASTING CROSS-SENSITIZATION Long-lasting cross-sensitization of behavioral responses has been extensively described for drugs and stressors (Antelman et al., 1980; MacLennan & Maier 1983; Wilcox et al., 1986; Robinson, Becker, Young, Akil, & Castaneda, 1987). Here, we will focus on cross-sensitization of HPA responses. As described in section 2, stimuli of a widely different nature can induce long-lasting sensitization of the HPA system, i.e., lead to exaggerated ACTH and corticosterone responses upon renewed exposure to the same or a similar stimulus. Interestingly, sensitization may also lead to exaggerated HPA responses to stimuli of a different nature, a phenomenon denoted as "crosssensitization". For example, single administration of IL-1 to adult rats not only

184

F. J. H. Tilders and E. D. Schmidt

enhanced ACTH and corticosterone responses to IL-1, but also to electric foot shocks (Schmidt et al., 1995a) and amphetamine (Schmidt et al., in press). Similarly, electric foot shocks induce long lasting sensitization of the ACTH response to a novel environment (fig IB, Van Dijken, De Goeij, Sutanto, Mos, De Kloet, & Tilders, 1993) and cross-sensitization of the HPA response to amphetamine (Schmidt et al., in press). This is not to say that long-lasting sensitization is always associated with cross-sensitization of the HPA system to other stimuli. Nonetheless, available data indicate that single exposure of rats to IL-1 is particularly powerful in increasing HPA responses to immune stimuli, emotional stressors, and psychostimulant drugs for prolonged periods of time (weeks), and leads to the concept that transient immune activation can alter the individual's HPA response to stressors, and therefore affect its stress-vulnerability for prolonged periods of time.

4. IMMUNE-INDUCED LONG-LASTING CHANGES IN THE HPA SYSTEM Let us consider the changes in the biological substrate that may underlie such long-lasting cross-sensitization. Because largely different brain circuits are involved in the activation of the HPA system by immune and emotional stressors (Sawchenko, Brown, Chan, Ericsson, Li, Roland, & Kovacs, 1996), it was obvious to search for functional alterations within the HPA system rather than in stressor-responsive neural projections to the PVN. So far, long lasting alterations have been found at the level of the hypothalamic CRH neurons, of the pituitary gland and in feedback mechanisms. Before going into details of these changes it should be noted that single immune activation induces little or no long-term alterations in the resting levels of ACTH and corticosterone in the blood or the content of these hormones in the pituitary gland and adrenal gland respectively. Therefore, immune induced sensitization of the HPA system appears to be associated with altered responsivity rather than with altered resting activity of the HPA system. Corticotropin Releasing Hormone (CRH) producing neurons situated in the parvocellular part of the paraventricular nucleus (PVN), are activated by IL-1 and play a crucial role in IL-1 induced ACTH and corticosterone secretion (Berkenbosch, Van Oers, Del Rey, Tilders, & Besedovsky, 1987; Rivest 1995; Ericsson et al., 1995; Gaykema et al., 1995). In collaboration with Aguilera's group in Bethesda, we found that 1-2 weeks after IL-1 priming, CRH neurons showed little or no changes in the expression levels of CRH mRNA, CRH peptide stores and CRH turnover (Schmidt et al., 1995a; Tilders et al., in prep). Another peptide that plays a prominent role in the control of ACTH secretion is vasopressin (AVP), which strongly potentiates the ACTH releasing effect of CRH. Although AVP is also produced by magnocellular neurons of the PVN and secreted from terminals in the posterior pituitary gland, various arguments support the view that AVP secreted from parvocellular CRH neurons represents the major potentiator of CRH mediated ACTH secretion (Whitnall, 1993; Kovacs & Sawchenko 1996). In rodents and humans, hypothalamic CRH neurons can occur in two phenotypes (Whitnall, 1993; Raadsheer, Sluiter, Ravid, Tilders, & Swaab, 1993; Raadsheer, Tilders, & Swaab, 1994b). Under resting conditions, only part of these CRH neurons belong to the AVP co-producing phenotype, i.e., co-express AVP mRNA, co-produce, co-store, and co-secrete AVP with CRH (De Goeij et al., 1992a; De Goeij, Dijkstra, & Tilders, 1992b; Whitnall, 1993; Bartanusz, Jezova, Bertini, Tilders, Aubry, & Kiss, 1993).

Cross-Sensitization between Immune and Non-Immune Stressors

185

In collaboration with Verwer in Amsterdam, we found that the numbers of presumed CRH neurons that express AVP mRNA had increased 1-2 weeks after IL-1 priming. This is in accordance with an increase in the fraction of CRH terminals that co-store AVP (Schmidt et al., 1995a) and shows that IL-1 induces long-lasting changes in the phenotypic expression of CRH neurons. Moreover, following a single IL-1 challenge, terminals of CRH neurons in the median eminence showed a delayed (>4 days) but long lasting (1-3 weeks) increase in AVP stores as illustrated in Fig 2. Therefore, the composition of ACTH secretagogues that are produced by hypothalamic CRH neurons is altered for prolonged periods of time after a single IL-1 exposure (Tilders & Schmidt, 1998). Peptide turnover studies (Schmidt et al., 1995a) further support the view that the signal which is produced by hypothalamic CRH neurons, had shifted towards AVP domination which is reminiscent of changes found after adrenalectomy, chronic stress (Whitnall, 1993; Tilders, Schmidt, & De Goeij, 1993; Aguilera, 1994), and depression (see below). At the level of the pituitary gland, IL-1 priming did not affect steady state levels of POMC mRNA, POMC hnRNA, CRH receptor mRNA, and of CRH and vasopressin receptors (Tilders et al., in prep). However, alterations were disclosed when IL-1 primed animals were challenged to mount an ACTH response. In vivo studies revealed that IL-1 priming increased the ACTH response to an AVP challenge but attenuated the ACTH response to CRH. Thus, although the levels of CRH receptors and CRH receptor mRNA and their kinetics in the pituitary gland seem relatively intact, CRH effects in the pituitary gland are suppressed, which may hint to alterations in post-receptor mechanisms. Re-exposure of IL-1 primed rats to IL-1 revealed that the rapid down regulation of AVP receptors as seen in vehicle-primed animals was obliterated (Tilders et al., in prep). Thus, the mechanisms that appear to terminate AVP signaling in the pituitary gland are shut down in IL-1 primed animals, and AVP signaling is enhanced. The alterations in the pituitary gland therefore do not seem to compensate for the increased AVP drive by hypothalamic CRH neurons, but rather amplify the impact of enhanced AVP exposure during the challenge (see Fig 3). In vivo studies further demonstrate that resting and stressor induced plasma

days Figure 2. Long-lasting immune stimulus induced effects on hypothalamic CRH neurons. Adult rats were given a single dose of human recombinant interleukin-1 (IL-1) or vehicle (control). At various time intervals thereafter (1-40 days) hypothalami of IL-1 primed and vehicle primed rats (n = 8) were processed for quantitative immunocytochemical determination of CRH and vasopressin (AVP) in the external zone of the median eminence (terminals of hypophysiotropic CRH neurons). Note the delayed (>4 days) and long lasting (7-20 days) increase in AVP signal. *p < 0.05; **p < 0.01 vs vehicle primed time-matched controls. For details see: Schmidt et al., 1995a.

F. J. H. Tilders and E. D. Schmidt

186 PRE-lL-1

13 WEEKS POSTtL-1

PVN

ZBME

CORTICOTROPH

ACTH ACTH

Figure 3. Schematic representation of immune-activation induced long-lasting alterations in hypothalamic CRH neurons and pituitary corticotrophs. PVN: paraventricular nucleus of the hypothalamus where cell bodies of hypophysiotropic CRH neurons are located. ZEME: external zone of the median eminence where the terminals of these neurons are located. Receptor mechanisms for CRH and AVP signaling are depicted as VIb and CRH-1 respectively. For details see text.

corticosterone levels of IL-1 primed rats were less sensitive to the suppressive actions of dexamethasone as compared to those in non-primed animals (Tilders et al., in prep), which reflects a partial glucocorticoid feedback deficiency for prolonged periods of time. The underlying processes may relate to alterations in glucocorticoid receptor mechanisms in the pituitary gland and/or the brain. Alternatively, because AVPmediated ACTH secretion as compared to CRH-mediated ACTH secretion is rather insensitive to glucocorticoid inhibition (Abou-Samra, Catt, & Aguilera, 1986; Aguilera, 1994), this may as well be a reflection of the increased contribution of AVP to the control of ACTH secretion. Although alterations at other levels cannot be excluded, the present findings strongly support the concept that long lasting alterations in hypothalamic CRH neurons and in pituitary corticotrophs are instrumental in IL-1 induced long-lasting sensitization of the HPA system.

5. STIMULI INDUCING CROSS-SENSITIZATION AND NEUROPLASTICITY In view of the above mentioned findings, we studied whether a similar relationship between long-lasting cross-sensitization and neuroplastic changes in hypothalamic

Cross-Sensitization between Immune and Non-Immune Stressors

187

CRH neurons as induced by IL-1 also occurs after other stimuli. In other words, do different stimuli that induce long-lasting cross-sensitization of HPA responses, all induce similar neuroplastic changes in hypothalamic CRH neurons and vice versa? This is most probably not the case. Let us consider the effects of different manipulations of adult male rats on AVP stores in CRH terminals in the median eminence as an index of AVP hyperproduction in CRH neurons. As summarized in Table 1, various stimuli lead to increased AVP stores 1-3 weeks after the primary event, whereas other stimuli, including cocaine (Schmidt et al., 1995b), amphetamine (Schmidt et al., in press) novel environment, ether, restraint, insulin induced hypoglycemia (Schmidt, Binnekade, Janszen, & Tilders, 1996), road transport from the animal supplier facilities to our labs (unpublished observations), and social defeat (Buwalda, De Boer, Schmidt, Sgoifo, Van Der Vegt, Tilders, Bohus, & Koolhaas, in press) do not. Earlier, we proposed that the power of the stimulus as reflected in the overall ACTH secretion (amplitude and duration) may determine long-lasting AVP hyper production in CRH neurons (Schmidt et al., 1996). Some novel observations are at odds with this view. For instance, social defeat which is a very strong stimulus for the HPA system (Buwalda et al., in press), does not affect AVP stores in CRH neurons. Based on all presently available data (see Table 1) we hypothesize that possibly not the intensity, but rather the nature of the stimulus primarily determines whether hypothalamic CRH neurons will increase their AVP production for prolonged periods of time and thereby produce an intrinsically more powerful ACTH releasing signal. Conditions that did result in long-lasting increases in AVP stores in CRH neurons include inflammatory stimuli that can stimulate the production of cytokines in the brain. For instance, peripheral administration of endotoxin has been reported to result in lL-1 expression in the brain (Van Dam, Bauer,Tilders, & Berkenbosch, 1995; Buttini & Boddeke, 1995; Gabellec, Griffais, Fillion, & Haour, 1995; Salta, Jacobs, Kaltsas, & Table 1. Long-lasting effects of various stimuli on CRH neurons and HRA sensitization. Adult male rats were subjected to a single exposure to one of various stimuli or to control conditions. Changes in CRH and AVP stores in CRH terminals in the external zone of the median eminence (ZEME) were measured 1-3 weeks later by quantitative immunocytochemistry (cf Schmidt et al., 1996). In separate experiments, groups of rats were exposed to a challenge of the same nature 1-3 weeks after the priming stimulus and ACTH and/or cort responses were monitored Peptide stores in Z E M E Stimulus

CRH

AVP

Sensitization of HPA-response

Footshocks IL-I i.p. IL-l i.c.v. Brain surgery LPS i.p. Social defeat Hypoglycaemia Restraint Ether Road transport Cocaine Amphetamine

—> ^/T -^ -> -^ —> -> -^ -^ ^1) —> ->

T T T t -^ -^ —> ^ -> 1) -^ ^

t

T '.' ? -^ 2) T T

-^: no change; T: increase; ? not determined; 1) control groups were sacrificed at the animal suppliers facilities; 2) all animals have this history. For references see text.

188

F. J. H. Tilders and E. D. Schmidt

Grossman, 1998). Similarly, peripheral and central administration of IL-1 has been reported to increase IL-1 transcription in the brain (Plata-Salaman & Ilyin 1997; Hansen, Taishi, Chen, & Krueger, 1998) and also brain injury/surgery is known to induce local IL-1 transcription (Fan, Young, Barone, Feuerstein, Smith, & Mcintosh, 1995). Under certain experimental conditions, also electric foot shocks can induce IL-1 protein in the brain (Nguyen, Deak, Owens, Kohno, Fleshner, Watkins, & Maier, 1998). Therefore, we speculate that cytokines that are produced in the brain in response to these stimuli may play a role in the cascade of events that result in long-lasting changes in the phenotype of hypothalamic CRH neurons.

6. D E P R E S S I O N - A S S O C I A T E D A L T E R A T I O N S I N T H E H P A SYSTEM Functional changes often referred to as dysregulations of the HPA system have been extensively documented in patients with depression (Akil & Ines Morano, 1995; Post & Weiss, 1995; Holsboer & Barden, 1996). These include increased episodes of ACTH and Cortisol secretion, elevated mean Cortisol levels in the blood and increased free urinary Cortisol levels. Various observations from both clinical and experimental animal studies support a causal link between dysregulation of the HPA system and psychopathology. Increased episodes of enhanced Cortisol levels will lead to overexposure of the brain to this glucocorticoid and will affect the functioning of limbic forebrain structures including the hippocampus that are involved in processing of cognitive stimuli and the regulation of mood, learning, and memory processes, as well as aminergic systems in the brain stem that play a role in arousal and appraisal of environmental stimuli (De Kloet & Joels, 1996; Joels, 1997). Glucorticoid induced alterations in the functioning of these brain structures as seen in chronic stress and depression in turn, are believed to facilitate activation of hypothalamic CRH neurons, and thereby sustain enhanced Cortisol exposure (McEwen & Sapolsky, 1995). There is an ongoing debate as to the initial changes that drive the hyperactivity of the HPA system in depressed patients. One of the hypotheses concerns an acquired or inherited glucocorticoid feedback resistance (De Kloet, Vreugdenhil, Oitzl, & Joels, 1997). In accordance with this hypothesis, deficient Cortisol responses to dexamethasone have been reported in up to 50% of patients with major depression. The finding that relatives of depressed patients that are at genetic risk for developing the disease, show excessive Cortisol responses in combined dexamethasone/CRH tests (Holsboer, Lauer, Schreiber, & Krieg, 1995) further supports the view that disturbances of glucocorticoid feedback (which is crucial for the normal functioning of the HPA system) represent a risk factor for the development of depression (Young, Haskett, Murphy-Weinburg, Watson, & Akil, 1991). In addition, increased AVP production and secretion by hypothalamic CRH neurons which is known to occur in depression (see below) may generate a less glucocorticoid suppressible signal for ACTH release (Aguilera 1994, Von Bardeleben, Holsboer, Stalla, & Muller, 1985). CRH provocation studies in depressed patients revealed blunted ACTH but normal Cortisol responses relative to controls. It is reasoned that the defective ACTH response may be due to reduced CRH receptor efficacy. Alternatively, elevated Cortisol levels may account for the blunted ACTH response by their feedback action at the pituitary gland (Von Bardeleben, Stalla, Muller, & Holsboer, 1988).The dissociation of ACTH and Cortisol responses to a CRH challenge points to yet another level

Cross-Sensitization between Immune and Non-Immune Stressors

189

of dysregulation in the system. Together with enhanced Cortisol responses to ACTH, these findings indicate that adrenal glands of depressed patients are hyper-responsive to circulating ACTH. In support of this, computer tomography studies revealed adrenal enlargement in major depressed patients (Nemeroff, Krishnan, Reed, Leder, Beam, & Dunnick, 1992) which disappears after successful treatment (Rubin, Phillips, Sadow, & McCracken, 1995). What exactly controls the changes in adrenal volume and responsiveness is not known but non-ACTH mechanisms including intra-adrenal CRH production (Van Oers, Hinson, Binnekade, & Tilders, 1992) and increased sympathetic drive (Dijkstra, Binnekade, & Tilders, 1996) are potential candidates. In major depression, dysregulation of the HPA system is associated with increased activity of hypothalamic CRH neurons. This is supported by increased CRH concentrations in the CSF of depressed patients (Nemeroff, Widerlov, Bissette, Walleus, Karlsson, Eklund, Kilts, Loosen, & Vale, 1984). Furthermore, CRH mRNA is markedly increased in the PVN of major depressives (Raadsheer, Van Heerikhuizen, Lucassen, Tilders, & Swaab, 1995) and the numbers of CRH immunoreactive neurons and the numbers and of CRH neurons that co-express AVP are strongly increased in depressed patients (Raadsheer, Hoogendijk, Stam, Tilders, & Swaab, 1994c; Swaab 1997). Several mechanisms may be involved in the hyper-activation of these neurons in depression including excessive stimulatory input from sensitized limbic or brainstem neurons, reduced input from inhibitory circuits, or intrinsic changes in CRH neurons such as reduced glucocorticoid receptor function in CRH neurons (De Kloet et al., 1997) or altered efficacy of efferent signal transfer to CRH neurons.

7. I M M U N E A C T I V A T I O N I N D U C E D H P A C H A N G E S A N D DEPRESSION Several of the long-lasting immune stimulus induced alterations found in rats are also seen in depressed patients (see Table 2). As outhned above, alterations at these levels are most probably instrumental for the sensitization of the HPA system as seen in both conditions. Therefore, we hypothesize that in humans as in rats, transient activation of the immune system may result in a period of HPA sensitization and changes in CRH neurons. During this period of HPA sensitization, the ACTH and Cortisol responses to internal or external stimuli are supposedly enhanced and therefore this period reflects a state of increased vulnerability to stressors. Based on animal Table 2. Long-lasting immune stimulus induced and depression associated changes in the HPA system. Note the similarities of alterations at hypothalamic and pituitary levels, but not at the level of the adrenal gland (for details see text) Variables AVP expression in CRH neurons CRH expression in CRH neurons AVP/CRH signal Pituitary response to CRH Pituitary response to AVP Glucocorticoid feedback Resting cortisol/corticosterone levels Adrenal response to ACTH Adrenal mass

Long-lasting immune induced

Depression associated

T -^/t T i T i —> -^ -^

T T T i ? i T T t

190

F. J. H. Tilders and E. D. Schmidt

Studies, we propose that this vulnerable state in humans is also transient, and will fade after some time, that is if no major second event occurs within this time domain. During this vulnerable period, CRH neurons and pituitary corticotrophs may show functional changes but these do not lead to altered resting Cortisol levels and readjust by the end of this period of stress sensitization. Although sensitization (Antelman et al., 1989) and increased Cortisol responses may be advantageous in some conditions, they represent a risk factor for the development of pathology (Munck, McGuire, & Holbrook, 1984; De Kloet & Joels, 1996; De Kloet et al., 1997). If during the period of stress sensitization the individual faces another sensitizing stimulus (immune or otherwise), the exaggerated HPA response may be followed by further alterations in CRH neurons and pituitary corticotrophs, further increasing the individual's vulnerability to stressors, as depicted in Fig 4. Therefore, immune activation induced sensitization of the stress response can be considered to represent a premorbid state. By cumulation or kindling like mechanisms, these reversible alterations may be driven into a more persistent mode during which the changes are accompanied by elevated resting Cortisol levels and adrenal hyperactivity. According to this scenario, it can be predicted that the exact timing of stressors relative to the immune activation induced stress vulnerable period is crucial for the development of adrenal hyperplasia and hyperactivity. We do not imply that this cascade of events is specifically activated by immune stimuli. In fact, major stimuli of other nature (e.g. loss of spouse) can precipitate depression in vulnerable individuals. Nonetheless, we propose that in humans as in rodents, immune stimuli may be particularly powerful triggers to induce long-lasting cross-sensitization to stimuli of a different nature and thereby represent a risk factor for the development of depression. Several observation indicate that immune activation may relate to a specific form of depression. For instance, multiple sclerosis patients show increased (re)activity of hypothal-

PATHOLOGICAL STATE

Figure 4. Long-lasting sensitization and its proposed role in the development of psychopathology. Transient activation of the immune system (first arrow) leads to transient activation of the HPA system (thin hne), which is followed by a period of adaptive changes in the biological substrate and stress sensitization (bold line). During this period, resting plasma levels of ACTH and Cortisol are not affected but hormone responses to other stimuli (second arrow) are exaggerated. This will result in further alterations in the neuroendocrine substrate and prolonged and/or increased stress vulnerability. When additional stimuli follow within the period of stress sensitization, a pathologic state can be reached during which adrenal Cortisol production becomes chronically activated. For details see text.

Cross-Sensitization between Immune and Non-Immune Stressors

191

amic CRH neurons as indicated by increased numbers of CRH expressing neurons and an increase in the fraction of these neurons that co-express AVP (Erkut, Hofman, Ravid, & Swaab, 1995; Purba, Raadsheer, Hofman, Ravid, Polman, Kamphorst, & Swaab, 1995; Swaab, 1997). Detailed cHnical studies revealed subtle HPA dysregulations e.g. increased resting Cortisol levels (Grasser, Moller, Backmund, Yassouridis, & Holsboer, 1996; Michelson, Stone, Galliven, Magaikou, Chrousos, Sternberg, & Gold, 1994) and blunted responses to CRH (Michelson et al., 1994) and AVP challenge tests (Wei & Lightman, 1997). These dysregulations in the HPA system show similarities but are not identical with those found in major depression indicating that partly distinct processes underlie these distortions. Accordingly, major depression is not likely to occur more often in multiple sclerosis patients as compared to the average population, but depressive symptoms categorized as depression due to general medical condition are frequently found in these patients (up to 50% lifetime risk, Schubert & Foliart, 1993). In addition, studies on first degree relatives of MS patients show that the genetic basis of depression in these patients is clearly different from that in major depression (Sadovnick, Remick, Allen, Swartz, Yee, Eisen, Farquhar, Hashimoto, Hooge, Kastrukoff, Morrison, Nelson, Oger, & Paty, 1996) supporting the view that other mechanisms, i.e., immune activation induced events, may play a role in MS associated depression. Another example refers to aging. In elderly populations, a high incidence of minor depression and depressive symptoms has been found and a low prevalence of major depression (Ernst, 1997). Evidence has been reported to show that the onset and remittance of depression is not related with age per se, but rather with physical illness, functional impairment and numbers of infectious and other disease episodes (Beekman, Kriegsman, Deeg, & Van Tilburg, 1995a; Beekman, Deeg, Smit, & Van Tilburg, 1995b). This points to the potential relevance of episodes of immune activation in the development of vulnerability for depression in the elderly. Interestingly, aging is associated with dysregulations of the HPA system (Huizenga, Koper, De Lange, Pols, Stolk, Grobbee, De Jong, & Lamberts, 1998). Studies on post mortem brains showed that aging is associated with indices of increased (re)activity of hypothalamic CRH neurons (increased numbers of CRH expressing neurons, increased fraction of vasopressin coproducing CRH neurons (Raadsheer et al., 1993, 1994b, Raadsheer, Oorschot, Verwer, Tilders, & Swaab, 1994a). Taken together, these findings support the notion that infections or inflammatory episodes that increase with age may induce cumulative long-lasting alterations in hypothalamic CRH neurons that are associated with sensitization to stressors and represent a risk factor for the development of depression. It should be noted that CRH neurons may play a role in depression by mechanisms other than those mediated by adrenal steroids. Studies with experimental animals show that symptoms resembling those seen in depression can be induced by intra-cerebral administration of CRH and are mediated by CRH receptors in various brainstructures (Chalmers, Lovenberg, Grigoriadis, Behan, & De Souza, 1996). Similarly, transgenic mice overproducing CRH show depression related phenotype (Stenzel-Poore, Heinrichs, Rivest, Koob, & Vale, 1994), whereas CRH deficient and CRH receptor deficient models show suppression of such symptoms (Muglia, Jacobson, & Mazjoub, 1996; Smith, Aubry, Dellu, Contarino, Bilezikjian, Gold, Chen, Marchuk, Hauser, Bentley, Sawchenko, Koob, Vale, & Lee, 1998). Since CRH neurons are known to project to various brain regions including autonomic motor, sensory, and limbic structures, it is conceivable that altered responsiveness of hypothalamic CRH

192

F. J. H. Tilders and E. D. Schmidt

neurons affect mood, cognition, and behavior. Following this line of reasoning infection, inflammation, surgery, and other immune challenges may induce neuroplastic changes that are associated with increased power and responsiveness of CRH neurons that may play a role in stress sensitization and consequently in the etiology of mood disorders. In this vein, strategies directed to prevent or correct immune activation induced long-lasting sensitization and neuroplasticity may prove beneficial.

ACKNOWLEDGMENTS The authors wish to thank drs P Eikelenboom and JG Hoogendijk, Dept of Psychiatry, Amsterdam for their valuable discussions on disease-associated depression, and drs AN Schoffelmeer and T De Vries, Dept of Pharmacology, Vrije Universiteit, Amsterdam for their stimulating discussions on drug sensitization.

REFERENCES Abou-Samra, A. B., Catt, K. J., & Aguilera, G. (1986). Biphasic inhibition of adrenocorticotropin release by corticosterone in cultured anterior pituitary cells. Endocrinology, 118, 972-977. Aguilera, G. (1994). Regulation of pituitary ACTH secretion during chronic stress. Frontiers in Neuroendocrinology, 15, 321-350. Akil, H. A. & Ines Morano, M. (1995). Stress. In: F. E. Bloom & D. J. Kupfer (Eds.), Psychopharmacology: The Fourth Generation of Progress (pp. 773-785). New York: Raven Press. Anisman, H., Kokkinidis, L., Borowski,T., & Merali, Z. (1998). Differential effects of interleukin (IL)-l-beta, IL-2, and IL-6 on responding for rewarding lateral hypothalamic stimulation. Brain Research, 779, 177-187. Antelman, S. M. & Caggiula A. R. (1996). Oscillation follows drug sensitization: implications. Neurobiology, 10, 101-117. Antelman, S. M., Eichler, A. J., Black, C , & Kocan, D. (1980). Interchangeability of stress and amphetamine in sensitization. Science, 207, 329-331. Antelman, S. M., DeGiovanni, L. A., & Kocan D. (1989). A single exposure to cocaine or immobilization stress provides extremely long-lasting, selective protection against sudden cardiac death from tetracaine. Life Sciences, 44, 201-207. Bartanusz, V., Jezova, D., Bertini, L. T., Tilders, E J. H., Aubry, M., & Kiss, J. (1993). Stress-induced increase in vasopressin and corticotropin releasing factor expression in hypophysiotropic paraventricular neurons. Endocrinology, 132, 895-902. Beekman, A. T. P., Kriegsman, D. M. W., Deeg, D. H. J., & Van Tilburg, W. (1995a). The association of physical health and depressive symptoms in the older population. Age and sex differences. Society of Psychiatry and Psychiatric Epidemiology, 30, 32-49. Beekman, A. T E, Deeg, D. H. J., Smit, J. H., & Van Tilburg, W. (1995b). Predicting the course of depression in an older population. Results from a community based study in the Netherlands. Journal of Affective Disorders, 34, 41-49. Berkenbosch, E, Van Oers, J. W. A. M., Del Rey, A.,Tilders, E J. H., & Besedovsky, H. A. (1987). Corticotropin releasing factor producing neurons in the rat activated by interleukin-1. Science, 238, 524-525. Besedovsky, H. O. & Del Rey, A. (1996). Immune-neuro-endocrine interactions: facts and hypotheses. Endocrine Reviews, 17, 64-102. Buttini, M. & Boddeke H (1995). Peripheral lipopolysaccharide stimulation induces interleukin-lbeta messenger RNA in rat brain microglial cells. Neuroscience, 65, 523-530. Buwalda, B., De Boer, S. E, Schmidt, E. D., Sgoifo, A., Van Der Vegt, B. J., Tilders, E J. H., Bohus, B., & Koolhaas, J. M. Long-lasting deficient dexamethasone suppression of HPA activation following peripheral CRH challenge in socially defeated rats. J Neuroendocrinology (in press). Chalmers, D. T, Lovenberg, T W., Grigoriadis, D. E., Behan, D. R, & De Souza, E. B. (1996). Corticotropinreleasing factor receptors: from molecular biology to drug design. Trends in Pharmacological Sciences, 77,166-172.

Cross-Sensitization between Immune and Non-Immune Stressors

193

Chowdrey, H. S., Larsen, P. J., Harbuz, M. S., Jessop, D. S., Aguilera, G., Eckland, D. J. A., & Lightman, S. L. (1995). Evidence for arginine vasopressin as the primary activator of the HPA axis during adjuvantinduced arthritis. British Journal of Pharmacology, 116, 2417-2424. De Goeij, D. C. E., Binnekade, R., & Tilders, E J. H. (1992a). Chronic stress enhances vasopressin but not corticotropin releasing factor secretion during hypoglycemia. American Journal of Physiology, 263, E394-E399. De Goeij, D. C. E., Dijkstra, H., & Tilders, F. J. H. (1992b). Chronic psychosocial stress enhances vasopressin but not corticotropin releasing factor in the external zone of the median eminence of male rats: relationship to subordinate status. Endocrinology, 131, 847-853. De Kloet, E. R. & Joels, M. (1996). Corticosteroid hormones in neuroprotection and brain damage. Current Opinion in Endocrinology, 3, 184-192. De Kloet, E. R., Vreugdenhil, E., Oitzl M. S., & Joels M. (1997). Glucocorticoid feedback resistance. Trends in Endocrinology and Metabolism, 8, 26-33. Denicoff, K. D., Durkin, T. M., Lotze, M. T., Quinlan, P. E., Davis, C. L., Listwak, S. J., Rosenberg, S. A., & Rubinkow, D. R. (1989). The neuroendocrine effects of interleukin-2 treatment. Journal of Clinical Endocrinology and Metabolism, 69,402-410. De Vries, T. J., Schoffelmeer, A. N. M., Binnekade, R., Mulder, A. H., & Vanderschuren, L. J. M. J. (1998). Drug-induced reinstatement of heroin- and cocaine-seeking behavior following long-term extinction is associated with expression of behavioral sensitization. European Journal of Neuroscience, 10, 3565-3571. Dijkstra, I., Binnekade, R., & Tilders, R J. H. (1996). Diurnal variation in resting levels of corticosterone is not mediated by variation in adrenal responsiveness to adrenocorticotropin but involves splanchnic nerve integrity. Endocrinology, 137, 540-547. Ericsson A., Liu C , Hart R. P., & Sawchenko, P. E. (1995). Type 1 interleukin-l receptor in the rat brain: distribution regulation and relationship to sites of IL-1-induced cellular activation. Journal of Comparative Neurology, 136, 681-698. Ericsson, A., Arias, C , & Sawchenko, P. E. (1997). Evidence for intramedullary prostaglandin-dependent mechanism in the activation of stress-related neuroendocrine circuitry by intravenous interleukin-l. Journal of Neuroscience, 17, 7166-7179. Erkut, Z. A., Hofman, M. A., Ravid, R., & Swaab D. F. (1995). Hyperactivity of hypothalamic corticotropinreleasing hormone (CRH) neurons in multiple sclerosis. J Neuroimmunology, 62, 27-33. Ernst, C. (1997). Epidemiology of depression in late Ufe. Current Opinion in Psychiatry, 10,107-112. Fan, L., Young, R R., Barone, F C , Feuerstein, G. Z., Smith, D. H., & Mcintosh, T. X. (1995). Experimental brain injury induces expression of interleukin-l beta mRNA in rat brain. Molecular Brain Research, 30,125-130. Gabellec, M. M., Griffais, R., Pillion, G., & Haour, F. (1995). Expression of interleukin-l alpha, interleukin1 beta, and interleukin-l receptor antagonist mRNA in mouse brain: regulation by bacterial Lipopolysacharide treatment. Molecular Brain Research, 31,122-130. Gaykema, R. P. A., Dijkstra, I., & Tilders, F. J. H. (1995). Subdiaphragmatic vagotomy suppresses endotoxininduced activation of hypothalamic corticotropin-releasing hormone neurons and ACTH secretion. Endocrinology, 136,4714-4720. Grasser, A., Moller, A., Backmund, H., Yassouridis, A., & Holsboer, F. (1996). Heterogeneity of hypothalamic-pituitary-adrenal system responses to a combined dexamethasone-CRH test in multiple sclerosis. Experimental and Clinical Endocrinology and Diabetes, 104, 31-37. Hansen, M. K.,Taishi, P., Chen, Z., & Krueger, J. M. (1998). Vagotomy blocks the induction of interleukin-l beta mRNA in the brain of rats in response to systemic interleukin-l beta. Journal of Neuroscience, 18, 2247-2253. Harbuz, M. S. & Lightman, S. L.(1992). Stress and the hypothalamo-pituitary-adrenal axis: acute, chronic, and immunological activation. Journal of Endocrinology, 134, 327-339. Hatzinger, M., Reul, J. M. H. M., Landgraf, R., Holsboer, F , & Neumann, I. (1996). Combined Dexamethasone/CRH test in rats: Hypothalamo-Pituitary-Adrenocortical system alterations in aging. Neuroendocrinology, 64, 349-356. Holsboer, F & Barden, N. (1996). Antidepressants and hypothalamic-pituitary-adrenocortical regulation. Endocrine Reviews, 17,187-205. Holsboer, F, Lauer, C. J., Schreiber, W., & Krieg, J. C. (1995). Altered hypothalamic-pituitary-adrenocortical regulation in healthy subjects at high familial risk for depression. Neuroendocrinology, 62, 340347. Huizenga, N. A., Koper, J. W., De Lang, P, Pols, H. A., Stolk, R. P, De Jong, F H., & Lamberts, S. W. (1998). Interperson variability but intraperson stability of baseline plasma Cortisol concentrations, and its

194

F. J. H. Tilders and E. D. Schmidt

relation to feedback sensitivity of the hypothalamo-pituitary-adrenal axis to a low dose of dexamethasone in elderly individuals. Journal of Clinical Endocrinology and Metabolism, 83, 47-54. Joels, M. (1997). Steroid hormones and excitability in the mammalian brain. Frontiers in Neuroendocrinology, 18, 2-48. Kalivas, P. W. & Steward, J. (1991). Dopamine transmission in the initiation and expression of drug- and stressinduced sensitization of motor activity. Brain Research Reviews, 16, 223-244. Kent, S., Bluthe, R. M., Kelly, K. W., & Dantzer R. (1992). Sickness behavior: a new target for drug development. Trends in Pharmacological Sciences, 13, 24-28. Kluger, M. J. (1991). Fever: Role of pyrogens and cryogens. Physiological Reviews, 71,93-127. Kovacs, K. J. & Sawchenko, P. E. (1996). Sequence of stress-induced alterations in indices of synaptic and transcriptional activation in parvocellular neurosecretory neurons. Journal of Neuroscience, 16, 262-273. McEwen, B. S. & Sapolsky R. M. (1995). Stress and cognitive function. Current Opinion in Neurobiology, 5, 205-217. MacLennan, A. J. & Maier, S. F. (1983). Coping and the stress-induced potentiation of stimulant stereotypy in the rat. Science, 219, 1091-1093. MacPhee, I. A., Antoni, E, & Mason, D. W. (1989). Spontaneous recovery of rats from experimental allergic encephalomyelitis is dependent on regulation of the immune system by endogenous adrenal corticosteroids. Journal of Experimental Medicine, 169, 431-445. Maier, S. E, Goehler, L. E., Fleshner, M., & Watkins, L. R. (1998). The role of the vagus nerve in cytokineto-brain communication. Annals of the New York Academy of Sciences, 840, 289-300. Meaney, M. J., Diorio, J., Francis, D., Widdowson, J., Laplante, P., Caldji, C , Sharma, S., Seckl, J. R., & Plotsky, P. M. (1996). Early environmental regulation of forebrain glucocorticoid receptor gene expression— implications for adrenocortical responses to stress. Developmental Neuroscience, 18,49-72. Meerlo, P., Overkamp, G. J., & Koolhaas, J. M. (1997). Behavioral and physiological consequences of a single social defeat in Roman high- and low-avoidance rats. Psychoneuroendocrinology, 22,155-68. Michelson, D., Stone, L., Galliven, E., Magiakou, M. A., Chrousos, G. P, Sternberg, E. M., & Gold, P W. (1994). Multiple sclerosis is associated with alterations in hypothalamic-pituitary-adrenal axis function. Journal of Clinical Endocrinology and Metabolism, 79, 848-853. Mugha, L., Jacobson, L., & Mazjoub, J. A. (1996). Production of corticotropin-releasing hormone deficient mice by targeted mutation in embryonic stem cells. Annals of the New York Academy of Sciences, 780, 49-59. Munck, A., McGuire, P. M., & Holbrook, N. J. (1984). Physiological functions of glucocorticoids in stress and their relation to pharmacological actions. Endocrine Reviews, 5, 25-44. Nemeroff, C. B., Widerlov, E., Bissette, G., Walleus, H., Karlsson, I., Eklund, K., Kilts, C. D., Loosen, P T, & Vale, W. (1984). Elevated concentrations of CSF corticotropin-releasing factor-hke immunoreactivity in depressed patients. Science, 226, 1342-1344. Nemeroff, C. B., Krishnan, K. R. R., Reed, D., Leder, R., Beam, C., & Dunnick, R. (1992). Adrenal gland enlargement in major depression. Archives of General Psychiatry, 49, 384-387. Nguyen, K. T., Deak, T., Owens, S. M., Kohno, T., Fleshner, M., Watkins, L. R., & Maier, S. E (1998). Exposure to acute stress induces brain interleukin-ip protein in the rat. Journal of Neuroscience, 18, 2239-2246. Paulson, P. E., Camp, D. M., & Robinson, T. E. (1991). Time course of transient behavioral depression and persistent behavioral sensitization in relation to regional brain monoamine concentrations during amphetamine withdrawal in rats. Psychopharmacology, 103, 480-492. Plata-Salaman, C. R. & Ilyin, S. E. (1997). Interleukin-ip (IL-ip)-induced modulation of the hypothalamic IL-ip system. Tumor Necrosis Factor-a, and Transforming Growth Factor-pi mRNAs in Obese (fa/fa) and Lean (Fa/Fa) Zucker rats: imphcations to IL-1[3 feedback systems and cytokine-cytokine interactions. Journal of Neuroscience Research, 49, 541-550. Post, R. M. & Weiss, S. R. B. (1995). The Neuro biology of treatment-resistant mood disorders. In: F. E. Bloom & D. J. Kupfer (Eds.), Psychopharmacology: The Fourth Generation of Progress (pp.1155-1170). New York: Raven Press. Purba, J. S., Raadsheer, E C , Hofman, M. A., Ravid, R., Polman, C. H., Kamphorst, W, & Swaab, D. E (1995). Increased number of corticotropin-releasing hormone expressing neurons in the hypothalamic paraventricular nucleus of patients with multiple sclerosis. Neuroendocrinology, 62, 62-70. Raadsheer, F. C , Sluiter,A.A., Rabid R.,Tilders, F. J. H., & Swaab, D. F. (1993). Localization of corticotropinreleasing hormone (CRH) neurons in the paraventricular nucleus of the human hypothalamus; agedependent colocalization with vasopressin. Brain Research, 615, 50-62. Raadsheer, F C , Oorschot, D. E., Verwer, R. W H., Tilders, E J. H., & Swaab, D. E (1994a). Age-related

Cross-Sensitization between Immune and Non-Immune Stressors

195

increase in the total number of corticotropin-releasing hormone neurons in the human paraventricular nucleus in controls and Alzheimer disease: Comparison of the dissector with an unfolding method. Journal of Comparative Neurology, 339, 447-457. Raadsheer, F. C , Tilders, F. J. H., & Swaab, D. F. (1994b). Similar age-related increase of vasopressin colocalization in paraventricular corticotropin-releasing hormone neurons in controls and Alzheimer patients. Journal of Neuroendocrinology, 6, 131-133. Raadsheer, F C , Hoogendijk, W. J. G., Stam, F. C , Tilders, F J. H., & Swaab, D. F (1994c). Increased numbers of corticotropin releasing hormone expressing neurons in the hypothalamic paraventricular nucleus of depressed patients. Neuroendocrinology, 60, 436-444. Raadsheer, F C.,Van Heerikhuizen, J. J., Lucassen, P. J., Tilders, F.J. H., & Swaab, D. F. (1995). Increased corticotrophin-releasing Hormone (CRH) mRNA in the paraventricular nucleus of patients with Alzheimer's disease and depression. American Journal of Psychiatry, 152, 1372-1376. Rivest, S. (1995). Molecular mechanisms and neural pathways mediating the influence of interleukin-1 on the activity of neuroendocrine CRF motor neurons in the rat. International Journal of Develop mental Neuroscience, 13, 135-146. Robinson, T. E, (1984). Behavioral sensitization: Characterization of enduring changes in rotational behavior produced by intermittent injections of amphetamine in male and female rats. Psychopharmacology, 84, 466-475. Robinson, T. E & Berridge K. C. (1993). The neural basis of drug craving: an incentive sensitization theory of addiction. Brain Research Reviews, 18, 247-291. Robinson, T. E., Becker, J. B., Young, E. A., Akil, H., & Castaneda, E. (1987). The effects of footshock stress on regional brain dopamine metabolism and pituitary beta-endorphin release in rats previously sensitized to amphetamine. Neuropharmacology, 26,679-691. Rothwell, N. J. & Hopkins S. J. (1995). Cytokines and the central nervous system 11: Actions and mechanisms of actions. Trends in Pharmacological Sciences, 18, 130-136. Rubin, R. T., Phillips, J. J., Sadow, T. F, & McCracken, J. T. (1995). Adrenal gland volume in major depression. Archives of General Psychiatry, 52, 213-218. Sadovnick, A. D., Remick. R. A., Allen, J., Swartz, E., Yee, I. M. L., Eisen, K., Farquhar, R., Hashimoto, S. A., Hooge, J., Kastrukoff, L. F , Morrison, W., Nelson, J., Oger, J., & Paty, D. W. (1996). Depression and multiple sclerosis. Neurology, 46, 628-632. Sarlis, N. J., Chowdrey, H. S., Stephanou, A., & Lightman S. L. (1992). Chronic activation of the hypothalamopituitary-adrenal axis and loss of circadian rhythm during adjuvant-induced arthritis in the rat. Endocrinology, 130, 1775-1779. Salta, M. A., Jacobs, R.-A., Kaltsas, G. A., & Grossman, A. B. (1998). Endotoxin induces Interleukin-1 (i and nitric oxide synthase mRNA in rat hypothalamus and pituitary. Neuroendocrinology, 67, 109-116. Sawchenko, R E., Brown, E. R., Chan, R. K. W., Ericsson A.. Li H-Y., Roland B. L., & Kovacs, K. J. (1996). The paraventricular nucleus of the hypothalamus and the functional neuroanatomy of visceromotor responses to stress. In: G. Holstege, R. Bandler, & C. B. Saper (Eds.), Progress in Brain Research Vol. 107 (pp.201-222). Elsevier Science: Amsterdam. Schmidt, E. D., Janszen, A. W. J, W., Wouterlood, F G, & Tilders, F J. H. (1995a). Interleukin-1 induced long-lasting changes in hypothalamic corticotropin-releasing hormone (CRH)-neurons and hyperresponsiveness of the hypothalamus-pituitary-adrenal axis. Journal of Neuroscience. 15, 7417-7426. Schmidt, E. D., Tilders, F J. H., Janszen, A. W. J. W., Binnekade, R., De Vries, T. J., & Schoffelmeer, A. N. M. (1995b). Intermittent cocaine exposure causes delayed and long-lasting sensitization of cocaineinduced ACTH secretion in rats. European Journal of Pharmacology, 285, 317-321. Schmidt, E. D., Binnekade, R., Janszen, A. W. J. W., & Tilders, F J. H. (1996). Short stressor induced long-lasting increases of vasopressin stores in hypothalamic CRH neurons in adult rats. Journal of Neuroendocrinology 8, 703-712. Schmidt, E. D., Tilders, F J. H., Binnekade, R., Schoffelmeer, A. N. M., & De Vries, T. J. Stressor or drug induced hypersecretion of corticosterone is not critically involved in the expression of long-term behavioral sensitization to amphetamine. Neuroscience (in press). Schubert, D. S. P. & Foliart, R. H. (1993). Increased depression in multiple sclerosis patients: a metaanalysis. Psychosomatics, 34, 124-130. Sehic, E. & Blatteis, C. M. (1996). Blockade of lipopolysacharide-induced fever by subdiaphragmatic vagotomy in guinea pigs. Brain Research, 726, 160-166. Shanks, N., Harbuz, M. S., Jessop D. S., Perks, P., Moore, P. M., & Lightman, S. L. (1998). Inflammatory disease as chronic stress. Annals of the New York Academy of Sciences, 840, 599-607. Sherman, A. D. & Petty, F. (1982). Specificity of the learned helplessness model of depression. Pharmacology Biochemistry and Behavior, 16, 449-454.

196

F. J. H. Tiiders and E. D. Schmidt

Smith, G. W., Aubry, J. M., Dellu, R, Contarino, A., Bilezikjian, L. M., Gold, L. H., Chen, R., Marchuk, Y., Hauser, C , Bentley, C. A., Sawchenko, P. E., Koob, G. R, Vale, W. W., & Lee, K. R (1998). Corticotropin releasing factor receptor 1-deficient mice display decreased anxiety, impaired stress response, and aberrant neuroendocrine development. Neuron, 20,1093-1102. Stenzel-Poore, M. P., Heinrichs, S. C , Rivest, S., Koob, G. P., & Vale, W. W. (1994). Overproduction of corticotropin releasing factor in transgenic mouse: a genetic model of anxiogenic behavior./owma/ ofNeuroscience, 14, 2579-2584. Stewart, J. & Badiani, A. (1993). Tolerance and sensitization to the behavioral effects of drugs. Behavioral Pharmacology, 4, 289-312. Swaab, D. R (1997). Neuro biology and neuropathology of the human hypothalamus. In: R E. Bloom, A. Bjorklund, & T. Hokfelt (Eds.), Handbook of Chemical Neuroanatomy Vol 13 (pp. 39-138). Amsterdam: Elsevier Science. Tiiders, P. J. H. & Schmidt, E. D. (1998). InterIeukin-1-induced plasticity of hypothalamic CRH neurons and long-term stress hyperresponsiveness. Annals of the New York Academy of Sciences, 840,65-80. Tiiders, R J. H., Schmidt, E. D., & De Goeij, D. C. E. (1993). Phenotypic plasticity of CRH neurons during stress. Annals of the New York Academy of Sciences, 697, 39-52. Tiiders, R J. H., De Rijk, R. H., Van Dam, A. M., Vincent, V. A. M., Schotanus, K., & Persoons, J. H. A. (1994). Activation of the hypothalamus-pituitary-adrenal axis by bacterial endotoxins: Routes and intermediate signals. Psychoneuroendocrinology, 19, 209-232. Van Dam, A.-M., Bauer, J., Tiiders, R J. H., & Berkenbosch, R (1995). Endotoxin-induced appearance of immunoreactive interleukin-ip in ramified microglia in rat brain: A hght and electron microscopical study. Neuroscience, 3, 815-826. Van Dam, A-M., De Vries, H. E., Kuiper, J., Zijlstra, R J., Tiiders, R J. H., & Berkenbosch, R (1996). Interleukin-1 receptors on rat brain endothelial cells: a role in neuro-immune interaction? FASEB Journal, 10, 351-356. Van Dijken, H. H. (1992). Once is enough. An animal study on temporal aspects of stress-induced behavioral and neuroendocrine changes. (Doctoral thesis). Monnickendam: Gravo Groep (pp. 1-175). Van Dijken, H. H., Mos, J., Van der Heijden, J. A. M., & Tiiders, R J. H (1992a). Characterization of stressinduced long-term behavioral changes in rats: evidence in favor of anxiety. Physiology and Behavior, 52,945-951. Van Dijken, H. H., Tiiders, R J. H., Olivier, B., & Mos, J. (1992b). Effects of anxiolytic and antidepressant drugs on long-lasting behavioral deficits resulting from one short stress experience in male rats. Psychopharmacology, 109, 395^02. Van Dijken, H. H., Van der Heijden, J. A. M., Mos, J., & Tiiders, R J. H. (1992c). Inescapable footshocks induce progressive and long-lasting behavioral changes in male rats. Physiology and Behavior, 51, 787-794. Van Dijken, H. H., De Goeij, D. C. E., Sutanto, W., Mos, J., De Kloet, E. R., & Tiiders, R J. H. (1993). Short inescapable stress produces long lasting changes in the pituitary-adrenal axis of adult male rats. Neuroendocrinology, 58: 57-64. Van der Meer, M. J. M., Sweep, C. G. J., Pesman, G. J.,Tiiders, R J. H., & Hermus, A. R. M. M. (1996). Chronic stimulation of the hypothalamus-pituitary-adrenal axis in rats by interleukin-lji: central and peripheral mechanisms. Cytokine, 8, 910-919. Van Oers, J. W. A. M., Hinson, J. P., Binnekade, R., & Tiiders, P. J. H. (1992). Physiological role of Corticotropin Releasing Pactor in the control of ACTH mediated corticosterone release from the rat adrenal gland. Endocrinology, 130, 282-288. Von Bardeleben, U., Holsboer, P., Stalla, G. K., & Mtiller, O. A. (1985). Combined administration of human corticotropin-releasing factor and lysine vasopressin induces Cortisol escape form dexamethasone suppression in healthy subjects. Life Sciences, 37,1613-1618. Von Bardeleben, U , Stalla, G. K., MuUer, O. A., & Holsboer, R (1988). Blunting of ACTH response to human CRH in depressed patients is avoided by metyrapone pretreatment. Biological Psychiatry,24,782-786. Watanabe, T., Makisuma, M., Tan, N., Nakamori, T., Hakaura, S., & Murakami, N. (1995). Pebrile responses induced in adrenalectomized rats by administration of interleukin-1 beta or prostaglandin E2. Journal of Physiology, 484, 767-775. Watkins, L. R., Wiertelak, E. P., Goehler, L. E., Mooney-Heiberger, K., Martinez, J., Purness, L., Smith, K. P., & Maier, S. P. (1994). Neurocircuitry of illness-induced hyperalgesia. Brain Research, 639, 283-299. Wei, T. & Lightman, S. L. (1997). The Neuroendocrine axis in patients with multiple sclerosis. Brain, 120, 1067-1076. Weinstock, M. (1997). Does prenatal stress impair coping and regulation of hypothalamus-pituitary-adrenal axis. Neuroscience and Biohavioral Reviews, 21,1-10.

Cross-Sensitization between Immune and Non-Immune Stressors

197

Whitnall, M. H. (1993). Regulation of the hypothalamic corticotropin releasing hormone neurosecretory system. Progress in Neurobiology, 40, 573-629. Wilcox, R. A., Robinson, T. E., & Becker, J. B. (1986). Enduring enhancement in amphetamine-stimulated striatal dopamine release in vitro produced by prior exposure to amphetamine or stress in vivo. European Journal of Pharmacology, 124, 375-376. Young, E. A., Haskett, R. E, Murphy-Weinburg, V., Watson, S., & Akil, H. (1991). Loss of glucocorticoid fast feedback in depression. Archives of General Psychiatry, 48, 693-699.

12

ANHEDONIC AND ANXIOGENIC EFFECTS OF CYTOKINE EXPOSURE Hymie Anisman' and Zul Merali^ ' Institute of Neuroscience, Carleton University, Ottawa, Canada ^ School of Psychology, and Department of Cellular and Molecular Medicine University of Ottawa, Ottawa, Canada

1. INTRODUCTION It is abundantly clear that interactions occur between the immune, endocrine, central, and autonomic nervous systems. Immunologic manipulations (or products of an activated immune system, e.g., cytokines) affect neuroendocrine and central neurotransmitter processes, and conversely, neuroendocrine and central neurotransmitter alterations may impact on immune activity (Anisman, Zalcman, & Zacharko, 1993; Blalock, 1994; Dunn, 1990; Rivier, 1993; Rothwell & Hopkins, 1995). It has been posited that, among other things, the immune system acts like a sensory organ informing the brain of antigenic challenge (Blalock, 1984, 1994). Furthermore, given the nature of the neurochemical changes elicited by antigens and cytokines, it was suggested that immune activation may be interpreted by the CNS as a stressor (Anisman et al., 1993; Dunn, 1990; Dunn, Powell, Meitin, & Small, 1989). To be sure, the effects of systemic stressors (e.g., those associated with viral insults, bacterial endotoxins, cytokines) are not entirely congruous with those elicited by processive stressors (i.e., those involving higher-order sensory processing, e.g., fear conditioning, exposure to a predator or novel environment) (Herman & Cullinan, 1997). Nevertheless, cytokines may be part of a regulatory loop that, by virtue of effects on CNS functioning, might influence behavioral outputs and may even contribute to the symptoms of behavioral pathologies, including mood and anxiety-related disorders (Anisman et al., 1993; Crnic, 1991). It is curious that while cytokines have found an increasingly greater role in immunotherapy, and may contribute to neurodegenerative processes (Rothwell, Luheshi, & Toulmond, 1996), limited attention has been devoted to behavioral analyses of these cytokines in animal studies. This is particularly remarkable given that humans undergoing immunotherapy (e.g., with IL-2) display numerous adverse behavioral effects, including neuropsychological, neurologic, and psychiatric disturbances of sufficient Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

199

200

H. Anisman and Z. Merali

severity to require discontinuation of treatment (Caraceni, Martini, Belli, et a l , 1992; Denicoff, Rubinow, Papa, Simpson, Seipp, Lotze, Chang, Rosenstein, & Rosenberg, 1987; Meyers & Valentine, 1995). While not dismissing the contribution of other modalities of communication between the immune system and the CNS (Dantzer, Bluthe, Aubert, Goodall, BretDibat, Kent, Goujon, Laye, Parnet, & Kelley, 1996; Maier & Watkins, 1998; Watkins, Maier, & Goehler, 1995), several macrophage-derived cytokines, such as interleukin (IL)-ip, IL-6, and tumor necrosis factor-a (TNFa), as well as IL-2 secreted from Thelper cells, may act as mediators (immunotransmitters) in this respect. These cytokines and their receptors are endogenous to the brain, may be affected by psychological factors (e.g., stressors) and by physical insults (concussive injury, seizure), as well as neurodegenerative processes (Hopkins & Rothwell, 1995; Rothwell et al., 1996). Moreover, when administered systemically or directly into the brain, cytokines influence hypothalamic-pituitary-adrenal (HPA) activity, and promote hypothalamic and extrahypothalamic neurotransmitter alterations, some of which are reminiscent of the effects of stressors. Furthermore, there is reason to believe that synergistic behavioral, neuroendocrine and neurotransmitter effects occur between cytokines and/or stressors. Finally, cytokines (most notably IL-lp and TNFa) may induce a sensitization effect, such that subsequent responses to cytokine treatment (and to stressor challenges) may be enhanced. In the present review we will discuss some of the neurochemical and behavioral sequelae of immune and cytokine challenge. It is our contention that interleukins act as messengers between the immune system and the CNS. Moreover, by virtue of the neurochemical and neuroendocrine variations imparted, they may contribute to illnesses involving an affective component, as well as pathologies that may be provoked or exacerbated by stressors. In addition to the well documented "sickness" profile elicited by endotoxins and some cytokines, there is reason to believe that, like stressors, cytokine treatment may engender anxiogenic and anhedonic effects.

2. B E H A V I O R A L I M P L I C A T I O N S O F C Y T O K I N E C H A L L E N G E 2.1. Sickness Behaviors In concert with immunological changes, pathogens promote several adaptive behavioral styles. Yet, the biological or behavioral strategies deployed may have both immediate and protracted adverse consequences, including increased vulnerability to psychological disturbances. Activation of the immune system (e.g., by bacterial endotoxins) or direct administration of products of an activated immune system (e.g., ILip, IL-6, and TNFa) induce an array of behavioral symptoms referred to as "sickness behavior" (Kent, Bluthe, Kelley, & Dantzer, 1992). As discussed in other chapters of this volume, in addition to fever and curled body posture, systemic IL-ip and TNFa provoke anorexia, increased sleep, and reduced locomotor activity, each of which minimizes energy expenditure and sustains body temperature (Bluthe, Dantzer, & Kelley, 1989; Dantzer et al., 1996; Elmquist, Scammell, & Saper, 1997; Hellerstein, Meydani, Meydani, Wu, & Dinarello, 1989; Holden & Pakula, 1996; Kent, Rodriguez, Kelley, & Dantzer, 1994; Kent, Bret-Dibat, Kelley, & Dantzer, 1996; Linthorst, Flachskamm, Muller-Preuss, Holsboer, & Reul, 1995; Moldofsky, 1995; Plata-Salaman, 1989, 1994, 1998; Sonti, Ilyin, & Plata-Salaman, 1996). Predictably, cytokines disrupt operant

Anhedonic and Anxiogenic Effects of Cytokine Exposure

201

responding for food reinforcement, diminish exploratory behaviors, as well as social interaction (Bluthe, Dantzer, & Kelley, 1997; Bluthe, Michaud, Kelley, & Dantzer, 1996; Bret-Dibat, Bluthe, Kent, Kelly, & Dantzer, 1995; Lacosta, Merali, & Anisman, 1998a). Some of the responses to IL-ip (e.g., diminished food & water intake, and social exploration; sleep and soporific effects) directly or indirectly involve central mechanisms (Bluthe et al., 1997; De Sarro, Gareri, SinopoU, David, & Rotiroti, 1997; Kent, BretDibat, Kelley, & Dantzer, 1996), although peripheral IL-ip effects on vagal afferents may contribute, as well (Dantzer et al., 1996; Raber, Koob, & Bloom, 1995; Watkins et al., 1995). While the complex behavioral changes observed may be secondary to the sickness-induced by cytokines, they may also reflect motivational actions (e.g., anhedonia) (Dantzer et al., 1996; Kent et al., 1992). 2.2. Depressive Illness While major depressive illness may be accompanied by suppression of some aspects of immune functioning, there have also been indications of immune activation, particularly in severely depressed (melancholic) patients (Maes, 1995). There is evidence suggesting that depression may be associated with variations of either circulating cytokine levels or cytokine production from stimulated cells. In this respect, it seems that the immune profile of severely depressive patients is not unlike an acute phase reaction. Among other things, depressed patients exhibit increased plasma concentrations of complement proteins, C3 and C4, and IgM, as well as positive acute phase proteins haptoglobin, al-antitrypsin, a l and a2 macroglobulin, coupled with reduced levels of negative acute phase proteins. As well, depression was associated with an increased number of activated T cells (CD25^ and HLA-DR"^), secretion of neopterin, prostaglandin E2 and thromboxane, as well as increased concentrations of soluble IL2 receptors (sIL-2R), IL-lp, IL-1 receptor antagonist (IL-lRa),IL-6, soluble IL-6 receptor (sIL-6R), and y-IFN (Connor & Leonard, 1998; Maes, 1995; Maes, Lambrechts, Bosmans, Jacobs, Suy, Vandervorst, DeJonckheer, Minner, & Raus, 1992; Maes, Stevens, Peeters, DeClerck, Bridts, Peeters, Schotte, & Cosyns, 1993; Maes, Bosmans, Suy, Vandervorst, DeJonckheere, Minner, & Raus, 1991; Maes, Meltzer, & Bosmans, 1995; Mullar & Ackenheil, 1998; Nassberger & Traskman-Bendz, 1993; Smith, 1991; Song, Dinan, & Leonard, 1994). While there have been reports that the elevated levels of ILip, IL-6 and al-acid glycoprotein normalized with antidepressant medication (Sluzewska, Rybakowski, Laciak, Mackiewicz, Sobieska, & Wiktorowicz, 1995), it was demonstrated that the upregulated production of sIL-2R, IL-6 and sIL-6R was not attenuated with antidepressant medication, leading to the suggestion that these factors may be trait markers of the illness (Maes et al., 1995). Although severity of depressive illness is likely fundamental in determining cytokine levels (Maes, 1995), the possibility could not be ignored that characteristics of the illness might also contribute in this respect. Indeed, we observed that IL-ip production in the supernatant of mitogen stimulated lymphocytes was increased in chronic depression (dysthymia), but not in major depressive disorder. This effect was observed irrespective of whether the patients exhibited typical or atypical depressive features (i.e., regardless of whether patients exhibited reduced sleep, appetite, and weight loss or reversed neurovegetative symptoms) (Anisman, Ravindran, Griffiths, & Merali, 1998). Interestingly, serum IL-lp levels were elevated in atypical major depression, but a comparable effect was not seen in either typical major depression, or in typical and atypical dysthymia. Moreover, it was observed that IL-ip concentrations normalized

202

H. Anisman and Z. Meraii

after successful pharmacotherapy (Griffiths, Ravindran, Merah, & Anisman, 1996). Taken together, these data indicate that a relationship may exist between circulating cytokines (or cytokine production) and the expression of depressive symptoms, but the nature of the effect may be dependent on specific characteristic of the illness. However, it should be noted that the data are largely correlational, and it is unclear whether the cytokine alterations seen in affective disorders are secondary to the illness (or the stress associated with the illness) or play an etiological role in the provocation of the disorder. Yet, high doses of IL-2, IFN-a, and TNFa in humans undergoing immunotherapy have been shown to induce neuropsychiatric symptoms, including depression, and that such effects were related to the cytokine treatment rather than the illness being treated (Meyers & Valentine, 1995). Of course, the doses administered in these studies were in the pathophysiological range, and thus their relevance to depression per se may be questioned. This notwithstanding, as IL-ip potently stimulates the release of corticotropin releasing hormone (CRH) (Dunn, 1995; Rivier, 1993), the possibility exists that this cytokine contributes to the altered HPA functioning characteristic of major depressive illness. While studies in infrahumans have suggested that cytokines may affect behavioral processes, it is uncertain whether the behavioral changes observed are, in fact, characteristic of the depressive symptomatology seen in humans. Since cytokines may promote a range of behavioral changes (as described earlier), including "sickness behaviors", it needs to be determined whether the depressive-like behaviors are simply an epiphenomenon.

3. CYTOKINES IN THE CNS Although peripheral cytokines do not have ready access to the brain, entry may occur where the blood-brain barrier is less restrictive (e.g., the organum vasculosum laminae terminalis), by a saturable transport system, and in some pathological conditions (e.g., seizure) (Banks, Ortz, Plotkin, & Kasten, 1991; Gutierrez, Banks, & Kastin, 1993; Imura, Fukata, & Mori, 1991). Although only small amounts of peripheral IL-ip gain entry to the brain, de novo synthesis may be induced centrally in microglia, and perhaps even neuronal tissues (Hopkins & Rothwell, 1995). The central circuitry associated with an acute phase response is fairly complex, involving several pathways and sites of entry, culminating in activation of neurons in numerous brain regions (Rivest, 1995). Moreover, following IL-ip or LPS challenge c-fos mRNA expression across brain regions may follow a biphasic temporal course (Brady, Lynn, Herkenham, & Gottesfeld, 1994; Rivest & Laflamme, 1995). The early set of changes may reflect the initial entry into the brain, promoting neuroendocrine changes and the neural transduction of peripheral signals, while the late phase may reflect a cascade of neurochemical alterations, cellular activation associated with IL-1 p (or bioactive products) along various diffusion routes or secondary to central de novo synthesis of cytokines (Brady et al., 1994; Gaillard, 1995). The bioactivity and mRNA expression of several cytokines and their receptors have been demonstrated in numerous brain regions (Gabellec, Griffais, Pillion, & Haour, 1996; Hanisch & Quirion, 1996; Haour, Marquette, Ban, Crumeyrolle-Arias, Rostene, Tsiang, & FiUion, 1995; Ilyin & Plata-Salaman, 1996; Kinouchi, Brown, Pasternak, & Donner, 1991; Schobitz, De Kloet, Sutanto, & Holsboer, 1993). In addition to being constitutively expressed, increased central bioactivity has been observed

Anhedonic and Anxiogenic Effects of Cytokine Exposure

203

following systemic or central bacterial endotoxin challenge, viral infection, brain injury, cerebral ischemia, seizure, and the pain/inflammation induced by formalin injection (Ban, Haour, & Lenstra, 1992; Buttini & Boddeke, 1995; De Simoni, Del Bo, De Luigi et al., 1995; Gabellec et al., 1996; Laye, Bluthe, Kent, Combe, Medina, Parnet, Kelly, & Dantzer, 1995; Liu, Kita, Tanaka, & Kinoshita, 1996; Minami, Kuraishi, Yamaguchi, Nakai, Hirai, & Satoh, 1990; Muaramami, Fukata,Tsukada, Kobayashi, Ebisui, Segawa, Muro, Imura, & Nakao, 1993; Pitossi, del Rey, Kabiersch, & Besedovsky, 1997; Quan, Sundar, & Weiss, 1994; Rajora, Boccoli, Burns, Sharma, Catania, & Lipton, 1997; Sato, Reiner, Jensen, & Roos, 1997;Taupin, Toulmond, Serrano, Benavides, & Zavala, 1993; van Dam, Brouns, Louisse, & Berkenbosch, 1992; Yabuuchi, Sharma, Catania, & Lipton, 1996). On the one side, it is possible that IL-ip may affect reparatory processes in neurological diseases or brain trauma (at least, in culture following low doses). Alternatively, IL-ip may actually promote neuronal damage following central insults (Kluger, 1991; Loddick, MacKenzie, & Rothwell, 1996; Strijbos & Rothwell, 1995), while IL-lRa or synthesis inhibition may be neuroprotective (Loddick, Wong, Bongiorno, Gold, Licinio, & Rothwell, 1997; Loddick et al., 1996; Rothwell, Loddick, & Stroemer, 1997). As icv IL-1(3 increased central mRNAs for the complete IL-ip panel (IL-ip, ILIRI, IL-IRII, IL-lRa, IL-1 accessory protein), and these were highly correlated with one another, fine regulation likely occurs between the ligand and receptors, and these may be fundamental in analyses of neurological diseases (Ilyin & Plata-Salaman, 1996).

4. NEUROCHEMICAL AND BEHAVIORAL EFFECTS OF CYTOKINES 4.1. Neurochemical Consequences of Cytokines: Stressor-like Effects There have been a number of reports showing that cytokines may influence central neurotransmitter activity and/or levels. Given the marked effects of cytokines on neuroendocrine factors, it is not surprising that particular attention was devoted to neurotransmitter variations within the hypothalamus. For instance, when systemically administered, both IL-ip and LPS increased the utilization of hypothalamic NE (Masana, Heyes, & Mefford, 1990; Mefford & Heyes, 1990; Mefford, Masters, Heyes, & Eskay, 1991; Terao, Oikawa, & Saito, 1993; Zalcman, Green-Johnson, Murray, Nance, Dyck, Anisman, & Greenberg, 1994). Similarly, like stressors, LPS increased NE turnover and tryptophan levels in hypothalamus, and these effects were modifiable by blocking sympathetic nervous system activity (Dunn & Welch, 1991). As in the case of the post-mortem amine changes, systemic IL-1 (3 influences the in vivo release of hypothalamic NE (Smagin, Swiergel, & Dunn, 1996) and central administration of IL-ip enhanced the in vivo release of hypothalamic NE, DA, and 5-HT (Mohankumar & Quadri, 1993; Mohankumar, Thyagarajan, & Quadri, 1991, 1993; Shintani, Kanba, Nakaki, Nibuya, Kinoshita, Suzuki, Vagi, Kato, & Asai, 1993). In addition to the hypothalamic amine variations, LPS and endotoxin affect tryptophan levels and NE turnover in the prefrontal cortex, parietal cortex and brain stem (Dunn & Welch, 1991), as well as NE and 5-HT turnover in hippocampus, and DA turnover within the prefrontal cortex (Zalcman et al., 1994). In vivo studies have also revealed that systemic IL-1 (3 and LPS lead to marked hippocampal 5-HT release which coincided with changes of locomotor activity, and these effects were attenuated by icv administration of IL-lra (Linthorst et al., 1995). As will be described shortly, results from our laboratory have

204

H. Anisman and Z. Merali

essentially confirmed these findings, and suggest that peripheral immune and cytokine challenges also influences NE and DA at several brain sites. It has been demonstrated that IL-2 influences secretion of several hormones, but few studies have assessed the central neurochemical consequences of IL-2 (see Hanisch & Quirion, 1996). Intravenous administration of IL-2 was shown to provoke hypothalamic unit activity (Bartholomew & Hoffman, 1993), and intraperitoneal IL-2 treatment increased the turnover of hypothalamic NE (Zalcman et al., 1994). It was Hkewise observed that IL-2 increased DA release from striatal sHces under basal conditions, as well as during potassium-evoked neuronal depolarization (Alonso, Chaudieu, Dorio, Krishnamurthy, Quirion, & Boksa, 1993; Lapchak, 1992), although biphasic effects were observed such that lower doses increased and higher doses inhibited veratrine-evoked DA release (Petitto, McCarthy, Rinker, Huang, & Getty, 1997). As in the case of DA, it has been demonstrated that in hippocampal slices the application of IL-2 in relatively low doses potentiated ACh release, but had an inhibitory effect in higher doses (Araujo, Lapchak, Collier, & Quirion, 1989; Seto, Kar, & Quirion, 1997). As well, when applied directly to the caudate or substantia nigra, IL-2 provoked ipsilateral turning, suggesting that the lymphokine may act through DA or opioid mechanisms, which induce circling when applied to nigrostriatal structures (Lapchak, 1992). Taken together, these data suggest a possible role for IL-2 in subserving both some of the central neurochemical and behavioral effects associated with antigenic challenge. Stressors, like antigenic challenge, increase IL-ip mRNA expression, as well as that of IL-lRa (Suzuki, Shintani, Kanba, Asai, & Nakaki, 1997). Interestingly, in the absence of the inhibitory effects of glucocorticoids (i.e., in adrenalectomized animals), stressor exposure markedly increased IL-ip protein in hypothalamus, hippocampus, cerebellum, and nucleus tractis solitaris (Nguyen, Deak, Owens Kohn, Fleshner, Watkins, & Maier, 1998). Inasmuch as IL-1(3 strongly influences CRH release from the PVN, thus provoking HPA activity (Ericsson, Arias, & Sawchenko, 1997; Ericsson, Ek, Wahlstrom, Kovacs, Liu, Hart, & Sawchenko, 1996; Rivest, 1995; Rivest & Rivier, 1994), the possibility has been explored that this cytokine may play a pivotal role in the HPA activation associated with stressful events. Indeed, it was demonstrated (Shintani et al., 1993; Shintani, Nakaki, Kanba, Sato, Kato, & Asai, 1995; Shintani, Nakaki, Kanba, Sato, Yagi, Shiozawa, Aiso, Kato, & Asai, 1995) that the increased hypothalamic NE and DA activity and pituitary ACTH release associated with a stressor was prevented by icv pretreatment with IL-lra. If the cytokine receptor antagonist was applied once the stressor had been administered, then there was no effect on either HPA functioning or the central amines. In effect, IL-lra acted in a prophylactic fashion in affecting HPA functioning and amine changes engendered by stressors; however, once these processes were initiated, the cytokine antagonist did not operate in a therapeutic capacity in attenuating the stressor effects. Although stressors influence IL-ip mRNA or protein levels within the brain and pituitary (Minami, Kuraishi, Yamaguchi, Nakai, Hirai, & Satoh, 1991; Nguyen et al., 1998; Shintani et al., 1995a, b; Takao, Tojo, Nishioka et al., 1994; Yabuuchi, Maruta, Minami, & Satoh, 1996), stressors and IL-ip may act through independent inputs to CRH neurons (Harbuz & Lightman, 1992; Whitnall, Perlstein, Mougey, & Neta, 1992). Indeed, the view was offered that while processive stressors influence HPA activity via the central amygdala, systemic stressors (including IL-ip) do so via an amygdalaindependent mechanism (Herman & Cullinan, 1997). This is not to say that cytokines are without effect on amygdaloid activity, as IL-ip increased Fos-immunoreactivity in the central amygdala and bed nucleus of the stria terminalis, as well as cetecholaminergic neurons of the nucleus tractis solitaris and ventrolateral medulla (the latter

Anhedonic and Anxiogenic Effects of Cytokine Exposure

205

projecting to regions of the PVN) (Day & Akil, 1996; Ericsson, Kovacs, & Sawchenko, 1994; Sawchenko, Brown, Chan, Ericsson, Li, Roland, & Kovacs, 1996). Furthermore, as will be discussed shortly, we have observed that systemic IL-lp increased the in vivo release of NE from the central amygdala (Merali, Michaud, Mclntyre, & Anisman, unpubHshed report). In addition to their immediate effects, both stressors and acute IL-ip may provoke persistent effects on HPA functioning (Schmidt, Janszen, Wouterlood, & Tilders, 1995; Tilders, Schmidt, & De Goeij, 1993). With the passage of time following stressor or IL-ip treatment, phenotypic variations occurred in CRH terminals within the external zone of the median eminence, such that they came to coexpress arginine vasopressin (AVP). As CRH and AVP synergistically promote ACTH release, phenotypic shifts of CRH and AVP coexpressing cells induced by IL-ip and by stressors may account for long-term and/or the sensitization effects (Bartanusz, Jezova, Bertini, Tilders, Aubry, & Kiss, 1993; Schmidt et al., 1995;Tilders et al., 1993). Interestingly, there is reason to suppose that chronic depressive illness, as well as the high rates of illness recurrence following successful pharmacotherapy may be related to enhanced coexpression of CRH and AVP (Gold, Chrousos, Kellner, Post, Roy, Augerinas, Schulte, Oldgield, & Loriaux, 1984; Holsboer, 1995; Nemeroff, 1996). As will be discussed shortly, we have likewise observed time-dependent sensitization with respect to the actions of TNFa on plasma ACTH and corticosterone concentrations, as well as NE, DA, and 5-HT turnover and levels in several brain regions (Hayley, Brebner, Staines, Merali, & Anisman, 1998).

4.2. Anhedonic and Anxiogenic Actions of Cytokines: Stressor-Iike Effects Since several cytokines have been shown to be potent stimulators of corticotropin-releasing hormone (CRH) release from the PVN (Dunn, 1995; Hopkins & Rothwell, 1995; Rivier, 1993), a peptide which induces anxiogenic effects (Gray, 1991; Heilig, Koob, Ekman, & Britton, 1994), we assessed whether various cytokines would provoke an anxiogenic response. Since the anxiogenic effects of various treatments may be evident in some test situations and not others, we have typically assessed anxiety in several different paradigms. In our experiments, which are still at early phase, we have assessed the effects of LPS and various cytokines on exploratory patterns, neophobia (as reflected by approach to novel stimuli), responding in an elevated plus-maze test and a light-dark box, and we are currently assessing startle and fear-potentiated startle. As indicated earlier, there is reason to believe that immune activation and cytokine changes may be associated with depressive illness. Furthermore, as anhedonia is a characteristic feature of depression, it was of interest to assess whether LPS or cytokine challenge would engender anhedonic-like effects. While consumption of palatable substances have been used to assess anhedonia, this type paradigm does not readily distinguish anorexic from anhedonic effects of a given treatment. This is not to disparage the use of feeding paradigms, as they certainly have been used effectively. Rather, it was felt that in this instance it would be advantageous to employ a test devoid of appetitive motivation. Accordingly, we opted to assess the effects of LPS and cytokine treatments on responding for rewarding brain stimulation in rats, a behavioral test previously shown to be sensitive to the effects of stressors (Anisman, Zalcman, Shanks, & Zacharko, 1991). Typically, when electrodes are implanted in the medial forebrain bundle, or any number of other DA rich brain regions (prefrontal cortex, nucleus accumbens, sub-

206

H. Anisman and Z. Merali

stantia nigra, ventral tegmentum), animals will emit operant responses to receive electrical stimulation. In view of the large number of responses animals emit, it is assumed that responding is rewarding. Treatments that reduce the apparent rewarding value of the stimulation may, however, provoke an increase of response rate. It seems that animals will compensate for the reduced value of the stimulation by increasing its response frequency. Further, a decline in responding may be secondary to motoric effects of a given treatment, independent of effects on the variations in the rewarding properties of the brain stimulation. Thus, rather than employing rate-dependent measures (e.g., number of responses emitted), we employ a titration paradigm in which animals are permitted to respond for rewarding brain stimulation (where the operant is a head dip response into a hole located on the floor of the test chamber) over the course of a descending series of current intensities (or current frequencies), after which responding is measured over an ascending series. The two curves are then combined to provide a single titration curve, unless the ascending and descending series are very different (in which case the two curves are both presented). In addition, in some of our experiments a simultaneous discrimination paradigm is employed, wherein animals are required to respond to an illuminated hole (+) and not to a nonilluminated hole (-). In this way it is possible to determine whether the change in the titration curve is accompanied by variations in the animals discrimination ability (i.e., altered responses to the + stimulus, without affecting responding to the— stimulus, or conversely high responses being maintained but in a less discriminating fashion). Predictably, in nonstressed animals responding declines as the current intensity is reduced, reflecting the diminished reward obtained for an operant response. Ordinarily, stressors provoke a marked reduction of responding for rewarding brain stimulation, such that responding declines at virtually all of the current intensities. That is, the stressor induces a right-ward shift of the dose response curve, so that higher current intensities are needed to elicit the same number of responses. However, it is important to note that with sufficiently high current intensities the blunted responding in stressed animals may be overcome. In effect, it seems that stressed animals are capable of responding, and motoric factors are likely not responsible for behavioral impairments. Parenthetically, the effects of the stressor are pronounced soon after stressor exposure, and are still evident at lengthy intervals afterward (e.g., 168hr), provided that the initial test session was administered soon after stressor exposure. Moreover, the stressorinduced disruption of responding for rewarding brain stimulation may be attenuated by a regimen of chronic antidepressant treatment (Zacharko & Anisman, 1991). It is important to emphasize, as well, that the effects of stressors on rewarding stimulation vary across brain regions. For instance, we observed that footshock (and other stressors) reliably disrupted responding from the medial forebrain bundle, as well as the prefrontal cortex, nucleus accumbens, and dorsal aspects of the ventral tegmentum, brain regions in which stressors influence DA activity. In contrast, responding for rewarding stimulation was unaffected from those brain regions where DA was not influenced by stressors (i.e., substantia nigra and caudate) (Zacharko & Anisman, 1991). The latter findings are consistent with the supposition that DA mechanisms, among others, contribute to the stressor-related performance changes (Fibiger & Phillips, 1988; Wise, 1985), and also indicate that the altered behavioral responsivity is not attributable to general malaise or motoric changes attributable to the stressor. After aU, if such factors were pertinent, then behavioral disturbances should have been evident irrespective of the site of electrode placements.

Anhedonic and Anxiogenic Effects of Cytokine Exposure

207

5. BEHAVIORAL AND NEUROCHEMICAL EFFECTS OF ENDOTOXIN AND CYTOKINES 5.1. Behavioral and Neurochemical Consequences of Lipopolysaccharide 5.1.1. Anxiogenic Effects of Lipopolysaccharide. In a series of recent experiments (Lacosta, Kulczycki, Merali, & Anisman, 1996; Borowski, Kokkinidis, Merali, & Anisman, 1997), some conducted in mice and others in rats, we assessed the effects of lipopolysaccharide (LPS) on anxiety-like responses and on responding for rewarding brain stimulation. In mice, administration of low doses of LPS provoked classic signs of illness, including apparent soporific effects, reduced open-field exploration, reduced consumption of a palatable food (chocolate milk) and increased latency to approach a novel stimulus. As seen in Figure 1, in an elevated plus-maze, which has been used as a test of anxiety, animals treated with LPS reduced the frequency of open-arm visits, whereas visits to the closed arms of the maze were unaffected. When animals ventured onto the open arms they did not remain there long, supporting the view that the reduced visits were not secondary to motor disturbances. Paralleling the effects seen in the plus-maze, in a light-dark box test of anxiety, where high anxious animals tend to avoid the illuminated region, it was found that LPS increased the apparent reluctance of mice to remain in the brightly lit area. This effect was seen with doses as low as 0.5 fig/mouse. As in the case of the plus-maze, it was noted that when animals that had been treated with LPS entered the illuminated area of the light-dark box, they quickly left this area. Moreover, if LPS treated animals were placed directly into the illuminated region, like their vehicle-treated counterparts they left this area immediately (<2sec). 5.1.2. Anhedonic Effects of LPS. In addition to provoking anxiety-like effects, endotoxin treatment markedly affected responding for rewarding brain stimulation. Figure 2 shows the effects of LPS (100 jig/rat) on responding for stimulation from the

SALINE

0.5 1 LPS Dose (ug)

SAUNE

0.5 1 LPS Dose (ug)

Figure 1. LPS provokes a dose-de pendent anxiogenic response in mice, as reflected by decreased visits to the open arms of an elevated plus-maze at 2hr following i.p. administration (left-hand panel). Entries into the closed arms was not affected by the endotoxin (right-hand panel) (From Lacosta et al., 1996).

H. Anisman and Z. Merali

208 Descending

Ascending - O - L P S (100 HE)

^HH ^

100

\ 0

5fl •

\

s^ 25

30

35

Current (uAf

50

25

30 35 Currenl (uA)

40

Figure 2. In rats treated with LPS (100|j.g) responding for rewarding stimulation from the lateral hypothalamus was impaired. Responding at each current intensity is shown for vehicle and LPS-treated rats over the course of a descending, followed by an ascending series of current presentations (FromBorowski et al., 1997).

lateral hypothalamus in Wistar rats 2hr following LPS administration (Borowski et al., 1997). It is clear that the endotoxin provoked a marked reduction of responding for rewarding stimulation. However, the profile of results was not characteristic of that induced by a stressor. In particular, it seemed that during the initial portion of the test (descending series of current intensities) responding was only modestly reduced. At the higher current intensities LPS-treated animals generally displayed proficient performance. At the lower current intensities, which are ordinarily associated with lower rates of responding, LPS retarded performance. Interestingly, during the ascending portion of the titration function, responding remained depressed even when higher intensities were applied. At a lower dose (50|ig), not shown in the figure, animals still emitted some responses, and hence had the opportunity to sample the rewarding value of the stimulation. Still, these animals showed depressed responding for rewarding stimulation during the ascending series of current presentations. Even at high doses, the animals displayed proficient discrimination ability (directing their responses to the + stimulus) suggesting that they could attend to the appropriate environmental cues. In effect, these data suggest that despite the malaise associated with the LPS treatment, given sufficient reward, rats will continue to emit responses, but as the benefit of responding wanes, the propensity to respond tails off precipitously. Moreover, once this occurs, it is less likely for animals to return to the operant readily. 5.1.3. Neurochemical Effects of LPS. Consistent with numerous earlier reports, LPS appreciably influenced HPA functioning, as reflected by increased plasma corticosterone and ACTH at 2hr following treatment. As alluded to earlier, there have been several reports showing that LPS will profoundly affect biogenic amines within hypothalamic sites. Our findings in post-mortem tissue (in mice) and in vivo (among rats) have been consistent with these reports, and have also revealed that LPS markedly influenced NE, DA, and 5-HT at extrahypothalamic sites. In particular, as seen in Figure 3, in mice low doses of LPS (1.0-5.0 |j,g) increased the utilization of NE within the PVN, as well as the hippocampus and prefrontal cortex (Lacosta et al, 1996). Given that DA variations within the nucleus accumbens are thought to play a pivotal role in subserving rewarding brain stimulation, we also assessed the effects of LPS (100 |ig) on accumbal DA release in vivo. As shown in Figure 4, LPS effectively increased the release of

Anhedonic and Anxiogenic Effects of Cytokine Exposure

209

11 11

It

I! O SI (ui!s;ojd 5uj/6u) OdHW y

cu

1%

a & ss

lis o « g PJ

OJ

1"

e s

(ujeiojd 6uj/6y> SdHrt

ra

-dj

"5 .a 0

ST

e

•J

tu

^ O

4>

VO

C f

rt

o

u
a c u U T*i <1J

Q

O

3

bi

b

tJl C

£^ ra xi

^

-n

o

Ui

^ ^^ a) ^ a

9 4-H BO O

(ujeiQjd 6ui;6y) r)dHtN

o

CU T3 O

H. Anisman and Z. Merali

210 200 180ieo-

I 1«^

120

g . 100-

< 80' 6040 10

14

16

Sample #

Figure 4. Percent change from baseline of extracellular DA from the nucleus accumbens over 30min samples. Using in vivo microdialysis, systemic LPS administration (lOO^ig) was found to increase DA release for about 1.5 hr (3 sampling periods). Exposure to mild air-puff presentations 4hr after LPS administration did not appreciably influence the response to LPS (from Borowski et al., 1997).

accumbal DA (Borowski et al., 1997). As will be seen shortly, this effect (as well as the behavioral actions of LPS) are distinct from those elicited by IL-ip. In Summary, our findings suggest that in addition to promoting sickness, LPS may engender an anxiogenic-like reaction. Of course, it is still unclear whether such an outcome was a direct result of the central amine or peptide variations induced by the treatment or were secondary to the malaise engendered by the endotoxin. Moreover, it appeared that LPS effectively influenced responding for otherwise rewarding brain stimulation, but the characteristic pattern observed differed from that ordinarily evoked by stressors. In particular, at higher current intensities rats readily responded for rewarding stimulation, but as the rewarding value of stimulation diminished, the drop-off of responding was particularly precipitous. Once these animals ceased responding, the operant did not reappear during the ascending series despite the presence of high current intensities. In effect, in the presence of sickness or anhedonia engendered by the LPS, the cost of responding exceeded the benefits derived, and hence responding did not progress. It remains to be established whether a similar profile would be evident in a progressive ratio situation (Richardson & Roberts, 1996), where current was held constant but cost of responding was increased (by progressively increasing the number of responses needed to obtain a reward). In a different type of paradigm which has been used to assess anhedonia, Yirmiya (1996) found that LPS (like stressors) reduced consumption of sucrose, and this effect was attenuated by chronic pretreatment with an antidepressant agent (suggesting a parallel to depressive illness). We have, as yet, not assessed the effects of chronic antidepressant treatment on LPS-provoked disturbances of self stimulation responding. 5.2. Behavioral and Neurochemical Consequences of IL-2 As indicated earlier, the few behavioral analyses of patients undergoing immunotherapy that have been conducted indicated that chronic IL-2 administration

Anhedonic and Anxiogenic Effects of Cytokine Exposure

211

markedly influenced cognitive processing, and promoted psychiatric disturbances. Thus, it was of interest to estabHsh whether systemic IL-2 administration would likewise influence anxiety or affect responding for rewarding stimulation in rodent studies. At the outset, it must be emphasized that in most of our experiments the effects of acute IL-2 treatment were assessed, and only recently have we begun to evaluate the effects of chronic IL-2 treatment. As in other reports (Hanisch & Quirion, 1996; Hanisch, Rowe, Sharma, Meaney, & Quirion, 1994; Hanisch, Neuhaus, Rowe, van Rossum, MoUer, Kettenmann, & Quirion, 1997), our studies revealed that the behavioral effects of IL-2, as well as central neurochemical changes induced by such treatments, were most notable after a chronic IL-2 regimen. 5.2.1. Anxiety Tests. The effects of acute IL-2 treatment could be distinguished from the actions of LPS, as well as those elicited by proinflammatory cytokines (IL-ip andTNFa). In CD-1 mice acutely injected with IL-2 (0.05-1.6|xg; 550-17,600lU) there was no evidence of any change of exploratory patterns in an open field, including locomotor activity, rearing, grooming, sitting digging, or approach to a novel object. Likewise, when mice were permitted to freely explore a symmetrical Y-maze, IL-2 did not influence spontaneous alternation (i.e., mice tended to enter those arms of the maze least recently visited). Thus, these data suggest that IL-2 treated animals reacted normally to environmental cues, at least within a nonthreatening context. In addition to the lack of an effect in an exploratory situation, we observed that in tests of anxiety, such as the light-dark box, and in the elevated plus-maze, acute systemic IL-2 treatment was absolutely without effect on performance (Lacosta, Merali, & Anisman, 1998b). It was likewise reported that murine IL-2 did not influence elevated plus-maze performance in mice (Petitto et al., 1997), while we observed that in rats icv administration of IL-2 also failed to affect plus-maze performance (Connor, Song, Leonard, Merali, & Anisman, 1998). Although we have not yet assessed the effects of chronic IL2 on performance in anxiety tests, we have observed that chronic IL-2 treatment over 7 days affected exploratory patterns. Specifically, while chronically treated mice did not display any signs of illness, they exhibited reduced levels of locomotor activity, and tended to make less frequent contact with a novel stimulus object (Lacosta, Merali, & Anisman, 1998b). Thus, it remains possible that chronic IL-2 (unlike acute treatment) impacts on anxiety. 5.2.2. Spatial Learning. Our initial experiments suggested that acute administration of hIL-2 did not provoke anxiety, nor did it affect several other behaviors in exploratory or habituation tests. Yet, it was reported that IL-2 administered repeatedly directly into brain disrupted Morris-water maze performance (Hanisch & Quirion, 1996), a finding commensurate with data indicating that IL-2 altered hippocampal ACh activity (Araujo et al., 1989; Hanisch, Seto, & Quirion, 1993) which is presumably involved in spatial learning. As well, humans undergoing IL-2 immunotherapy also exhibited disturbances in tests involving spatial learning or performance (Capuron, Ravaud, Radat, Dantzer, & Goodall, 1998). Accordingly, we conducted several experiments in which we assessed the effects of acute and chronic systemically administered IL-2 on Morris water-maze performance. In the Morris water-maze, animals are placed in a water-filled pool (often with milk powder added to make the water opaque), and required to learn the position of a platform submerged just beneath the surface. As each trial begins from a different location, animals need to use extraneous cues to localize and to find the platform.There

H. Anisman and Z. Merali

212

40 30 20-

10

— 0 — Saline - _ ^ [L.2

I

Ai ^^NN-t-r^

V-^ Days

Figure 5. Mice received 7 days of systemic IL-2 (l.O^ig/mouse; 1.1 x 10"lU) or vehicle and then tested in a Morris water-maze where the location of the safe platform varied over days. Mice received 8 trials a day over 8 days, and continued to receive the IL-2 or vehicle treatments 1 hr before test on each test day. The left hand panel shows that performance in IL-2 treated mice over days was disrupted by the chronic IL-2 treatment. The right hand panel indicates that on the first trial of each day vehicle and IL-2 mice exhibited comparable performance. However, while rapid improvement was seen over trials in vehicle animals, the betweentrial improvement was limited in IL-2 treated mice (from Lacosta et al., 1998a).

are several variants of this procedure: in some paradigms the location of the platform is always the same (e.g., N.E. quadrant), while in other paradigms a new location is used on each day. In the latter paradigm, on the first trial of each day the animal is unaware of the location of the platform, although the animal will have learned the concept that there is, in fact, a platform located within the pool (reference memory), and hence latencies decline over days. Having found the platform on any given day, subsequent trials evaluate the animal's ability to recall the position of the platform (working memory) (Whishaw, 1985). When animals were tested in a simple Morris water-maze (fixed location of the platform) animals acutely treated with hIL-2 (1.0 ^ig; 1.1 x 10^ lU, 1 hr before test on each of 5 successive days), exhibited proficient performance. In subsequent studies we assessed the effects of chronic IL-2 on performance. However, as the fixed location paradigm is so quickly acquired, the potential effects of IL-2 may be obfuscated. Accordingly, mice were tested in the more complex Morris water-maze in which the position of the platform varied over days. Mice received 7 days of i.p. IL-2 treatment (1.0|ig; 1.1 X lO^'IU) or saline administration, after which training commenced. On each of the ensuing 8 days mice received i.p IL-2 treatment and tested 1 hr afterward (8 trials/day). As seen in Figure 5, performance in the spatial learning test was impaired following chronic IL-2 treatment. This disruption was evident as early as the first test day. Importantly, when the data were assessed over the 8 daily trials, it was noted that on the first trial of each session the groups did not differ, suggesting that they were equally capable of swimming and employing a search strategy to find the platform. However, while saUne-treated mice gained from the first trial experience, showing marked between-trial improvements, the IL-2 treated mice seemed less disposed to gaining from the experience of the first trial. Thus, it seems that the animal's working spatial memory was impaired by the chronic IL-2 treatment. 5.2.3. Anhedonia. Acute, systemic IL-2 treatment influenced responding for rewarding brain stimulation in both rats and mice. In the initial experiments rats had electrodes implanted within the lateral hypothalamus (medial forebrain bundle) and trained to respond on the discriminated titration paradigm until stable responding was achieved (<10% between days variation, coupled with >95% choice accuracy in the dis-

Anhedonic and Anxiogenic EfTects of Cytokine Exposure 200'

)L-2 1.0 ug

213 IL-2 1,0 ug

£ 150 E loo-

se

15 20

25 30 35 40 Current (uA)

45 50

0.25 hr 24 hr 1 week 2 weeks Time Post-Treatment

Figure 6. Following the establishment of stable baseline responding (less than 10% between-days variability) rats received acute administration of hIL-2 (l.Ong). The cytokine disrupted responding for rewarding stimulation from the lateral hypothalamus in a current titration paradigm. Relative to baseline, responding was reduced at the intermediate and higher current intensities, and this effect was evident as long as 1 week following cytokine treatment (left-hand panel). The current threshold necessary to provoke half-maximal responding was increased commensurately (right-hand panel) (from Anisman et al., 1998).

crimination component). Under these conditions a single administration of IL-2 (1.0 |ig; 2.4 X 10^ units from R&D Systems) provoked modest, but statistically significant reductions of responding for rewarding brain stimulation (Anisman, Kokkinidis, & Merali, 1996; Anisman, Kokkinidis, Borowski, & Merali, 1998) (see Figure 6). Interestingly, this effect was still evident 24-168 hr afterward. As in the case of stressor effects, if animals received the cytokine, but were not tested until some time afterward (e.g. 7 days), then responding for rewarding brain stimulation was unimpaired. In effect, in order for long-term deficits to be evident, it was necessary for animals to experience the stimulation in the presence of the cytokine effects (possibly reflecting either an extinction-like response due to lack of reward, or a conditioned suppression like that seen when responding is aversive). In contrast to the effects of the cytokine on responding for rewarding brain stimulation, discrimination performance was unaffected by the treatment. That is, rats continued to emit the response to the -t- stimulus and not to the—stimulus. Coupled with the data indicating that acute IL-2 treatment did not affect exploration, or spontaneous alternation, these data suggest that animals were capable of appropriately attending to environmental stimuli, and maintaining a learned response. Thus, these data are consistent with the view that IL-2 may promote anhedonic effects. As in our studies with rats, it was observed that icv IL-2 administration in CD-I mice (5.0ng) impaired responding for rewarding stimulation from the dorsal aspect of the ventral tegmentum (Hebb, Zacharko, & Anisman, 1998). The VTA contains the cell bodies of the mesocorticolimbic system, and is thought to be fundamental in subserving rewarding brain stimulation. As the dorsal and ventral aspects of the VTA are differentially influenced by stressors (Zacharko & Anisman, 1991), we assessed the impact of icv IL-2 on responding for rewarding stimulation from both these regions. It was observed that in a titration paradigm (in this experiment frequency rather than current intensity was altered in order to minimize current spread) IL-2 impaired responding from the dorsal VTA. Although there were few animals with electrodes located in the ventral aspect of the VTA, it appeared that IL-2 did not disrupt responding from this

214

H. Anisman and Z. Merali

region. Together, these data provisionally suggest that (a) IL-2 may induce an anhedonic effect, as reflected by altered responding from both the dorsal VTA and the lateral hypothalamus, (b) these effects likely reflect the central actions of the cytokine, and (c) the effects engendered by IL-2 are reminiscent of those induced by stressors in terms of the protracted behavioral changes observed, and the regionspecific effects noted. 5.2.4. Neurochemical alterations. In evaluating the neuroendocrine effects of IL-2, it became clear that the effects of this cytokine were readily distinguishable from the effects of stressors. For instance, while stress potently activates HPA activity, we observed across a range of doses that acute i.p. administration of IL-2 did not influence plasma corticosterone or ACTH levels (Lacosta et al., 1998a). Moreover, we observed that even after 7 days of IL-2 treatment, which it will be recalled influences exploration and Morris water-maze performance, plasma corticosterone levels were unaltered. Across a series of doses IL-2 (0.1-L6|ig; LI x 10'*IU/fig) did not appreciably influence NE, DA, 5-HT or their respective metabolites in several mouse brain regions, including the hypothalamus, locus coeruleus, hippocampus, prefrontal cortex, nucleus accumbens, substantia nigra, and caudate (Lacosta et al., 1998). Of course, static measurements of amines and metabolites in post mortem tissues may not provide a clear picture of transmitter utilization, and thus we also evaluated in vivo variations of amine release in response to 1.0|j.g of systemically administered IL-2 (Anisman et al., 1996; Song, Merali, & Anisman, 1998). In this experiment we assessed DA released from the nucleus accumbens, since this region was thought to be essential in subserving the rewarding value of electrical brain stimulation. In contrast to the effects of stressors which ordinarily enhanced DA release from the shell of the nucleus accumbens (Kalivas & Duffy, 1995), DA release was greatly inhibited by IL-2 treatment, and this effect was further exacerbated by application of a stressor (mild air-puff) (see Figure 7). It ought to be underscored, however, that IL-2 may have dose-dependent biphasic effects on DA and ACh activity (Araujo et al., 1989; Lapchak, 1992). Thus, while IL-2 inhibited DA release from the accumbens, it is essential to conduct more detailed doseresponse analyses in order to understand the actions of acute IL-2 treatment on motivational systems. As well, as stressors differentially influence DA release from the core and the shell of the nucleus accumbens, it may be important to distinguish the effects of IL-2 from both these regions. We have recently observed that the differential effects of acute and chronic IL-2 noted in several behavioral paradigms, are also evident with respect to brain amine variations. In particular, contrary to the lack of effect seen after acute treatment, repeated administration of IL-2 (1.0 |ig over each of 7 days) resulted in marked regionspecific biogenic amine variations. In mice that had received chronic IL-2 (1.0|ig; 1.1 X lO'^IU), the level and/or utilization of NE was elevated in the median eminence (although not in the PVN), hippocampus and prefrontal cortex. Hippocampal 5-HT utilization was likewise elevated, while in the prefrontal cortex 5-HT levels declined. It is still premature to ascribe specific cognitive impairments associated with IL-2 immunotherapy in humans to such neurochemical changes in the rodent brain, but it is significant that the spatial disturbances seen in human patients is paralleled by spatial disturbances in the Morris water-maze test. It is conceivable that the behavioral changes described earlier, as well as the cognitive disturbances in humans, may represent the extrahypothalamic amine changes associated with a chronic IL-2 regimen.

Anhedonic and Anxiogenic Effects of Cytokine Exposure

215

140-1

120-

100

£

80-

60-

40'

-•

1 8

•-

Sample*

Figure 7. Following 4 baseline dialysate samples (30min/sample) being taken, rats received either IL-2 (l.Ojxg; 1.1 X 10''IU) or saline treatment. Release of DA from the nucleus about 2.5hr following treatment declined by approximately 25%. Upon presentation of an air-puff the decline was more pronounced, reaching 50% of baseline (from Anisman et al., 1996).

5.3. Behavioral and Neurochemical Consequences of IL-lp The majority of studies assessing the behavioral effects of IL-ip have focused on sickness behavior (including sleep and anorexia) and the processes subserving these effects; however, assessments have also been made of IL-ip effects on other behaviors, including exploration, and acquisition of learned appetitive responses (Bluthe et al., 1997; Dantzer et al., 1996; Linthorst et al., 1995; Moldofsky, 1995; Plata-Salaman, 1998). In the main, these studies have shown that IL-ip readily disrupted behavioral outputs. Given the profound illness induced by IL-lp, it is difficult to discern to what extent the behavioral effects of the cytokine reflect changes of motivational or anxiety processes. Yet, as indicated earlier, IL-lra has been shown to attenuate the central amine and neuroendocrine changes induced by stressors (Shintani et al., 1995a, b), and to antagonize the shuttle-escape performance disturbances otherwise provoked by prior exposure to inescapable shock (Maier & Watkins, 1995). Thus, there is reason to suppose that the effects of IL-1(3 might reflect the stress-like actions, rather than sickness induced by the cytokine. Accordingly, we conducted several experiments to assess the actions of ILip, alone or in conjunction with either TNFa or IL-6, on behavioral outputs and on central neurochemical processes. 5.3.1. Palatable Food Intake. Behaviorally, IL-1 produced a dose dependent disruption in the consumption of a highly palatable substance (chocolate milk), and in an open field situation this cytokine reduced active responding and reduced exploratory tendencies (Lacosta, Merali, & Anisman, 1998b). Interestingly, although IL-lp (0.21.6|ig) substantially reduced locomotor activity (and accompanying exploration), at

H. Anisman and Z. Merali

216

low doses IL-ip (0.2 and 0.4|j,g) induced a transient increase of exploratory tendencies directed towards a novel object. When animals were tested in a free-running spontaneous alternation task, they exhibited high levels of alternation, comparable to that seen in saline treated mice. Indeed, this was the case even if mice received 7 days of treatment (0.5|ig/mouse/day). As spontaneous alternation performance is thought to reflect habituation of exploratory tendencies (Kokkinidis & Anisman, 1980), these data, like those involving LPS, suggest that IL-lp treated mice are able to attend and respond appropriately to the environmental cues, at least when placed in a novel situation (Lacosta et al., 1998b). 5.3.2. Anxiogenic Effects. The behavioral profile of mice in anxiety-related situations differed from that seen in following IL-2 treatment. Systemic administration of IL-ip (at low doses) provoked a decline in the number of entries into the open arms of an elevated plus maze, without an appreciable change of closed-arm entries. Figure 8 shows that while IL-ip induced an anxiogenic-like effect (reduced open arm entries in a plus maze), IL-2 did not provoke such an effect. We have observed that the anxiogenic effect of IL-ip tended to appear at low doses, and was less remarkable at doses above 0.8 (ig, possibly owing to the masking of anxiety by the sickness (Lacosta et al., 1996). In addition to the effects of systemic IL-lp, central IL-lp administration also decreased entries into the open arms of a plus maze (Connor et al., 1998). Thus, these data suggest that IL-ip induces anxiogenic-like effects, stemming from the central actions of this cytokine. However, a note of caution must be introduced at this juncture. We have observed that the plus-maze test is generally associated with a high degree of between-animal variability, and we have not uniformly found IL-ip to induce anxiogenic-like effects in this paradigm. In one study, for instance, we observed that the cytokine did not promote a statistically significant decline in open-arm entries. However, in this study animals engaged in a great increase of risk-assessment (i.e., the animal sits at the choice point, and extends its forelimbs and body onto the open arm, but without fully entering on to this arm). Thus, while there is reason to suppose that IL-ip induces an anxiogenic-hke action, such a conclusion cannot be considered as being unequivocal. 5.3.3. Anhedonic Effects. With respect to the potential anhedonic effects of ILip, we assessed the effects of this cytokine (as well as that of other cytokines) on the consumption of a highly palatable food substance (chocolate milk). After stable consumption was established (less than 10% between days variability) mice were injected with IL-ip and I h r later their consumption was monitored over a I h r period. As seen in Figure 9, the cytokine provoked a dose dependent reduction of chocolate milk con-

VEHICLE

IL-1

IL-2

Figure 8. Administration of IL-ip (0.5 jig) reduced the number of open-arm entries emitted by mice tested on an elevated plusmaze ( I h r after treatment), while IL-2 (0.5^ig) was without effect. Neither treatment influenced the frequency of open arm entries.

Anhedonic and Anxiogenic Effects of Cytokine Exposure

217

IL-1 DDSB

BasBliie

Test

Recovery

Baseline

Test

Recovery

Figure 9. Relative to baseline rates of chocolate milk consumption, mice that received intraperitoneal administration of IL-ip (left-hand panel) or TNFa (right-hand panel) exhibited dose-dependent reductions of consumption (measured over 2hr commencing 1 hr after treatment). Upon re testing 24 hr later, in the absence of further treatment, consumption levels returned to baseline values.

sumption (as did TNFa), which returned to basal levels upon subsequent retesting 24 hr later (Brebner, Hayley, Lacosta, Zacharko, Merali, & Anisman, 1998). While IL-ip reliably reduced consumption, such an outcome may have been related to the illness induced by the cytokine, or even a change in the perceived palatability of the food substance, rather than an anhedonic effect. Indeed, several experiments conducted in rats provisionally suggested that IL-ip has limited effects on reward processes. In rats with electrodes positioned within the medial forebrain bundle, treatment with IL-ip (1.0)j.g) did not alter responding for rewarding brain stimulation (Figure 10). To rule out the possibility that the lack of effects were attributable to the source of the cytokine, we assessed the effects of IL-1J3 obtained from R&D Systems and from the National Cancer Institute (Bethesda, Md). In both cases IL-ip was without effect on responding for brain stimulation reward. Moreover, this occurred irrespective of whether rats were initially tested 1 hr after cytokine administration (corresponding to peak ACTH and corticosterone changes) or at earlier times (15min post

IL-1 1.0 ug

IL-1 1.0 ug

<

15 20

25

30

35 40

Current (uA)

45

50

30

weeks

Figure 10. Following the establishment of stable rates of responding for rewarding stimulation from the lateral hypothalamus on a current titration schedule (baseline), rats received intraperitoneal injection of IL1 (3 and then were repeatedly tested for self-stimulation performance at various times afterward. Treatment with the cytokine did not influence responding in this paradigm. The left hand panel depicts the response/min (±SEM), while the right-hand panel shows the response thresholds (from Anisman et al., 1998).

218

H. Anisman and Z. Merali

injection) (Anisman et al., 1996, 1998). When animals were tested with a higher dose of IL-ip (2.0jxg/rat), a modest reduction of responding was observed. However, at this dose pronounced signs of illness were evident. As well, marked variability of performance was noted in response to the cytokine treatment. Two of the 7 rats tested ceased responding entirely, while the remaining rats were only modestly affected. Thus, while the overall response rate was not significantly different from baseHne, responding during the ascending portion of the titration schedule was significantly impaired. This profile was somewhat similar to that elicited by LPS, and clearly distinguishable from that provoked by IL-2.Thus, it seems that the altered responding induced by IL-ip was more appropriately attributed to illness induced by the cytokine than direct effects on reward processes (Anisman et al., 1998). 5.3.4. Neurochemical Effects. As observed by other investigators (Dunn, 1995; Rivier, 1993), we have reliably found systemic IL-ip treatment to provoke marked increases of plasma ACTH and corticosterone, with the latter effects being evident at doses as low as 0.025 and 0.05 |ig. Systemic IL-lp also provoked region-specific alterations of biogenic amines in the brain (see Figure 11). Although NE levels and utilization were unaltered in the entire hypothalamus, NE levels were reduced in the PVN, while the utilization of NE (as reflected by MHPG accumulation) was increased in the arcuate plus median eminence. Furthermore, IL-ip increased NE utilization within the locus coeruleus, and induced somewhat less of an effect in the prefrontal cortex. Within nigrostriatal regions, as well as in the nucleus accumbens, there was no indication of any amine variations (Lacosta et al., 1998b). It will be recalled that a static measurement of amine levels or metabolites may not provide a sufficiently accurate index of amine activity, and thus experiments were conducted in rats to determine in vivo amine variations. These experiments revealed that IL-ip (1.0|ig/rat) had little effect on DA release from the nucleus accumbens or the prefrontal cortex. Interestingly, as seen in Figure 12, IL-1(3 affected the release of accumbal 5-HT, as reflected by an accumulation of 5-HIAA, and these effects were exacerbated by the introduction of a stressor (air-puff) (Merali, Lacosta, & Anisman, 1997; Song et al., 1998). IL-ip was also found to influence 5-HT release at the hippocampus, but the interaction of the IL-ip and stressor effects were less profound. As in the case of IL-2, the amine variations were sustained for a protracted period of time, even though the half-life of IL-ip is fairly brief. Thus, it is possible that IL-ip sets into motion a series of peripheral and/or central effects that sustain the altered amine variations. Overall, it appears from these in vivo experiments that IL-lp influences 5-HT turnover, and that the neurochemical response associated with the cytokine may be influenced by the introduction of a relatively mild stressor. In effect, it may be important to consider that the behavior and neurochemical changes associated with cytokine and endotoxin challenge may be dependent on the stimulus context in which animals are tested. Essentially, as will be discussed shortly, following introduction of sicknessinducing agents the behavior of animals in their normal surroundings may be very different from that apparent in a novel, and thus somewhat threatening environment. Although it has been suggested that systemic stressors (such as cytokine administration) would not affect HPA functioning via an amygdaloid mechanism, it ought to be noted that IL-ip may augment NE release from the central amygdala. In particular, in a recent experiment, we assessed the effects of IL-1 treatment in two strains of rats that had originally been bred for Fast or Slow development of generalized convulsive seizures elicited by amygdala stimulation (kindling) (Merali, Mclntyre,

Anhedonic and Anxiogenic Effects of Cytokine Exposure

(ujaiojd jUJ/Bu) SdHlM

219

H. Anisman and Z. Merali

220

I50n Accumbal 5-HIAA

\ 140-

1±

- O - I L - 1 + Stress V fL-1, no Stress ~ n — S a l i n e + Stress

130-

120-

110-

100-

stress baseline 0H

1

r—I

.—

—I—

10

-r12

14

16

Figure 12. Percent change (+S.E.M) of extracellular 5-hydroxyindole acetic acid (5-HIAA) at the nucleus accumbens over 30min samples. After 4 samples rats received intraperitoneal IL-lp (1.0|ig) or saline and 5 dialysate samples collected. Saline treated rats (open squares), and one group of rats that received IL-ip (closed circles) then received a series of 5 air puff stressors, after which 5 further samples were collected. Rats that received IL-ip, but were not exposed to the air puff are represented by open triangles. Treatment with IL-ip increased accumbal 5-HIAA relative to baseline and relative to saline treated rats. While the airpuff hardly affected 5-HIAA in saline-treated rats, a marked increase of 5-HIAA was evident in rats that had received IL-ip (from MeraH et al., 1997).

Michaud, & Anisman, unpublished report). However, it ought to be noted that these rats also differ in their response to stressors. The Fast rats tend to be more reactive (struggle more when restrained) and exhibit poorer inhibitory response tendencies in a passive avoidance test, habituation in an exploratory situation, as well as in a foodmotivated delayed alternation test (Anisman, Lu, Song, Kent, Mclntyre, & Merali, 1997; Mohapel & Mclntyre, 1998). As shown in Figure 13, we observed that in the Fast kindling rats the release of NE at the central amygdala was not affected by the IL-ip (1.0 |ig) treatment. In contrast, in the Slow kindling rats a marked rise of NE release was evident as early as 0.5 h following IL-1 administration, and persisted over the ensuing 2.5 hr test. Ordinarily, NE at the amygdala acts to inhibit kindling-induced seizure (Mclntyre, 1980), raising the possibility that the Slow strain is less prone to seizures owing to greater reactivity of amygdala NE neurons. Of course, at this juncture the relationship between amygdala NE and kindling in these strains is highly speculative. For the present purposes, the main point is that IL-ip may influence amygdala NE release, but organismic factors, possibly related to NE neuronal lability, may determine the central actions of IL-1 (3. Inasmuch as the amygdala Hkely plays a prominent role in subserving anxiety (Lee & Davis, 1997), and CRH and NE may contribute in this

Anhedonic and Anxiogenic Effects of Cytokine Exposure

221

600 n

^•VT baselinf

saline inj

500 i

400 -•-Slow

I

300-

^

200-

100-

-0-Fast

Izt:^. 7

8

9

n 1 1 1 1 1 r10 11 12 13 14 15 16

Sample a

Figure 13. Levels of NE at the central amygdala (as a percent of baseline ± S.E.M) in rats selectively bred for Fast and Slow seizure development in response to stimulation of the amygdala (kindling). Following 5 baseline samples (30min/sample), rats of both strains received intraperitoneal injection of saline to assess the effects of a mild stressor. After 5 additional samples were collected rats received intraperitoneal IL-ip treatment (1.0|j.g). A marked increase of NE release was evident in the "Slow" rats, while NE release was unaffected in the "Fast rats" (from Merali, Michaud, Mclntyre, & Anisman, unpublished report).

respect, the data are consistent with the proposition that IL-ip may act in an anxiogenic capacity. 5.3.5. Concluding Remarks Regarding IL-ip. In addition to promoting sickness behaviors, IL-ip may contribute to the promotion of an anxiogenic-Uke response. Moreover, paralleUng the effects of stressors, IL-ip treatment influenced HPA activity, and affected NE release at the PVN, locus coeruleus and central amygdala, all of which have been associated with anxiety. We have, as yet, not determined whether the behavioral changes induced by IL-ip reflect the direct effects of the cytokine on anxiogenic processes, such behaviors are secondary to illness, or the behavioral and amine consequences associated with IL-ip are modifiable by anxiolytics or antidepressant agents. Unlike the effects of IL-2, it did not appear that IL-ip readily affected responding for rewarding brain stimulation. Moreover, unlike the effects of stressors, which provoke an anhedonic response, IL-ip did not alter DA release at the nucleus accumbens or prefrontal cortex. Thus, while IL-lp has characteristic effects reminiscent of those elicited by stressors, the induction of an anhedonic reaction is not one of these. In considering the neurochemical and behavioral effects of IL-ip treatment, one additional point warrants mention. We have reliably observed that following IL-ip treatment, mice in a famihar environment (e.g., home cage) tend to exhibit a rapid decline of active behaviors, followed by the typical sickness profile. However, when these animals are transferred to a novel situation they exhibit a marked increase in their activity and exploration (although perhaps not to the level of control animals) and appear to respond appropriately to environmental cues (as judged by their proficiency in exploring the novel areas of a Y maze in a spontaneous alternation test). Likewise, it appeared that the effects of IL-ip on central amine activity could be appre-

222

H. Anisman and Z. Merali

ciably enhanced by the appHcation of a nonpainful stressor (air-puff), and indeed the two manipulations were found to act synergistically. A similar outcome was observed with respect to the actions of IL-1(3 coupled with novelty in affecting plasma ACTH and corticosterone concentrations (Zhou, Shanks, Riechman, Liang, Kusnecov, & Rabin, 1996). In effect, although it may be adaptive for animals to adopt the "sickness behaviors" described by Kent et al. (1992), it may be the case that in novel, potentially threatening situation, cytokine treated animals will give up this behavior pattern and will instead respond to environmental stimuli in a manner like that of untreated animals. It may be profitable to consider that the context in which the cytokine is apphed may appreciably influence behavioral outcomes and may act upon neuroendocrine and neurotransmitter processes, as well.

5.4. Behavioral and Neurochemical Consequences of TNFa We have recently begun to explore the possibility that TNFa may promote either anxiogenic or anhedonic effects. Admittedly, our data in this respect are limited, but they do provide some early clues as to the behavioral effects of this proinflammatory cytokine. Before outlining our findings, we make particular note of the fact that we observed marked differences between the effects of TNFa obtained from different sources (R&D Systems vs PeproTech). In the studies that are described, human TNFa was obtained from R & D systems. 5.4.1. Anxiety and Anhedonia. We have, as yet, not undertaken a detailed analysis of the anxiogenic and anhedonic effects of systemic TNFa administration. However, in preliminary studies (Hayley, Merali, & Anisman, unpublished report) we observed that TNFa, administered i.p. (4.0 fig/mouse), markedly reduced open-arm entries in an elevated plus maze test, without appreciable changes of closed-arm entries. Likewise, we have found that when administered icv an anxiogenic-like effect was observed in a plus-maze test (Connor et al., 1998). In an effort to assess the effects of TNFa on reward processes, we began by evaluating the effects of the cytokine on consumption of chocolate milk in mice that had not been food-deprived. As before, mice were habituated to this diet for approximately 10 days. Once, the between-day variability was less than 10%, the effects of cytokine was assessed. Administration of TNFa, like that of IL-1|3, reduced consumption of the palatable substance 1 hr following cytokine exposure. Furthermore, as previously reported (e.g., Sonti et al., 1996; Yang, Koseki, Meguid, Gleason, & Debonis, 1994), subthreshold doses of the two treatments acted synergistically to reduce chocolate milk consumption (see Figure 14) (Brebner et al., 1998). It is premature at this juncture to make firm conclusions concerning the effects of TNFa on responding for rewarding stimulation from the lateral hypothalamus, as we have only examined the effects in a small number of mice (Brebner, Merali, & Anisman, unpublished report). However, at doses that markedly influenced chocolate milk consumption (4.0 "ug), responding for rewarding stimulation remained unaffected. Only under conditions where TNFa induced a marked illness (e.g., upon reexposure to TNFa 2 weeks after initial administration, as will be described shortly) was performance in the self stimulation test affected. At this point the relative contribution of anhedonia and sickness is difficult to gauge. 5.4.2. Neurochemical Effects. In many respects the effects of TNFa on neurochemical processes were aligned with variations induced by IL-ip. In particular, TNFa

Anhedonic and Anxiogenic Effects of Cytokine Exposure

223

4.5

Figure 14. Following the establishment of stable chocolate milk consumption (baseline) mice received treatment with either vehicle, subthreshold levels of IL-ip (0.05^g) or TNFa (l.Ong), or a combination of the two cytokine treatments. Consumption was assessed over 2hr commencing Ihr after intraperitoneal injection. While neither cytokine alone affected intake, mice that received both cytokines displayed a marked reduction of consumption. Upon retesting 24 hr later, intake had returned to baseline levels.

2.5

1.5

Baseline

Test

Recovery

dose-dependently increased plasma corticosterone, as did low doses of IL-ip. Paralleling the consumption of a palatable substance, co-administration of modest does of TNFa and IL-ip induced corticosterone variations that were greater than those elicited by the additive effects of the individual treatments (see Figure 15) (Brebner et al., 1998). In contrast, we did not observe synergistic effects between IL-lp and IL-6 or between TNFa and IL-6, although other investigators reported that IL-lp acted synergistically with IL-6 in promoting ACTH release (Zhou et al., 1996). Of course, it is premature to exclude the possibility of IL-ip and IL-6 or IL-ip and TNFa acting synergistically in our paradigm, since such an outcome could be dependent on the specific cytokine doses used, the timing between the two cytokine treatments, as well as the time at which blood was sampled. Systemic administration of TNFa induced marked variations of neurotransmitter levels and functioning in post-mortem tissues. In several respects, the effects of TNFa were more pronounced than those induced by IL-lp, and could also be distinguished from the actions of the latter cytokine. The nature of the changes observed varied across brain regions, and also as a function of the transmitter in question. Among other things, TNFa dose dependently reduced NE in the hippocampus, and locus coeruleus, while increasing NE utilization within the amygdala. As well, 5-HT levels were reduced in the PVN, but were increased in the PFC and amygdala. Finally, in contrast to the lack of effect of IL-ip on accumbal DA, TNFa increased the utilization of DA within the nucleus accumbens, as reflected by increased DOPAC accumulation (Hayley et al., 1998).

Figure 15. Mice received i.p. injection of either vehicle or subthreshold doses of IL-ip (0.05jag) or TNFa (l.O^g), or both cytokines. Plasma corticosterone levels (Mean ± S.E.M) (collected Ihr after cytokine treatment) in mice that received both cytokines were greater than in mice that received vehicle or only one of the cytokines.

Veh + Veh IL-1 + T^F IL-l + Veh TNF + Veh

H. Anisman and Z. Merali

224 Reexposure Treatment ^ Saline D TNF{1.0ug)

Saline 1.0 2,0 Initial TNF Treatment (ug)

4.0

Figure 16. Mean (+S.E.M) consumption of chocolate milk among mice that received either saline or TNFa (1.0,2.0 or 4.0 (ig) and were then reexposed to either saline or a low dose of TNFa (1.0 ug) 2 weeks later. Consumption was assessed over a 2hr period beginning 1 hr after the second injection. The low dose reexposure treatment provoked a markedly greater reduction of consumption in animals that had previously received TNFa than in mice that had received saline. If the animals received the second injection 1 or 7 days following the initial TNFa treatment, then a sensitization effect was not evident (data not shown).

5.4.3. Sensitization Effects. In addition to it's immediate effects, systemic TNFa may engender the sensitization of central amine functioning, and may provoke protracted effects on behavior. Moreover, as in the case of IL-ip (Schmidt et al., 1995; Tilders et al., 1993), the appearance of at least some of these consequences may be time-dependent. It was observed that in a test of chocolate milk consumption, marked alterations of behavior may be induced upon reexposure to TNFa in animals previously treated with the cytokine. However, this depended on the interval between the two cytokine treatments. If animals were treated with TNFa (4.0 ug) and then tested 2 weeks later behavior was hardly affected. However, in animals that were retested following treatment with a low dose of TNFa (1.0|j,g), profound behavioral changes were observed (see Figure 16). These animals appeared much hke animals that received a much higher dosage of TNFa (i.e., beyond that of 4.0|j,g). In addition to reduced chocolate milk consumption, these animals exhibited clear signs of illness, including reduced motor activity, and marked soporific effects. In fact, reexposure to a somewhat higher dosage (2.0 |ig), which in itself had modest effects, was lethal to about one-third of the mice. Precisely the same effects were evident irrespective of whether testing was undertaken 2 or 4 weeks after the initial TNFa treatment. Interestingly, if the second administration took place at a fairly short time following the initial treatment (1 or 7 days), then the second treatment was without behavioral effect. Thus, it appeared that treatment with TNFa induced a sensitization effect that was dependent on the passage of time. Paralleling the behavioral sensitization, when mice were treated with TNFa (4.0 |j.g) and then exposed to a low dose of this cytokine, a marked time-dependent sensitization effect was evident with respect to plasma corticosterone concentrations. At brief intervals following the initial treatment (1 day) reexposure to TNFa (1.0|ag) provoked a modest rise of corticosterone levels (mean ± SEM = 16.05 ± 2.30|xg/dl) relative to control values (4.58 ± 0.81 M-g/dl). The magnitude of this became progressively greater as the interval between the two TNFa tests increased (1,2 and 4 weeks equaled 22.11 ± 3.33,29.94 ± 1.97, and 30.46 ± 3.85|Lig/dl, respectively). Sensitization effects were likewise apparent with respect to central neurotransmitter activity. Once more, the time-dependent nature of the sensitization was brain region-specific and also varied as a function of the particular neurotransmitter being assessed. As seen in Figure 17, in animals reexposed to the 1.0)Lig dose of TNFa 4 weeks following initial treatment there

Anhedonic and Aiudogenic Effects of Cytokine Exposure

225

PVN

I Sat Sal

Sat TNF

TNF Sal

TNF TNF TNF TNF TNF-1 TNF-7 TNF-14 TNF-23

Days Between TNF Reaxposure

PFC

15 129 63

^ Sa) Sd

Wy mnm Sal TNF

TMF Sal

mmmm.

TNF TNF TNF TNF TMF-t TNF-7 TWF-14TNF-28

Days Between TNF Reexposure

Figure 17. Mean (+S.E.M) concentrations of MHPG within the PVN (upper panel) and prefrontal cortex (lower panel) as a function of TNFa treatments. Mice received 2 intraperitoneal injections. The 3 groups at the left received either vehicle treatment only, vehicle followed two weeks later by a low dose of TNFa (l.O^g) or TNFa (4.0^g) followed by vehicle. The four groups on the right received two injections of TNFa; the first was a 4.0|a,g dose, while the second was 1.0|j.g. The two injections were administered either 1, 7,14 or 28 days apart. Note that a sensitization effect was evident with respect to PVN MHPG in mice that received the second TNFa treatment 28 days after initial exposure, but not at shorter intervals. In contrast, within the prefrontal cortex (as well as several other limbic sites not shown) the sensitization was only evident 1 day following initial treatment (from Hayley et al., 1998).

was an increase of NE utilization in the PVN relative to that seen in animals that received only a single TNFa injection. A comparable amine change was not evident at earlier intervals following the initial TNFa treatment. Inasmuch as the plasma corticosterone changes also increased over time, these two changes may be related to one another. In contrast to the PVN NE variations, an entirely different profile of changes was seen in other brain regions. In particular, both NE and 5-HT utilization in the hippocampus, amygdala, and PFC were very much increased in animals reexposed to TNFa 1 day following the initial treatment. At the longer intervals (7,14, and 28 days), however, there was no indication of a sensitization effect (see Figure 17). Finally, as in the case of the aforementioned amine changes, accumbal DA utilization was evident upon reexposure to TNFa 1 day after initial treatment, but not at longer intervals. We have, as yet, not had the opportunity to assess peptide changes in detail in animals treated with TNFa. However, in some recent experiments (Hayley, Staines, Mclntyre,

226

H. Anisman and Z. Merali

Merali, & Anisman, unpublished report) we observed that 14 days after TNFa treatment there was a marked increase of cFOS expression in the PVN, preoptic nucleus, central amygdala, and the PVN of the thalamus. Moreover, at this time CRH and AVP coexpression was increased within CRH neurons in the external zone of the median eminence. The latter finding, it will be recognized, is commensurate with reports concerning the effects of IL-1 and stressors (Bartanusz et al., 1993; Schmidt et al, 1995; Tilders et al., 1993), although it remains to be determined whether this apparent sensitization increases with time following initial treatment. It is important to underscore at this juncture that the nature of the sensitization effects associated with TNFa lead to the suggestion that independent circuits seem to be involved in the augmented HPA functioning, and those involved in limbic activity. Although it is often assumed that hippocampal 5-HT or NE activity may be involved in PVN activation, it is clear that effects on the hippocampus (as well as the amygdala and PFC) are not necessarily translated as either PVN NE or corticosterone variations. Conversely, PVN NE and plasma corticosterone changes are not necessarily linked to variations of limbic amines. Furthermore, it is abundantly clear that the amine sensitization effects seen 1 day after initial TNFa treatment are independent of any malaise animals suffered from the treatment, which did not occur until 2 weeks after initial treatment. Of course, it is possible that the elevated PVN NE activity, as well as the increases of plasma corticosterone observed at the longer intervals were, in fact, secondary to illness associated with the reexposure treatment.

6. SUMMARY Systemic interleukin IL-ip, TNFa, and IL-2 profoundly influenced central monoamine activity, as well as behavioral outputs. The effects of the various cytokines were clearly distinguishable from one another, although synergistic effects were detected between several of these cytokines and between the actions of cytokines and stressors. Acutely applied IL-2 appeared to affect reward processes, but did not affect anxiety. When chronically administered, this cytokine markedly influenced working memory in a spatial learning test. In contrast to IL-2, both IL-1 [3 and TNFa appeared to provoke an anxiogenic action, and provoked clear signs of illness. While these cytokines induced anorexia, they did not appear to affect reward processes. IL-1|3 and TNFa were found to act synergistically, and the TNFa provoked a sensitization with respect to the action of subsequent TNFa treatment. The findings indicated that cytokine treatments profoundly influence extrahypothalamic neurochemical functioning and may thus impact on behavioral outputs. Analyses of the behavioral and neurochemical changes elicited by cytokines, and particularly TNFa, need to consider not only the immediate impact of such treatments, but also the proactive effects that may be engendered.

ACKNOWLEDGMENTS The research was supported by a grant from the Medical Reseach Council of Canada. We are indebted to Tom Borowski, Karen Brebner, Jenna Griffiths, Pam Kent, Shawn Hayley, Larry Kokkinidis, Jerzy Kulyczki, Susan Lacosta, Dan Mclntyre, Dave Michaud, and Marilee Zaharia for their assistance in the various facets of this work.

Anhedonic and Anxiogenic Effects of Cytokine Exposure

227

REFERENCES Alonso, R., Chaudieu, I., Dorio, J., Krishnamurthy, A., Quirion, R., & Boksa, P. (1993). Interleukin-2 modulates evoked release of [''HJdopamine in rat cultured mesencephalic cells, Journal of Neurochemistry, 61, 1284-1290. Anisman, H., Zalcman, S., Shanks, N., & Zacharko, R. M. (1991). Multisystem regulation of performance deficits induced by stressors: An animal model of depression. In A. Boulton, G. Baker, & M. MartinIverson (Eds.), Neuromethods, vol. 19: Animal Models of Psychiatry, 11. (pp. 1-59). New Jersey: Humana Press. Anisman, H., Zalcman, S., & Zacharko, R. M. (1993). The impact of stressors on immune and central neurotransmitter activity: Bidirectional communication. Reviews in the Neurosciences, 4,147-180. Anisman, H., Kokkinidis, L., & Merali, Z. (1996). Influence of interleukin-2 on responding for electrical brain stimulation from the hypothalamus. Brain Research, 731,1-11. Anisman, H., Lu, Z. W., Song, C , Kent, P., Mclntyre, D. C , & Merali, Z. (1997). Influence of psychogenic and neurogenic stressors on endocrine and immune activity: differential effects in fast and slow seizing rat strains. Brain, Behavior, and Immunity, 11, 63-74. Anisman, H., Kokkinidis, L., Borowski,T., & Merah, Z. (1998a). Differential effects of interleukin (IL)-l, IL2, and IL-6 on responding for rewarding lateral hypothalamic stimulation. Brain Research, 779, 177-187. Anisman, H., Ravindran, A. V., Griffiths, J., & Merah, Z. (1998b). Behavioral, endocrine, and cytokine correlates of major depression and dysthymia with typical or atypical features. Molecular Psychiatry, in press Araujo, B. M., Lapchak, P. A., Collier, B., & Quirion, R. (1989). Localization of interleukin-2 immunoreactivity and interleukin-2 receptors in the rat brain: interaction with the cholinergic system. Brain Research, 498, 257-266. Ban, E. M., Haour, E, & Lenstra, R. (1992). Brain interleukin-1 gene expression induced by peripheral lipopolysaccharide administration. Cytokine, 4, 48-54. Banks, W. A., Ortz, L., Plotkin, S. R., & Kasten, A. J. (1991). Human interleukin (IL)-l alpha, murine IL-1 alpha, and murine IL-2 beta are transported from blood to brain in the mouse by a shared saturable mechanism. Journal of Pharmacological and Experimental Therapeutics, 259, 988-996. Bartanusz, v., Jezova, D., Bertini, L.T.,Tilders, F. J., Aubry, J. M., & Kiss, J. Z. (1993). Stress-induced increase in vasopressin and corticotropin-releasing factor expression in hypophysiotrophic paraventricular neurons. Endocrinology, 132, 895-902. Bartholemew, S. A. & Hoffman, S. A. (1993). Effects of peripheral cytokine injections on multiple unit activity in the anterior hypothalamic area of the mouse. Brain, Behavior, and Immunity, 7, 301-316. Blalock, J. E. (1984). The immune system as a sensory organ. Journal of Immunology, 132, 1067-1070. Blalock, J. E. (1994).The syntax of immune-neuroendocrine communication. Immunology Today, 15,504-511. Bluthe, R. M., Dantzer, R., & Kelley, K. W. (1989). Corticotropin releasing hormone is not involved in the behavioral effects of peripherally injected interleukin-1 in the rat. Neuroscience Research Communincations, 5, 149-154. Bluthe, R. M., Michaud, B., Kelley, K. W., & Dantzer, R. (1996). Vagotomy attenuates behavioral effects of interleukin-1 injected peripherally but not centrally. Neuroreport, 7, 1485-1488. Bluthe, R. M., Dantzer. R., & Kelley, K. W. (1997). Central mediation of the effects of interleukin-1 on social exploration and body weight in mice. Psychoneuroendocrinology, 22, 1-11. Borowski,T., Kokkinidis, L., Merali, Z., & Anisman, H. (1997). Behavioral and central monoamine alterations induced by interleukin-1, interleukin-6, and lipopolysaccharide. Society for Neurosciences Abstract, 23, 1508. Brady, L. S., Lynn, A. B., Herkenham, M., & Gottesfeld, Z. (1994). Systemic interleukin-1 induces early and late patterns of c-fos mRNA expression in brain. Journal of Neuroscience, 14, 4951-4964. Brebner, K., Hayley, S., Lacosta, S., Zacharko, R., Merali, Z., & Anisman, H. (1998). Synergistic actions of interleukin (IL)-ip, IL-6, and TNFa: Central neurochemical, neuroendocrine and behavioral alterations. Society for Neurosciences Abstract, in press. Bret-Dibat, J. L., Bluthe, R. M., Kent, S., Kelley K. W., & Dantzer, R. (1995). Lipopolysaccharide and interleukin-1 depress food-motivated behavior in mice by a vagal-mediated mechanism. Brain, Behavior, and Immunity, 9, 242-246. Buttini, M. & Boddeke, H. (1995). Peripheral lipopolysaccharide stimulation induces interleukin-1|3 messenger RNA in rat brain microglial cells. Neuroscience, 65, 523-530. Capuron, L., Ravaud, A., Radat, P., Dantzer, R., & Goodall, G. (1998). Effects of interleukin-2 and alpha

228

H. Anisman and Z. Merali

interferon cytokine immunotherapy on the mood and cognitive performance of cancer patients. Neuroimmunomodulation, 5, 9. Caraceni, A., Martini, C , Belli, F., Mascheroni, L., Rivoltini, L., Arienti, E, & Cascinelli, N. (1992). Neuropsychological and neurophysiological assessment of the central effects of interleukin-2 administration, European Journal of Cancer, 29A, 1266-1269. Connor, T. J. & Leonard, B. E. (1998). Depression, stress and immunological activation: The role of cytokines in depressive disorders. Life Sciences, 62, 583-606. Connor, T., Song, C , Leonard, B. E., Merali, Z., & Anisman, H. (1998). An assessment of the effects of central interleukin-ip, -2, -6 and tumor necrosis factor-a administration on some behavioural, neurochemical, endocrine, and immune parameters in the rat. Neuroscience, 84,923-933. Crnic, L. S. (1991). Behavioral consequences of viral infection. In R. Ader, D. L. Felten, & N. Cohen (Eds.), Psychoneuroimmunology (pp. 749-770). San Diego: Academic Press. Dantzer, R., Bluthe, R. M., Aubert, A., Goodall, G, Bret-Dibat, J-L., Kent, S., Goujon, E., Laye, S., Parnet, P., & Kelley, K. W. (1996). Cytokine actions on behavior. In N. J. Rothwell (Ed.), Cytokines and the Nervous System, (pp. 117-144). London: Landes. Day, H. E. & Akil, H. (1996). Differential pattern of c-fos mRNA in the rat brain following central and systemic administration of interleukin-1-beta: implications for mechanism of action. Neuroendocrinology,63,207-218. Denicoff, K. D., Rubinow, D. R., Papa, M. Z., Simpson, L., Seipp, L. A., Lotze, M. T., Chang, A. E., Rosenstein, D., & Rosenberg, S. A. (1987). The neuropsychiatric effects of treatment with interleukin-2 and lymphokine-activated killer cells. Annals of Internal Medicine, 107, 293-300. De Sarro, G., Gareri, P., Sinopoli, V. A., David, E., & Rotiroti, D. (1997). Comparative, behavioural, and electrocortical effects of tunor necrosis factor-alpha and interleukin-1 microinjected into the locus coeruleus of rat. Life Sciences, 60, 555-564. De Simoni, M. G , Del Bo, R., De Luigi, A., Simard, S., & Forloni, G. (1995). Central endotoxin induces different patterns of interleukin (IL)-lb and IL-6 messenger robonucleic acid expression and IL-6 secretion in the brain and periphery. Endocrinology, 136, 897-902. Dunn, A. J. (1990). Interleukin-1 as a stimulator of hormone secretion. Progress in NeuroEndocrinlmmunology, 3, 26-34. Dunn, A. J. (1995). Interactions between the nervous system and the immune system. In F. E. Bloom & D. J. Kupfer (Eds.), Psychopharmacology: The Fourth Generation of Progress (pp. 719-731). New York: Raven Press. Dunn, A. J. & Welch, J. (1991). Stress- and endotoxin-induced increases in brain tryptophan and serotonin metabolism depend on sympathetic nervous system activity.ioMrna/ of Neurochemistry, 57,1615-1622. Dunn, A. J., Powell, M. L., Meitin, C , & Small, P. A. (1989). Virus infection as a stressor: Influenza virus elevates plasma concentrations of corticosterone, and brain concentrations of MHPG and tryptophan. Physiology and Behavior, 45, 591-594. Elmquist, J. K., Scammell, T. E., & Saper, C. B. (1997). Mechanisms of CNS response to systemic immune challenge: the febrile response. Trends in Neurosciences, 20, 565-570. Ericsson, A., Kovacs, K. J., & Sawchenko, P. E. (1994). A functional anatomical analysis of central pathways subserving the effects of interleukin-1 on stress-related neuroendocrine neurons. Journal of Neuroscience, 14, 89-913. Ericcson, A., Ek, M., Wahlstrom, I., Kovacs, K., Liu, C-L., Hart, R., & Sawchenko, P. E. (1996). Pathways and mechanisms for interleukin-1 mediated regulation of the hypothalamo-pituitary-adrenal axis. In R. McCarty, G. Aguilera, E. L. Sabban, & R. Kvetnancky (Eds.), Stress: molecular genetic and neurobiological advances (pp. 101-120). New York: Gordon & Breach. Ericsson, A., Arias, C , & Sawchenko, P. E. (1997). Evidence for an intermedullary prostaglandin-dependent mechanism in the activation of stres-related neuroendocrine circuitry by intravenous interleukin-1. Journal of Neuroscience, 17, 7166-7179. Fibiger, H. C. & Phillips, A. G. (1988). Mesocorticolimbic dopamine systems and reward. In P. Kalivas & C. Nemeroff (Eds.), The Mesocorticolimbic System, (pp. 206-215). New York: New York Academy of Science. Gabellec, M-M., Griffais, R., Fillion, G , & Haour, F. (1996). Interleukin-1 receptors type I and type II in the mouse brain: kinetics of mRNA expressions after peripheral administration of bacterial lipopolysaccharide. Journal of Neuroimmunology, 66, 65-70. Gaillard, R. C. (1995). Immunoendocrine interactions at the hypothalamo-pituitary level. Annales D'Endocrinologie, 56, 561-566. Gold, P W., Chrousos, G. P, Kellner, C , Post, R. M., Roy, A., Augerinas, P, Schulte, H., Oldgield, E. H., & Loriaux, D. L. (1984). Psychiatric implications of basic and clinical studies with corticotropin-releasinp factnr. Ameriran Jnurnni nf Psvrhintrv 141 M9-f01

Anhedonic and Anxiogenic Effects of Cytokine Exposure

229

Gray, T. S. (1991). Amygdala: Role in autonomic, and neuroendocrine responses to stress. In J. A. McCubbin, P. G. Kaufmann, & C. B. Nemeroff (Eds.), Stress, Neuropeptides, and Systemic Disease (pp. 37-53). New York: Academic Press. Griffiths, J., Ravindran, A. V., Merali, Z., & Anisman, H. (1996). Immune and behavioral correlates of typical and atypical depression. Society for Neuroscieces Abstract, 22, 1350. Gutierrez, E. G., Banks, W. A., & Kastin, A. J. (1993). Murine tumor necrosis factor alpha is transported from blood to brain in the mouse. Journal of Neuroimmunology, 47, 169-176. Hanisch, U-W. & Quirion, R. (1996). Interleukin-2 as a neuroregulatory cytokine. Brain Research Reviews, 21, 246-284. Hanisch, U-W, Seto, D., & Quirion, R. (1993). Modulation of hippocampal acetylcholine release: a potent central action of interleukin-2. Journal of Neuroscience, 13, 3368-3374. Hanisch, U-W, Rowe, W., Sharma, S., Meaney, M. J., & Quirion, R. (1994). Hypothalamicpituitary-adrenal activity during chronic central administration of interleukin-2. Endocrinology, 135, 2465-2472. Hanisch, U-W, Neuhaus, J., Rowe, W, van Rossum, D., Moller, T., Kettenmann, H., & Quirion, R. (1997). Neurotoxic consequences of central long-term administration of interleukin-2 in rats. Neuroscience, 79,799-818. Haour, E, Marquette, C , Ban, E., Crumeyrolle-Arias, M., Rostene, W.,Tsiang, H. & Pillion, G. (1995). Receptors for interleukin-1 in the central nervous system and neuroendocrine systems. Annates d'Endocrinologie,56, 173-179. Harbuz, M. S. & Lightman, S. L. (1992). Stress and the HPA axis: Acute, chronic and immunological activation. Journal of Endocrinology, 134, 327-339. Hayley, S. Brebner, K., Staines, W, Merali, Z., & Anisman, H. (1998). Sensitization effects of tumor necrosis factor-a: behavioral, neuroendocrine and neurochemical alterations. Society for Neurosciences Abstract, in press. Hebb, A. L. O., Zacharko, R. M., & Anisman, H. (1998). Self-stimulation from the mesencephalon following intraventricular interleukin-2 administration. Brain Research Bulletin, 45, 549-556. Heilig, M., Koob, G. E, Ekman, R., & Britton, K. T. (1994). Corticotropin releasing factor and neuropeptide Y: role in emotional integration. Trends in Neurosciences, 17, 80-84. Hellerstein, M. K., Meydani, S. N , Meydani, M., Wu, K., & Dinarello, C. A. (1989). Interleukin-1-induced anorexia in the rat. Influence of prostaglandins. Journal of Clinical Investigation, 84, 228-235. Herman, J. P. & Cullinan, W. E. (1997). Neurocircuitry of stress: Central control of hypothalamo-pituitaryadrenocortical axis. Trends in Neurosciences, 20,78-84. Holden, R. J. & Pakula, I. S. (1996). The role of tumor necrosis factor-alpha in the pathogenesis of anorexia and bulimia nervosa, cancer cachexia and obesity. Medical Hypotheses, 47,423-438. Holsboer, F. (1995). Neuroendocrinology of Mood Disorders. In F. E. Bloom & Kupfer, D. J. (Eds.), Psychopharmacology: The Fourth Generation of Progress (pp. 957-969). New York: Raven Press. Hopkins, S. J., Rothwell, N. J. (1995). Cytokines and the nervous system. Trends in Neurosciences, 18, 83-88. Ilyin, S. E. & Plata-Salaman, C. R. (1996). In vivo regulation of the IL-1 beta system (hgand, receptors I and II, receptor accessory protein, and receptor antagonist) and TNF-alpha mRNAs in specific brain regions. Biochemical and Biophysical Research Communications, 227, 861-867. Imura, H., Fukata, J-I., & Mori, T. (1991). Cytokines and endocrine function: an interaction between the immune and neuroendocrine systems. Clinical Endocrinology, 35,107-115. Kalivas, P. W. & Duffy, P. (1995). Selective activation of dopamine transmission in the shell of the nucleus accumbens by stress. Brain Research, 675, 325-328. Katsumata, S. & Satoh, M. (1993). In situ hybridization study of interleukin-1 (3 mRNA induced by kainic acid in the rat brain. Molecular Brain Research, 20, 153-161. Kent, S., Bluthe, R-M., Kelley, K. W., & Dantzer, R. (1992). Sickness behavior as a new target for drug development. Trends in Pharmacological Sciences, 13, 24-28. Kent, S., Rodriguez, E, Kelley, K. W., & Dantzer, R. (1994). Reduction in food and water intake induced by microinjection of interleukin-1 beta in the ventromedial hypothalamus of the rat. Physiology and Behavior, 56, 1031-1036. Kent, S., Bret-Dibat, J. L., Kelley, K. W, & Dantzer, R. (1996). Mechanisms of sickness-induced decreases in food-motivated behavior. Neuroscience and Biobehavioral Reviews, 20, 171-175. Kinouchi, K., Brown, G., Pasternak, G., & Donner, D. B. (1991). Identification and characterization of receptors for tumor necrosis factor-a in the brain. Biochemical and Biophysical Research Communications, 181, 1532-1538. Kluger, M. J. (1991). Fever: Role of endogenous pyrogens and cryogens. Physiologal Reviews, 71, 93-127. Kokkinidis, L. & Anisman, H. (1980). Amphetamine models of paranoid schizophrenia: An overview and

230

H. Anisman and Z. Merali

Lacosta, S., Kulczycki, J., Merali, Z., Anisman, H. (1996). Behavioral, endocrine and central neurochemical alterations induced by interleukin-1. Society for Neurosciences Abstract, 22, 2059. Lacosta, S., MeraH, Z., & Anisman, H. (1998a). Behavioral and neurochemical consequences of acute and chronic interleukin-2 administration. Society for Neurosciences Abstract, in press, (a) Lacosta, S., Merah, Z., & Anisman, H. (1998b). Influence of interleukin-1 on exploratory behaviors, plasma ACTH and Cortisol, and central biogenic amines in mice. Psychopharmacology, in press, (b) Lapchak, P. A. (1992). A role for interleukin-2 in the regulation of striatal dopaminergic function. Neuroreport, 3, 165-168. Laye, S., Bluthe, R. M., Kent, S., Combe, C , Medina, C , Parnet, R, Kelly, K., & Dantzer, R. (1995). Subdiaphragmatic vagotomy blocks induction of IL-lbeta mRNA in mice brain in response to peripheral LPS. American Journal of Physiology, 37, R1327-R1331. Lee, Y. & Davis, M. (1997). Role of the hippocampus, bed nucleus of the stria terminalis, and the amygdala in the excitatory effect of corticotropin-releasing hormone on the acoustic startle reflex. Journal of Neuroscience, 17, 6434-7446. Linthorst, A. C. E., Flachskamm, C , MuUer-Preuss, P Holsboer, E, & Reul, J. M. H. M. (1995). Effect of bacterial endotoxin and interleukin-1 b on hippocampal serotonergic neurotransmision, behavioral activity, and free corticosterone levels: An in vivo microdialysis study. Journal of Neuroscience, 15, 2920-2934. Liu, L., Kita, T., Tanaka, N., & Kinoshita, Y. (1996). The expression of tumor necrosis factor in the hypothalamus after treatment with lipopolysaccharide. International Journal of Experimental Pathology, 77, 37-44. Loddick, S. A., MacKenzie, A., & Rothwell, N. J. (1996). An ICE inhibitor, z-VAD-DCB attenuates ischaemic brain daage in the rat. Neuroreport, 7,1465-1468. Loddick, S. A., Wong, M. L., Bongiorno, R B., Gold, P W., Licinio, J., & Rothwell, N. J. (1997). Endogenous interleukin-1 receptor antagonist is neuroprotective. Biochemical and Biophysical Research Communications, 234, 211-215. Maes, M. (1995). Evidence for an immune response in major depression: A review and hypothesis. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 19, 11-38. Maes, M., Bosmans, E., Suy, E., Vandervorst, C , DeJonckheere, C , Minner, B., & Raus, J. (1991). Depressionrelated disturbances in mitogen-induced lymphocyte responses, interleukin-1 (3, and soluble interleukin-2-receptor production. Acta Psychiatrica Scandinavica, 84, 379-386. Maes, M., Lambrechts, J., Bosmans, E., Jacobs, J., Suy, E., Vandervorst, C , DeJonckheer, C , Minner, B., & Raus, J. (1992). Evidence for a systemic immune activation during depression: Results of leukocyte enumeration by flow cytometry in conjunction with monoclonal antibody staining. Psychological Medicine, 22, 45-53. Maes, M., Stevens, W. J., Peeters, D., DeClerck, L. S., Bridts, C. H., Peeters, D., Schotte, C , & Cosyns, R (1993). Significantly increased expression of T-cell activation markers (interleukin-2 and HLA-DR) in depression: further evidence for an inflammatory process during that illness. Progress in NeuroPsychopharmacology and Biological Psychiatry, 17, 241-255. Maes, M., Meltzer, H. Y, Bosmans, E., Bergmans, R., Vandoolaeghe, E., Ranjan, R., & Desnyder, R. (1995). Increased plasma concentrations of interleukin-6, soluble interleukin-6, soluble interleukin-2, and transferrin receptor in major depression. Journal of Affective Disorders, 34, 301-309. Maier, S. E & Watkins, L. R. (1995). Intracerebroventricular interleukin-1 receptor antagonist blocks the enhancement of fear conditioning and interference with escape produced by inescapable shock. Brain Research, 695, 279-286. Maier, S. E & Watkins, L. R. (1998). Cytokines for psychologists: Implications of bidirectional immune-tobrain communication for understanding behavior, mood, and cognition. Psychological Review, 105, 83-107. Masana, M. I., Heyes, M. P., & Mefford, I. N. (1990). Indomethacin prevents increased catecholamine turnover in rat brain following systemic endotoxin challenge. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 74,609-621. Mclntyre, D. C. (1980). Amygdala kindling in rats: facihtation after local amygdala norepinephrine depletion with 6-hydroxydopamine. Experimental Neurology, 69, 395-407. Mefford, I. N. & Heyes, M. P. (1990). Increased biogenic amine release in mouse hypothalamus following immunological challenge: antagonism by indomethacin. Journal of Neuroimmunology, 27, 55-61. Mefford, I. N., Masters, C. E, Heyes, M. P., & Eskay, R. L. (1991). Cytokine-induced activation of the neuroendocrine stress axis persists in endotoxin-tolerant mice. Brain Research, 557, 327-330. Merali, Z., Lacosta, S., & Anisman, H. (1997). Effects of interleukin-ip and mild stress on alterations of central monoamines: A regional microdialysis study. Brain Research, 761, 225-235.

Anhedonic and Anxiogenic Effects of Cytokine Exposure

231

Meyers, C. A. & Valentine, A. D. (1995). Neurological and psychiatric adverse effects of immunological therapy, CNS Drugs, 3, 56-68. Minami, M., Kuraishi, Y., Yamaguchi, T., Nakai, S., Hirai, Y., & Satoh, M. (1990). Convulsants induce interleukin-lb messenger RNA in rat brain. Biochemical and Biophysical Research Communications, 171, 832-837. Minami, M., Kuraishi, Y., Yamaguchi, T., Nakai, S., Hirai, Y, & Satoh, M. (1991). Immobilization stress induces interleukin-lp mRNA in rat hypothalamus. Neuroscience Letters, 123, 254-256. Mohankumar, P. S. & Quadri, S. K. (1993). Systemic administration of interleukin-1 stimulates norepinephrine release in the paraventricular nucleus. Life Sciences, 52, 1961-1967. Mohankumar, P. S.,Thyagara3an, S., & Quadri, S. K. (1991). Interleukin-1 stimulates the release of dopamine and dihydroxyphenylacetic acid from hypothalamus in vivo. Life Sciences, 48, 925-930. Mohankumar, P. S., Thyagarajan, S., & Quadri, S. K. (1993). Interleukin-ip increases 5-hydroxyindoleacetic acid release in the hypothalamus in vivo. Brain Research Bulletin, 31, 745-748. Mohapel, P. & Mclntyre, D. C. (1998). Amygdala kindling-resistant (SLOW) or -prone (FAST) rat strains show differential fear responses. Behavioral Neuro sciences, in press. Moldofsky, H. (1995). Sleep and the immune system. International Journal of Immunopharmacology, 17, 649-665. Muaramami, N., Fukata, J., Tsukada, T., Kobayashi, H., Ebisui O. Segawa, H., Muro, S., Imura, H., & Nakao, K. (1993). Bacterial lipopolysaccharide expression of interleukin-6 messenger ribonucleic acid in the rat hypothalamus, pituitary, adrenal gland, and spleen. Endocrinology, 133, 2574-2578. Mullar, N. & Ackenheil, M. (1998). Psychoneuroimmunology and the cytokine action in the CNS: Implications for psychiatric disorders. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 22, 1-33. Nassberger, L. & Traskman-Bendz, L. (1993). Increased soluble interleukin-2 receptor concentrations in suicide attempters. Acta Psychiatrica Scandinavica, 88, 48-52. Nemeroff, C. B. (1996). The coricotropin-releasing factor (CRF) hypothsesis of depression: New findings and new directions. Molecular Psychiatry, 1, 336-342. Nguyen, K. T, Deak, T, Qwens, S. M., Kohno, T, Fleshner, M., Watkins, L. R., & Maier, S. F (1998). Exposure to acute stress induces brain interleukin-1 P protein in the rat. Journal Neuroscience, 18, 2239-2246. Petitto, J. M., McCarthy, D. B., Rinker, C. M., Huang, Z., & Getty, T (1997). Modulation of behavioral and neurochemical measures of forebrain dopamine function in mice by species-specific interleukin-2. Journal of Neuroimmunology, 73, 183-190. Pitossi, F., del Rey, A., Kabiersch, A., & Besedovsky, H. (1997). Induction of cytokine transcripts in the CNS and pituitary following peripheral administration of endotoxin to mice. Journal of Neuroscience Research, 48, 287-298. Plata-Salaman, C. R. (1989). Immunomodulators and feeding regulation: A humoral link between the immune and nervous systems. Brain, Behavior and Immunity, 3,193-213. Plata-Salaman, C. R. (1994). Meal patterns in response to the intracerebroventricular administration of interleukin-1 in rats. Physiology and Behavior, 55, 727-733. Plata-Salaman, C. R. (1998). Cytokines and anorexia: A brief overview. Seminars in Oncology, 25, 64-72. Quan, N., Sundar, S. K., & Weiss. J. M. (1994). Induction of interleukin-1 in various brain regions after peripheral and central injections of lipopolysaccharide. Journal of Neuroimmunology, 49, 125-134. Raber, J., Koob, G. F , & Bloom, F. E. (1995). Interleukin-2 (IL-2) induces corticotropin-releasing factor (CRF) release from the amygdala and involves a nitric oxide-mediated signaling: Comparison with the hypothalamic response. Journal of Pharmacology and Experimental Therapeutics, 272, 815-824. Rajora, N., Boccoli, G., Burns, D., Sharma, S., Catania, A. P., & Lipton, J. M. (1997). a-MSH modulates local and circulating tumor necrosis factor-a in experimental brain inflammation. Journal of Neuroscience, 15, 2181-2186. Richardson, N. R. & Roberts, D. C. S. (1996). Progressive ratio schedules in drug self-administration studies in rats: a method to evaluyate reinforcing efficacy. Journal of Neuroscience Methods, 66,1-11. Rivest, S. (1995). Molecular mechanisms and neural pathways mediating the influence of interleukin-1 on the activity of neuroendocrine CRF motoneurons in the rat. International Journal of Developmental Neuroscience, 13, 135-146. Rivest, S. & Rivier, C. (1994). Stress and interleukin-1 beta-induced activation of c-Fos, Ngfi-B, and CRF gene expression in the hypothalamic PVN—Comparison between Sprague-Dawley, Fisher-344, and Lewis rats. Journal of Neuroendocrinology, 6,101-117. Rivest, S. & Laflamme, N. (1995). Neuronal activity and neuropeptide gene transcription in the brains of immune-challenged rats. Journal of Neuroendocrinology, 7, 501-525.

232

H. Anisman and Z. Merali

Rivier, C. (1993). Effect of peripheral and central cytokines on the hypothalamic-pituitary-adrenal axis of the rat. Annals of the New York Academy of Sciences, 697, 97-105. Rothwell, N. J. & Hopkins, S. J. (1995). Cytokines and the nervous system II: Actions and mechanisms of action. Trends in Neurosciences, 18,130-136. Rothwell, N. J., Luheshi, G., & Toulmond, S. (1996). Cytokines and their receptors in the central nervous system: physiology, pharmacology, and pathology. Pharmacology and Therapeutics, 69, 85-95. Rothwell, N. J., Loddick, S. A., & Stroemer, P. (1997). Interleukins and cerbral ischemia. International Review of Neurobiology, 40,291-298. Sato, S., Reiner, S. L., Jensen, M. A., & Roos, R. P. (1997). Central nervous system mRNA expression following Tlieiler's murine encephalomyelitis virus infection. Journal of Neuroimmunology, 76,213-223. Sawchenko, R E., Brown, E. R., Chan, R. K. W., Ericsson, A., Li, H.-Y, Roland, B. L., & Kovacs, K. J. (1996). The paraventricular nucleus of the hypothalamus and the functional neuroanatomy of visceromotor responses to stress. Progress in Brain Research, 107, 201-222. Schmidt, E. D., Janszen, A. W. J. W., Wouterlood, F. G., & Tilders, R J. H. (1995). Interleukin-1 induced longlasting changes in hypothalamic corticotropin-releasing hormone (CRH) neurons and hyperresponsiveness of the hypothalamic-pituitary-adrenal axis. Journal of Neuroscience, 15,7417-7426. Schobitz, B., De Kloet, E. R., Sutanto, W., & Holsboer, F. (1993). Cellular localization of interleukin-6 mRNA and interleukin 6 mRNA in rat brain. European Journal of Neuroscience, 5,1426-1435. Seto, D., Karr, S., & Quirion, R. (1997). Evidence for direct and indirect mechanisms in the potent modulatory action of interleukin-2 on the release of acetylcholine in rat hippocampal slices. British Journal of Pharmacology, 120,1151-1157. Shintani, E, Kanba, S., Nakaki, T., Nibuya, M., Kinoshita, N., Suzuki, E., Yagi, G., Kato, R., & Asai M. (1993). Interleukin-1 P augments release of norepinephrine, dopamine and serotonin in the rat anterior hypothalamus. Journal of Neuroscience, 13, 3574-3581. Shintani, R, Nakaki, T , Kanba, S., Sato, K., Kato, R., & Asai, M. (1995a). Role of interleukin-1 in stress responses. Molecular Neurobiology, 10,47-71. a Shintani, E, Nakaki, T., Kanba, S., Sato, K., Yagi, G., Shiozawa, M., Aiso, S., Kato, R., & Asai, M. (1995b). Involvement of interleukin-1 in immibilization stress-induced increase in plasma adrenocorticotropic hormone and in release of hypothalamic monoamines in the rat. Journal of Neuroscience, 15, 1961-1970.b Sluzewska, A., Rybakowski, J. K., Laciak, M., Mackiewicz, A., Sobieska, M., & Wiktorowicz, K. (1995). Interleukin-6 serum levels in depressed patients before and after treatment with fluoxetine. Annals of the New York Academy of Sciences, 762, 474-476. Smagin, G. N,, Swiergiel, A. H., & Dunn, A. J. (1996). Peripheral administration of interleukin-1 increases extracellular concentrations of norepinephrine in rat hypothalamus: Comparison with plasma corticosterone. Psychoneuroendocrinology, 21, 83-93. Smith, R. S. (1991). The macrophage theory of depression. Medical Hypotheses, 35,298-306. Song, C , Dinan, T , & Leonard, B. E. (1994). Changes in immunoglobulin, complement and acute phase protein levels in depressed patients and normal controls. Journal of Affective Disorders, 30,283-288. Song, C , Merali, Z., & Anisman, H. (1998). Variations of nucleus accumbens dopamine and serotonin following systemic interleukin-1, interleukin-2 or interleukin-6 treatment. Neuroscience, in press. Sonti, G., Ilyin, S. E., & Plata-Salaman, C. R. (1996). Anorexia induced by cytokine interactions at pathophysiological concentrations. American Journal of Physiology, 270, R1394-R1402. Strijbos, P. J. & Rothwell, N. J. (1995). Interleukin-1 beta attenuates excitatory amino acid-induced neurodegeneration in vitro: Involvement of nerve growth factor. Journal of Neuroscience, 15, 3468-3478. Suzuki, E., Shintani, E, Kanba, S., Asai, M., & Nakaki, T. (1997). Immobilization stress increases mRNA levels of interleukin-1 receptor antagonist in various brain regions. Cellular Molecular Neurobiology, 17, 557-562. Takao, T., Tojo, C , Nishioka, T, Kurokawa, H., Takemura, T, Hashimoto, K., & De Souza, E. B. (1994). Reciprocal modulation of corticotropin-releasing factor and interleukin-1 receptors following etherlaparotomy stress in the mouse. Brain Research, 660,110-11 A. Taupin, v.,Toulmond, S., Serrano, A., Benavides, J., & Zavala, F. (1993). Increase in IL-6, IL-1, and TNF levels in rat brain following traumatic lesion. Journal of Neuroimmunology, 42,177-186. Terao, A., Oikawa, M., & Saito, M. (1993). Cytokine-induced change in hypothalamic norepinephrine turnover: involvement of corticotropin-releasing hormone and prostaglandins. Brain Research, 622, 257-261. Tilders, E J. H., Schmidt, E. D., & De Goeij, D. C. E. (1993). Phenotypic plasticity of CRF neurons during stress. Annals of the New York Academy of Sciences, 697, 39-52. van Dam, A-M., Brouns, M., Louisse, S., & Berkenbosch, F. (1992). Appearance of interleukin-1 in

Anhedonic and Anxiogenic Effects of Cytokine Exposure

233

macrophages and in ramified microglia in the brain of endotoxin-treated rats: a pathway for the induction of non-specific symptoms of sickness? Brain Research, 588, 291-296. Watkins, L. R., Maier, S. E, & Goehler, L. E. (1995). Cytokine-to-brain communication: A review & analysis of alternative mechanisms. Life Sciences, 11,1011-1026. Whishaw, I. Q. (1985). Eormation of a place learning-set by the rat: a new paradigm for neurobehavioral studies. Physiology & Behavior, 35,139-143. Whitnall, M. H., Perlstein, R. S., Mougey, E. H., & Neta, R. (1992). Effects of interleukin-1 on the stress responsive and nonresponsive subtypes of corticotropin-releasing hormone neurosecretory axons. Endocrinology, 131, 37-44. Wise, R. A. (1985). The anhedonia hypothesis: Mark III. Behavioural Brain Science, 8,178-186. Yabuuchi, K., Sharma, S., Catania, A. P., & Lipton, M. (1996). Induction of interleukin-1 p mRNA in the hypothalamus following subcutaneous injections of formalin into the rat hind paws. Neuroscience Letters, 207,109-112. Yang, Z-J., Koseki, M., Meguid, M. M., Gleason, J. R., & Debonis, D. (1994). Synergistic effect of rhTNF-a and rhIL-la in inducing anorexia in rats. American Journal of Physiology, 36, R1056-R1064. Yirmiya, R. (1996). Endotoxin produces a depressive-like episode in rats. Brain Research, 711,163-174. Zacharko, R. M. & Anisman, H. (1991). Stressor-provoked alterations of intracranial self-stimulation in the mesocorticolimbic system: An animal model of depression. In P. Willner & J. Scheel-Kruger (Eds.), The Mesolimbic Dopamine System: From Motivation to Action (pp. 411-442). New York: Wiley. Zalcman, S., Green-Johnson, J. M., Murray, L., Nance, D. M., Dyck, D., Anisman, H., & Greenberg, A. (1994). Cytokine-specific central monoamine alterations induced by interleukin (IL)-l, IL-2, and IL-6. Brain Research, 643,40^9. Zhou, D. H., Shanks, N., Riechman, S. E., Liang, R. M., Kusnecov, A. W, & Rabin, B. S. (1996). Interleukin-6 modulates interleukin-1 and stress-induced activation of the hypothalamic-pituitary-adrenal axis in male rats. Neuroendocrinology, 63,227-236.

13

STRESS, LEARNED HELPLESSNESS, AND BRAIN INTERLEUKIN-lp Steven F. Maier, Kien T. Nguyen, Terrence Deak, Erin D. Milligan, and Linda R. Watkins University of Colorado, Boulder, Colorado

1. I N T R O D U C T I O N Similarities between the behavioral, endocrine, and neurochemical sequelae of exposure to psychological stressors and agents that activate peripheral immune cells (viruses, lipopolysaccharide, arthritis inducing agents, etc.) have often been noted (Dunn, 1993). Central cytokines such as IL-1P play an important role in mediating many of these responses that follow immune activation, and the purpose of the present chapter is to summarize work that is directed at determining whether central cytokines also play a role in mediating the consequences of exposure to stressors. In addition, there will be a focus on whether central cytokines are involved in the production of a particular stress-related phenomenon, learned helplessness.

1.1. Brain Cytokines and Sickness Infectious and inflammatory agents produce a characteristic pattern of behavioral, physiological, endocrine, and neurochemical alterations that we will collectively call "sickness" (Dantzer, Bluthe, Kent, & Goodall, 1993; Dunn, 1995; Hart, 1988). These sequelae of infection, tissue injury, and inflammation include fever, reduced activity and exploration, reduced social interaction, reduced food and water intake, reduced sexual behavior, anhedonia, increased sleep, shifts in pain sensitivity/reactivity, shifts in liver metabolism away from the production of carrier proteins and towards the production of acute phase proteins (the acute phase response), leukocytosis, increases in sympathetic nervous system activity, increases in hypothalamo-pituitary-adrenal activity, and activation of monoaminergic neurons in discrete regions of the brain. Pro-inflammatory cytokines such as interleukin-1 (IL-1) a and p, IL-6, tumor necrosis factor (TNF), and IL-8, released by activated immune cells such as macrophages, play a key role in at least 3 different levels of the host response to Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

235

236

S. F. Maier et al

infection or injury. First, they are involved in initiating the cascade of events that occurs following entry of a pathogen or inflammatory agent into the body and have direct actions on peripheral organs such as the liver, here initiating the production of acute phase proteins. However, many sickness responses are mediated by the central nervous system, rather than merely by immune-derived products such as cytokines acting at the periphery (Maier & Watkins, 1998). Secondly, cytokines are key signals to the central nervous system that infection or injury is present (Dinarello & Thompson, 1991; Ericsson, Kovacs, & Sawchenko, 1994), thereby initiating the neural cascade involved in sickness. And thirdly, there is evidence that cytokines, most notably IL-ip, are also important in the neural processes themselves that mediate sickness responses. There is dispute concerning whether there is constitutive expression of IL-ip in brain, although there is increasing evidence that under basal conditions there is IL-1(3 bioactivity (Quan, Zhang, Emery, Bosnall, & Weiss, 1996), immunoreactivity (Hagan, Poole, & Bristow, 1993), protein (Nguyen, Deak, Owens, Kohno, Fleshner, Watkins, & Maier, 1998), and even basal release (Tringali, Mirtella, Mancuso, Guerriero, Preziosi, & Navarra, 1996). However, it is clear that the administration of immune activating agents such as lipopolysacharide (LPS) which release IL-1 and the administration of IL-1 itself are followed by regionally specific increases in brain IL-ip mRNA (Ban et al., 1992), immunoreactivity (Van Dam, Brouns, Louisse, & Berkenbosch, 1992), bioactivity (Quan et al, 1994), and protein levels (Nguyen et al., 1998). The cellular source(s) of IL-1 certainly includes glial cells, but probably neurons as well (Tringali et al., 1996). This is not a critical issue for the present paper since IL-1 released from either glia or neurons would be available to bind to IL-1 receptors on neurons, thereby altering neuronal activity. The fact that peripheral immune activation and cytokine administration increases brain IL-ip does not prove that brain IL-ip plays a role in the neural mediation of sickness. This is, however, suggested by a) the blockade or reduction in some sickness responses to peripheral LPS and IL-1 produced by blockade of IL-1 receptors in brain (Kakucksa, Qi, Clark, & Lechan, 1993; Kent, Bluthe, Kelley, & Dantzer, 1992; Klir, McClellan, & Kluger, 1994,Takahashi, Kapas, Fang, Seyer, Wang, & Krueger, 1996), and b) the induction of sickness responses produced by central injection of IL-1 (Bluthe et al., 1996; Kent et al., 1994; Kluger, 1991; Krueger et al., 1984; Plata-Salaman, Oomura, 8L Kai, 1988; Sundar, Cierpial, Kilts, Ritchie, & Weiss, 1990). Interestingly, central injection of IL-1 produces not only sickness responses that are directly mediated in the brain such as fever, but also the production of acute phase proteins by the liver and leukocytosis (Rothwell, 1989), a process more often thought to be mediated in the periphery without CNS involvement.

2. LEARNED HELPLESSNESS The similarity in the consequences of infection/inflammation and stressor exposure noted above leads naturally to an inquiry into whether IL-1 in brain is involved in "stress". There is very little information in the published literature. However, the intracerebroventricular (ICV) administration of IL-1 has been reported to produce some consequences similar to those produced by stressors, including increases in plasma ACTH and corticosteroids (Weiss, Sundar, Cierpial, & Ritchie, 1991), increased activity in monoaminergic neurons (Linthorst et al., 1994), reduced motor activity (Linthorst, Flachskamm, Holsboer, & Reul, 1994), and reduced contact with novel

Stress, Learned Helplessness, and Brain Interleukin-ip

237

Stimuli (Spadaro & Dunn, 1990). Furthermore, Minami, Kuraishi, Yagaguchi, Nakai, Hirai, & Satoh. (1991) reported that immobilization increased mRNA for IL-ip in hypothalamus, Shintani, Nakaki, Kanba, Sato., Yagi, Shiozawa, Aiso, Kato, & Asai (1995) found that the central administration of interelukin-1 receptor antagonist (ILIra) blunted neurochemical and endocrine responses to restraint, and Saperstein, Brand, Audhya, Nobriski, Hutchinson, Rosenzwieg, & Hollander (1992) found ICV IL-lra to blunt the immunosuppressive effects of footshock. However, the generality of these results are not known. A number of psychological variables modulate the behavioral, neurochemical, and physiological impact of stressors. It is unknown whether these variables might modulate brain IL-1, and whether brain IL-1 might be a mediator of the influence of these variables. The degree of behavioral control which an organism has over a stressor is among the most potent of these variables (Maier, 1993, for review). Control refers to the organism's ability to alter the onset, termination, duration, intensity, or temporal pattern of an event by behavioral responses or adjustments. Exposure to aversive events over which the organism has no control produces a constellation of behavioral, neurochemical, and physiological changes that do not occur if the organism can control the stressor (Maier, 1993). Effects of stressors that occur only if the stressor is uncontrollable have been called "learned helplessness effects", and the state induced by uncontrollable stressors has been called "learned helplessness" (Maier & Seligman, 1976). It is important to understand that not all behavioral (Woodmansee, Silbert, & Maier, 1993) and physiological (Maier, Ryan, Barksdale, & Kalin, 1986) consequences of stressor exposure are affected by the controllability of the stressor—some are simply a product of exposure to the stressor per se and so are not learned helplessness effects. It is also important to note that the differential impact of uncontrollable and controllable stressors cannot be understood as simply reflecting "more" and "less" stress. It is not possible to mimic the effects of uncontrollable stressors by merely increasing the intensity or duration of controllable stressors. The two types of stressors are very different events and have distinctive consequences.

3. INESCAPABLE SHOCK, SICKNESS, AND THE ACUTE PHASE RESPONSE Most investigators have studied the stressor controllability dimension by comparing escapable (the termination of each shock is controllable) electric shock with equal intensities and durations of inescapable (uncontrollable) electric shock; inescapable shock (IS) is generally regarded as inducing learned helplessness (Maier & sehgman, 1976). IS produces many of the same consequences that have been shown to be produced by other stressors that are similar to sequelae of infection (e.g., subsequent reduced activity and exploration), but does it produce other aspects of sickness that are produced by infection or the central administration of IL-1 that have not ordinarily been thought to be produced by stressors? Fever is perhaps the key physiological adjustment to infection, and it has been shown that even relatively mild stressors such as placement in a novel environment produce increases in core body temperature (Morrow, McClellan, Conn, & Kluger, 1993). Furthermore, this increase in core body temperature has been demonstrated to reflect a true fever in that it is a regulated rise in thermoregulatory setpoint in the brain,

S. F. Maier et al

238

rather than only a hyperthermic response reflecting metabolic changes associated with the stress response itself (Long, Vander, & Kluger, 1990). However, fever has been examined only during or very shortly after a brief stressor exposure, and so whether there would be a prolonged fever such as observed during infection is unknown. Thus we exposed loosely restrained rats to a single session of inescapable tailshock (IS) with parameters typical in our laboratory (100 5sec ISs of 1.0mA in intensity, occuring on the average of once a min) and measured core body temperature using implanted thermisters during IS or control treatment as well as for the periods 1-4 and 20-24 hr after IS. The data are shown in Fig. 1. It is no surprise that core body temperature increased during IS. However, as can be observed, IS produced a long lasting fever. In subsequent work we have found this fever to persist for 4, but not for 7 days. Behavioral changes such as reduced social interaction and physiological responses such as fever are mediated by the CNS. It is therefore not surprising that these would be produced by stressors as well as by infection. However, there are also aspects of sickness that are mediated in the periphery. Leukocytosis and shifts in liver metabolism away from the production of carrier proteins towards the production of acute phase proteins are examples. The "stressor" surgery has been examined for its ability to increase acute phase proteins, but surgery confounds the inevitable introduction of endotoxin, tissue damage, and other direct immune activating agents with

A. FEVER DURING TAILSHOCK 40.0 39,5 39.0 3a.5 39.0 37,5-

36.5

B. FEVER AFTER TAILSHOCK 3B.Q

i >•

•a o

37.0

H-i-r'

e

o o

1

2

3

4

20

21

22

23

24

Hours After Taflahock Figure 1. Core body temperature measured during inescapable tailshock and home cage control (HCC) treatment, as well as 1-4 and 20-24 hr after treatment. BL = Baseline, SK = Shock.

Stress, Learned Helplessness, and Brain Interleukin-lp

239

the psychological nature of stress. Thus we have measured white blood cells, the carrier protein corticosteroid binding globulin (CBG), seromucoids (a measure of acute phase proteins that is predominantly al-acid glycoprotein), and the acute phase protein haptoglobin in blood samples taken 24hr after IS or control treatment. As can be seen in Fig. 2, exposure to IS led to the typical pattern produced by pathogens (Deak, Meriweather, Fleshner, Spencer, Moldawer, Grahn, Watkins, & Maier., 1997). A final aspect that we have studied involves basal levels of plasma corticosteroids. We have shown that the administration of inflammatory agents increases basal corticosteroid levels for a number of days and IS also increases basal corticosteroids measured 24 and 48 hr after IS (Fleshner, Deak, Spencer, Laudenslager, Watkins, & Maier, 1995). It has already been noted that the administration of IL-1 into brain induces the outflow of some product or process from brain that can directly, or through the induction of further products, impact on liver function. The results described above suggest that this is so for IS as well. The cascade of processes initiated by pathogens begins with the macrophage and other immune cell types, and the sickness responses are driven by cytokines and other substances that are produced by these cells during infection. The 2hr IS session produces fever, acute phase protein increases, and other aspects of sickness that persist for a number of days, well beyond the duration of the stressor. A variety of mechanisms that would lead to this persistent response could be envisaged, but one would involve a prolonged activation of macrophages. This would lead to the long-term production of macrophage products that drives sickness during "normal infection". Activated macrophages produce a variety of products. Nitric oxide (NO) has

E

f Seromucoids

Haptoglobin

C8Q

White Blood Celts

Figure 2. Blood levels of seromucoids, corticosteroid binding globulin, haptoglobin, and white blood cells 24hr after inescapable shock (IS) or home cage control (HCC) treatment. Panel A depicts seromucoids. Panel B depicts corticosteroid binding globulin. Panel C depicts haptoglobin, and Panel D depicts white blood cell count.

S. F. Maier et al.

240

48

96

Hours after IS (10.0 X 10'^6 spleen cells)

Figure 3. Nitrite accumulation in spleen cells taken immediately, 48 hr, or 96 hr after inescapable shock (IS) or home cage control (HCC) treatment.

received increasing attention in this regard because of its adaptive role in inhibiting microbial growth. Fig. 3 shows nitrite levels (the stable end-product of NO) measured in culture from stimulated spleen cells taken at various times after IS or control treatment. Clearly, cells taken from rats that had received IS produce enhanced levels of NO ex vivo, and this effect is both delayed, requiring a number of hours after IS to occur, and persistent over days after IS. The active cells producing this NO are likely to be macrophages (Fleshner, Nguyen, Cotter, Watkins, & Maier, 1998). Similar results have been obtained for IL-ip production in culture, although we have not yet completed a full timecourse. IS cannot "directly bind" to macrophages or other peripheral cells as can pathogens, LPS, etc. Therefore, the conclusion is that IS leads to an outflow product from brain that alters macrophage activity. The identity of this product or process is not known. However, again, the similarity between the effects of IL-1 injection into the brain and IS are striking. ICV IL-1 also produces increases in acute phase proteins, white blood cell counts, and basal corticosteroid levels, and reduces levels of CBG.

4. IS AND BRAIN IL-ip As already noted, the ICV administration of IL-1 produces behavioral, endocrine, physiological, and neurochemical sequelae that resemble those induced by stressors, as well as the acute phase responses produced by IS described above (DeSimoni, Seroni, DeLuigi, Manfridi, Mantovani, & Ghezzi, 1990; Morimoto, Sakata, Watanabe, & Murakami, 1989). It is therefore of interest to determine whether IS might induce ILi p in the brain. This issue is complex for a variety of reasons. One is that adrenal corticosteroids (CORT) destabilize IL-ip mRNA (Amano, Lee, & Allison, 1993) and inhibit IL-ip gene transcription (Lee, Tsou., Chan, Thomas, Petrie, Eugui, & Allison, 1988) as well as a number of translational and post-translational processes (Kern, Lamb, Reed, Daniele, & Nowell, 1988) involved in the production of IL-ip protein (see Watkins, Nguyen, & Maier this volume). Because IS increases adrenal CORT levels, we sought to determine whether IS would increase brain IL-ip protein levels in both adrenalectomized and sham surgery subjects. Basal CORT was maintained at normal levels by the addition of CORT to the drinking water. Thus the ADX subjects had normal basal CORT and a normal circadian rhythm of CORT (the rats drink

Stress, Learned Helplessness, and Brain Interleukin-l^

241

much more during the dark part of the cycle), but could not increase CORT in response to IS. IL-ip protein levels were measured by ELISA. Several methodological issues should be noted at the outset. The procedure involves administration of IS or control treatment, sacrifice, dissection of brain into regions, and separation of protein from brain by sonication and centrifugation. It is therefore possible that any or all of the detected IL-ip derives from IL-ip circulating in the cerebral vasculature or bound to IL-1 receptors on the inside of the vascular endothelium. To assess this possibility comparisons between IL-ip levels measured in saline-perfused and non-perfused brains have been compared, and the resulting data do not differ. Data to be described were obtained following perfusion. Second, ELISA uses an antibody (Ab) specifically directed against rat IL-1 (3 to detect IL-1. However, there is always the possibility that the Ab that is used is capable of detecting another protein as well. We have used a number of approaches in an attempt to overcome this potential interpretive difficulty, a) Three different Abs directed against rat IL-ip have been used. Although absolute values have differed between Abs, the direction and magnitude of the changes have been the same for all Abs. b) Specificity of Ab binding was assessed by Western blot analysis. Strong bands appeared at approximately IVkDa, the molecular weight of mature (active) IL-ip. However, faint bands were also detected at roughly 33kDa, the molecular weight of the precursor (inactive) form of IL-1 p. Several laboratories have now reported that Ab directed against 17kDa IL-ip recognizes 33kDa pro-IL-ip with approximately 10-fold less affinity, likely due to a conformational change conferred by amino acid sequences present in pro-, but not active IL-lp (Dinarello, 1992). c) Total protein from the brain sonication supernatants was seperated on a Sephadex G-50 column, and fractions assayed with the ELISA. The signal was restricted to the fraction in the molecular weight range of IL-ip. d) The rat IL-ip Ab was verified for its ability to recognize pure rat IL-ip by immunoprecipitation. Details of these Ab verification procedures can be found in Nguyen et al. (1998). Fig. 4 shows IL-ip protein levels in hypothalamus and hippocampus, while Fig. 5 shows data for the nucleus tractus solitarius and pituitary gland. Samples were taken either immediately, 2hr, 24 hr, or 48 hr after IS or home cage control treatment in adrenalectomized (ADX) and sham surgery subjects.. The general pattern is that IL-lp protein levels are elevated immediately and 2hr after IS in adrenalectomized subjects. Although appearing small in the figure because of the scale used, IL-ip increased by 2-3 fold in the hypothalamus immediately and 2hr after IS in sham subjects. It should be noted that the increase in IL-lp was regionally specific and failed to occur in other regions tested. For example, IL-1 levels did not increase in posterior cortex, in either sham or adrenalectomized subjects. The fact that brain IL-ip protein increases were detectable immediately after the 2hr IS session deserves comment. It is not yet known whether IS increases mRNA for IL-ip, nor at what point during the stress session that this might occur. IL-ip gene transcription and translation have been reported to occur as rapidly as within 1 hr (Abel & Czop, 1992), and mRNA increases have been observed within 15min of stimulation in vitro and in vivo (Enk, Angeloni, Udey, & Katz, 1993). Thus the IL-ip observed could reflect newly synthesized protein. Alternatively, the IL-ip measured could be the result of post-translational processing of preformed IL-lp precursor. As noted above, IL-ip is synthesized as a 33kDa precursor (pro-IL-lp) that is biologically inactive, with mature IL-ip being formed from pro-IL-ip by proteolytic cleavage.

S. F. Maier et al.

242

Tlmepofnt of IL-1 after JS Hypothalamus 1.5 n

2 hr

24 hf

4S)ir

ADX-cort Hippocampus

Immecf.

Figure 4. IL-ip protein levels measured in hypothalamus and hippocampus of sham (SHAM) or adrenalectomized (ADX) rats given either inescapable shock (IS) or home cage control treatment (HCC). Samples were taken either immediately, 2hr, 24 hr, or 48 hr after treatment.

The proteolytic processing is accomplished by IL-1 converting enzyme (ICE), and ICE activity can be regulated in a number of different ways (Keane, Giegel, Lipinski, Callahan, & Shivers, 1995). There is evidence that there is constitutive expression of preformed pro-IL-l(3 in the brain, and in the hypothalamus in particular (Tringali et al. 1996). The active forms of ICE also result from proteolytic cleavage from an ICE precursor, and stimulation by LPS has been shown to produce the rapid processing of ICE, leading to the production of mature IL-ip. It is possible that IS also increases ICE activity in this fashion, thereby leading to the rapid production of mature IL-ip from existing pro-IL-l|3. Interestingly, ATP induces a rapid increase in ICE activity and ILIp processing such that mature IL-ip is released within 7.5 min of ATP administration (Griffiths, Stam, Downs, & Otterness, 1995), and IS is known to increase ATP levels (Minor & Saade. 1997). Furthermore, neurons that utilize ATP as a transmitter project from the A l cell group via the ventral noradrenergic bundle to the hypothalamus (Sperlagh, Sershen, Lojtha, & Vizi, 1998), the region in which IL-1 increases after IS are most readily detectable.

Stress, Learned Helplessness, and Brain Interleukin-l^

243

Timepoint of IL-1 after IS Nucleus Tractus Solitarius

2 hr

24 hr

4a hr

Immed,

2 hr

24 hr

4B hr

ADX-cort

Sham Pituitary

2 hr

24 hr

SHAM

16 hr

Immed.

2hr

24 hr

ADX-cort

Figure 5. IL-ip protein levels measured in nucleus tractus solitarius and pituitary of sham (SHAM) or adrenalectomized (ADX) rats given either inescapable shock (IS) or home cage control treatment (HCC). Samples were taken either immediately, 2hr, 24 hr, or 48 hr after treatment.

4.1. IL-1 Antagonism The research described above gives some support for the idea that IS increases IL-1 in brain. However, such findings do not indicate that IL-1 in brain plays a role in mediating any or all of the sequelae of IS. To examine this question we have blocked or antagonized IL-1 in brain and examined various outcomes that normally follow IS. 4.1.1. a-Melanocyte Stimulating Hormone, a-melanocyte-stimulating hormone (a-MSH) is a tridecapeptide occurring in widespread regions of the CNS, in the circulation, and in peripheral tissues (Eberle, 1988). It is a derivative of proopiomelanocortin, shares the 1-13 amino acid sequence with ACTH, and multiple subtypes of melanocortin receptors are distributed in brain in regionally selective fashion (Tatro, 1993). These facts are noted because a-MSH can have potent anti-cytokine action (see Catania and Lipton, 1994, for review), including potent blockade of IL-1 effects. The intracerebroventricular (ICV) administration of a-MSH has been reported

244

S. F. Maier et al.

to reduce or block a variety of effects of IL-1 administration including fever (Glyn «& Lipton, 1981), reduced food and water intake (Uehara et al., 1992), pituitary adrenal response (Weiss et al., 1991), and suppressed natural killer cell activity (Weiss et al., 1991). a-MSH also blocks a number of effects of IL-1 in vitro, such as the stimulation of CRH release from hypothalamic explants (Zelazowski, Patchev, Zelazowska, Chrousos, Gold, & Sternberg, 1993) and the augmentation of thymocyte proliferation (Cannon, Tatro, Reichlin, & Dinarello, 1986). Furthermore, brain and plasma levels of a-MSH increase after administration of IL-1, and depletion of a-MSH exaggerates responses to IL-1 (Catania & Lipton, 1994). These data have led to the suggestion that a-MSH acts as an endogenous IL-1 antagonist, although the mechanism by which antagonism occurs remains unresolved. With the idea that ICV a-MSH can antagonize the effects of IL-1 in brain, and because it is more readily available than IL-lra, we have conducted a series of experiments designed to determine whether ICV a-MSH would block the effects of IS on sickness or acute phase responses. For example, Fig. 6 shows the plasma levels of the carrier protein CBG measured at baseline and 24hr after either IS or control treatment. Separate groups received either vehicle or 0.5 }Ag aMSH injected ICV before the IS. As can be seen, IS produced a robust reduction in plasma CBG 24hr later, and this effect was completely blocked by ICV a-MSH. We have also found ICV a-MSH to block many of the other sickness-type responses produced by IS including fever, reduced food and water intake, and shifts in basal corticosteroid levels (Milligan, Nguyen, Deak, Fleshner, Watkins, & Maier, unpublished observations). 4.1.2. IL-lra. We have conducted fewer experiments with IL-lra because it has been less readily available. We have tested IL-lra for its ability to modulate some of the more classic behavioral effects of IS. IS has often been shown to produce later failure to learn to escape shock in a different situation in which escape is possible (the "original" learned helplessnes effect) and to potentiate new fear conditioning (Maier, 1990). These effects only follow IS, and do not follow exposure to escapable shock. They

1000

VEH + HC MSH -I- HC

n •

VEH + IS MSH + IS

900 o

E

SOO 700 600 500 BASELINE

24 MRS AFTER IS

Figure 6. Plasma corticosteroid binding globulin at baseline and 24 hr after experiment treatment in subjects given a-MSH (MSH) or vehicle (VEH) before either inescapable shock (IS) or home cage control (HC) treatment.

Stress, Learned Helplessness, and Brain Interleukin-lp

245

are therefore clearly learned helplessness effects, and both were completely blocked by the ICV administration of IL-lra (Maier & Watkins, 1995) . 4.2. Stressor Controllability The IS experiments described in this chapter did not manipulate the controllability of the tailshock stressor. Thus, the relevance of many of these experiments to learned helplessness, and the relevance of brain IL-lp, is not clear. We have therefore conducted a number of experiments in which the controllability of the tailshocks has been manipulated. Rats were first either adrenalectomized or received sham surgery, and CORT was replaced in the drinking water as before. They then received either escapable tailshock, yoked inescapable tailshock, restraint in the apparatus, or remained as home cage controls, with parameters typically used in our laboratory (e.g., Grahn & Maier, 1995). Brains were removed 2hr later and assayed for IL-ip protein levels as above. IL-1(3 protein levels were examined in hypothalamus, hippocampus, nucleus tractus solitarius, posterior cortex, amygdala, cerebellum, frontal cortex, and pituitary. The general pattern of the results for IS was similar to that described above. Importantly, IL-lp increases were generally not modulated by the controllability of the shock and were just as large after escapable shock as they were for IS. Only the amygdala revealed a controllability effect, and we have not yet replicated this result. The effects of restraint could also be examined here since both restrained and home cage controls were included. Restraint did not have a detectable effect on brain IL-lp. These results led us to determine whether stressor controllability modulates the sickness/acute phase consequences of IS that were described above. We have now examined plasma carrier proteins, basal CORT shifts, plasma acute phase proteins, and white blood cell counts. None were modulated by stressor controllability. Thus, effects known to be produced by either peripheral immune stimulation and ICV IL-lp administration were not sensitive to stressor controllabihty. We therefore sought to determine whether behavioral effects that are sensitive to stressor controllability would be produced by ICV IL-ip injection. ICV IL-ip at a number of different doses was administered, and fear conditioning and shuttlebox escape learning was tested 24 hr later. These measures were chosen because only IS potentiates fear conditioning and interferes with escape learning. ICV IL-ip failed to potentiate fear conditioning or interfere with shuttlebox escape 24 hr later. Thus, ICV IL-ip was not sufficient to mimic these effects that are selective to IS.

5. CONCLUSIONS The conclusions to be drawn from the literature and the pattern of our own data are far from clear, and any suggestion can only be made in very tentative fashion. The data tend to suggest that brain IL-ip may be involved in mediating some consequences of exposure to stressors per se. Tailshock did increase brain IL-ip protein levels, most notably in the hypothalamus where rapid increases could be detected in nonadrenalectomized subjects. In addition, antagonizing IL-1 either with a-MSH or IL-lra blocked a variety of effects of tailshock including fever, increased acute phase proteins, increased white blood cells, decreased CBG, basal CORT shifts, and a number of behavioral effects. Furthermore, ICV IL-ip has been shown to produce many of these changes. Interestingly, restraint did not increase brain IL-1 P protein, nor does it produce

246

S. F. Maier et al.

outcomes charactersitic of the acute phase response such as increases in acute phase proteins, prolonged fever, and the Uke. Thus not all stressors are likely to engage the brain IL-1 machinery, and perhaps only intense stressors such as tailshock will do so. But what of learned helplessness? Stressor controllability did not modulate the increase in brain IL-ip produced by tailshock. ICV IL-ip produces many of the same consequences as does IS, but the consequences that it produces do not depend on the controllability of the shock—they follow both IS and escapable shock. The outcomes of shock that do depend on shock controllability, such as shuttlebox escape, were not induced by ICV IL-1|3. However, these controllability-sensitive sequelae of shock were blocked by ICV IL-lra. One way to rationalize this pattern would be to suggest that the induction of learned helplessness requires the activation of two distinct neural circuitries. One of these circuitries may be engaged by the physical stressor or aversive event itself, by the somatosensory input. It is intriguing here to note that laminae I and V of the spinal cord dorsal horn (which contains neurons that respond to noxious somatosensory stimuli) terminate in the nucleus tractus solitarius, the brain region at which the neural cascade in response to infection has also been reported to begin. IL-ip may be important in this circuit. The other circuit would be one that is involved in learning about psychological aspects of the stressor, such as whether it is controllable or not. This circuitry would, presumably, utilize "higher" structures in the brain. Here IL-ip may be unimportant. Learned helplessness effects may require both that the physical stress circuit has been engaged, and that the organism has learned that the stressor is uncontrollable. Under this scenario brain IL-ip in regions such as the hypothalamus would not be sensitive to stressor controllability, and brain IL-lp would be necessary, but not sufficient, to produce learned helplessness.

REFERENCES Abel, G. & Czop, J. K. (1992). Stimulation of human monocyte beta-glucan receptors by glucan particles induces production of TNF-alpha and IL-1 beta. International Journal of Immunopharmacology, 14, 1363-1373. Amano, Y., Lee, S. W., & Allsion, A. C. (1993). Inhibition by glucocorticoids of the formation of interleukin1 alpha, interleukin-1 beta, and interleukin-6: Mediation by decreased mRNA stability. Molecular Pharmacology, 43,176-182. Ban, E., Haour, R, & R., L. (1992). Brain interleukin-1 gene expression induced by peripheral lipopolysaccharide administration. Cytokine, 4, 48-54. Bluthe, R.-M., Michaud, B., Kelley, K. W., & Dantzer, R. (1996). Vagotomy attenuates behavioural effects of interleukin-1 injected peripherally but not centrally. NeuroReport, 7, 1485-1488. Cannon, J. G.,Tatro, J. B., Reichlin, S., & Dinarello, C. A. (1986). a melanocyte stimulating hormone inhibits immunostimulatory and inflammatory actions of interleukin 1. Journal of Immunology, 137, 2232-2236. Catania, A. & Lipton, J. M. (1994). The neuropeptide alpha-melanocyte-stimulating hormone: A key component of neuroimmunomodulation. Neuroimmunomodulation, 1, 93-99. Dantzer, R., Bluthe, R.-M., Kent, S., & Goodall, G. (1993). Behavioral effects of cytokines: An insight into mechanisms of sickness behavior. In E. G. DeSouza (Ed.), Neurobiology of cytokines (pp.130-151). San Diego, CA: Academic Press. Deak, T, Meriweather, J. L., Fleshner, M., Spencer, R. L., Moldawer, L. L., Grahn, R. E., Watkins, L. R., & Maier, S. F. (1997). Evidence that brief stress may induce the acute phase response in rats. American Journal of Physiology, 273, R1998-R2004. DeSimoni, M. G., Seroni, M., DeLuigi, A., Manfridi, A., Mantovani, A., & Ghezzi, P. (1990). Intracerebroventricular injection of IL-1 induces high circulating levels of IL-6. Journal of Experimental Medicine, 171, 1773-1778.

Stress, Learned Helplessness, and Brain Interleukin-lp

247

Dinarello, C. A. (1992). ELISA kits based on monoclonal antibodies do not measure total IL-1 beta synthesis. Journal of Immunological Methods, 148, 255-259. Dinarello, C. A. & Thompson, R. C. (1991). Blocking IL-1: Interleukin-1 receptor antagonist in vivo and in vitro. Immunology Today, 12, 404-410. Dunn, A. J. (1993). Role of cytokines in infection-induced stress. Annals of the New York Academy of Science, 697, 189-202. Dunn, A. J. (1995). Interactions between the nervous system and the immune system: Implications for psychopharmacology. In F. E. Bloom & D. J. Kupfer (Eds.), Psychopharmacology: The fourth generation of progress (pp. 719-733). New York: Raven Press. Eberle, A. N. (1988). The melanotropins: Chemistry, physiology, and mechanisms of action. Basel: Karger. Enk, A. H., Angeloni, V. L., Udey, M. C , & Katz, S. J. (1993). An essential role for Langerhans cell derived IL-1 beta in the initiation of primary immune responses in skin. Journal of Immunology, 150, 3698-3704. Ericsson, A., Kovacs, K. J., & Sawchenko, P. E. (1994). A functional anatomical analysis of central pathways subserving the effects of interleukin-1 on stress-related neuroendocrine neurons. Journal of Neuroscience, 14, 897-913. Fleshner, M., Deak,T., Spencer, R. L., Laudenslager, M. L., Watkins, L. R., & Maier, S. F. (1995). A long-term increase in basal levels of corticosterone and a decrease in corticosteroid-binding globulin following acute stressor exposure. Endocrinology, 136, 5336-5342. Fleshner, M., Nguyen, K. T., Cotter, C. S., Watkins, L. R., & Maier, S. F. (in press). Acute stressor exposure both suppresses acquired immunity and potentiates innate immunity. American Journal of Physiology. Glyn, J. R. & Lipton, J. M. (1981). Hypothermic and antipyretic effects of centrally administered ACTH(l-24) and a-melanotropin. Peptides, 2,177-187. Grahn, R. E. & Maier, S. F. (1995). The elevated plus-maze is not sensitive to the effects of stressor controllability in rats. Pharmacology, Biochemistry, & Behavior, 52, 565-570. Griffiths, R. J., Stam, E. J., Downs, J. T, & Otterness, I. G. (1995). ATP induces the release of IL-1 from LPSprimed cells in vivo. Journal of Immunology, 154, 2821-2828. Hagan, P., Poole, S., & Bristow, A. F (1993). Endotoxin-stimulated production of rat hypothalamic interleukin1(3 in vivo and in vitro, measured by specific immunoradiometric assay. Journal of Molecular Endocrinology, 11, 31-36. Hart, B. L. (1988). Biological basis of the behavior of sick animals. Neuroscience and Biobehavioral Reviews, 12,123-137. Kakucska, I., Qi, Y.,Clark, B. D.,& Lechan, R. M. (1993). Endotoxin-induced corticotropin-releasing hormone gene expression in the hypothalamic paraventricular nucleus is mediated centrally by interleukin-1. Endocrinology, 133,815-821. Keane, K. M., Giegel, D. A., Lipinski, W. J., Callahan, M. G., & Shivers, B. D. (1995). Cloning, tissue expression and regulation of rat interleukin 1 beta converting enzyme. Cytokine, 7,105-110. Kent, S., Bluthe, R.-M., Kelley, K. W., & Dantzer, R. (1992). Sickness behavior as a new target for drug development. Trends in Pharmacological Science, 13, 24-28. Kent, S., Rodriguez, F , Kelley, K. W., & Dantzer, R. (1994). Anorexia induced by microinjection of interleukin-ip in the ventromedial hypothalamus of the rat. Physiology and Behavior, 56,1031-1036. Kern, J. A., Lamb, R. J., Reed, J. C , Daniele, R. P, & Nowell, R C. (1988). Dexamethasone inhibition of interleukin 1 beta production by human monocytes: Posttranscriptional mechanisms. Journal of Clinical Investigation, 81, 237-244. KHr, J. J., McClellan, J. L., & Kluger, M. J. (1994). Interleukin-1 causes the increase in anterior hypothalamic interleukin-6 during LPS-induced fever in rats. American Journal of Physiology, 266, 1845-1848. Kluger, M. J. (1991). Fever: The role of pyrogens and cryogens. Physiological Review, 71, 93-127. Krueger, J. M., Walter, J., Dinarello, C. A., Wolf, S. M., & Chedid, L. (1984). Sleep promoting effects of endogenous pyrogen (interleukin-1). American Journal of Physiology, 246, R994-R999. Lee, S. W.,Tsou, A. P, Chan, H., Thomas, J., Petrie, K., Eugui, E. M., & Allison, A. C. (1988). Glucocorticoids selectively inhibit the transcription of the interleukin 1 beta gene and decrease the stability of interleukinl beta mRNA. Proceedings of the National Academy of Science, 85, 1204-1208. Linthorst, A. C. E., Flachskamm, C , Holsboer, F , & Reul, J. M. H. M. (1994). Local administration of recombinant human interleukin-1 beta in the rat hippocampus increases serotonergic neurotransmission, hypothalamic-pituitary-adrenocortical axis activity, and body temperature. Endocrinology, 135, 520-532. Long, N. C , Vander, A. J., & Kluger, M. J. (1990). Stress-induced rise of body temperature in rats is the same in warm and cool environments. Physiology and Behavior, 47, 773-775.

248

S. F. Maier et al

Maier, S. F. (1990). The role of fear in mediating the shuttle escape learning deficit produced by inescapable shock, yowrnfl/ of Experimental Psychology: Animal Behavior Processes, 16, 137-150. Maier, S. F. (1993). Learned helplessness, fear, and anxiety. In C. Stanford & P. Salmon (Eds.), Stress: From synapse to syndrome (pp. 207-248). London: Academic Press Ltd. Maier, S. F. & Seligman, M. E. P. (1976). Learned helplessness: Theory and evidence. Journal of Experimental Psychology: General, 105, 3-46. Maier, S. F., Ryan, S. M., Barksdale, C. M., & Kalin, N. H. (1986). Stressor controllability and the pituitaryadrenal system. Behavioral Neuroscience, 100, 669-678. Maier, S. F. & Watkins, L. R. (1995). Intracerebroventricular interleukin-1 receptor antagonist blocks the enhancement of fear conditioning and interference with escape produced by inescapable shock. Brain Research, 695, 279-286. Maier, S. F. & Watkins, L. R. (1998). Cytokines for psychologists: Implications of bidirectional immune-tobrain communication for understanding behavior, mood, and cognition. Psychological Review, 105, 83-107. Minami, M., Kuraishi, Y., Yagaguchi, T, Nakai, S., Hirai, Y., & Satoh, M. (1991). Immobilization stress induces interleukin-ip mRNA in the rat hypothalamus. Neuroscience Letters, 123, 254-256. Minor, T. R. & Saade, S. (1997). Poststress glucose mitigates behavioral impairments in rats in the "learned helplessness" model of psychopathology. Biological Psychiatry, 42, 324-334. Morimoto, A., Sakata, Y, Watanabe, T, & Murakami, N. (1989). Characteristics of fever and acute-phase response induced by IL-1 and TNF. American Journal of Physiology, 363, R35-R41. Morrow, L. E., McClellan, J. L., Conn, C. A., & Kluger, M. J. (1993). Glucocorticoids alter fever and IL-6 responses to psychological stress and to lipopolysaccharide. American Journal of Physiology, 264, R1010-R1016. Nguyen, K. T, Deak, T , Owens, S. M., Kohno, T, Fleshner, M., Watkins, L. R., & Maier, S. F (1998). Exposure to acute stress induces brain interleukin-1 beta protein in the rat. Journal of Neuroscience, 18, 2239-2246. Plata-Salaman, C. R., Oomura, Y, & Kai, Y (1988). Tumor necrosis factor and interleukin-1 beta: Suppression of food intake by direct action in the central nervous system. Brain Research, 448,106-114. Quan, N., Sundar, S. K., & Weiss, J. M. (1994). Induction of interleukin-1 in various brain regions after peripheral and central injections of lipopolysaccharide. yowrna/ of Neuroimmunology, 49. 125-134. Quan, N., Zhang, Z., Emery, M., Bosnall, R., & Weiss, J. M. (1996). Detection of interleukin-1 bioactivity in various brain regions of normal healthy rats. Neuroimmunomodulation, 3, 47-55. Rothwell, N. M. (1989). CRF is involved in the pyrogenic and thermogenic effects of interleukin 1 beta in the rat. American Journal of Physiology, 256, E111-E115. Saperstein, A., Brand, H., Audhya, T, Nobriski, D., Hutchinson, B., Rosenzwieg, J., & Iiollander, C. S. (1992). Interleukin-1 beta mediates stress-induced immunosuppression via CRH. Endocrinology, 130, 152-158. Shintani, F , Nakaki, T, Kanba, S., Sato, K., Yagi, G., Shiozawa, M., Aiso, S., Kato, R., & Asai, M. (1995). Involvement of interleukin-1 in immobilization stress-induced increase in plasma adrenocorticotropic hormone and in release of hypothalamic monoamines in the rat. Journal of Neuroscience, 15, 1961-1970. Spadaro, F. & Dunn, A. J. (1990). Intracerebroventricular administration of interleukin-1 to mice alters investigation of stimuli in a novel environment. Brain, Behavior, and Immunity, 4, 308-322. Sperlagh, B., Sershen, H., Lojtha, A., & Vizi, E. S. (1998). Co-release of endogenous ATP and pH] noradrenaline from rat hypothalamic slices: Origin and modulation of alpha 2-adrenoceptors. Neuroscience, 52,511-528. Sundar, S. K., Cierpial, M. A., Kilts, C , Ritchie, J. C , & Weiss, J. M. (1990). Brain interleukin-1-induced immunosuppression occurs through activation of both pituitary-adrenal axis and sympathetic nervous system by corticotropin-releasing factor. Journal of Neuroscience, 10, 3701-3706. Takahashi, S., Kapas, L., Fang, J., Seyer, J. M., Wang, Y, & Krueger, J. M. (1996). An interleukin-1 receptor fragment inhibits spontaneous sleep and muramyl dipeptide-induced sleep in rabbits. American Journal of Physiology, 271, 101-108. Tatro, J. B. (1993). Brain receptors for central and peripheral melanotropins. Annals of the New York Academy of Science, 680, 621-625. Tringali, G., Mancuso, C , Mirtella, A., Pozzoli, G., Parente, L., Preziosi, P., & Navarra, P. (1996). Evidence for the neuronal origin of immunoreactive interleukin-1 p released by hypothalamic explants. Neuroscience Letters, 219, 143-146. Tringali, G., Mirtella, A., Mancuso, C , Guerriero, G., Preziosi, P., & Navarra, P. (1997). The release of

Stress, Learned Helplessness, and Brain Interleukin-lp

249

immunoreactive interleukin-1 beta from rat hypothalamic explants is modulated by neurotransmitters and corticotropin-releasing hormone. Pharmacological Research, 36, 269-273. Uehara, Y., Shimizu, H., Sato, N., Tanaka, Y., Shimomura, Y, & Mori, M. (1992). Carboxyl-terminal tripeptide of a-melanocyte-stimulating hormone antagonizes interleukin-1-induced anorexia. European Journal of Pharmacology, 220, 119-122. Van Dam, A.-M., Brouns, M., Louisse, S., & Berkenbosch, F. (1992). Appearance of interleukin-1 in macrophages and in ramified microglia in the brain of endotoxin-treated rats: A pathway for the induction of non-specific symptoms of sickness? Brain Research, 588, 291-296. Weiss, J. M., Sundar, S. K., Cierpial, M. A., & Ritchie, J. C. (1991). Effects of interleukin-1 infused into brain are antagonized by a-MSH in a dose-dependent manner. European Journal of Pharmacology, 192, 177-179. Woodmansee, W. W., Silbert, L. H., & Maier, S. F. (1993). Factors that modulate inescapable shock-induced reductions in daily activity in the rat. Pharmacology, Biochemistry, & Behavior, 45, 553-559. Zelazowski, P., Patchev, V. K., Zelazowska, E. B., Chrousos, G. P, Gold, P W., & Sternberg, E. M. (1993). Release of hypothalamic corticotropin-releasing hormone and arginine-vasopressin by interleukin ip and aMSH: Studies in rats with different susceptibility to inflammatory disease. Brain Research, 631, 22-26.

14

STRESS, DEPRESSION, AND THE ROLE OF CYTOKINES B. E. Leonard' and Cai Song^ ' Department of Pharmacology, National University of Ireland Galway, Ireland ^ Life Sciences Research Institute, Carlton University Ottawa, Ontario, Canada

INTRODUCTION The concept of an inter-relationship between the psychological state of a depressed patient and the immune status can be traced back to Galen who, in 200 AD, suggested that melancholic women are more susceptible to breast cancer than sanguine women (Leonard, 1987). Over the past 15 years it has become apparent that the central nervous system (CNS) and the immune system are intimately connected and that a functional bidirectional communication exists between these systems (Ballieux, 1992). Indeed, it may be possible to conceive of the nervous, endocrine, and immune systems as being part of a single integrated network rather than three separate systems. The study of the interactions between these systems has given rise to the new discipline of psychoimmunology, a term first coined by Ader, Felten, and Cohen in 1987. It is widely accepted that stress and psychiatric illness can compromise immune function (Leonard, 1990; 1995). Furthermore, soluble mediators released by immune cells can influence brain function and cause changes in behaviour in both man and lower animals. In addition to the behavioural changes that occur in depressed patients, there are also profound alterations in the endocrine and immune systems (Connor and Leonard, 1998). Most of the initial studies of the immune changes in depression indicated that a suppression of immune function occurs as indicated by an impaired zymosan induced neutrophil phagocytosis (O'Neill and Leonard, 1990), mitogenstimulated lymphocyte proliferation (Kronfol and House, 1989) and natural killer cell (NKC) activity (Irwin, Lacher, and Caldwell, 1992). In addition to the immune changes that occur in depressed patients, a number of studies have concentrated on indices of immune function in those who have been exposed to stressful life events such as bereavement, divorce, and academic examinations. Exposure to such stressful events has also been reported to cause impairment in Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

251

252

B. E. Leonard and C. Song

various aspects of cellular immune function that qualitatively resemble those changes reported to occur in depression (Irwin, 1995). Despite these changes demonstrating that immunosuppression occurs in those exposed to psychological stress or to depression, there is also evidence that activation of some aspects of the immune system can also arise (Maes, Smith, and Scharpe, 1995). This serves to emphasize the complex interrelationship that exists between external or internal stressors, activation of the pituitary adrenal axis and the subsequent changes in immune function. The purpose of this review is to present the evidence that impHcates the proinflammatory cytokines as the causal factors in depression and in the impact of stress that may trigger the onset of depression.

1. STRESS, ITS DEFINITION AND CLASSIFICATION Stress may be defined as the physical or psychological stimulus which, when impinging upon an individual, produces strain or disequilibrium. A more complex definition proposed by Sklar and Anisman (1981), defines stress as the reactions of the body to forces of a deleterious nature, infections, and various abnormal states that tend to disturb its normal physiological equilibrium or homeostasis. The stimulus that causes such a disruption is referred to as the stressor. From such definitions, it is clear that stress is a very broad concept that must be clarified in terms of the severity of the stressor, its duration and the age, gender and genetic composition of the subject upon whom the stress impinges. Thus in regard to the classification of stress, it is essential to determine whether the stress is acute or chronic, whether it is unavoidable or avoidable, physical or psychological. Furthermore, many types of physical stress will have direct metabolic consequences (for example, cold or heat exposure) which will have an impact on the general metabolism in addition to causing psychophysiological changes. Knowledge of the effects of stress on the immune, neurotransmitter, and endocrine systems has been derived from animal experiments. The pioneering studies of Rasmussen (1969) and Solomon (1969) produced evidence to indicate that certain forms of stress could cause decreased immunocompetence. Such research led other investigators to study the relationship between stress induced immunity and cancer in animals (Gottesfeld and Liehr, 1987; Tejwani, Gudehithlu, Gienapp, Malarkey, and Hanissian, 1991) and man (Shekelle, Raynor, and Ostfeld, 1981). The connection between predisposition to certain types of cancer, stress, and depression was based upon studies indicating that those individuals with a certain personality were more prone to cancer, an event which appeared to be linked to a supression of some aspects of cellular immunity (Cooper, 1984; Cox and MacKay, 1982). Many studies have subsequently shown that the predisposition to cancer, and more recently the development of AIDS following a prolonged HIV infection, is related to the severity and duration of the stress to which the individual is subjected (Levy, Herberman, Lippman, and d'Angelo, 1987). Given the complexity of the stressors that have been used to investigate the changes in the immune system, and the differences in the responses of the individual to these stressors depending on the age, sex, genetic composition, etc., it is perhaps not surprising to find a wide variation in their results that are expressed in the literature regarding the causal relationship between stress and the immune system.

Stress, Depression, and the Role of Cytokines

253

With regard to the endocrine system, the impact of stress is generally measured by determing the adrenal glucocorticoid concentrations in the plasma. Stress and the increase in plasma glucocorticoid concentrations are closely related (Owens and Nemeroff, 1991). However, the changes in glucocorticoid release from the adrenals is partly dependent on the duration of the stress. Thus acute stress usually increases glucocorticoid secretion in both animals and man whereas following chronic stress, decreased concentrations are frequently observed. In addition to the changes in adrenal cortical function that may be directly associated with acute stress caused by pain or restraint for example, it is essential to determine other physiological parameters (such as food and water deprivation) which may also indirectly impact upon the immune and endocrine systems following the application of stressors. Thus renal function should be assessed following water deprivation and blood glucose concentrations estimated in food deprivation stress.

2. STRESS AND THE IMMUNE SYSTEM Because stress activates both the hypothalamic-pituitary-adrenal-axis (HPA axis) and the sympathetic nervous system (SNS), it is not surprising to find that most acute stressors can modify the immune response. It is well known that plasma catecholamines released from the adrenals in response to stress, in addition to the adrenal glucocorticoids, can cause immuno-suppression. There are numerous experimental studies showing that various types of externally applied acute stressors (for example, electric shocks, social defeat, maternal separation, immersion in cold water) suppress some aspects of immune function. Similarly chronic stressors such as overcrowding, have been shown to suppress aspects of cellular and humoral immunity. Indeed, it is difficult to consider any aspect of cellular and humoral immunity that is not altered by some stressor (Maier, Watkins, and Fleshner, 1994). It is evident from the results of both the experimental and clinical studies of the effects of stress on the immune system that all stressors do not produce identical changes in the immune and endocrine systems. It has long been known that different stressors produce different degrees of SNS and endocrine activation (Mason, 1971). For example, one stressor might strongly activate the SNS but have relatively little effect on the HPA axis whereas another may have the converse effect. In addition, the time course of these changes will differ for different stressors and for the psychological state of the individual. Moreover coping strategies are important in that the individual can learn to modify the adverse impact of the stressor (Mormede, Dantzer, Michaul, Kelley, and Le Maul, 1988). It must also be emphasizd that the specific immune response involves a complex cascade of events that may extend over several days. As the catecholamines, endorphins, glucocorticoids, etc. play crucial roles in modulating this cascade, the effects of stress will, of necessity, be variable. Thus experimental and clinical situations will arise in which stressors may have an effect, have no effect or even an enhanced effect on immune function (see Croiset, Heijnen, Veldhuis, de Wied, and Balheus, 1987). Fleshner and co-workers have shown, for example, that a stressor will interfere with antibody synthesis, determined several weeks following the antigen administration, only if the stressor is applied near the time of the antigen exposure (Fleshner, Bellgrau, Watkins, Laudenslager, and Maier, 1995). Such findings serve as a caution from studies that have determined one aspect of immune function at one time point

254

B. E. Leonard and C. Song

only from which conclusions are drawn that stress suppresses immune function. Many immune parameters are non specific and assess some intermediate aspect of the immune response (for example, the synthesis of the interleukins or proliferative response of T-cells to mitogens) rather than an effector end point that detects and destroys antigen, recognizes virus infected cells, etc. It must be remembered that the immune system contains a high degree of redundancy and therefore the changes in part of the immune cascade are not by themselves evidence that the final end point of the immune process (for example, the production of a specific antibody) is affected (Cunnick, Lysle, Aronfield, and Rabin, 1991). There is a large and relatively consistent literature on the effects of stressful life events on predisposition to both physical illness and infections. While the correlations between such life events and illness are not large, generally accounting for only about 10% of the variance (Weisse 1992), the effects are consistent across populations and different types of life events. Bereavement stress has been the subject of several important studies. There is also evidence that risks to health associated with separation and divorce are greater than with bereavement (Kiecolt-Glaser, Garner, Speicher, Trask, and Glaser, 1987). For example, Kiecolt-Glaser et al. (1987) showed that separated or divorced women had a poorer immune function on five of the six immunological variables studies than matched married women. Somewhat similar findings were reported for separated or divorced men (Kiecolt-Glaser and Glaser, 1988). It should be emphasized that the sample size in these studies was quite small, but such data does serve to emphasize the impact of severe life events on the immunological state and consequence health of normal individuals. The effect of chronic stress on individuals caring for parients with Alzheimer's disease has also been the subject of several studies. It has been shown, for example, that such individuals show a high risk of depression (Crook and Miller, 1985; Fiore, Becker, and Cooppel, 1983). In addition to the greater physical and emotional distress shown by the carers, there is also evidence of impaired immune function (KiecoltGlaser, Dura, Speicher, Trask, and Glaser, 1991). Other studies of those individuals subject to chronic environmental stress (for example, living in the vicinity of Three Mile Island in the USA, the site of a nuclear power plant accident some years ago) showed that the residents had fewer T-suppressor cells, B-lymphocytes, and natural killer cells than a comparable group living in a normal environment (Davidson and Baum, 1986). The conclusions of these studies is that chronic stress in man does not necessarily lead to immunological adaptation. Clearly, there are marked differences between the stress induced changes in rodents and those reported in man. Thus, in rodents, acute stress appears to be immunosuppressive, whereas chronic stress is associated with adaptive changes or even enhancement of the immune response (Cohen and Crnic, 1982; Monjan and Collector 1977). Examination stress in University students has been the subject of several studies in the United States. Thus, a decrease in natural killer cell number and function has been reported by several groups of investigators (Glaser, Rice, and Speicher 1986; Kiecolt-Glaser and Glaser (1988), effects that were not attributable to poor nutritional status. In addition, academic stress has been associated with significant changes in antibody titers to latent herpes viruses suggesting changes in cellular immunity. In particular, elevated antibody titers to the Epstein Barr virus (the causal agent for

Stress, Depression, and the Role of Cytokines

255

infective mononucleosis), herpes simplex virus type 1 (that causes cold sores), and cytomegalovirus (which causes the monucleosis syndrome) were raised prior to examinations but returned to normal levels following the examination (Glaser, Kiecolt-Glaser, and Stout, 1985). There were additional changes in mitogen stimulated lymphocyte replication associated with academic stress. Thus, the incidence of self-reported infectious illness was also increased in these individuals. The effect of relaxation techniques on these immune parameters was also studied and showed that although the percentage of helper T-cells did not decline so markedly in those subjects that were given relaxation exercises, natural killer cell activity was unaffected by such an intervention. It may be concluded from these studies of the effects of stress and adverse life events that adaptive changes in the immune system are not pronounced in man. One of the major problems arising from the clinical studies lies in the difficulty in adequately defining stress, because the same event may have different effects on different individuals. Furthermore, most of the components of the immune system normally vary within wide limits, thereby making the small, but important changes difficult to detect. Added to these problems is the difficulty in deciding which parameter accurately reflects the true status of the individuals immune defences.

3. I N T E R A C T I O N B E T W E E N T H E B R A I N A N D I M M U N E SYSTEM IN STRESS A N D D E P R E S S I O N The monoamine hypothesis of depression proposes that a deficit of brain noradrenaline (NA) and/or serotonin (5-HT) may be causally involved in the symptoms of illness (Baldessarini (1975). Another theory of depression suggests that the disorder in hypothalamus-pituitary-adrenal (HPA) axis causes an increase in secretion of corticotropin-releasing factor (CRF), which stimulates adrenocorticotrophin hormone (ACTH) and Cortisol release (Bateman Singh, Krai, and Solomon, 1989). Recently, the macrophage theory of depression, which will be discussed in detail later in this review, has also been proposed. In this hypothesis, the abnormal secretion of some cytokines such as interleukin-1 (IL-1) and interferon-alpha (INF-alpha), results in disordered secretions of CRF, ACTH, prolactin, and Cortisol, together with a depressive state (Smith, (1991). These three hypothesis may be linked. Whatever changes in the central nervous system (CNS) or in the endocrine system occur, different aspects of immune function are affected. It is known that noradrenergic and cholinergic terminals innervate the thymus gland and bone marrow and there are different neuropeptide, neurotransmitter and hormone receptors on lymphocytes and monocytes. In addition, cytokines produced by immune cells and microglia exert different effects on the brain and on the endocrine and immune systems (Bost, 1988; Farrar, 1988). A number of studies have shown that stress and depression are often associated with an impairment in immune function (Kronfol and House, 1989). At the cellular level, a reduction in mitogen-stimulated lymphocyte proliferation and neutrophil phagocytosis and elevated monocyte activity has been reported in stress and depression (Leonard, 1990; McAdams and Leonard, 1993; O'Neill and Leonard, 1990). An increase in total white cell number and an abnormality in differential white blood cell (WBC) count (for example, increase in the percentage of neutrophil and decrease in lymphocytes) has also been found after stress or in depressed patients (Maes, Planken,

256

B. E. Leonard and C. Song

and Stevens, 1992). Recently, leucocyte adhesiveness/aggregation (LAA) has also been reported to be increased during stress; this has been suggested as a marker of stress (Arber, Berliner, and Tamir, 1991). At the subcellular level, it has been reported that serum and plasma concentrations of immunoglobulin (Ig) complement (C) and acute phase proteins are changed in depressed patients (Healy, 1991; Kronfol and House, 1989; Maes et al. 1992). For example, IgA, IgM, complement C3, C4, and positive acute phase proteins are increased in depression, while negative acute phase proteins are decreased (Song, Dinan, and Leonard, 1994). The concentrations of cytokines IL-1, INF-alpha, and tumor necrosis factor are raised and IL-2 is reduced in the depressed patient (Katila, Rimon, Cantwell, Appelberg, and Nikkila, 1991; Nathan, 1987; Smith, 1991). At the organ and system level, it has been reported that the weights of thymus gland and spleen are reduced and adrenal is increased during stress (Dohmus and Metz, 1991). Histological studies reveal that a stress-induced rise in corticosterone, or an exogenous injection of corticosterone, causes cortical atrophy and lymphocyte necrosis in the thymus gland of rats (Dohmus and Metz (1991). However, despite the circumstantial evidence implicating an interaction between neurotransmitter, endocrine, and immune changes, a causal relationship between these changes has yet to be proven.

4. DO GLUCOCORTICOIDS INHIBIT OR STIMULATE IMMUNE FUNCTION IN DEPRESSION OR FOLLOWING CHRONIC STRESS? The efficacy of synthetic glucocorticoids in the treatment of inflammatory diseases is usually attributed to their potent anti-inflammatory action, primarily due to the inhibition of the synthesis and release of proinflammatory cytokines. These effects are well documented (Munck and Naray-Fejes-Toth, 1994; Almawi, Beyhum, Rahme, and Rieder, 1996). However, not all cytokines are suppressed by glucocorticoids. For example, macrophage colony-stimulating factor (M-CSF) and transforming growth factor beta (TGF-beta) are not inhibited (Munck and Naray-Fejes-Toth, 1994), while interleukin 4 may be increased (Wu, Fargeas, Nakajima, and Delespesse, 1991) or decreased (Daynes and Araneo, 1989). In addition to the dual action of glucocorticoids on cytokine concentrations, it is also known that they can act synergistically with cytokines. For example, in cultures of hepatic cells, glucocorticoids strongly potentiate the ILl and IL-6 induced expression of acute phase proteins (Baumann and Gauldie, 1994). Furthermore the expression of many cytokine receptors is strongly upregulated by glucorticoids. These include receptors for IL-1, IL-2, IL-4, IL-6, INF gamma and TNF alpha (see Wiegers and Reul, 1998). Thus while glucocorticoids may inhibit cytokine synthesis and release, they may also act synergistically with some cytokines and enhance the expression of a number of cytokine receptors and a common signal transducer, gpl30 (Wiegers and Renl, 1998). The apparent diverse effects of the glucocorticoids on immune function may be explained by the fact that by increasing the number of high affinity IL-2 receptors on the surface of T-cells, the prohferative response of these cells is increased, even though the production of IL-2 is decreased (Wiegers, Labeur, Stec, Klinkert, Holsboer, and Reul 1995). Thus it would appear that glucocorticoids enhance, rather than reduce the biological response by increasing the sensitivity of the target cell for certain cytokines to the point where the concentration of the cytokine becomes rate-limiting. In this way.

Stress, Depression, and the Role of Cytokines

257

the glucocorticoids appear to optimise the course of a biological response that, if delayed, could be potentially detrimental to the organism. As an example of the pathological interactions that can occur between glucocorticoids and cytokines, it is possible that in some allergies glucocorticids induce T-helper 2 cellular response (Daynes and Araneo, 1989) that involve the release of IL-4 and other cytokines. There is evidence that some allergies involve inappropriate T-helper cell responses. Thus while it is generally accepted that glucocorticoids are effective in treating most allergic diseases following chronic administration, in some cases they may exacerbate the condition by inducing T-helper cells to produce cytokines such a IL-4 (Blotta, De Krugff, and Umestsu, 1997). It would also appear that glucocorticids may act synergistically with IL-4 to induce immunoglobulin E (IgE), a principal mediator of allergic diseases (Wu, Nakajima, and Delespessa 1991). Such observations may be relevant to explaining the increased incidence of allergies in patients with depression. Thus it appears that glucocorticoids exert distinct, even paradoxical effects on cytokine expression, cytokine receptor expression and cytokine regulated biological responses. An understanding of these complex interrelationships appears to have been hampered by the fragmented information that has been obtained from clinical and experimental studies in which, for example, the changes in the cytokine concentrations only have been determined. Studies such as those by Wiegers et al. (1995) for example, have demonstrated the importance of a comprehensive approach that involves assessing all the primary immune parameters (such as IL-2, IL-2 receptors, T-cell proliferative responses) over time so that the effect of glucocorticoids on specific aspects of cellular immunity can be fully assessed.

5. THE MACROPHAGE THEORY OF DEPRESSION Of all the endocrine changes that are reported to occur in stress and depression, Cortisol hypersecretion is the most frequently observed. Although hypercortisolaemia can arise as a consequence of an acute stressor, there is a qualitative difference between the circadian pattern of Cortisol secretion in the depressed patient and that observed following exposure to a stressful stimulus. Thus the nadir of the plasma Cortisol concentration occurs some 6 hours earlier in the depressed patient than in the stressed, non-depressed patient. The ACTH concentration is also elevated while the concentration of corticotorphin releasing factor (CRF) in the cerebrospinal fluid is raised in depressed patients (Linkowski, Mendlewicz, and Kerklofs, 1987; Nemeroff, Widerlov, and Bissette, 1984). As there is a close association between plasma glucocorticoids and immune function, it is often assumed that the immune changes are a direct consequence of the raised plasma, and brain concentrations of the adrenal glucocorticoids. However, this seems unlikely as the prolonged increase in glucocorticoids in depression leads to a decrease in the sensitivity of the CRF and glucocorticoid receptors in the brain, pituitary, and on immune cells (Leake, Perry, and Ferrier, 1980; Dinan, 1994) thereby reducing the inhibitory effect to these steroid hormones on cellular immunity. What causes the adrenal steroid abnormalities in depression? One possibility is that the elevated proinflammatory cytokine IL-1 is at least partly responsible. It is known that IL-1 has a direct action on the hypothalamus leading the increased release of CRF. In vitro, IL-1 has been shown to stimulate ACTH and growth hormone secretion (Goetzl, Screedharan, and Harkonen, 1988); these effects are not shared by the

258

B. E. Leonard and C. Song

Other major proinflammatory cytokines, IL-6, and TNF alpha. In addition to the direct effect of IL-1 on the HPA axis, there is controversial experimental evidence that macrophages may also secrete ACTH which could directly stimulate the adrenals to synthesize Cortisol. Thus an activated macrophage system in depression could both directly and indirectly contribute to hypercortisolaemia. The question arises regarding the mechanism whereby an increase in peripheral IL-1 can precipitate changes in the immune, endocrine, and neurotransmitter systems in the brain. Indeed, an important and unresolved question is whether cytokines released peripherally gain access to the brain in concentrations that are biologically effective. Cytokines are large hydrophilic molecules (Hamblin, 1994), their size and structure being such that passive diffusion across the blood brain barrier (BBB) is likely to be minimal (Hopkins and Rothwell, 1995). Currently, it is postulated that cytokines produced in the periphery can act on one or other circumventricular organs, such as the median eminence (ME) and the organum vasculosum laminae terminalis (OVLT) that lack a functional BBB (Hopkins and Rothwell, 1995). It has been suggested that cytokines from the periphery bind directly to glial cells on the OVLT, which in turn produce cytokines and other mediators such as prostaglandins, particularly prostaglandin E2 (PGE2). This hypothesis is consistent with the observations that peripheral IL-1 beta administration elevates PGE2 concentrations in many brain structures as assessed by in vivo microdialysis, which is maximal and most rapid at the OVLT and the medial preoptic area, and that the central increase in PGE2 precedes the onset of fever (Komaki, Arimura, and Koves, 1991). This hypothesis is further strengthened by the fact that many cytokine- and endotoxin-induced neurochemical (Lavicky and Dunn, 1995; Mefford and Heyes, 1990) and behavioural (Crestani, Seguy, and Dantzer, 1991; Hellerstein, Meydani, Meydani, Wu, and Dinarello, 1989; Osaka, Kannan, Kawano, Veta, and Yamashita, 1992) responses are attenuated by cyclooxygenase inhibitors such as indomethacin. Moreover, certain neurons in the preoptic nucleus have receptors for IL-1, IL-6, and TNF-alpha (Schettini, 1990). Finally, it has been reported in rodents that peripheral injection of an endotoxin increases the production of cytokines in the brain. Thus peripheral infections might affect the activity of the human brain at least in part through a similar mechanism (Rivier and Rivest, 1993). Not only does IL-1 act at the OVLT, but it can cross the BBB by an active transport system (Banks, Kastin, and Durham, 1989). For instance, an active transport mechanism for TNF-alpha has been described (Gutierrez, Banks, and Kastin, 1993). The concentration of these cytokines crossing the BBB by such mechanisms may be so low that they are physiologically insignificant (Hopkins and Rothwell, 1995), but this may be an important route of entry into the brain when plasma concentrations of cytokines are very high (Banks et al., 1989). In addition to the hypothesis that peripherally produced cytokines affect CNS function via the circumventricular organs, there is also evidence suggesting the existence of neurally mediated mechanisms of communication between peripherally produced cytokines and the CNS (Dantzer, 1994). This hypothesis that postulates communication between peripherally produced cytokines and the CNS via a neural afferent pathway is supported by the fact that subdiaphramatic vagotomy attenuates endotoxin induced depressive effects on behaviour, c-fos expression in the CNS and IL-1 beta expression in the hypothalamus (Dantzer, 1994). In addition, subdiaphramatic vagotomy has been reported to block HPA-axis activation produced by peripheral IL-1 beta and TNF-alpha administration (Fleshner, Bellgrau, Watkins, Laudenslager,

Stress, Depression, and the Role of Cytokines

259

and Maier, 1995) and also hypothalamic noradrenaline depletion produced by peripheral IL-1 beta administration. The findings from the vagotomy studies are important because they indicate that the brain is able to respond to cytokines that have been released at the periphery during the course of an infection or an inflammatory response and to respond to this stimulus by a local synthesis of cytokines. The mechanisms that are responsible for the transformation of the immune message into a neuronal message at the periphery, and the transduction of this neuronal message back into an immune message in the central nervous system, still need to be determined (Dantzer, Bluthe, Aubert, Goodall, Parnet, and Kelley, Bret-Dibat, Kent, Goujon, and Laye, 1996). Thus, in addition to cytokines produced from microglia and other macrophages within the CNS, peripherally produced cytokines can also affect the brain and produce many physiological, behavioural, endocrine, and neurochemical changes following most immunological challenges. There is now substantial evidence to show that major depression is accompanied by an acute phase protein response, an increased secretion of prostaglandins and by an excessive secretion of proinflammatory cytokines. These and other changes suggest that immune activation may play a role in the pathogenesis of depression and provide the basis for the macrophage theory of depression (Smith, 1991). Thus inflammatory cytokines or lipopolysaccharide (LPS) administered to animals or man provoke an extensive set of symptoms that are also identical to those found in major depression. These changes affect not only the psychological state of the individual but are also associated with changes in the activity of the HPA similar to those seen in depression. There is also evidence that the increase in the circulatory concentrations of IL-1 and IL-6 mediate the acute phase protein response which is characterized by elevated positive acute phase proteins (Song, et al., 1994) and a reduction in the serum tryptophan concentration. The precise mechanism whereby pro-inflammatory cytokines such as IL-1 beta modulate central serotonergic function is uncertain but recent evidence from in vitro studies on JAR cells, components of a choriocarcinoma cell line derived from human placenta, have shown that IL-1 activates the serotonin transporter directly (Ramamoorthy, Ramamoorthly, Prasad, Bhat, and Mahesh, 1995). If a similar effect occurs on central serotonergic neurons, it could lead to an increased removal of the amine from the synaptic cleft thereby leading to a reduction in serotonergic function. Receptors for IL-1 beta occur on serotonergic neurons (Cunningham and de Souza, 1996) and it is well established that this cytokine is synthesized by neurons and glial cells. Furthermore, raphe neurons may also respond to IL-1 delivered by white blood cells penetrating endothelial barriers during an inflammatory process (Cunningham and de Souza, 1996). Thus a reduction in the serum tryptophan concentration associated with elevated acute phase proteins, and an enhanced reuptake of serotonin from the synaptic cleft caused by the action of IL-1 on the serotonin transporter, may contribute to a malfunction of the serotonergic system that is causally associated with depression. Indirect support for this hypothesis comes from the observation that antidepressants suppress the pro-infammatory cytokines (Xia, De Piere, and Nassberger, 1996). This suggest that the proinflammatory cytokines can act as common mediators for the action of external (for example, psychosocial) and internal (for example, infections and toxins) stressors that are known to play a crucial role in the aetiology of depression. One advantage of the macrophage theory of depression is that it brings together the disparate changes in the immune, endocrine, and neurotransmitter system with the

260

B. E. Leonard and C. Song

clinical and epidemiological observations and also provides direct predictions that may be tested experimentally and/or clinically. For example, Maes, Smith, Christophe, Cosyns, Desnyder, and Meltzer (1996) have shown that the concentration of omega 3 fatty acids in the erythrocyte membranes of depressed patients is significantly decreased. This could suggest an inbalance between the omega 3 and omega 6 fatty acid pathway and reflects an increased synthesis of prostaglandins due to the relatively high intake of vegetable oil (a source of omega 6 fatty acids). Thus an increase in omega 6 and/or a decrease in omega 3, fatty acids could contribute to the changes that cause depression. A diet rich in omega 3 fatty acids (fish oil) might therefore have some immunoprotective function (Smith, 1991; Maes and Smith, 1998). 6. A R E C H A N G E S I N I M M U N E F U N C T I O N C A U S A L L Y O R COINCIDENTALLY C O N S E Q U E N C E S OF STRESS OR DEPRESSION? The association between cancer, autoimmune diseases, myocardial infarction, stroke and dementia with depression and the activation of the immune system, particularly involving the proinflammatory cytokines, has been the subject of considerable discussion in recent years. The initial studies indicating that patients suffering from major depression had decreased cellular immune response compared to healthy controls (Kronfol and House, 1985; Schleifer, Keller, Siris, Davis, and Stein, 1985) helped to lay the scientific basis whereby psychosocial factors could profoundly affect the development of physical and psychiatric disease. However, in well over 30 studies in the last 15 years, the consistency of the immune changes in depression is uncertain, with some investigators findings impaired immunocompetence while others do not. This situation led Miller, Spencer, McEwen, and Stein (1993) to review all the published studies regarding the changes in differential white blood cell counts, mitogen induced proliferation of T cells and changes in NK cell activity in depression. The results of their survey failed to find significant differences between depressives and their controls in the majority of the 30 studies assessed. For example, with regard to mitogen-induced lymphocyte proliferation, approximately half the studies demonstrated a significant decrease in lymphocyte proliferation whereas half the studies found no differences. Finally, of the 10 studies of changes in NK cell activity in depressed patients, 6 reported decreased activity and 4 found no difference. Thus of the 3 parameters of immune function that are frequently evaluated in studies of depression and stress, it does not appear that alterations in the immune system are specific or reproducible correlates of the psychological state, but may be associated with other variables that characterize the patients, such as the age, gender, severity, and duration of the stress or depressive episode. The following factors appear to contribute to the equivocal outcome of the reported changes in cellular immunity in depression. i. The heterogeneity of the patients used and the controls selected for comparison. In many of the studies cited, the patients were older than their controls and had been hospitalized for several weeks. It is well established that age, gender, and hospitalization status can profoundly affect the immune status that is frequently not taken into account (Evans, Folds, and Pettito, 1991; Miller, Asmis, and Lackner, 1991, Schleifer, Keller, Bond, Cohen, and Stein, 1989). ii. The variability of the immune assays used. This applies particularly to the

Stress, Depression, and the Role of Cytokines

261

mitogen-induced lymphocyte proliferation assay where it has been shown that up to a 50% variability in the results can be obtained in the same laboratory using the same method (Schleifer et al. 1989). iii. The relevance of the assays used. The relationship between the number of immune cells, or NKC activity in the peripheral blood, and the competence of the immune system in otherwise healthy individuals is unclear. Given that the immune system is a complex network of multiple cell types with various specialized functions, the number and location of these cells in any one immune compartment of the immune system may not reflect their activity in other compartment (Keller, Weiss, Schleifer, Miller, and Stein, 1981). Furthermore, the endocrine and peripheral sympathetic system can vary in their effects on the different immune compartments. For example, while the inhibitory effects of stress on mitogen induced proliferation, in the spleen is mediated by catecholamines in the blood it is mediated by glucocorticoids (Rabin, Cunnick, and Lysle, 1990). iv. The variation in an immune parameter may be statistically significant between the patients and their controls but still lie within the normal range for immune function. Thus patients may not be immunocompromised from the clinical viewpoint. This could be relevant to the interpretation of the data showing that there is a greater risk of cancer for individuals with high depressive scores (Persky, Kempthone-Rawson, and Shekelle, 1991). However, further examination of the increased mortality rate in patients with affective disorders indicates that this is due to suicide and accidents rather than to causes related to a disorganized immune system (Martin, Cloninger, Guze, and Clayton, 1985). V. It is evident that both stressful and non stressful environmental events can profoundly affect the immune systems. The degree to which stressor can cause suppression of the immune system can be influenced by the perception of the stressful event. Thus there is evidence that the immune system can be classically conditioned (Ader and Cohen, 1991). Animals can learn to immunosuppress, or more dramatically immunoenhance, their T-cell, B-cell, NKC, and most immune cell functions (Dark, Peeke, Ellman, and Salfi, 1987; Solvason, Ghanta, and Hiramoto, 1988). Such factors are frequently ignored in both clinical and experimental studies that seek to explore the relationships between the effects of different types of stress on the immune system.

CONCLUSIONS There are exciting developments relating to the interactions between the immune system and the brain in patients subject of chronic stress or suffering from depression. However, many of the reported studies have concentrated on determining changes in relatively non-specific immune variables in the peripheral blood of those who are otherwise healthy; such findings are therefore conceptuafly of hmited value. The future of psychoneuroimmunology will probably depend on research into the basic physiological processes that underlie the neurendocrine-immune interactions and the relevance of these interactions to the development and outcome of psychiatric illness. Undoubtedly a better understanding of the cytokines, cytokine receptors, and their neuromodulators may lead to the development of novel drugs that have psychotropic actions of value in the treatment of mental illness. This could be a most exciting area of neurobiological research in the next decade.

262

B. E. Leonard and C. Song

REFERENCES Ader, R., Felten, A., & Cohen, N. (1987). Brain, Behavior, Immunity. Brain, Behavior, and Immunity. 1987 (1) 1-6. Ader, R. & Cohen, N. (1991). Conditioning the immune system. Netherlands Journal of Medicine, 39,263-276. Almawi, W. Y., Beyhum, H. N., Rahme, A. A., & Rieder, M. J. (1996). Regulation of cytokine and cytokine receptor expression by glucocorticoids. Journal of Leucocyte Biology, 60,5563-5572. Arber, N., Berhner, S., & Tamir, A. (1991). The state of leucocyte adhesiveness/aggregation in the peripheral blood: a new independent marker of stress. Stress Medicine, 7, 75-78. Baldessarini, R. J. (1975). The basis for amine hypothesis in affective disorder: a critical evaluation. Archives of General Psychiatry, 32,1087-1093. Ballieux, R. E. (1992). Bidirectional communication between the brain and the immune system. European Journal of Clinical Investigation, 22, 1, 6-9. Banks, W. A., Kastin, A. J., & Durham, D. A. (1989). Bidirectional transport of interleukin-1 alpha across the blood brain barrier. Brain Research Bulletin, 23, 77-84. Bateman, A., Singh, A., Krai, T , & Solomon, S. (1989). The immune-hypothalamic-pituitary adrenal axis. Endocrine Reviews, 10, 92-112. Baumann, H. & Gauldie, J. (1994). The acute phase response. Immunology Today, 15, 74-80. Blotta, M. H. (1997). Corticosteroids inhibit IL-12 production in human monocytes and enhance their capacity to induce IL-4 synthesis in CD4+ lymphocytes. Journal of Immunology, 158, 5589-5595. Borysenko, M. & Borysenko, J. (1982). Stress, behaviour and immunity: animal models and mediating mechanisms. General Hospital Psychiatry, 4, 59-67. Bost, K. L. (1988). Hormone and neuropeptide receptors on mononuclear leucocytes. Progress Allergy, 43, 68-83. Cohen, J. J. & Crnic, L. S. (1982). Glucocorticoids, stress, and the immune response. In D. R. Webb (Ed.). Immunopharmacology and the Regulation of Leucocyte Function. New York, Marcel Dekker, pp. 61-91. Connor, T. J. & Leonard, B. E. (1998). Depression, stress, and immunological activation: the role of cytokines in depressive disorders. Life Sciences, 62, 583-606. Cooper, C. L. (1984). The social-psychological pressors to cancer. Journal of Human Stress, 10, 4-11. Cox, T. M. & MacKay, C. (1982). Psychosocial factors and psychophysiological mechanism in the aetiology and development of cancers. Social Science and Medicine, 16, 381-396. Crestani, R, Seguy, R, & Dantzer, R. (1991). Behavioural effects of peripherally injected interleukin-1: role of prostaglandin. Brain Research, 542, 330-335. Croiset, G., Heijnen, C. J., Veldhuis, H. D., de Wied, D., & Ballieux, R. E. (1987). Modulation of the immune response by emotional stress. Life Sciences, 40,775-782. Crook, T. H. & Miller, N. W. (1985). The challenge of Alzheimer's disease. American Psychology, 40, 1245-1250. Cunnick, J. E., Lysle, D. T, Aronfield, A., & Rabin, B. S. (1991). Stressor-induced changes in mitogenic activity are not associated with decreased IL-2 production or changes in lymphocyte subsets. Clinical Immunology and Immunopathology, 60, 419-429. Cunningham, E. T. & de Souza, E. B. (1996). Interleukin 1 receptors in the brain and endocrine tissue. Immunology Today, 14, 171-176. Dantzer, R. (1994). How do cytokines say hello to the brain? Neural versus humoral mediation. European Cytokine Network, 5,271-273. Dantzer, R., Bluthe, R. M., Aubert, A., Goodall, G., Bret-Dibat, J. L., Kent, S., Goujon, E., Laye, S., Parnet, P., & Kelley, K. W. (1996). Cytokine actions on behaviour. In Cytokines in the Nervous System. (Ed. Rothwell NJ), Chapman and Hall, London, 117-140. Dark, K., Peeke, H. V. S., EUman, G., & Salfi, M. (1987). Behaviourally conditioned histamine release. Annals of the New York Academy of Sciences, 496, 578-582. Davidson, L. M. & Baum, A. (1986). Chronic-stress and post traumatic stress disorder. Journal of Consultants in Clinical Psychology, 54, 303-308. Daynes, R. A. & Araneo, B. A. (1989). Contrasting effects of glucocorticoids in the capacity of T cells to produce the growth factors interlukin 2 and interleukin 4. European Journal of Immunolgy. 19, 2319-2325. Dinan, T. (1994). Glucocorticoids and the genesis of depressive illness—a psychobiological model. British Journal of Psychiatry, 164, 365-371. Dohmus, J. E. & Metz, A. (1991). Stress mechanisms of immunosuppression. Veterinary Immunology and Immunopathology, 30, 89-109.

Stress, Depression, and the Role of Cytokines

263

Farrar,W. L. (1988). Evidence for the common expression of neuroendocrine hormones and cytokines in the immune and central nervous system. Brain Behaviour and Immunity, 2, 322-327. Fiore, J., Becker, J., & Cooppel, D. B. (1983). Social network interactions: a buffer or a stress? American Journal of Communications in Psychology, II, 423-429. Fleshner, M., Bellgrau, D., Watkins, L. R., Laudenslager, M. L., & Maier, S. F. (1995). Stress induced reduction in the rat mixed lymphocyte reaction is due to macrophages and not to changes in T cell phenotypes. Journal of Neuroimmunology, 56, 45-52. Glaser, R., Kiecolt-Glaser, J. K., & Stout, J. C. (1985). Stress related impairments in cellular immunity. Psychiatric Research, 16, 233-239. Glaser, R., Rice,T., & Speicher, C. E. (1986). Stress depresses interferon production by leucocytes concomitant with a decrease in natural killer cell activity. Behavioural Neuroscience, 100, 675-678. Gottesfeld, A. & Liehr, J. G. (1987). Chronic exposure to random restraint stress retards the development of oestrogen-induced pituitary prolactinoma in rats. Neuroscience Letters, 80, 44-48. Gutierrez, E. G., Banks, W. A., & Kastin, A. J. (1993). Murine tumour necrosis factor alpha is transported from blood to brain in the mouse. Journal of Neuroimmunology, 47,169-176. Hambhn, A. S. (1994). Cytokines. In Textbook of Immunopharmacology. 3rd Edition. M. M. Dale, M. M. J. C. Foreman,!. D. Fan (Eds.), Blackwell Scientific Publication, Oxford, pp. 179-192. Healy, D., Calvin, J., Whitehouse, A. M., White, W., Wilton-Cox, H., Theodorou, A. E., Lawrence, K. M., Hawton, R. W, & Paykel, E. S. (1991). Alpha-1-acid glycoprotein in major depressive and eating disorders. J. Affective Disorders (22), 13-20. Hellerstein, M. K., Meydani, S. N., Meydani, M., Wu, K., & Dinarello, C. A. (1989). Interleukin-1 induces anorexia in the rat. Journal of Clinical Investigations, 84, 28-235. Hopkins, S. J. & Rothwell, N. J. (1995). Cytokines in the nervous system I: Expression and recognition. Trends in Neuroscience, 18, 83-88. Irwin, M. (1995). Psychoneuroimmunology and depression In Psychopharmacology. The Fourth Generation of Progress. (Eds. Bloom, FE, Kupfer, DJ). Raven Press, New York, pp. 983-998. Irwin, M., Lacher, U. B., & Caldwell, C'. (1992). Depression and reduced natural killer cytotoxicity: a longitudinal study of depressed patients and control subjects. Psychological Medicine, 22, 1045-1050. Katila, H., Rimon, R., Cantwell, K., Appelberg, B., & Nikkila, H. (1991). Interferon productions in acute psychiatric disorder. Psychiatric and Biological Factors, 16,191-196. Keller, S. E., Weiss, J. M., Schleifer, S. J., Miller, N. E., & Stein, M. (1981). Suppression of immunity by stress: effect of a gradient series of stressors on lymphocyte stimulation in the rat. Science, 213,1397-1400. Kiecolt-Glaser, J. K., Garner, W., & Speicher, C. E. (1984). Psychosocial modifiers of immune competence in medical students. Psychosomatic Medicine, 46, 7-14. Kiecolt-Glaser, J. K., Fisher, L., & Ogrocki, P. (1987). Marital quality, marital disruption, and immune function. Psychosomatic Medicine, 49, 13-34. Kiecolt-Glaser, J. K. & Glaser, R. (1988). Methodological issue in behavioural immunology research in humans. Brain Behaviour and Immunity, 2, 67-78. Kiecolt-Glaser, J. K., Dura, J. R., Speicher, C. E.,Trask, O. J., & Glaser, R. (1991). Spousal caregivers of dementia victims: longitudinal changes in immunity and health. Psychosomatic Medicine, 53, 345-362. Komaki, G., Arimura, A., & Koves, K. (1991). Effect of intraveous injection of IL-IB on PGE2 levels in several brain areas as determined by microdialysis. Endocrinology and Metabolism, 143, 220-227. Kronfol, Z. & House, J. D. (1985). Depression, hypothalamic pituitary adrenal cortical activity and lymphocyte function. British Journal of Psychiatry, 148, 70-73. Kronfol, Z. & House, J. D. (1989). Depression, HPA activity, and lymphocyte function. >lcfa Psychiatrica Scandinavica, 80, 142-147. Lavicky, J. & Dunn, A. J. (1995). Endotoxin administration stimulates cerebral catecholamine release in hypothalamus and prefrontal cortex in freely moving rats as assessed by microdialysis. Journal of Neuroscience Research, 40, 407^13. Leonard, B. E. (1987). Stress, the immune system and mental illness. Stress Medicine, 3, 257-258. Leonard, B. E. (1990). Stress and the immune system: immunological aspects of depressive illness. International Reviews of Psychiatry, 2, 321-330. Levy, S., Herberman, R., Lippman, M., & d'Angelo,T. (1987). Correlation of stress factors with sustained suppression of NKC activity and predicted prognosis in patients with breast cancer. Journal of Clinical Oncology, 5, 348-353. Linkowski, P., Mendlewicz, J., & Kerklofs, M. (1987). 24hr profiles of ACTH, Cortisol, and growth hormone in major depressed illness: effect of AD treatment. Journal of Clinical Endocrinology and Metabolism, 65,141-146. McAdams, C. & Leonard, B. E. (1993). Neutrophil and monocyte phagocytosis in depressed patients. Progress in Npuronsvrhonharmaroloev

and Biological

Psvchiatrv.

17. 971-984.

264

B. E. Leonard and C. Song

Maes, M., Smith, R., & Scharpe, S. (1995). The monocyte-T-lymphocyte hypothesis of major depression. Psychoneuroendocrinology, 20, 111-116. Maes, M., Planken, V. D., & Stevens, W. J. (1992). Leucocytosis, monocytosis, and neutrophiha: hallmarks of severe depression. Journal of Psychiatric Research, 261, 125-134. Maes, M., Smith, R. S., Christophe, A., Cosyns, P., Desnyder, R., & Melzer, H. Y. (1996). Fatty acid composition in major depression: decreased omega 3 fractions in cholesteryl esters and increased C20: 4W6/C20:5W3 ratio in cholesteryl esters and phosphohpids. Journal of Affective Disorders, 38, 35-46. Maes, M. & Smith, R. S. (1998). Fatty acids, cytokines, and major depression. Biological Psychiatry, 43, 313-314. Maier, S. R, Watkins, L. R., & Fleshner, M. (1994). Psychoneuroimmunology—the interface between behaviour, brain, and immunity. American Psychologist, 49, 1004-1017. Martin, R. C , Cloninger, R., Guze, S., & Clayton, P. (1985). Mortality in a follow-up to 500 psychiatric outpatients. I. Total mortality and II cause specific mortality. Archives of General Psychiatry, 47-54 and 58-66. Mason, J. W. (1971). A re-evaluation of the concept of "non-specificity" in stress theory. Journal of Psychiatric Research, 8,123-140. Mefford, I. N. & Heyes, M. P. (1990). Increased biogenic amine release in mouse hypothalamus following immunological challenge: antagonism by indomethacin. Journal of Neuroimmunology, 27, 55-62. Miller, A. H., Asmis, G. M., & Lackner, C. (1991). Depression, natural killer cell activity and Cortisol secretion. Biological Psychiatry, 29, 878-886. Miller, A. H., Spencer, R. L., McEwen, B. S., & Stein, M. (1993). Depression, adrenal steroids, and the immune system. Annuals of Medicine, 25, 481-487. Monjan, A. A. & Collector, M. I. (1977). Stress-induced modulation of immune response. Science, 196, 307-308. Mormede, P., Dantzer, R., Michaul, B., Kelly, K., & Le Maoal, M. (1988). Influence of stressor predictability and behavioural control on lymphocyte reactivity, antibody response and neuroendocrine activation in rats. Physiology and Behaviour, 43, 577-583. Munck, A. & Naray-Fejes-Toth, A. (1994). Glucocorticoids and stress: permissive and suppressive actions. Annals of the New York Academy of Sciences, 746, 115-130. Nathan, C. F. (1987). Secretory products of macrophages. Journal of Clinical Investigation, 79, 319-324. Nemeroff, C. B., Widlerlov, E., & Bissette, G. (1984). Elevated concentration of CSF-CRF-like immunoreactivity in depressed patients. Science, 226,1342-1348. O'Neill, B. & Leonard, B. E. (1990). Abnormal zymosan-induced neutrophil chemiliuminescence as a marker of depression. Journal of Affective Disorders, 19, 265-272. Owens, M. J. & Nemeroff, C. B. (1991). Physiology and pharmacology of corticotropin releasing factor. Pharmacological Reviews, 43, 425-473. Persky,V.W., Kempthone-Rawson, J., & Shekelle, R. B. (1991). Personality and risk of cancer: 20 year followup of the Western Electric Study. Psychosomatic Medicine, 49, 435^47. Rabin, B. S., Cunnick, J. E., & Lysle, D. T (1990). Stress induced alteration of immune function. Progress in Neuroendocrinology and Immunology, 3, 116-124. Ramamoorthy, S., Ramamoorthy, J. D., Pradad, R, Bhat, G. K., & Mahesh, V. B. (1995). Regulation of the human serotonin transporter by interleukin-1 beta. Biochemical Biophysics Research Communications, 216, 560-567. Rasmussen, A. F. (1969). Emotions and immurniy. Annals of the New York Academy of Sciences, 764,458-^62. Rivier, C. & Rivest, S. (1993). Mechanisms mediating the effects of cytokines on neuroendocrine functions in the rat. In Corticotrophin-releasing factor. W. W. Vale (Ed.), Ciba Foundation Symposium, 172, John Wiley & Sons, Chichester, UK. pp. 204-225. Schettini, G. (1990). Interleukin in the neuroendocrine system: from gene to function. Progress in Neuroendocrinimmunology, 3, 157-166. Schleifer, S. J., Keller, S. E., Siris, S. G., Davis, K. L., & Stein, M. (1985). Depression and Immunity: lymphocyte stimulation in ambulatory depressed patients. Archives of General Psychiatry, 42,129-133. Schleifer, S. J., Keller, S. E., Bond, R. N., Cohen, J., & Stein, M. (1989). Major depressive disorder: role of age, sex, severity, and hospitalization. Archives of General Psychiatry, 46, 81-87. Shekelle, R. B., Raynor, W. J., & Ostfeld, A. M. (1981). Psychological depression and 17 years risk of death from cancer. Psychosomatic Medicine, 43, 117-125. Sklar, L. S. & Anisman, H. (1981). Stress and cancer. Psychological Bulletin, 89, 369^06. Smith, R. S. (1991). The macrophage theory of depression. Medical Hypothesis, 35, 298-306. Solomon, G. F. (1969). Stress and antibody response in rats. International Archives of Allergy and Applied Immunology, 35, 97-104.

Stress, Depression, and the Role of Cytokines

265

Solvason, H. B., Ghanta, Y. K., & Hiramoto, R. N. (1988). Conditioned augmentation of Natural Killer Cell activity: independence from nociceptive effects and dependence on interferon beta. Journal of Immunology, 140, 661-665. Song, C , Dinan, T, & Leonard, B. E. (1994). Changes in immunoglobulin, complement and acute phase protein concentrations in depressed patients and normal controls. Journal of Affective Disorders, 30, 283-288. Tejwani, G. A., Gudehithlu, K. P., Hanissian, S. H., Gienapp, I. E., Whitacre, C. C , & Malarkey, W. B. (1991). Facilitation of dimethylbenaz (a) anthracine induced rat mammory tumorigenesis by restraint stress: role of beta endorphin, prolactin, and naltrexone. Carcinogenesis, 12, 637-641. Weisse, C. S. (1992). Depression and immunocompetence: a review of the literature. Psychological Bulletin, / 77,475-487. Wiegers, G. J. & Reul, J. M. H. M. (1998). Induction of cytokines receptors by glucocorticoids: functional and pathological significance. Trends in Pharmacological Sciences, 19, 317-321. Wiegers, G. J., Labeur, M. S., Stec, I. E., Klinkert, W. E., Holsboer, E, & Reul, J. M. (1995). Glucocorticoids accelerate anti- T cell receptor induced T cell growth. Journal of Immunology, 155, 1893-1902. Wu, C. Y., Fargeas, C , Nakajima, T., & Delespesse, G. (1991). Glucocorticoids suppress the production of interleukin 4 by human lymphocytes. European Journal of Immunology, 10, 2645-2647. Wu Sarfati, M., Heusser, C , Fourneir, S., Rubio-Trujillo, M., Peleman, R., & Delespesse, G. (1991). Glucocorticoids increase the synthesis of immunoglobulin E by interleukin 4-stimulated human lymphocytes. Journal of Clinical Investigations, 87, 870-877. Xia, Z., De Piere, J. W, & Nassberger. L. (1996). TCA's inhibit IL-6, IL-1 and TNF release in human blood monocytes and IL2 and interferon in T cells. Immunopharmacology, 34, 27-37.

15

IS THERE EVIDENCE FOR AN EFFECT OF ANTIDEPRESSANT DRUGS ON IMMUNE FUNCTION? Pierre J. Neveu and Nathalie Castanon Neurobiologie Integrative, INSERM U 394, Institut Francois Magendie rue Camille Saint-Saens, 33077 Bordeaux, France

1. INTRODUCTION Depression is a complex disease which is likely to involve several pathophysiological pathways. There is clear evidence that depression is associated with neurochemical and neuroendocrine alterations. Reduced activity of the serotoninergic (5-HT) and noradrenergic (NA) central systems are observed in a majority of patients with major depression (Garver & Davis, 1979). Depressed patients usually also exhibit an alteration of the hypothalamus-pituitary-adrenal (HPA) axis activity characterized by an hyperproduction of corticotropin-releasing hormone (CRH), which stimulates adrenocorticotropic hormone (ACTH) and Cortisol release (Holsboer, Bardeleben, Gerken, Stalla, & Muller, 1984). Therefore, the biochemical activity of most antidepressants, including selective 5-HT reuptake inhibitors, monoamine oxidase inhibitors, and tricyclic antidepressants, has been assessed on the basis of their ability to reverse the alterations of monoamine and/or HPA axis activities (HoUister, 1986). However, the metabohc activity of these drugs is not necessarily related directly to their clinical efficacy (Harden, Reul, & Holsboer, 1995; Blier & de Montigny, 1994). Despite repeated attempts, the neuro-hormonal abnormalities observed in depression have never been shown to predict therapeutic response, nor can they account for the symptomatic profile of the patients. Furthermore, depletion of 5-HT or NA in healthy individuals does not induce clinically significant depressive symptomatology (Young, Smith, Pihl, & Erwin, 1985). In addition, there are also some atypical antidepressants with known experimental and clinical therapeutic effects, but devoid of the classic antidepressant actions on central monoamine activity (Guelfi, 1992; Van Riezen & Leonard, 1990). Because of the growing interest for brain-immune interactions, many authors Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

267

268

P. J. Neveu and N. Castanon

have also looked for a possible defect of immune responsiveness in depression (Anderson, 1996; Stein, Miller, & Trestman, 1991), and considerable attention is now paid to the possible role of immunological dysregulation in the pathogenesis of this disease. For decades, the neuro-hormonal changes observed in depression were believed to be exclusively associated with an immunosuppression, especially because of the well known inhibitory influence of glucocorticoids on the immune system (Munck, Guyre, & Meltzer, 1984), as well as the regulatory role of central monoamines on immune activity (see for review Deleplanque & Neveu, 1995). As detailed below, there is both clinical and experimental evidence that various aspects of the immune system, including mitogen-induced cell proliferation, natural killer cell activity, and variations of lymphocyte subsets, are severely compromised in depression (Irwin, Daniels, Bloom, Smith, & Weiner, 1987; Irwin, Patterson, Smith, Caldwell, Brown, Gillin, & Grant, 1990). However, recent studies have provided some evidence that major depression may be also accompanied by an immune activation in the form of an acute phase reaction mediated by the synthesis and release of proinflammatory cytokines (mainly interleukin-1: IL-1, interleukin-6: IL-6, tumor necrosis factor: TNF) by activated macrophages (Connor & Leonard, 1998; Maes, 1995; Maes, Smith, & Scharpe, 1995; Smith 1991; Ur, White, & Grossman, 1992). Interestingly, these results were proposed by some authors to provide a possible explanation for the neuroendocrine abnormaUties observed in depressed patients, since cytokines, i.e. the key factors for immune activation, exert important changes in the HPA axis activity (Besedovsky & Del Rey, 1996). By extension, it has also been suggested that the immune system, and more specifically cytokines, may play a role in the onset of depressive symptoms.

2. C Y T O K I N E - I N D U C E D D E P R E S S I V E - L I K E S Y M P T O M S Interestingly, a few recent studies appear to strengthen the hypothesis of a causal relationship between activation of the immune system and depression, since immunological activation can induce depressive-like symptoms in humans, as well as in laboratory animals. In humans, proinflammatory cytokines (e.g. interferon alpha, IFN or TNFa) induced a depressed mood and other symptoms found in major depressive episodes (McDonald, Mann, & Thomas, 1987). After a viral infection by influenza, some individuals also show depressive symptoms, including depressed mood, reduced appetite, sleep disturbances, and sense of guilt (Meijer, Zakay-Rones, & Morag, 1988). It has been also reported that some patients, infected with hepatitis C and treated with IFNa, develop depression, that can be successfully treated with the antidepressant fluoxetine (Levenson & Fallon 1993). Moreover, diseases during which macrophage activation is known to occur, for example rheumatoid arthritis or atherosclerosis, have high rates of depression (Mitchinson & Ball, 1987). For example, 41% of patients with multiple sclerosis treated with IFN-ip, self-report new or increased depression within 6 months of initiating IFN-therapy (Mohr, Goodkin, Likosky, Gatto, Baumann, & Rudick, 1997). In fact, depressive symptoms are exceedingly common in a wide range of diseases (Katon & Sullivan, 1990). However, it is not known whether depression is associated with illness severity or whether it is directly linked to the cytokine concentrations measured in these patients. An answer to this question may give some additional insights about the real role of cytokines in depression. These clinical observations have been confirmed by experiments performed in rodents. Injection of lipopolysacharide (LPS), the active fragment of endotoxin from

Is There Evidence for an Effect of Antidepressant Drugs on Immune Function?

269

Gram-negative bacteria, or injections of proinflammatory cytokines whose synthesis and release are induced by this bacterial fragment, affect the 5-HT and NA neurochemical systems (Delrue, Deleplanque, Rouge-Pont, Vitiello, & Neveu, 1994; Dunn, 1992; Linthorst, Flachskamm, Holsboer, & Reul, 1995; Zalcman, Green-Johnson, Murray, Nance, Dyck, Anisman, & Greenberg, 1994). Cytokines also induce an activation of the HPA axis (Besedovsky & Del Rey, 1996), comparable to the activation observed in response to stress and in depressed patients. Interestingly, pretreatment with interleukin-1 receptor antagonist (IL-lra) blocked acute stress-induced HPA axis activation and in vivo monoamine changes in the hypothalamus (Shintani, Nakaki, Kanba, Sato, Yagi, Shiozawa, Aiso, Kato, & Asai, 1995). IL-lra also prevents the development of the behavioral deficits seen in the learned helplessness model of depression in rats (Maier & Watkins, 1995). Cytokine administration also elicits a large set of behavioral changes, that is known as "sickness behavior", and includes lethargy, anorexia, adipsia, reduced social activities and interest in the environment (Kent, Bluthe, Kelley, & Dantzer, 1992), i.e. parameters considered as depressive-like symptoms. It is noteworthy that some of these LPS-induced behavioral changes are attenuated or completely blocked by chronic treatment with the tricyclic antidepressant imipramine (Yirmiya, 1996). In addition, in vitro, incubation of monocytes with clomipramine, imipramine or citalopram markedly inhibits of LPS-induced interleukin (IL) IL-ip, TNF and IL-6 production. IL-2 and IFN-y release from T-cells is also inhibited by preincubation with these antidepressants (Xia, De Pierre, & Nassberger, 1996). These interesting results raise the question of the effects of antidepressants on cytokine production and/or function. However, two important issues need to be pointed out before proceeding. First, it should be noted that in all animal studies described above, depression-like symptoms are observed only after an acute LPS or cytokine treatment, whereas only experiments with chronic LPS or cytokine administration should be considered in relation to depression-like situation. Second, the behavioral effects observed after LPS administration are mediated by central cytokines, and this is not necessarily accompanied by a significant change in peripheral levels of cytokines (Kent et al., 1992). However, in depressed patients only peripheral cytokines are measured. This point should be kept in mind to account for the apparent discrepancy we will see about cytokine production and both severity of depression and Cortisol levels in depressed patients.

3. IMMUNE ALTERATIONS IN DEPRESSED PATIENTS Until now, very few investigations have been carried out to assess the effects of antidepressant medications on the immune alterations associated with depression and more specifically to assess whether antidepressants normalize immunity and/or abrogate the production and release of proinflammatory cytokines. In order to answer these questions, it is first necessary to depict the immune status that is usually observed during depression. The immune parameters studied in depressed patients usually include mitogen-induced lymphoproliferation, natural killer cell activity, peripheral blood leukocyte counts and cytokine plasma levels. Lymphoproliferation induced by phytohemagglutinin (PHA), concanavalin A (ConA) or pockweed mitogen (PWM) has been usually found to be decreased in depressed patients. For example, Kronfol, Silva, Greden, Dembinski, Gardner, & Carroll (1983) found that this lymphoproliferation was reduced in depressed in-patients

270

P. J. Neveu and N. Castanon

compared to healthy out-patient controls, although this reduction was not correlated with severity of depression as reflected by Hamilton scores. Despite the well known inhibitory action of glucocorticoids on immune activity, the decrease in lymphoproliferation was not related to the HPA axis hyperactivity (Cosyns, Maes, Vandewoude, Stevens, De Clerk, & Schottel, 1989; Kronfol & House 1985), as assessed by the dexamethasone suppression test (DST). Indeed, depression of lymphoproliferation was equally found in both DST positive and DST negative patients (Albrecht, Helderman, Schlesser, & Rush, 1985). However, others authors (Maes, Bosmans, Suy, Minner, & Raus, 1989; Maes, Bosmans, Suy, Vandervorst, Dejonckheere, Minner, & Raus, 1991) did find an inverse correlation between lymphocyte proliferation and Cortisol levels. The decrease of mitogenesis has also been found to be inversely correlated to the turn-over of norepinephrine and the age of the patients (Maes et al., 1989). Likely, Schleifer, Keller, Bond, Cohen, & Stein (1989) described a reduced mitogenesis in old but not in young patients. In fact, an increase in lymphoproliferation induced by optimal doses of PHA has been even reported in young patients (Altshuler, Plaeger-Marshall, Richeimer, Daniels, & Baxter, 1989). A small decrease of mitogenesis induced by a low dose of ConA was observed in a group of depressed patients, whereas lymphoproliferation and IL-2 production induced by PHA remained unchanged, eventhough, there was no change in plasma levels of Cortisol, ACTH, growth hormone, and prolactin (Darko, Lucas, Gilhn, Risch, Golshan, Hamburger, Silverman, & Janowsky, 1988). Finally, other studies showed normal mitogenesis induced by ConA, PHA, or PWM irrespective of the severity of the disease (Albrecht et al., 1985). Classically, natural killer cell (NK) activity is found to be reduced in major depression (Irwin et al., 1987; 1990). Nerozzi, Santoni, Bersani, Magnani, Bressan, Pasini, Antonozzi, & Frajese (1989) found that the slightly reduced NK activity measured in depressed patients was not inversely correlated with plasma Cortisol. It has also been shown that NK activity was unchanged in male depressed patients but enhanced in depressed women. This increase was not correlated with Cortisol levels nor the severity of depression (Miller, Asnis, Lackner, Halbreich, & Norin (1991). As pointed out by Schleifer et al. (1989), such a variability of results could be related to age, severity of depression, and hospitalization. In depressed patients, the percentage of neutrophils was increased and the percentage of lymphocytes was decreased, and these modifications were correlated with plasma levels of Cortisol (Kronfol & House, 1989). Furthermore, these authors observed a decrease of CD3 and CD4 positive cells. Another longitudinal study showed that the number of leukocytes, granulocytes, eosinophils, basophils, and thrombocytes was elevated over the 6 week-period study (Seidel, Arolt, Hunstiger, Rink, Behnisch, & Kirchner, 1996). Sub-classes of lymphocytes were similar to those of controls, but macrophage counts, high at the onset, slightly decreased with time. Schleifer, Keller, Bartlett, Eckholdt, & Delaney (1996) reported that in young adults with major depression, there was an increase in granulocytes, a decrease in NK cells, but no modification in the percentage of monocytes, and T, CD4+, CD8-t-, and B lymphocytes. In this study no correlation was found between immune alterations and the gravity of depression. In aged depressed women, the number of T lymphocytes and helper T cells was decreased compared to aged controls, whereas the percentage of B cells and suppressor T cells was unaffected (Targum, Marshall, Ficshman, & Martin, 1989). Interestingly, these alterations were more pronounced in DST resistant patients, and may be related to modifications in cell distribution induced by high level of circulating Cortisol. By contrast, no difference in lymphocyte subsets, including CD3+, CD4-I-, CD8+ was found.

Is There Evidence for an Effect of Antidepressant Drugs on Immune Function?

271

whereas an increased CD4/CD8 ratio was observed in patients with endogenous depression. Interestingly, this ratio increase was correlated with the score of anxiety but not with the plasma level of Cortisol (Charles, Machowski, Brohee, Wilmotte, & Kennes, 1992). On the contrary, Darko et al. (1988) only found an increase of CD4+ cells in depressed patients, with no modification in the percentage of T, B, CD8+ cells, and NK cell activity. Finally, other authors, measuring the percentage of lymphocyte subsets in blood, concluded that depression is associated with an immune activation instead of a suppression (Maes, Lambrechts, Bosmans, Jacobs, Suy, Vandervorst, De Jonckheere, Minner, & Raus, 1992; Perini, Zara, Carraro, Tosin, Cava, Santucci, Valverde, & De Franchis, 1995). In fact, the signs of immune activation were reported to be more pronounced in patients with melancholic-type depression than in patients with other types of major depression or with minor depression. More recently it has been proposed, in the light of the monocyte-T-lymphocyte hypothesis of major depression, that activation of the HPA axis and brain metabolic alterations may result from an hyperproduction of cytokines (for a review, see the chapter by Maes in this volume). Monocyte phagocytosis was enhanced during the acute phase of depression and returned to normal values with recovery. These changes in phagocytosis were not due to hypercortisolemia or to the direct effects of antidepressant drugs. Neutrophil phagocytosis was depressed during the acute phase of depression and returned to normal values with recovery (McAdams & Leonard, 1993). There are some data in favor of an increase of plasma levels of inflammatory cytokines in depressed patients. High levels of acute phase proteins (Maes, Bosmans, Meltzer, Scharpe, & Suy, 1993a) were observed, together with increased number of IL-2 receptor-bearing cells and levels of soluble IL-2 receptors (Maes et al., 1991). These studies also revealed, in patients with major depression, a positive correlation between mitogen-stimulated production of IL-ip by peripheral blood mononuclear cells and post-DST Cortisol values. Moreover, the Cortisol non-suppressor patients showed higher IL-ip production than the Cortisol suppressor ones. However, the correlation between IL-ip production and post-DST Cortisol levels was also found in normal controls (Maes et al., 1993a). In a controlled longitudinal study, induced levels of several cytokines (IL-1 p, IL-2, IL-10, and IFNy) were measured in whole blood of unmedicated depressed in-patients. On admission, the patients had higher levels of cytokines in the supernatants of mononuclear blood cells stimulated by PHA, but the levels decreased over a 6-week period (Seidel, Arolt, Hunstiger, Rink, Behnisch, & Kirchner, 1995). However, the differences were not very large and substantial interindividual differences were observed. IL-ip and IL-6 levels were higher than in controls, but not significantly, and decreased thereafter. Similarly, the non-significant increase of IL-2, IFNy, and IL-2R dissipated with time. Likewise, the level of IL-10 was lower at day 43 as compared to day 1. These results clearly show that the production of both proinflammatory and anti-inflammatory cytokines may be enhanced during the onset of a depressive episode (Seidel et al., 1995). Even though this enhanced IL-6 production was observed in DST-suppressor patients, there was a correlation between IL-6 production and plasma levels of post-DST Cortisol (Maes, Scharpe, Meltzer, Bosmans, Suy, Calabrese, & Cosyns, 1993b). Likewise, plasma levels of IL-6 were higher in the acute state of depression and similar to those of controls after remission (Frommberger, Bauer, Haselbauer, Fraiilin, Riemann, & Berger, 1997). By contrast, IL-1 (3, IL-2, and IL-3 were found to be lowered in major depression (Weizman, Laor, Podliszewski, Notti, Djaldetti, & Bessler, 1994). In another report, plasma levels of IFNy and in vitro production of TNF were studied in DST-responder patients with unipolar depression and long-lasting disease.

272

P. J. Neveu and N. Castanon

The values measured before or after a DST were similar to those observed in controls (Landmann, Schaub, Link, & Wacker, 1997). In accordance with the findings of increased production of cytokines during depression, plasma levels of acute phase proteins, including C reactive protein, haptoglobin, a2 macroglobulin, al-antitrypsine, and ceruloplasmin were higher in depression, and this increase was related to the disease severity (Maes et al., 1992; Seidel et al., 1995). In another study, some acute phase proteins Uke haptoglobin and al-antichimotrypsin were also found to be increased, whereas others like transferrin, al-antitrypsin, ceruloplasmin, C reactive protein or al-acid glycoprotein remained unchanged (Joyce, Hawes, Mulder, Sellman, Wilson, & Boswell, 1992). Plasma levels of haptoglobin were increased in major depression, associated or not with melancholia, but not in minor depression, whereas levels of transferrin were decreased in all types of depression. These variations were correlated with the production of IL-6 (Maes et al., 1993b). However, the increased plasma levels of IL-6 observed in acute depression was not correlated with age, severity of the disease, and concentrations of C reactive protein (Frommberger et al., 1997). The discrepancies observed in the enhanced production of cytokines during the acute phase of depression may result from the fact that this increase is not observed in all patients. An increase of plasma level of IL-6, associated with an increase of acute phase proteins and monocyte counts was observed only in 6 out of 22 patients. These increases were not correlated with the severity of the disease (Sluzewska, Rybakowski, Sobieska, Bosmans, Pollet, & Wiktorowicz, 1995b). The increase of plasma IL-6 associated with increase in plasma levels of a-1-acid glycoprotein and C reactive protein was observed in patients resistant but not in patients sensitive to antidepressant treatment. Furthermore, the increase of a-l-acid glycoprotein was higher in DST non-suppressor patients (Sluzewska et al., 1995b). The increase of cytokines during depression may have an adaptive value as suggested by the data of Bauer, Hohagen, Gimmel, Bruns, Lis, Krieger, Ambach, Guthmann, Grunze, Fritsch-Montero, Weissbach, Ganter, Frommberger, Riemann, & Berger (1995). In patients with major depression, administration of endotoxin induced an increase of TNF and IL-6 plasma levels, as well as a higher expression of IL-ip mRNA in LPS-stimulated mononuclear blood cells. Interestingly, these cytokine changes were associated with an increase in body temperature, a transient suppression of REM sleep, and an improvement of depression, even though this improvement was only transient and usually followed by a relapse. Moreover, a correlation was observed between maximal plasma concentrations of IL-6 reached after endotoxin adminstration and the decrease of Hamilton scores. From the results summarized above, it appears that no clear conclusion can be drawn on the immunological profile of depressed patients. As already pointed out, this reflects some methodological problems in the psychoimmunological research on depression. The investigation of the immune system in depression has been primarily embedded in the conceptual framework of the role of immune function in the maintenance of health and the development of physical disease. The functional measures of the immune system used in the studies of depressive disorders have been based on in vitro correlates of immune system activity. It has not been established, however, that mitogen-induced lymphocyte proliferation is related to in vivo immune responses as might be expected in response to infections or tumors. There is some evidence suggesting a relationship between NK activity and in vivo viral infections. However, no study has demonstrated concomitantly, i.e. in the same individuals, depression-related alterations in NK activity, mitogenic responses, cytokine production

Is There Evidence for an Effect of Antidepressant Drugs on Immune Function?

273

or any immune measures for that matter, and changes in health and illness, thereby limiting any causal inferences. Finally, the observations that depression is associated with an immune suppression or an immune stimulation appear to be contradictory. This results is due, in part, to a misuse of the terms of suppression and stimulation concerning the immune system as a whole. Indeed, any single immune measure cannot be used as a marker for immune responsiveness, since different immune parameters may change in opposite directions during an immune response. In depression, an enhanced production of cytokines appears to be associated with a decrease in mitogen-induced lymphoproliferation and NK cell activity. Studies with experimental animals favor this hypothesis. For example, an intraperitoneal injection of LPS induces both an increase of cytokine production in the periphery and in the brain (Laye, Parnet, Goujon, & Dantzer, 1994) and an inhibition of mitogen-induced lymphoproHferation (Delrue et al., 1994). Likely, IL-1 injected intracerebroventricularly decreases NK cell activity (Sundar, Cierpial, Kilts, Ritchie, & Weiss, 1990). These effects of cytokines appear to result from the activation of the HPA axis. On the clinical side, the diagnosis of major depression possibly includes different pathophysiological subtypes of depressive illness that show different immune alterations, which may be epiphenomena or contribute to pathophysiological processes. It has been proposed that an immune suppression may occur in psychogenic types of depression and an immune activation in endogenous types (Muller, 1995). Furthermore, differences are mainly marked in studies of older patients and those hospitalized for more severe disorders, suggesting that age and/or severity of depression are independent determinants of impaired immunity, or that a higher-order central nervous system disturbance is a cause of both the endogenous subtype and immune dysfunctions (Hickie, 1990). Future studies concerned with psychological influences, such as depression, on immunocompetence should use more clinically relevant and specific immune measures and/or the actual disease end points (Stein et al., 1991).

4. IMMUNE EFFECTS OF ANTIDEPRESSANTS The immune effects of antidepressants can be studied using various protocols, including study of immune parameters in treated patients or in animal models of depression, although examples of this last approach are really scarce. In patients, the improvement of immune alterations may be directly related to the effects of antidepressants, but also to the improvement of the disease state. The immune effects of antidepressants can also be studied in normal animal or in vitro experiments.

4.1. Immune Effects of Antidepressants in Depressed Patients Very few investigations have been carried out to assess the effect of antidepressant medication on immune parameters. When comparing various immune parameters, including percentage of T and B lymphocytes and mitogenesis, to psychological markers in two groups of treated and untreated patients, there was no clear relationship between psychological and immune parameters (Udelman & Udelman, 1985). However, some observations in the literature indicate that in patients treated with antidepressants, a normalization of immune parameters occured; increased immune parameters were lowered whereas depressed immune parameters were restored (Table 1). Mitogenesis induced by ConA or PHA was reduced in depressed patients treated

274

P. J. Neveu and N. Castanon Table 1. Effects of antidepressants on immune functions in depressed patients

Drugs

Immune alterations in depressed patients

Effects of antidepressants

TCA TCA

Normal mitogenesis Increased monocyte counts

i i

Moclobemide (MAOI)

Normal expression of antigen class II, CD 14,TNF production Increased plasma levels of IL-6

-^

Albreicht et al., 1985 Seidel et al., 1995, 1996 Landmann et al., 1997

i

Sluzewska et al., 1995

Fluoxetine (SSRI)

(in 6 out of 22 patients) Clomipramine (TCA) Imipramine (TCA)

Reduced cytokine production (IL-lp, IL-3, IL-2) HIV infection and major depression

T

Weizman et al., 1994

i

Rabkin and Harrisson, 1990

decline of T4

Abbreviations: TCA: tricyclic antidepressants. MAOI: monoamine oxydase inhibitors. SSRI: selective serotonin reuptake inhibitors.

with tricyclic antidepressants, whereas it was normal before treatment. It should be noted that in this study, depression of mitogenesis was also observed after electroconvulsive therapy (Albrecht et al., 1985). The elevation of monocyte counts observed at the onset of a depressive episode decreased after six weeks of treatment with tricyclics, and this decrease was associated with clinical improvement (Seidel et al., 1996). In contrast, when monocyte functions, such as class II and CD14 antigens expression, TNF production and plasma IFNy, were found to be unaltered at the onset of depression, these functions remained normal after a 4-12 week moclobemide therapy (Landmann et al., 1997). It has also been reported that the increased plasma levels of IL-6 observed during acute depression was normalized after an 8-week period of fluoxetine treatment. However, this normalization was observed only in 6 out of 22 patients (Sluzewska, Rybakowski, Laciak, Mackiewicz, Sobieska, & Wiktorowicz, 1995a). As described previously, the use of cytokines to treat some diseases, such as IFN-ip for multiple sclerosis, could be associated with the onset of a depressive episode (Mohr et al., 1997). Antidepressant medication as well as psychotherapy improve adherence of patients to IFNp therapy (Mohr et al., 1997). It has been suggested that tricyclic antidepressants may be successful in alleviating the neuropsychiatric disturbances induced by INFa therapy including depression (Goldman, 1994). In a recent study in ten patients with major depression, a 4-week treatment with clomipramine, a tricyclic antidepressant, increased the production of IL-lp, IL-3, and to a lesser extent IL-2, by peripheral blood mononuclear cells in depressed patients presenting reduced cytokine production before treatment (Weizman et al., 1994). Cytokine production before and after drug treatment did not correlate with severity of depression as assessed by the Beck Depression Inventory (Weizman et al., 1994). In addition, lithium treatment induced in breast cancer patients an increase of TNFa production (Merendino, Mancuso, Tomasello, Gazzara, Cusumano, Chillemi, Spadaro, & Mesiti, 1994), and this increase could be confirmed in vitro. Furthermore, a preliminary report in a small group of homosexual men with HIV infection and major depression suggests that treatment with imipramine mght minimize the decUne in T4 cell count (Rabkin & Harrison, 1990). From these clinical studies, it is not possible to conclude whether antidepressants

Is There Evidence for an Effect of Antidepressant Drugs on Immune Function?

275

have a direct effect on the immune system or whether their putative immune effects result from improved mood. Further experiments, comparing the effects of antidepressants and other antidepressive therapy Hke psychotherapy or electroconvulsive therapy, would be necessary to answer this question.

4.2. Immune Effects of Antidepressants in Experimental Models Investigation of the effects of antidepressant treatment on immune activity, and especially on cytokine production and/or central actions, is still in its infancy, and not much data are so far available (Table 2). In normal mice, chronic administration of maprotiline, a highly selective noradrenergic reuptake blocker, but not of desipramine, suppressed NK activity both in vivo and in vitro. Delayed hypersensitivity and T cell proliferation in vitro were not affected by this treatment, suggesting selectivity of the maprotiline effect (Eisen, Irwin, Quay, & Livnat, 1989). In another experiment, it was shown that, in mice, imipramine depressed delayed hypersensitivity to sheep erythrocytes when this antidepressant was injected i.p. as a single dose of 4 or 8mg/kg on the day of immunization or of challenge, or daily from immunization to challenge (Descotes, Tedone, & Evreux, 1985). In rats subjected to chronic mild stress model of depression, the enhanced proliferation of lymphocytes, as well as IL-1 and IL-2 production by stimulated splenic cells, were reversed by a chronic treatment with imipramine (Kubera, Symbirtsev, Basta-Kaim, Borycz, Roman, Papp, & Claesson,

Table 2. Effects of antidepressants on immune functions in laboratory animals Drugs

Immune parameters in animals

Effects of antidepressants

Maprotiline (MAOI)<" not Imipramine (TCA)"* Maprotiline (MAOI)<"

NK cell activity in normal mice

i

Eisen et al., 1989

DTH and T cell proliferation in normal mice DTH in normal mice Inflammation induced by carrageenin Migration of rat macrophages chemiotaxis of human PMBC Enhanced lymphoproliferation and IL-1, IL-2 production by splenocytes in rats subjected to a chronic mild stress model of depression Increase of alpha-1 acid glycoprotein after olfactory bulbectomy in rats IL-1 beta, IL-lRa mRNA in the brain (hypothalamus. hippocampus, frontal cortex, brain stem)

-4

Eisen et al., 1989

i i

i

Descotes et al., 1985 Bianchi et al., 1994, 1995 Sacerdote et al., 1994, 1997 Kubera et al., 1996

i

Song & Leonard, 1994

T

Suzuki et al., 1996

Imipramine (TCA)'" Fluoxetine (SSRI)<-' Clomipramine (TCA)'"' Clomipramine (TCA)'^* Imipramine (TCA)*''

Sertraline (SSRI)''>

Imipramine (TCA)'^' Fluvoxamine (SSRI)"' Maprotiline (MAOI)*"

(1): Daily injections during 6 days (2): Single injection (3): Daily injections during 5 weeks (4): Daily injections during 3 weeks (5): Daily per os during 4 weeks

i

276

P. J. Neveu and N. Castanon

1996). An inhibitory effect of monoamine reuptake inhibitors on immune activation has also been found in rats with experimental allergic neuritis (Bengtsson, Zhu,Thorell, Olsson, Link, & Walinder, 1992; Zhu, Bengtsson, Mix,Thorell, Olsson, & Link, 1994). Moreover, olfactory bulbectomy in rats, a classic animal model of depression (Jancsar & Leonard, 1983), increased a-l-acid glycoprotein 5 to 9 weeks after surgery (Arnold & Meyerson, 1990), and this increase of acute phase proteins response was reversed by treatment with selective 5-HT reuptake inhibitors (Song & Leonard, 1994). In addition, acute treatment with fluoxetine (a 5-HT reuptake inhibitor) and clomipramine (a tricyclic antidepressant) reduced the inflammation induced by carrageenin or brewer's yeast in rats (Bianchi, Sacerdote, & Panerai, 1994 a, b; Bianchi, Rossoni, Sacerdote, Panerai, & Berti, 1995). However, different mechanisms appear to be involved in mediating the effects of these compounds, which belong to two different families. The anti-inflammatory activity of fluoxetine in inflammation induced by brewer's yeast is associated with activation of the HPA axis (Bianchi et al., 1994a), whereas activity of clomipramine in carrageenin-induced inflammation may be due to an inhibition of substance P production (Bianchi et al., 1995). Moreover, clomipramine depressed the migration of rat macrophages both in vivo and in vitro (Sacerdote, Bianchi, & Panerai, 1997). Clomipramine also reduced human polymorphonuclear cell chemiotaxis in vitro (Sacerdote, Bianchi, & Panerai, 1994), whereas non-tricyclic antidepressants like fluoxetine had no effect on this chemiotaxis. These results indicate that the antiinflammatory activities of tricyclic antidepressants are independent of their antidepressive activity and result from a particular effect linked to their molecular structure. The anti-inflammatory effects of tricychc antidepressants partly result from the inhibition of NA reuptake induced by these drugs. Indeed, this anti-inflammatory effect can be attenuated by pharmacological inhibition of catecholamine synthesis or medulloadrenalectomy (Arrigoni Martelh, Toth, Segre, & Corsico, 1967). In addition, chronic (28 days), but not acute treatment with different antidepressants (imipramine, fluvoxamine, and maprotiline) administered per os, induces a marked increase of both IL-1|3 and IL-lra mRNA expression in different brain stuctures, including the hypothalamus, hippocampus, frontal cortex, and brain stem (Suzuki, Shintani, Kanba, Asai, & Nakaki, 1996). It is noteworthy that the increase of the expression of IL-lra, the natural antagonist for IL-1, is more intense than that of IL-1, and its time-profile in response to antidepressants coincides with the time required for therapeutic efficiency. These interesting results suggest that antidepressants may either inhibit cytokine production at the periphery or block the central action of IL-1 via the increase of brain IL-lra.

5. POSSIBLE MECHANISMS FOR IMMUNE EFFECTS OF ANTIDEPRESSANTS In view of all the results presented in this review, it is clear that further research is required to fully elucidate the effect of antidepressant drugs on monocyte function and to specify the role of cytokines in the onset of depression. This line of research would help to extend our understanding of the pathogenesis of depression and may allow the development of new antidepressant drugs. However, different hypotheses can be already suggested with respect to the possible mechanisms underlying the immune effects of antidepressants. Antidepressants may modify immune reactivity by acting on the neural substrate which is involved in neuroimmunomodulation. Lymphocytes and macrophages are

Is There Evidence for an Effect of Antidepressant Drugs on Immune Function?

277

innervated by the sympathetic and parasympathetic nervous systems and this innervation modulates immune reactivity. Furthermore, immune cells are known to have receptors for neurotransmitters and antidepressants may modify the activity of immune cells by acting at the level of these receptors. Antidepressants may also have direct effects on immune cells. The antidepressant rolipram, a selective type IV phosphodiesterase inhibitor, suppresses the production of TNF and to a lesser extent that of IFNy in human and rat MBP (myelin-basic protein)specificT cell lines. This suppression may be related to an increase of intacellular cyclic adenosyl monophosphate (Sommer, Loschmann, Northoff, Weller, Steinbrecher, Steinbach, Lichtenfels, Meyerman, Riethmiiller Fontana, Dichgans, & Martin, 1995). Binding sites for imipramine and desmethylimipramine have been documented on lymphocyte membranes as well as on thrombocytes and brain tissues (Audus & Gordon, 1982). Binding of antidepressants was not modified in the presence of 5-HT and NA suggesting that these binding sites are not related to neurotransmitter receptor sites. Bmax of tricyclic receptors was enhanced in depressed patients but Kd values remained normal (Krulik, Sliva, Sikora, Farska, & Fuksova, 1988). The functional significance of these receptors remains unknown. In fact, tricychcs have been shown to depress NK activity (Xiao & Eneroth, 1995), IL-6, IL-ip, and TNFa release in human blood monocytes, and IL-2 and IFNy in T lymphocytes (Xia et al., 1996). However, these effects were observed when using high doses of tricyclics and are not mediated by the specific high affinity binding sites. These immune alterations of immune reactivity may rather result from alterations in lipid or hydrophobic domains of lymphocyte membranes (Audus & Gordon, 1984). Maximal P-2 adrenoreceptor binding (Bmax) on blood mononuclear cells was reduced in unipolar depression but not in bipolar or dysthymic patients. There was a strong inverse correlation between Bmax values and severity of depression in bipolar but also in the whole population of depressed patients (Jeanningros, Mazzola, Azorin, Samuelian-Massa, & Tissot, 1991).

6. CONCLUSION Cytokines, which appear to be altered in some depressed patients, have been hypothesized to be involved in the pathogenesis of depression. Nevertheless, numerous methodological problems need to be solved before this hypothesis can be confirmed. In animals, the role of cytokines in the activation of both the HPA axis and brain metaboHsm has only been demonstrated in acute situations, whereas chronic experiments should be more appropiate to reproduce a depression-like situation. Furthermore, cytokines act as a complex network that has to be studied in more details in depression. If cytokines play a role in the pathphysiology of depression, antidepressants should be active on cytokine production and/or action. However, there is no evidence so far to conclude whether depressed mood is associated with severity of illness or whether it is directly linked to the cytokine concentrations measured in these patients. The comparison of the efficacy of antidepressant drugs with alternative therapies (e.g. electroconvulsive therapy or psychotherapy) should help to answer this question. Furthermore, depending on their molecular heterogeneity, antidepressants could have different effects on cytokine production, and a systematic study of the effects of different classes of antidepressants needs to be carefully done. Last, but not the least, if cytokines are involved in the pathogenesis of depression, molecules known to modulate the

278

P. J. Neveu and N. Castanon

inflammatory cytokine production, like receptor-antagonists, anti-cytokine antibodies, or anti-inflammatory cytokines, should improve depressive disorders.

REFERENCES Albrecht, J., Helderman, J. H., Schlesser, M. A., & Rush, A. J. (1985). A controlled study of cellular immune function in affective disorders before and during somatic therapy. Psychiatry Research, 15,185-193. Altshuler, L. L., Plaeger-Marshall, S., Richeimer, S., Daniels, M., & Baxter, L. R. (1989). Lymphocyte function in major depression. Acta Psychiatrica Scandinavia, 80,132-136. Anderson, J. L. (1996). The immune system and major depression. Advanves in Neuroimmunology, 6,119-129. Arnold, F. J. & Meyerson, L. R. (1990). Olfactory bulbectomy alters alpha-1 acid glycoprotein levels in rat plasma. Brain Research Bulletin, 25, 259-262. Arrigoni Martelli, E.,Toth, E., Segre, A. D., & Corsico, N. (1967). Mechanism of inhibition of experimental inflammation by antidepressant drugs. European Journal of Pharmacology, 2, 229-233. Audus, K. L. & Gordon, M. A. (1982). Characteristics of murine tricyclic antidepressant binding sites associated with murine lymphocytes from spleen. Journal of Immunopharmacology, 4,1-12. Audus, K. L. & Gordon, M. A. (1984). Effect of tricyclic antidepressant drugs on lymphocyte membrane structure. Journal of Immunopharmacology, 6,105-132. Harden, N., Reul, J. M. H. M., & Holsboer, R (1995). Do antidepressants stabilize mood through actions on hypothalamic-pituitary-adrenocortical system? Trends in Neurosciences, 18,6-11. Bauer, J., Hohagen, E, Gimmel, E., Bruns, P., Lis, S., Krieger, S., Ambach, W., Guthmann, A., Grunze, H., Fritsch-Montero, R., Weissbach, A., Ganter, U., Frommberger, U , Riemann, D., & Berger, M. (1995). Induction of cytokine synthesis and fever suppresses REM sleep and improves mood in patients with major depression. Biological Psychiatry, 38, 611-621. Bengtsson, B. O., Zhu, J., Thorell, L. H., Olsson, T, Link, H., & Walinder, J. (1992). Effects of zimeldine and its metabolites, clomipramine, imipramine, and maprotiline in experimental allergic neuritis in lewis rats. Journal of Immunology, 39,109-112. Besedovsky, H. O. & Del Rey, A. (1996). Immune-neuro-endocrine interactions: facts and hypotheses. Endocrine Reviews, 17, 64-102. Bianchi, M., Rossoni, G., Sacerdote, R, Panerai, A. E., & Berti, F. (1995). effects of clomipramine and fluoxetine on subcutaneous carrageenin-induced inflammation in the rat. Inflammation Research,44, 466-469. Bianchi, M., Sacerdote, P., & Panerai, A. E. (1994a). Fluoxetine reduces inflammatory edema in the rat: involvement of the pituitary-adrenal axis. European Journal of Pharmacology, 263, 81-84. Bianchi, M., Sacerdote, P., & Panerai, A. E. (1994b). Clomipramine differently affects inflammatory edema and pain in the rat. Pharmacology Biochemestry and Behavior, 4,1037-1040. BHer, P. & de Montigny, C. (1994). Current advances and trends in the treatment of depression. Trends in Pharmacological Sciences, 15, 220-226. Charles, G., Machowski, R., Brohee, D., Wilmotte, J., & Kennes, B. (1992). Lymphocyte subsets in major depressive patients. Influence of anxiety and corticoadrenal overdrive. Neuropsychobiology, 25,94-98. Connor, T. J. & Leonard, B. E. (1998). Depression, stress, and immunological activation: the role of cytokines in depressive disorders. Life Sciences, 62, 583-606. Cosyns, P., Maes, M., Vandewoude, M., Stevens, W. J., De Clerck, L. S., & Schotte, C. (1989). Impaired mitogeninduced lymphocyte responses and the hypothalamic-pituitary-adrenal axis in depressive disorders. Journal of Affective Disorders, 16, 41-48. Darko, D. F , Lucas, A. H., Gillin, J. C , Risch, S. C , Golshan, S., Hamburger, R. N., Silverman, M. B., & Janowsky, D. S. (1988). Cellular immunity and the hypothalamic-pituitary axis in major affective disorder: a preliminary study. Psychiatry Research, 25, 1-9. Deleplanque, B. & Neveu, P. J. (1995). Immunological effects of neuropsychotropic substances. In M. Guenounou (Ed.). Forum on immunomodulators (pp. 287-302). Paris: John Libbey Eurotext. Delrue, C , Deleplanque, B., Rouge-Pont, F , Vitiello, S., & Neveu, P. J. (1994), Brain monoaminergic, neuroendocrine, and immune responses to an immune challenge in relation to brain and behavioral lateralization. Brain Behavior and Immunity, 8,137-152. Descotes, J., Tedone, R., & Evreux, J. C. (1985). Different effects of psychotropic drugs on delayed hypersensitivity responses in mice. Journal of Neuroimmunology, 9, 81-85. Dunn, A. J. (1992). Endotoxin-induced activation of cerebral catecholamine and serotonin metabolism: comparison with interleukin-1.7ournfl/ of Pharmacology and Experimental Therapeutics, 261,964-969.

Is There Evidence for an Effect of Antidepressant Drugs on Immune Function?

279

Eisen, J. N., Irwin, J., Quay, J., & Livnat, S. (1989). The effect of antidepressants on immune function in mice. Bioliogical Psychiatry, 26, 805-817. Frommberger, U. H., Bauer, J., Haselbauer, P., Fraulin, A., Riemann, D., & Berger, M. (1997). Interleukin-6(IL-6) plasma levels in depression and schizophrenia: comparison between the acute state and after remission. European Archives of Psychiatry and Clinical Neuroscience, 247, 228-233. Garver, D. L. & Davis, J. M. (1979). Biogenic amine hypothesis of affective disorders. Life Sciences, 24, 283-394. Goldman, L. S. (1994). Successful treatment of interferon alfa-induced mood disorder with nortriptyline. Psychosomatics, 35, 412-413. Guelfi, J. D. (1992). Efficay of tianeptine in comparative trials versus reference antidepressants. An overview. British Journal of Psychiatry, 160, 72-75. Hickie, I. (1990). Is there immune dysfunction in depressive disorders? Psychololical Medecine, 20,755-761. Holhster, L. E. (1986). Current antidepressants. Annual Review of Pharmacology and Toxicology, 26,23-37. Holsboer, P., Bardeieben, U. von, Gerken, A., Stalla, G. K., & Muller, O. A. (1984). Blunted corticotropin and normal Cortisol response to human corticotropin-releasing factor in depression. New England Journal of Medecine, 311,1127-1132. Irwin, M. R., Daniels, M., Bloom, E., Smith, T. L., & Weiner, H. (1987). Life events, depressive symptoms, and immune function. American Journal of Psychiatry, 144, A31-4AI. Irwin, M., Patterson, T, Smith, T. L., Caldwell, C , Brown, S. A., Gillin, J. C , & Grant, I. (1990). Reduction of immune function in life stress and depression. Biological Psychiatry, 27, 22-30. Janscar, S. & Leonard, B. E. (1983). The olfactory bulbectomized rat as a model of depression. In: Frontiers in Neuropsychiatric ressearch (E. Usdin, M. Goldstein, A. J. Friedhoff, A. Geogatas, Eds.) pp 357-372, MacMillan, New-York Jeanningros, R., Mazzola, P., Azorin, J. M., Samuelian-Massa, C , & Tissot, R. (1991). B-adrenoreceptor density of intact mononuclar leukocytes in subgroups of depressive disorders. Biological Psychiatry, 29, 789-798. Joyce, P R., Hawes, C. R., Mulder, R. T , Sellman, J. D., Wilson, D. A., & Boswell, D. R. (1992). Elevated levels of acute phase plasma proteins in major depression. Biological Psychiatry, 32,1035-1041. Katon, W. & Sullivan, M. D. (1990). Depression and chronic medical illness. Journal of Clinical Psychiatry, 57,3-11. Kent, S., Bluthe, R. M., Kelley, K. W., & Dantzer, R. (1992). Sickness behavior as a new target for drug development. Trends in Pharmacological Sciences, 13, 24-28. Kronfol, Z. & House, J. D. (1985). Depression, hypothalamic-pituitary-adrenocortical activity, and lymphocyte function. Psychopharmacology Bulletin, 21, 476-478. Kronfol, Z., Silva, J., Greden, J., Dembinski, S., Gardner, R., & Carroll, B. J. (1983). Lymphocyte function in depressive illness. Life Sciences, 33, 241-247. Krulik, R., Sliva, D., Sikora, J., Farska, I., & Fuksova, K. (1988). Tricyclic antidepressant binding to lymphocyte membranes and changes during depression. European Journal of Pharmacology, 149, 357-361. Kubera, M., Symbirtsev, A., Basta-Kaim, A., Borycz, J., Roman, A., Papp, M., & Claesson, M. (1996). Effect of chronic treatment with imipramine on interleukin 1 and interleukin 2 production by splenocytes obtained from rats subjected to a chronic mild stress model of depression. Polish Journal of Pharmacology, 48, 503-506. Landmann, R., Schaub, B., Link, S., & Wacker, H. R. (1997). Unaltered monocyte function in patients with major depression before and after three months of antidepressive therapy. Biological Psychiatry, 41, 675-681. Laye, S., Parnet, P., Goujon, E., & Dantzer, R. (1994). Peripheral administration of hpopolisaccharide induces the expression of cytokine transcripts in the brain and pituitary in mice. Molecular Brain Research, 27,157-162. Levenson, J. L. & Fallon, H. J. (1993). Fluoxetine treatment of depression caused by interferon-a. American Journal of Gastroenterology, 88, 760-761. Linthorst, A. C. E., Flachskamm, C , Holsboer, E, & Reul, J. M. H. M. (1995). Intraperitoneal administration of bacterial endotoxin enhances noradrenergic neurotransmission in the rat preoptic area: relationship with body temperature and hypothalamic-pituitary-adrenocortical axis activity. European Journal of Neuroscience, 7,2418-2430. Maes, M., Bosmans, E., Suy, E., Vandervorst, C , Dejonckheere, C , Minner, B., & Raus, J. (1991). Depressionrelated disturbances in mitogen-induced lymphocyte responses, interleukin-ip, and soluble interleukin-2-receptor production. Acta Psychiatrica Scandinavia, 84, 379-386. Maes, M. (1995). Evidence for an immune response in major depression: a review and hypothesis. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 19, 11-38.

280

P. J. Neveu and N. Castanon

Maes, M., Bosnians, E., Meltzer, H. Y., Scharpe, S., & Suy, E. (1993a). Interleukin-ip: a putative mediator of HPA axis hyperactivity in major depression? American. Journal of Psychiatry, 150, 1189-1193. Maes, M., Bosmans, E., Suy, E., Minner, B., & Raus, J. (1989). Impaired lymphocyte stimulation by mitogens in severely depressed patients. A complex interface with HPA-axis hyperfonction, noradrenergic activity, and the ageing process. British. Journal of Psychiatry, 155, 793-798. Maes, M., Lambrechts, J., Bosmans, E., Jacobs, J., Suy, E., Vandervorst, C , De Jonckheere, C , Minner, B., & Raus, J. (1992). Evidence for a systemic immune activation during depression: results of leukocyte enumeration by flow cytometry in conjunction with monoclonal antibody staining. Psychological Medecine, 22, 45-53. Maes, M., Scharpe, S., Meltzer, H. Y., Bosmans, E., Suy, E., Calabrese, J., & Cosyns, P. (1993b). Relationship between interleukin-6 activity, acute phase proteins, and function of the hypothalamic-pituitaryadrenal axis in severe depression. Psychiatry Research, 49,11-27. Maes, M., Smith, R., & Scharpe, S. (1995). The monocyte-T-lymphocyte hypothesis of major depression. Psychoneuroendocrinology, 20, 111-116. Maier, S. R & Watkins, L. R. (1995). Intracerebroventricular interleukin-1 receptor antagonist blocks the enhancement of fear conditioning and interference with escape produced by inescapable shock. Brain Research, 695, 279-282. McAdams, C. & Leonard, B. E. (1993). Neutrophil and monocyte phagocytosis in depressed patients. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 17, 971-984. McDonald, E. M., Mann, A. H., & Thomas, H. C. (1987). Interferon as mediatiors of psychiatric morbidity. Lancet, 2, 1175-1178. Meijer, A., Zakay-Rones, Z., & Morag, A. (1988). Post influenza psychiatric disorder in adolescents. Acta Psychiatica Scandinavia, 78, 176-181. Merendino, R. A., Mancuso, G., Tomasello, E, Gazzara, D., Cusumano, V., Chillemi, S., Spadaro, P., & Mesiti, M. (1994). Effects of lithium carbonate on cytokine production in patients affected by breast cancer. Journal of Biological Regulators and Homeostatic Agents, 8, 88-91. Miller, A. H., Asnis, G. M., Lackner, C , Halbreich, U., & Norin, A. J. (1991). Depression, natural killer cell activity, and Cortisol secretion. Biological Psychiatry, 29, 878-886. Mitchinson, M. & Ball, R. (1987). Macrophages and atherogenesis. Lancet, 1, 146-148. Mohr, D. C , Goodkin, D. E., Likosky, W., Gatto, R , Baumann, K. A., & Rudick, R. A. (1997). Treatment of depression improves adherence to interferon beta-1(3 therapy for multiple sclerosis. Archives of Neurology, 54, 531-533. Muller, N. (1995). Psychoneuroimmunology: implications for the drug treatment of psychiatric disorders. Central Nervous System Drugs, 4, 125-140. Munck, A., Guyre, P. M., & Meltzer, M. S. (1984). Physiological functions of glucocorticoids in stress and their relation to pharmacological actions. Endocrine Reviews, 1, 25-44. Nerozzi, D., Santoni, A., Bersani, G., Magnani, A., Bressan, A., Pasini, A., Antonozzi, I., & Frajese, G. (1989). Reduced natural killer cell activity in major depression: neuroendocrine implications. Psychoneuroendocrinology, 14, 295-301. Perini, G. I., Zara, M., Carraro, C., Tosin, C., Gava, R, Santucci, M. G., Valverde, S., & De Franchis, G. (1995). Psychoimmunoendocrine aspects of panic disorder. Human Psychopharmacology, 10, 461-465. Rabkin, J. G. & Harrison, W. M. (1990). Effect of imipramine on depression and immune status in a sample of men with HIV infection. American Journal of Psychiatry, 147, 495^97. Sacerdote, P., Bianchi, M., & Panerai, A. E. (1997). In vivo and in vitro clomipramine treatment decreases the migration of macrophages in the rat. European Journal of Pharmacology, 319, 287-290. Sacerdote, P., Bianchi, M., & Panerai, A. E. (1994). Chlorimipramine and nortriptyhne but not fluoxetine and fluvoxamine inhibit human polymorphonuclear cell chemotaxis in vitro. General Pharmacol., 25, 409-412. Schleifer, S. J., Keller, S. E., Bond, R. N., Cohen, J., & Stein, M. (1989). Major depressive disorder and immunity: role of age, sex, severity, and hospitalisation. Archives of General Psychiatry, 46, 81-87. Schleifer, S. J., Keller, S. E., Bartlett, J. A., Eckholdt, H. M., & Delaney, B. R. (1996). Immunity in young adults with major depressive disorder. American Journal of Psychiatry, 153, 477^82. Seidel, A., Arolt, V., Hunstiger, M., Rink, L., Behnisch, A., & Kirchner, H. (1996). Major depressive disorder is associated withe elevated monocyte counts. Acta Psychiatrica Scandinavia, 94,198-204. Seidel, A., Arolt, V., Hunstiger, M., Rink, L., Behnisch, A., & Kirchner, H. (1995). Cytokine production and serum proteins in depression. Scandinavian Journal of Immunology, 41, 534-538. Shintani, R, Nakaki, T, Kanba, S., Sato, K., Yagi, G., Shiozawa, M., Aiso, S., Kato, R., & Asai, M. (1995). Involvement of interleukin-1 in immobilization stress-induced increase in plasma adrenocorticotropic

Is There Evidence for an Effect of Antidepressant Drugs on Immune Function?

281

hormone and in release of hypothalamic monoamines in the rat. Journal of Neuroscience, 15, 1961-1970. Sluzewska, A., Rybakowski, J. K., Laciak, M., Mackiewicz, A., Sobieska, M., & Wiktorowicz, K. (1995a). Interleukin-6 serum levels in depressed patients before and after treatment with fluoxetine. Annals of the. New York Academy of Sciences. 762, AlA-All. Sluzewska, A., Rybakowski, J. K., Sobieska, M., Bosmans, E., PoUet, H., & Wiktorowicz, K. (1995b). Increased levels of alpha-1-acid glycoprotein and interleukin-6 in refractory depression. Depression, 3,170-175. Smith, R. S. (1991). The macrophage theory of depression. Medecine Hypotheses, 35, 298-306. Sommer, N., Loschmann, P.-A., Northoff, G. H., Weller, M., Steinbrecher, A., Steinbach, J. P., Lichtenfels, R., Meyerman, R., Riethmiiller, A., Fontana, A., Dichgans, J., & Martin, R. (1995). The antidepressant rolipram suppresses cytokine production and prevents autoimmune encephalomyelitis. Nature Medecine, 7,244-248. Song, C. & Leonard, B. E. (1994). An acute phase protein response in the olfactory bulbectomized rat: effect of sertraline treatment. Medical Science Research, 22, 313-314. Stein, M., Miller, A. H., & Trestman, R. L. (1991). Depression, the immune system, and health and illness: findings in search of meaning. Archives of General Psychiatry, 48, 171-177. Sundar, S. K., Cierpial, M. A., Kilts, C , Ritchie, J. C , & Weiss, J. M. (1990). Brain IL-1 induced immunosuppression occurs through activation of both pituitary-adrenal axis and sympathetic nervous system by corticotropin-releasing factor. Journal of Neuroscience, 10, 3701-3706. Suzuki, E., Shintani, E, Kanba, S., Asai, M., & Nakaki,T. (1996). Induction of interleukin-ip and interleukin1 receptor antagonist mRNA by chronic treatment with various psychotropics in widespread area of rat brain. Neuroscience Letters, 215, 201-204. Targum, S. D., Marshall, L. E., Ficshman, P., & Martin, D. (1989). Lymphocyte subpopulations in depressed elderly women. Biological Psychiatry, 26, 581-589. Udelman, D. L. & Udelman, H. D. (1985). A preliminary report on anti-depressant therapy and its effects on hope and immunity. Social Science and Medicine, 20, 1069-1072. Ur, E., White, P. D., & Grossman, A. (1992). Hypothesis: cytokines may be activated to cause depressive illness and chronique fatigue syndrome. European Archives of Psychiatry and Clinical Neuroscience, 241, 317-322. Van Riezen, H. & Leonard, B. E. (1990). Effects of psychotropic drugs on the behavior and neurochemistry of olfactory bulbectomized rats. Pharmacology and Therapeutics, 47, 21-34. Weizman, R., Laor, N., Podliszewski, E., Notti, I., Djaldetti, M., & Bessler, H. (1994). Cytokine production in major depressed patiens before and after clomipramine treatment. Biololical Psychiatry, 35, 42-47. Xia, Z., De Pierre, J. W., & Nassberger, L. (1996). Tricyclic antidepressants inhibit lL-6, IL-1 beta, and TNFalpha release in human blood monocytes and IL-2 and interferon-gamma in T cells. Immunopharmacology, 34, 27-37. Xiao, L. & Eneroth, P. (1995). Tricyclic antidepressants inhibit human natural killer cells. Toxicology and Applied Pharmacology, 137, 157-162. Yirmiya, R. (1996). Endotoxin produces a depresive-like episode in rats. Brain Research, 711, 163-174. Young, S. N., Smith, S. E., Pihl, R. O., & Ervin, F. R. (1985). Tryptophan depletion causes a rapid lowering of mood in normal males. Psychopharmacology, 87,173-177. Zalcman, S., Green-Johnson, J. M., Murray, L., Nance, D. M., Dyck, D., Anisman, H., & Greenberg, A. H. (1994). Cytokine-specific central monoamine alterations induced by interleukin-1, -2, and -6. Brain Research, 643, 4 0 ^ 9 . Zhu, J., Bengtsson, B. O., Mix, E., Thorell, L. H., Olsson, T, & Link, H. (1994). Effect of monoamine reuptake inhibiting antidepressants on major histocompatibility complex expression on macrophages in normal rats and rats with experimental allergic neuritis (EAN). Immunopharmacology. 27, 225-244.

16

CYTOKINES, "DEPRESSION DUE TO A GENERAL MEDICAL CONDITION," AND ANTIDEPRESSANT DRUGS Raz Yirmiya,' Joseph Weidenfeld,^ Yehuda PoUak,^ Michal Morag,^ Avraham Morag,^ Ronit Avitsur,' Ohr Barak,' Avraham Reichenberg,' Edna Cohen,' Yehuda Shavit,' and Haim Ovadia^ ' Department of Psychology, Mount Scopus The Hebrew University of Jerusalem ^Departments of Neurology ^Department of Clinical Virology Hadassah-Hebrew University Hospital, Ein Karem, Jerusalem, Israel

1. INTRODUCTION Activation of the immune system during various medical conditions produces neural, neuroendocrine, and behavioral effects. The psychological and physiological effects of immune activation resemble many characteristics of depression. The essential features of depression are depressed mood and loss of interest or pleasure in all, or almost all activities (anhedonia). Several associated symptoms are also present, including, appetite disturbance, change in body weight, sleep disturbance, psychomotor disturbance, fatigue, loss of energy, and difficulty in thinking or concentrating (DSM-IV, 1994). Depression is also characterized by specific alterations in the functioning of neurochemical and neuroendocrine systems, including monoaminergic systems and the hypothalamic-pituitary-adrenal (HPA) axis (Brown, Steinberg, & van Praag, 1994; Holsboer, 1995). Most of these psychological and neuroendocrine symptoms appear both in humans and animals during diseases that involve immune activation. Based on these findings, and on several additional lines of evidence that will be presented below, we have recently argued that immune activation is involved in the etiology and symptomatology of depression associated with various medical conditions (Yirmiya, 1997). The physiological and psychological effects of immune activation (collectively termed "sickness behavior") are mediated by cytokines (Connor & Leonard, 1998; Dantzer, Bluthe, Aubert, Goodall, Bret-Dibat, Kent, Goujon, Laye, Parnet, & Cytokines, Stress, and Depression, edited by Dantzer et al. Kluwer Academic / Plenum Publishers, New York, 1999.

283

284

R. Yirmiya et al.

Kelley, 1996; Hart, 1988; Kent, Bluthe, Kelley, & Dantzer, 1992a; Maier & Watkins, 1998 Yirmiya, 1997). Most immune challenges produce their initial effects in the periphery, but information regarding their presence is almost immediately transmitted to the brain, in a sensory-like process. Communication between the immune system and the brain is at least partly mediated by proinflammatory cytokines, particularly IL-ip, TNFa, IL-6, and interferons. Two major communication pathways, a humoral and a neural one, have been described (Dantzer et al., 1996; Maier & Watkins, 1998; Watkins, Maier, & Goehler, 1995a). Blood-borne cytokines do not cross the blood brain barrier (BBB), but can still penetrate the brain via the circumventricular organs (which lack BBB), or be transported into the brain by carrier molecules, or bind to endothelial cells of the cerebral vasculature and induce the release of secondary mediators within the brain parenchyma (Watkins et al., 1995a). In addition to the humoral mechanisms, a neural pathway has been recently discovered: Locally-released cytokines activate receptors on peripheral nerves (primarily the vagus), which, in turn, transmit messages to the brain (Dantzer et al., 1996; Maier & Watkins, 1998; Watkins et al., 1995a). Within the brain, this immune-related information activates several areas, including the solitary nucleus, the paraventricular nucleus of the hypothalamus, and the central nucleus of the amygdala (Day & Akil, 1996). The neural activity within these areas is at least partly mediated by cytokines (e.g., ILIp and TNFa), which are released locally by glia cells and neurons, and serve as neurotransmitters and neuroregulators (Dantzer et al., 1996; Wong, Bongiorno, Rettori, McCann, & Licinio, 1997). The aim of the present chapter is to review the current knowledge on cytokinemediated depressive-like symptoms that accompany various medical conditions in humans, and experimental models of these conditions in animals. In addition, our recent efforts to evaluate the relationship between depressive behavior and sickness symptoms, by investigating the effects of antidepressants on sickness behavior in rats, will be presented and discussed.

2. PSYCHOLOGICAL EFFECTS OF IMMUNE ACTIVATION IN HUMANS Depression is a common, disturbing concomitant of medical conditions. The reported prevalence of major depression episodes in physically-ill varies from 5% to more than 40% (Chochinov, Wilson, Enns, & Lander, 1994). Because depression is often unrecognized and undertreated in sick patients, the prevalence reported in most studies is probably underestimated (Katon & Sullivan, 1990; Laghrissi-Thode, Pollock, Szanto, & Reynolds, 1996). Depression associated with medical conditions has serious implications. For example, the majority (74%) of men committing suicide during an episode of major depression were receiving treatment for a medical condition at the time of death (Isometsa, Aro, Henriksson, Heikkinen, & Lonnqvist, 1994). Several particular medical conditions have been described in humans, which are associated with both immune activation and high prevalence of behavioral symptoms characteristic of both sickness behavior and depression. In most of these conditions the behavioral symptoms are not explained by direct effects of pathogens on neural tissues. Thus, the effects of pathogens on brain and behavior are usually mediated by indirect mechanisms, such as immune factors.

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

285

2.1. Psychological Effects of Acute and Chronic Infectious Diseases Acute infectious illness, such as influenza, upper respiratory tract infections, gastroenteritis, Epstein-Barr virus, and cytomegalovirus, are associated with a range of depressive symptoms, including fatigue, psychomotor retardation, anorexia, somnolence, lethargy, muscle aches, cognitive disturbances, and depressed mood (Hickie & Lloyd, 1995). The evidence for these alterations is mainly anecdotal and only few studies examined these symptoms systematically. Experimentally-induced viral infections (e.g., common cold, influenza) are associated with decreased psychomotor performance of simple reaction-time tasks and memory impairments (Smith, Thomas, Brockman, Kent, & Nicholson 1993; Smith,Tyrrell, Al-Nakib, Coyle, Donovan, Higgins, & Willman, 1987; Smith,Tyrrell, Al-Nakib, Coyle, Donovan, Higgins, & Willman, 1988). In addition, they are often associated with long term psychiatric effects, particularly depression. Experimentally-induced influenza (but not infections with coronavirus or other minor cold viruses; see Smith, Tyrrell, & Barrow, 1992), as well as natural occurrence of upper respiratory tract illness (Hall & Smith, 1996a), produce a general negative mood state. Moreover, following infection with influenza, subjects showed depressive symptoms, including depressed mood, reduced appetite, sleep disturbances, sense of guilt, marked difficulty in decision making and memory impairments; these symptoms could be detected even months after the onset of the infection (Meijer, Zakay-Rones, & Morag, 1988). Similar disturbances have also been reported following herpesvirus infections (Greenwood, 1987), mononucleosis (Hall & Smith, 1996b; Hendler, 1987), and infections with Borna Disease virus (Bode, Zimmerman, Ferszt, Steinbach, & Ludwig, 1995) and HIV (Maj, 1996). For example, the results of a WHO Neuropsychiatric AIDS study, conducted in 5 geographical locations with a total of 955 subjects, strongly suggest that HIV infection is associated with an increased prevalence of depressive symptoms, and in some locations also with higher incidence of major depression (Maj, 1996). There is also a preliminary evidence for a direct relationship between brain TNFa production and cognitive deterioration in AIDS patients (Seilhean, Kobayashi, He, Uchihara, Rosenblum, Katlama, Bricaire, Duyckaerts, & Hauw, 1997). We have recently used a double-blind prospective design to investigate the immediate and prolonged psychological and physiological effects of a specific viral infection in humans (Morag, Yirmiya, Lerer, & Morag, in press). Subjects were teenager girls who were vaccinated with live attenuated rubella virus. Based on analysis of levels of antibodies to rubella, subjects were divided into two groups: An experimental group (n = 60), comprised of subjects who were initially seronegative and were infected following vaccination, and a control group (n = 180), comprised of subjects who were already immune to rubella before vaccination. Compared to control subjects, and to their own baseline, subjects from low socioeconomic status (SES) within the experimental group exhibited more severe depressed mood, as measured by the Children Depression Inventory (CDI) (Fig. 1). In addition, the same subjects exhibited more social and attention problems and delinquent behavior, as measured by the Achenbach Child Behavior Checklist. Subjects from high and middle SES did not show these disturbances (Morag et al., in press). The particular vulnerability of low SES subjects to immunization-induced depression is consistent with previous epidemiological studies, demonstrating that people of low SES have higher rates of major, as well as chronic and recurrent depression (Anderson & Armstead, 1995; Bruce, Takeuchi, & Leaf, 1991; Murphy, Olivier,

R. Yirmiya et al.

286

4 T

•Control • Experimental

1 -o

s Baseline

2

10

Time From Vaccination (in weeks) B 14 y

• Control

12 --

• Experimental

10 8

6+ 4 o

2+ 0 Baseline

2

10

Time From Vaccination (in weeks) Figure 1. Effect of vaccination with live-attenuated rubella virus on depression scores, measured by the Children Depression Inventory in 12-years old girls with low socioeconomic status.

Monson, Sobol, Federman, & Leighton, 1991). This vulnerability may be associated with several characteristics of low SES, including higher Incidence of stressful life events, and fewer sources of social support (Adler, Boyce, Chesney, Cohen, Folkman, Khan, & Syme, 1994; Anderson & Armstead, 1995; Dohrenwend, 1973; Ranchor, Bouma, & Sanderman, 1996). As demonstrated (Cohen, 1995; Kiecolt-Glaser & Glaser, 1991), these factors modulate the responsiveness to immune challenges. Thus, even a mild viral infection can produce prolonged increase in depressive symptomatology in vulnerable individuals. Mean (+S.E.M.) emotional depression score (A) and total depression score (B) were assessed before, and 2 and 10 weeks after vaccination. Subjects who were initially seronegative and were infected following vaccination (experimental group) had higher depression scores than subjects who were already immune to rubella before vaccination (control group), at both 2 and 10 weeks post-vaccination.

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

287

2.2. Psychological Effects of Non-Infectious Conditions Associated with Immune Activation Chronic activation of the immune system and enhanced secretion of cytokines may be associated with several types of non-infectious conditions (Dinarello & Wolff, 1993). High incidence of depression is observed in such cases too. Autoimmune diseases: Cytokines play an important role in the etiology and pathology of many autoimmune diseases (Cavallo, Pozzilli, & Thorpe, 1993; Dinarello & Wolff, 1993), v^hich are also associated with a high prevalence of depression. Particularly high incidence of depression has been demonstrated in patients with multiple sclerosis (Foley, Traugott, LaRocca, Smith, Perlman, Caruso, & Scheinberg, 1992; Minden & Schiffer, 1990; Schiffer & Babigian, 1984; Schubert & Foliart, 1993; Whitlock & Siskind, 1980). According to several estimates, the prevalence of depression in MS patients is in the range of 42-54% (Joffe, Lippert, Gray, Sawa, & Horvath, 1987; Minden, Orav, & Reich, 1987; Sadovnick, Remick, Allen, Swartz, Lee, Eisen, Farquhar, Hashimoto, Hooge, Kastrukoff, Morrison, Nelson, Ogar, & Paty, 1996). Other autoimmune conditions associated with high prevalence of depression, include rheumatoid arthritis (Parker, Smarr, Anderson, Hewett, Walker, Bridges, & Caldwell, 1992; Pincus, Griffith, Pearce, & Isenberg, 1996), systemic lupus erythematosus (Denburg, Carbotte, & Denburg, 1997; Hutchinson, Nehall, & Simeon, 1996; Lim, Ron, Ormerod, David, Miller, Logsdail, Walport, & Harding, 1988; Magner, 1991; Schneebaom, Singleton, & West, 1991), and allergy (Marshal, 1993). Detailed studies of some of these conditions suggest that, rather than psychological reactions to the medical condition per-se, illness-associated depression is causally related to immune activation (see below). Stroke and trauma: Stroke and some other types of brain traumas are associated with increased secretion of cytokines. For example, the levels of TNFa and IL-ip are dramatically increased in the brain following stroke or head trauma (Arvin, Neville, Barone, & Feuerstein, 1996). IFNa and its receptors have been identified in cerebral infarct tissues (Yamada & Yamanaka, 1995). Depression is the most common neuropsychiatric consequence of stroke, affecting up to 40% of the patients (Robinson, 1997; Schwartz, Speed, Brunberg, Brewe, Brown, & Greden, 1993). A relationship between the appearance and severity of depression to the location of the injury within the brain has been demonstrated in several cases (Robinson, 1997). However, most depressive symptoms are not explained by a direct localized neural impairment, suggesting an involvement of a more general mechanism, such as immune activation. Alzheimer's disease: Immune activation and cytokine secretion is associated with brain lesions of Alzheimer's disease and other neurodegenerative diseases (McGeer & McGeer, 1995; Rothwell, Luheshi, & Toulmond, 1996). Activated microglia cells and astrocytes that are associated with neuritic plaques produce IL-1. Moreover, IL-1 can upregulate the expression of p-amyloid precursor proteins and various other plaqueassociated proteins. Moreover, this process is self propagating, because P-amyloid directly activates microglia, thus inducing further IL-1 production (Mrak, Sheng, & Griffin, 1995). Depression is very common in Alzheimer's patients. In fact, mild to severe depression is the most prevalent psychiatric symptom in these patients (Mendez, Martin, Smyth, & Whitehouse, 1990). For example, a recent study on 109 patients identified major depression in 22% and minor depression in another 27% of the patients (Lyketsos, Steele, Baker, Galik, Kopunek, Steinberg, & Warren, 1997).

288

R. Yirmiya et al.

Menstrual cycle and post-partum period: Women exhibit higher levels of immune activation than men (Grossman, 1985), and a high incidence of depression (DSM-IV, 1994; Parry, 1995). This relation may be attributed to many factors, but it should be noted that plasma and urinary levels of IL-1 are much higher in women than in men (Cannon & Dinarello, 1985). Moreover, in vitro production of IL-1 is much greater in unstimulated mononuclear cells derived from women is in cells derived from men (Lynch, Dinarello, & Cannon, 1994). This effect depends on the phase of the menstrual cycle; compared to men's cells, women's cells isolated during the luteal and follicular phases secreted 5-10 fold and 13-28 fold more IL-1 a, IL-lp, and IL-1 receptor antagonist (IL-lra), respectively (Lynch et al., 1994). Although greater absolute amounts of each species of IL-1 were secreted during the follicular phase, the ratio of agonist to antagonist secreted was greater in the luteal phase (Lynch et al., 1994). This finding is in agreement with the in vivo data, which reflected greater IL-1 bioactivity in the plasma during the luteal phase (Cannon & Dinarello, 1985). The time course of IL-1 secretion and bioactivity is correlated with the onset of depressive episodes in women, which is highest during the luteal phase, particularly several days before the menstrual flow (Abramowitz, Baker, & Fleischer, 1982; Parry, 1995). Finally, child delivery, which in some women causes post-partum depression (Parry, 1995), also triggers a marked increase in cytokine secretion (Cox, King, Casey, & Macdonald, 1993).

2.3. Psychological Effects of Cytokine Administration Administration of cytokines in humans produces marked behavioral and neuroendocrine symptoms that are similar to those induced by viral infection. Administration of alpha interferon (IFNa) was found to cause flu-like symptoms as well as depressive symptoms, including depressed mood, dysphoria, anhedonia, helplessness, mild to severe fatigue, anorexia, and weight loss, hypersomnia, psychomotor retardation, decreased concentration, and confusion (Fent & Zbinden, 1987; McDonald, Mann, & Thomas, 1987; Meyers & Valentine, 1995; Niiranen, Laaksonen, livanainen, Mattson, Fakkila, & Cantell, 1988; Okanoue, Sakamoto, Itoh, Minami, Yasui, Sakamoto, Nishioji, Katagishi, Nakagawa, Tada, Sawa, Mizuno, Kagawa, & Kashima, 1996; Pavol, Meyers, Rexer, Valentine, Mattis, & Talpaz, 1995; Renault, Hoofnagle, Park, Mullen, Peters, Jones, Rustgi, & Jones, 1989; Valentine, Meyers, Kling, Richelson, & Hauser, 1998). In some studies, severe depressed mood has been reported in the majority of the cytokine-administered patients (McDonald et al., 1987; Niiranen et al., 1988; Valentine et al., 1998). Depressive symptoms increased with the dose and duration of IFNa treatment (Pavol et al., 1995; Renault et al, 1989; Valentine et al., 1998), and disappeared completely within 2-3 weeks after termination of the treatment. Patients receiving IL-2 or TNFa also exhibited flu-like symptoms and some depressive symptoms, including depressed mood, severe fatigue, weakness, lethargy, decreased concentration and confusion (Fent & Zbinden, 1987; Meyers, Valentine, Wong, & Leeds, 1994; Spriggs, Sherman, Michie, Arthur, Imamura, Wilmore, Frei, & Kufe, 1988; Walker, Walker, Heys, Lolley, Wesnes, & Eremin, 1997; Walker, Wesnes, Heys, Walker, Lolley, & Eremin, 1996). It should be noted that the effects of these cytokines on depressive symptomatology may be mediated by a cascade of other cytokines; for example, IFNa induces the expression and secretion of IL-1 and other cytokines in the periphery (Arenzana-Seisdedos & Virelizer, 1983) and within the CNS (Licinio, Kling, & Hauser, 1998). The findings on cytokine-induced depression have important theoretical and clinical implications. Clinical research in the "50 demonstrated that the monoamine

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

289

depleting drug reserpine, which was then used as an antihypertensive drug, produces severe depression. This observation stimulated the monoamine hypothesis of depression and the development of new and effective antidepressant drugs. In the same way, the finding that exogenous administration of cytokines produces depression can stimulate the cytokinergic hypothesis of depression, and may result in the development of a new generation of effective antidepressant therapeutic procedures.

2.4. The Role of Immune Activation in Depression 2.4.1. Immune Activation and ''Depression Due to a General Medical Condition." Studies on depression in randomly selected general medical inpatients indicate that more than one third report some degree of depression (Laghrissi-Thode et al., 1996; Rodin & Voshart, 1986). As discussed earlier, depression rates can be as high as 50% in certain medical conditions that are specifically associated with high levels of immune activation (e.g., autoimmune diseases, allergy, stroke) (Minden & Schiffer, 1990; Parker et al., 1992; Rodin & Voshart, 1986; Schwartz et al., 1993). The high prevalence of depression in various medical conditions is reflected by the special psychiatric diagnostic entity of "depression due to a general medical condition" (DSM IV, 1994). To diagnose this condition "the clinician should establish the presence of a general medical condition, and determine that the depression is etiologically related to the general medical condition through a physiological mechanism" (DSM-IV, 1994, p. 367). The depressive symptomatology that is associated with physical illness in humans may be produced directly by immune factors or may constitute a psychological reaction to the incapacitation, pain, and losses that accompany the physical disease process. It is difficult to design experiments that will directly differentiate between these two possibilities. However, several lines of evidence support the hypothesis that the direct influence of immune activation on mood and cognition is independent of, and possibly additive to, the maladaptive depressive response to the distress of having a general medical condition: 1) In a study on the neuropsychological effects of experimentally-induced influenza, cognitive disturbances were found to occur not only in sick individuals, but also in subjects with laboratory evidence of viral infection who did not have any clinical symptoms (Smith et al., 1988). 2) In some recurrent infectious conditions, the depressive symptoms precede the clinical manifestations of the disease. For example, in patients with recurrent herpes infections the depressed mood and other psychological alterations are reported by the patients 24-48 hr prior to recurrence of the peripheral skin or genital lesions (Hickie & Lloyd, 1995). This finding had been interpreted as evidence for the effects of depression on the recurrence of the virus, but the temporal relationship between the immunological, psychological, and somatic alterations is more consistent with the hypothesis that viral-induced immune activation is responsible for the psychological changes. 3) The induction of depressed mood and other depressive symptoms in cytokines-treated patients, and the fact that these symptoms appear almost immediately after cytokine administration and usually disappear shortly after termination of the cytokine treatment (Fent & Zbinden, 1987; McDonald et al., 1987; Meyers & Valentine, 1995; Niiranen et al., 1988; Spriggs et al., 1988), strongly suggests a causal role for cytokines in producing the depressive symptoms. Although cytokines are usually administered in the context of a medical condition (e.g., cancer), which by itself could

290

R. Yirmiya et al.

account for some of the depressive symptoms, at least part of the depressive symptomatology can be directly ascribed to the cytokine treatment. For example, in a controlled study on patients with chronic myelogenous leukemia, 50% of the patients had elevated levels of depression following the onset of IFNa therapy, compared to 25% following the onset of chemotherapy (Pavol et al., 1995). 4) Several studies on depression associated with autoimmune diseases suggest that the depressive symptoms reflects the action of a basic physiological mechanism, such as immune activation, rather than a psychological reaction to the consequences of the disease (e.g. functional losses). Compared to patients with other neurological diseases, MS patients showed higher levels of depressive symptoms at the time of the diagnostic interview, and higher number of depressive episodes since their diagnosis had been made (Schiffer & Babigian, 1984; Whitlock & Siskind, 1980). Furthermore, depressed MS patients are frequently characterized by the presence of vegetative symptoms and diurnal variations in mood and energy (Whitlock & Siskind, 1980). This quality of depression may suggest an "organic" rather than a "reactive" depression. In many cases the onset of depression precedes the neurological diagnosis (Schiffer & Babigian, 1984). A prospective study of MS depressed patients revealed that immune dysregulation preceded the development of depression (Foley et al., 1992). Similarly, a study that addressed the relationship of helplessness and depression to disease activity in rheumatoid arthritis (RA) patients revealed that immunological activation might moderate this relationship (Parker et al., 1992). Studies in patients with systemic lupus erythematosus (SLE) are somewhat less supportive of the immune activation hypothesis, since no differences were found in the magnitude and quality of SLE-associated depression, compared to depression associated with other chronic medical conditions (Denburg et al., 1997; Hutchinson et al., 1996; Lim et al., 1988; Magner, 1991). The authors interpreted these results as indicating reactive depression (Denburg et al., 1997). However, since many of their control patients suffered from autoimmune diseases, the involvement of immune mechanisms cannot be ruled out, both in control and in SLE patients. In fact, one study found a direct relationship between depressive manifestations and autoantibodies to ribosomal P proteins in SLE patients (Schneebaom et al., 1991). Together, these findings suggest that at least some of the depressive symptoms that accompany physical illness are not merely a reaction to the medical condition, but are at least partly produced by immune changes preceding and coincide with the appearance of clinical symptoms. 2.4.2. Immune Activation and Other Depressive Syndromes. Immune activation may be involved in depressive syndromes other than "depression due to a general medical condition." Although depression has traditionally been associated with suppression of specific immune functions (Herbert & Cohen, 1993), recent evidence indicates that several components of the immune system are activated in patients suffering from major depression (Connor & Leonard, 1998; Maes, 1995; Maes, Smith, & Scharpe, 1995c). Depression-associated immune activation includes: 1) Increased number of blood lymphocytes, neutrophils, monocytes, and activated Tcells (Maes, Lambrechts, Bosmans, Jacobs, Suy, Vandervorst, De Jockheere, Minner, & Raus, 1992a; Muller, Hofschuster, Ackenheil, Mempel, & Eckstein, 1993; Seidel, Arolt, Hunstiger, Rink, Behnisch, & Kirchner, 1996); 2) Increased serum levels of several soluble indicators of activated immune cells, including interleukin-2 receptor (Maes, Meltzer, Bosmans, Bergmans, Vandoolaeghe, Ranjan, & Desnyder, 1995a; Sluzewska,

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

291

Rybakowski, Bosnians, Sobieska, Berghmans, Maes, & Wiktorowicz, 1996), neopterin (Dunbar, Hill, Neale, & Mellsop, 1992; Maes, Scharpe, Meltzer, Okayli, D'Hondt, & Cosyns, 1994) and prostaglandin E2 (Lieb, & Karmali, 1983; Linnoila, Whorton, Rubinow, Cowdry, Ninan, & Waters, 1983); 3) Increased serum concentrations of positive acute phase proteins (APPs) and decreased levels of negative APPs (Maes, Scharpe, Neels,Wauters, Van Gastel, & Cosyns, 1995b; Maes, Scharpe, Van Grootel, Uyttenbroeck, Cooreman, Cosyns, & Suy, 1992b; Seidel, Arolt, Hunstiger, Rink, Behnisch, & Kirchner, 1995; Sluzewska et al, 1996; Song, Dinan, & Leonard, 1994); and 4) Increased secretion of cytokines, both in vivo (particularly IL-6) (Maes et al., 1995a; Maes, Bosmans, De Jongh, Kenis, Vandoolaeghe, & Neels, 1997a; Sluzewska, Rybakowski, Laciak, Mackiewicz, Sobieska, & Wiktorowicz, 1995; Sluzewska et al., 1996), and following in vitro induction by mitogens (particularly IL-ip, IL-6, and IFNy) (Maes, Bosmans, Meltzer, Scharpe, & Suy, 1993; Maes et al., 1994; Seidel et al, 1995; Seidel et al., 1996). Furthermore, immune activation is positively correlated with specific depressive symptoms and with the impaired feedback regulation of the HPA axis, found in major depression patients (Maes et al., 1993; Maes, 1995). Based on these findings, Maes and his colleagues hypothesized that production of interleukins contributes to the HPA hyperactivity and to the vegetative symptoms of severe major depression (Maes, 1995). Clearly, immune activation and cytokine secretion do not account for all types of depressive disorders. However, immune factors may be involved in the pathophysiology of certain subtypes of depression (e.g., melancholia), characterized by a constellation of symptoms that are often found in virus-infected and cytokineinjected individuals. The source of immune activation in any depressive disorder other than "depression due to a general medical condition" has not been identified yet. It is possible, however, that subclinical infectious processes or viral reactivation induce immune reaction, which in turn contributes to the depressive symptomatology. This hypothesis is supported by studies reporting increased antibody titers to several viruses, particularly herpes simplex virus (HSV) in patients with major depression (Cappel, Gregoire, Thiry, & Sprecher, 1978; Lycke, Norrby, & Roos, 1974). Moreover, in a recent study (Zorzenon, Colle, Vecchio, Bertoh, Giavedoni, Degrassi, Lavaroni, & Aguglia, 1996) clear evidence was provided for active viral multiplication and elevated antibody titers to HSV in 4 1 % of the patients with major depression. Similarly, higher serum antibodies to Borna disease virus, as well as isolation of this virus from patients with depression has been reported (Bode, Durrwald, Rantam, Ferszt, & Ludwig, 1996). 3. B E H A V I O R A L E F F E C T S O F I N F E C T I O U S D I S E A S E S A N D CYTOKINE ADMINISTRATION IN A N I M A L S In animals, the association between acute illness and suppression of general activity, appetite and grooming has been reported by animal handlers and veterinarians long ago. Experimental studies, using systemic protozoan, bacterial or viral infections, provided more conclusive evidence for sickness-induced behavioral changes (Hart, 1988). In addition, exogenous administration of proinflammatory cytokines produces sickness behavior, whereas cytokine antagonists, as well as cytokine synthesis blockers, attenuate the behavioral effects of pathogens. These findings indicate that sickness behavior

292

R. Yirmiya et al.

is mediated by immune-derived cytokines, rather than being produced by the pathogen itself. Several components of sickness behavior, which resemble the characteristics of depression, have been described in animal research:

3.1. Anorexia and Body Weight Loss Anorexia and body weight loss are among the most robust effects of acute as well as chronic illness (Plata-Salaman, 1996). Such effects have been recently demonstrated in experimental models of disease, including influenza virus in-fection (Swiergiel, Smagin, & Dunn, 1996; Swiergiel, Smagin, Johnson, & Dunn, 1997), local inflammation induced by subcutaneous injection of turpentine (Kozac, Poli, Soszynski, Conn, Leon, & Kluger, 1996; Kozac, Soszynski, Rudolph, Conn, & Kluger, 1997), and exogenous administration of pathogen products, including LPS (Kozac et al., 1997; Yirmiya, 1996), and heat-inactivated Mycoplasma fermentans (Yirmiya, Barak, Avitsur, Gahlly, & Weidenfeld, 1997). Several lines of evidence suggest that anorexia and body weight loss are mediated by cytokines: 1) Similar effects were observed following exogenous administration of cytokines, particularly IL-ip, TNFa, and IL-8, which act centrally and synergistically to suppress feeding (PlataSalaman, 1998; Sonti, Ilyin, & Plata-Salaman, 1996); 2) The anorexic effects of influenza virus infection or LPS could be attenuated by pretreatment with IL-lra (Swiergiel et al., 1997); 3) Increase in dietary N-3 fatty acids, which are known to reduce cytokine secretion, abolished the effects of local inflammation and LPS administration (Kozac et al., 1997); and 4) Mice deficient in IL-6 (IL-6 knockout) exhibited attenuated anorexia and weight loss following local inflammation or influenza pneumonitis (Kozac et al., 1996).

3.2. Hypersomnia Sleepiness and altered sleep patterns are among the earliest signs of infection. Increased somnolence is produced by specific pathogen products, such as muramyl peptides, LPS, and viral double-stranded RNA (Krueger & Majde, 1994). Changes in sleep patterns are characterized by an increase in slow-wave sleep (SWS) and inhibition of rapid eye movement (REM) sleep. Exogenous administration of IL-l,TNFa or IFN-a, either peripherally or into the brain, also induce somnolence (Krueger, Takahashi, Kapas, Bredow, Roky, Fang, Floyd, Renegar, Guha-Thakurta, Novitsky, & Obal, 1995). Moreover, brain-derived IL-1 has recently been implicated in sleep responses following systemic, as well as central bacterial infection (Takahashi, Kapas, Fang, Seyer, Wang, & Krueger, 1996).

3.3. Psychomotor Retardation, Fatigue, and Reduced Exploratory Behavior Psychomotor retardation, fatigue, and reduced exploratory behavior were found in various models of infectious disease, as well as following cytokine administration. For example, a decrease in locomotor and exploratory activity was produced by influenza pneumonitis, turpentine abscess (Kozac et al., 1996, 1997), Mycoplasma fermentans (Yirmiya et al., 1997), LPS (Dunn, Chapman, & Antoon, 1992; Yirmiya, 1996; Yirmiya, Rosen, Donchin, & Ovadia, 1994), IL-1 (Spadaro & Dunn, 1990), and

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

293

IFN-a (Segall & Crnic, 1990). Similar symptoms, as well as increased anxiety behavior were also reported, using a mouse model of autoimmune disease (systemic lupus erythematosus) (Schrott & Crnic, 1996), indicating that behavioral changes can also accompany non-infectious conditions.

3.4. Impaired Cognitive Abilities The presence of IL-1 and IL-6 receptors in the hippocampus, and the findings that LPS, IL-1 and TNFa modulate neural processes thought to be involved in learning (including synaptic inhibition and long-term potentiation) (Cunningham, Murray, O'Neill, Lynch, & O'Connor, 1996; Katsuki, Nakai, Hirai, Akaji, Kiso, & Satoh, 1990), suggest a role for cytokines in cognitive functions. Such a role has recently been demonstrated by showing that bacterial, parasitic or viral infections, and LPS or IL-1 administration, impair learning in various paradigms, including autoshaping (Aubert, Vega, Dantzer, & Goodall, 1995), and spatial navigation learning in the Morris maze and radial arm maze (Beers, Henkel, Kesner, & Stroop, 1995; Gibertini, 1996; Gibertini, Newton, Friedman, & Klein, 1995; Kavaliers, Colwell, & Galea, 1995; Oitzl, van Oers, Schobitz, & de Kloet, 1993). These effects are not confounded by pyrogenic or general performance effects of immune challenges. These findings are particularly interesting because immune activation and cytokine release in the brain accompany many neurodegenerative disorders (Rothwell et al., 1996). A critical role for cytokines in cognitive deficits associated with neurodegenerative disorders was recently demonstrated in transgenic mice expressing interleukin-6 in the brain; inflammatory neurodegeneration was accompanied in these mice by progressive decline in avoidance learning (Heyser, Masliah, Samimi, Campbell, & Gold, 1997).

3.5. Impaired Social Behavior The effect of immune activation on social behavior has been mainly studied using the behavioral test of olfactory exploration of a conspecific juvenile. The time spent by adult rats or mice in social exploration of a conspecific juvenile was profoundly reduced following systemic or intracerebral administration of LPS (Yirmiya, 1996; Johnson, Propes, & Shavit, 1996), Mycoplasma fermentans (Yirmiya et al., 1997), IL-1 (Bluthe, Dantzer, & Kelley, 1997; Kent, Bluthe, Dantzer, Hardwick, Kelley, Rothwell, & Vannice, 1992b), and TNFa (Bluthe, Dantzer, & Kelley, 1991). LPS administration also reduced aggressive attacks in mice selectively bred for high aggressive behavior (Granger, Hood, Ikeda, Reed, Jones, & Block, 1997).

3.6. Altered Pain Perception Alterations in pain perception accompany inflammation, infection, autoimmune diseases and nerve injury. Proinflammatory cytokines, associated with these conditions, were found to activate neural circuits which modulate pain perception (Watkins, Maier, & Goehler, 1995b). Typically, cytokine secretion results in an immediate phase of hyperalgesia (increased responsiveness to painful stimuli) (Watkins et al., 1995b), which is later followed by a prolonged analgesic phase (Romanovsky, Kulchitsky, Akulich, Koulchitsky, Simons, Sessler, & Gourine, 1996; Yirmiya et al., 1994).

294

R. Yirmiya et al.

3.7. Anhedonia Anhedonia means "without (an) pleasure (hedonia)" and can be defined as the diminished capacity to experience pleasure of any sort, i.e., activities that previously brought enjoyment, such as eating, sex, social, and recreational activities, provide little or no gratification in the depressed patient. Anhedonia is considered as one of the two essential features of a major depressive episode, and is even further emphasized in the subclassification criteria of a major depressive episode with melanchoHa (DSM-IV, 1994). The hypothesis that immune activation produces anhedonia has been studied in experimental animals using several paradigms, including the consumption of and preference for sweet solutions, intracranial electrical self stimulation (ICSS), and tests of libido and sexual activity. 3.7.1. Reduced Consumption of and Preference for Sweet Solutions. The consumption of and preference for sweet solutions can serve as a model for hedonic processes, because when presented with such solutions, animals will drink more fluid than they usually drink, and if given a choice, they will prefer these solutions over water. Moreover, non-hungry or thirsty animals will perform operant tasks to obtain sweet solutions as a rewarding stimulus, in the same way that they perform for ICSS and other rewards. Our studies demonstrated that various immune challenges attenuated the consumption of and preference for sweet solutions, while having minimal effects on water drinking (Table 1). LPS significantly decreased saccharin preference in fluid-deprived rats, and suppressed free consumption of saccharin solution in nondeprived rats; water consumption was not affected under these circumstances (Yirmiya, 1996). A similar decrease in saccharin-, but not water-consumption was demonstrated following intracerebral administration of Mycoplasma fermentans (Yirmiya et al., 1997) or HIV-1 gpl20 (unpublished results). In addition, we have recently demonstrated decreased preference for a dilute sucrose solution in rats with experimental allergic encephalomyelitis (EAE), an established model of multiple sclerosis in humans.

Table 1. Effects of various immune challenges on the consumption of sweet solutions Immune challenge LPS(50ng/kg,i.p.) Saline IL-l(50ng,i.c.v) Saline Mycoplasma fermentans (i.c.v.) Saline HIV gpl20 (i.c.v.) Saline EAE Control

Sweet solution (ml)* 25.3 (10.5) 55.6 (7.2) 17.2 (7.0) 45.6 (7.9) 16.0 (6.9) 39.1 (4.6) 38.9 (5.1) 51.5(3.1) 0.78 (0.20) 12.26 (0.22)

Water (ml)* 17.5 (2.8) 21.9(3.1) 1.7(0.5) 0.8 (0.3) 4.2(1.4) 1.3 (0.5) 0.6 (0.3) 1.6(1.1) 0.56 (0.10) 0.54 (0.07)

*For each immune challenge, animals were presented with two graduated tubes containing the sweet solution or water. Data represents the mean (+ S.E.M.) fluid consumption over a 24 hr period in non-deprived rats, following administration of LPS, IL-ip, Mycoplasma fermentans or gpl20. Data for the acute immune challenges referes to the consumption of a 10 mM saccharin solution by rats. Data for the EAE model refers to the consumption of a dilute sucrose solution by SJL mice over a period of 6 hr (averge of 2 sessions).

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

295

Interestingly, the decrease in sucrose preference preceded the appearance of clinical symptoms in this model, demonstrating the independence of anhedonia from the physical disease symptoms (unpublished observation). Studies in other laboratories corroborate these findings, demonstrating that immune challenges reduced the intake of sweetened milk (Swiergiel et al., 1996,1997), and abolished the reinforcing effect of cocaine (Suzuki, Funada, Sugano, Misawa, Okutomi, Soma, & Mizuno, 1994). Moreover, mice that spontaneously develop systemic autoimmune lupus-like disease, also show blunted sensitivity to sucrose, which can be reversed by an immunosuppressive treatment (Sakic, Denburg, Denburg, & Szechtman, 1996). Several previous studies have used the consumption of sweet solutions as a model of anhedonia, describing a decrease in the consumption of and preference for sucrose and saccharin solutions in an animal model of depression (exposure to chronic mild unpredictable stress) (Willner, 1997). However, the use of consumption of sucrose or other calorie-rich highly palatable diets as a measure of cytokine-induced anhedonia is problematic, because of the marked anorexia and loss of body weight that accompany immune activation. Thus, a decrease in the total amount of sucrose or sweetened milk consumption may reflect either anhedonia and/or anorexia and reduced calorie intake. Because saccharin has a sweet taste but no calories, our findings that immune activation induced reduction in saccharin preference provide further support for the anhedonic effect of immune challenges. It should be noted, however, that the validity of the saccharin preference test as a model of anhedonia is controversial. Previous research, demonstrated that although the major component of saccharin's taste is sweet, there is also an aversive, bitter-like component (Dess, 1993). Aubert & Dantzer (1998). have recently studied the effects of LPS on taste reactivity (TR) to saccharin, sucrose and quinine solutions. They showed that following LPS administration, TR to quinine and sucrose were unaltered, but the responses to saccharin or to a sucrose solution adulterated with quinine were changed, such that more aversive responses and less ingestive responses were exhibited by LPStreated animals. These findings indicate that LPS accentuates the aversive component of the rewarding saccharin stimulus, suggesting that LPS produces increased finickiness rather than anhedonia. Nevertheless, it should be noted that changes in TR do not necessarily represent the overall hedonic response of the animal or its inclination to consume a palatable solution. TR responses are mediated by hedonic mechanisms in the lower brain stem taste centers, as evidenced by their preservation in decerebrate animals (Grill & Norgren, 1978). Hedonic evaluation assessed by the TR test can be at odds with consummatory or instrumental behavior. For example, in lines of rats that were genetically selected for high and low saccharin consumption, the high-consuming rats manifested more rejection responses to saccharin in the TR test (Badia-Elder, Kiefer, & Dess, 1996). Thus, it is possible that the effects of LPS and other immune challenges on saccharin preference is not solely explained by increased finickiness, but may be also related to their effects on incentive-motivational systems in brain centers higher than the brain stem. In conclusion, studies on the intake of and preference for palatable solutions as a model for anhedonia have some methodological problems, and therefore should employ both nutritive and non-nutritive solutions. 3.7.2. Decreased Responding for Rewarding Intracranial Self Stimulation. Further evidence for the anhedonic effects of immune activation is provided by investigation of the effects of immune challenges and cytokines on rewarding intracranial self stimulation (ICSS). Suppression of ICSS is a very useful animal model to study

296

R. Yirmiya et al.

anhedonia (Willner, 1994). Endotoxin-induced suppression of ICSS has been demonstrated more than 3 decades ago (Miller, 1964). This report was recently corroborated by the findings that exogenous administration of LPS (Borowski, Kokkinidis, Merali, & Anisman, 1997) or IL-2 (Anisman, Kokkinidis, Borowski, & Merali, 1998; Anisman, Kokkinidis, & Merali, 1996) produce a dramatic and long-lasting decrease in ICSS. Antigenic challenge with sheep red blood cells also reduced ICSS from rat nucleus accumbens at times that approximated the peak of immune response (Zacharko, Zalcman, Macneil, Andrews, Mendella, & Anisman, 1997). As shown in the above studies, the effects of immune challenges on ICSS represent a specific motivational change, rather than motoric, soporific, attentional or cognitive deficit. In contrast to the effects of LPS and IL-2, administration of IL-1|3 had no clear effect on ICSS: No effect of IL-1|3 on ICSS was detected following administration of a dose comparable to an effective dose of IL-2 (1 |ig/rat), whereas a higher dose (2 fig/rat), which induced overt signs of illness, attenuated ICSS (to a lesser extent than IL-2), with some of the animals ceasing responding almost completely and the others affected only slightly (Anisman et al., 1998). The authors conclude that whereas IL-2 reduce ICSS by producing anhedonia, the effects of IL-1 in this paradigm are secondary to illness. It should be noted, however, that these findings do not entirely rule out the possibility that IL-1 is involved in hedonic processes. It is possible that in contrast to the "pure" anhedonia produced by IL-2 (i.e., disruption of ICSS with no other signs of illness), illness and anhedonia are tightly coupled in IL-1-injected animals. The anhedonic effect of LPS is also coupled with sickness behavior; in fact, the dose of LPS (100 |j.g/kg), which was effective in decreasing ICSS, probably produced more illness than the ineffective dose of IL-1 (1 |ig/rat). Thus, the possibility that IL-1 is at least partially involved in mediating the anhedonic effects of LPS can not be ruled out. Furthermore, several investigators have demonstrated that IL-1 acts synrgistically with other cytokines in producing sickness behavior symptoms. Thus, in future research the possibility that IL-1 influences hedonic processes via interactions with other cytokines should be examined. 3.7.3. Reduced Libido and Impaired Sexual Activity. A significant reduction in sexual interest or desire, as well as difficulties in sexual functioning are commonly associated features of depression (DSM-IV, 1994; Mathew & Weinman, 1982). These features are viewed as manifestation of the general loss of interest and pleasure in activities that were previously considered pleasurable (DSM-IV, 1994). Indeed, in animal studies, sexual contact has been shown to act as a reinforcing stimulus (Meyerson & Lindstrom, 1973), i.e., its motivational effects are similar to those of other reinforcing stimuli such as ICSS or palatable food. To further explore this anhedonic aspect of immune activation, we have recently examined the effects of LPS and IL-ip on sexual behavior in male and female rats, using various behavioral tests. Low doses of LPS, administered either peripherally or centrally, significantly reduced sexual motivation, proceptive (soliciting) behavior and receptivity in females. Higher doses of LPS completely abolished all aspects of female sexual behavior. In contrast, male sexual behavior was affected only by high doses of LPS (Avitsur, PoUak, & Yirmiya, 1997; Yirmiya, 1996) . Similarly, administration of IL-ip had no effect on any component of male sexual behavior, whereas even low doses of IL-ip, administered either peripherally or centrally, markedly decreased sexual motivation, proceptive behavior, receptivity, and attractivity in females (Avitsur et al., 1997; Avitsur, Cohen, & Yirmiya, 1998).

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

297

In conclusion, various immune challenges induce anhedonia as well as many behavioral alterations, which resemble the vegetative symptoms of depression. These findings suggest that immune activation produces a depression-like syndrome in animals.

4. IMMUNE ACTIVATION AND DEPRESSION ARE BOTH ASSOCIATED WITH SIMILAR NEUROENDOCRINE ALTERATIONS Depression is characterized by marked alterations in neuroendocrine function. In particular, dysregulation of the HPA axis has been suggested as an important aspect of the pathophysiology of depression. Patients with major affective disorders exhibit elevated serum, urinary, and cerebrospinal fluid (CSF) levels of Cortisol (CaroU, Curtis, & Mendels, 1976; Sachar, Hellman, Fukushima, & Gallaghaer, 1970;Traskman,Tybring, Asberg, Bertilsson, & Schalling, 1980), abnormal 24-hr Cortisol secretory patterns (Linkowski, Mendlewicz, Leclercq, Brasseur, Hubain, Goldstein, Copinschi, & van Cauter, 1985), and non-suppression of serum Cortisol following dexamethasone administration (Carroll, Martin, & Davies, 1968; Holsboer & Harden, 1996). The excessive adrenal Cortisol secretion in depressed patients probably reflects abnormal limbic-hypothalamic activation, which results in increased production and secretion of corticotrpin-releasing hormone (CRH) (Gold, Chrousos, Kellner, Post, Roy, Auerginos, Schulte, Oldfield, & Loriaux, 1984; Holsboer & Harden, 1996; Yehuda & Nemeroff, 1994). This hypothesis is supported both by preclinical studies, demonstrating that intracerebroventricular administration of CRH induces behaviors in animals that resemble symptoms of depression (Owens & Nemeroff, 1993), and by clinical studies indicating that CRH concentrations in the CSF and CRH mRNA in the hypothalamus are increased in depressed patients (Holsboer & Harden, 1996; Owens & Nemeroff, 1993). Immune activation produces neuroendocrine alterations, which are very similar to the alterations found in most depressed patients. Viral infection, in vivo antibody production, and direct administration of cytokines, including TNFa, IL-ip, and IL-6, have all been shown to activate the HPA axis (Besedovsky & Del Rey, 1996;Turnbull & Rivier, 1995). Such activation is mainly produced by the release of hypothalamic CRH (Rivier & Rivest, 1993;Tilders, Derijk, Van Dam, Vincent, Schotanus, & Persoons, 1994; Besedovsky & Del Rey, 1996). The findings that increased production and release of CRH are concomitant of both depression and immune activation, and the increased production of cytokines in depressed patients (Maes, 1995; Maes et al., 1995c) suggest a role for cytokines in mediating HPA axis hyperactivity in depression. Support for this hypothesis was provided by Maes et al., (1993), who described a significant correlation between IL-ip production and post-dexamethasone (post-DEX) Cortisol levels in depressed patients. According to this report, depressed patients who showed higher mitogen-induced in vitro production of IL-lp had higher post-DEX Cortisol levels. This finding may be related to the results of two recent studies on the effects of immune challenges on glucocorticoid (GC) negative feedback regulation in animal models. In one study, endotoxin or IL-1 administration decreased the affinity of corticosteroid receptors, which mediate feedback inhibition of the HPA system (Schobitz, De Kloet, & Holsboer, 1994). In another study (Weidenfeld & Yirmiya,

R. Yirmiya et al.

298 SO]

Pretreatment: Saline X 5 Days

Pretreatment: LPS X 5 Days

I

40-

• Basal EJ Photic Stress

X

30-

O

f^'--'

V

DEX

r?W'

DEX

Figure 2. Effects of LPS on the glucocorticoid negative feedback regulation of the adrenocortical response.

1996), we examined the functional significance of this phenomenon. Rats were administered with either LPS or saline for 5 days. Two days following this treatment, rats were administered with either vehicle or dexamethasone, and 3.5 h later they were exposed to an acute stressful photic stimulation. In saline-pretreated rats, photic stimulation caused a marked elevation of serum corticosterone levels, and pretreatment with DEX completely abolished this response (Fig. 2). In LPS-pretreated animals, corticosterone levels following photic stimulation were significantly greater than in saline-treated animals, and DEX was ineffective. The effects on the negative feedback were specific to LPS, since other treatments, which mimic various aspects of prolonged LPS exposure, such as corticosterone administration or chronic stressful exposure, did not produce the same impairment of the GC negative feedback (Weidenfeld & Yirmiya, 1996). Together, these findings indicate that immune activation produces a neuroendocrine alteration, which is similar to one of the central abnormalities found in most depressed patients. Serum corticosterone (CS) levels were measured under basal conditions and following exposure to stressful photic stimulation, in rats treated with either dexamethasone (DEX, 25 |Xg/Kg, i.p.) or vehicle, 3.5 hr prior to the onset of the stressor. The photic stimulation was applied 2 days following the 5th daily exposure to either saline or LPS. Pretreatment with LPS impaired the glucocorticoid negative feedback, reflected by elevated stress-induced CS secretion and insensitivity to DEX.

5. ADAPTIVE AND MALADAPTIVE ASPECTS OF IMMUNOLOGICALLY-INDUCED DEPRESSION Immunologically-induced depression-like syndrome may be adaptive during acute, and possibly even chronic diseases. During infection, inflammation, or neurological disease processes, functioning of the organism is disturbed, and it may be advantageous for the organism to refrain from most activities, and postpone active coping until recuperation begins to take place. Thus, depression-like sickness behavior serves to "keep the organism out of troubles" during periods of high vulnerability. The behavioral effects of cytokines may serve several more specific functions: Psychomotor retardation serves to save body energy levels and maintain the highly

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

299

adaptive febrile response (Hart, 1988); Anorexia reduces the motivation of the organism to look for food, which may involve expenditure of energy. It also reduces the consumption of nutrients that are essential for the growth and proliferation of many pathogens (e.g., iron) (Hart, 1988); Anhedonia may further ensure the lack of motivation to engage in goal-directed behaviors, including social activity and sexual behavior; Finally, the inhibition of female sexual behavior following activation of the immune system may be adaptive in preventing conception when the animal is sick, thus reducing the risk of spontaneous abortion, prenatal infection and abnormal development (Avitsur et al., 1997; Avitsur et al., 1998; Yirmiya, Avitsur, Donchin, Cohen, 1995). A depressive-like response to infection in the same way that pain is adaptive in response to tissue injury. Whereas pain serves to protect and attend to a specific tissue or organ, which usually displays local inflammation, "depression due to a general medical condition" serves to protect the whole organism when the disease and inflammation are systemic, involving multiple sites and systems, and/or when the inflammatory processes occur within the brain. It is also possible that many of the depressive symptoms associated with disease and immune activation are not adaptive in terms of the individual; rather, they are adaptive for the population. Sociobiologists argue that complex social behaviors may be selected for, in the course of evolution, even though the effect of the behavior on its bearer is to reduce its own personal fitness (Wilson, 1975). For example, an animal that utters a loud alarm call is drawing attention to itself, increasing the likelihood that it will be captured by the predator. Theories of inclusive fitness and kin selection provide convincing evidence for the adaptive value of such "altruistic" behaviors (Wilson, 1975). A similar phenomenon may explain the behavioral effects of infectious diseases, because such diseases may be transmitted from the individual to its kins and family members, thus reducing the individual's "inclusive fitness" (i.e., the net genetic representation in succeeding generations, including other relatives in addition to offspring). According to this view, the infected subject's withdrawal from regular activities, particularly those which involve social interactions, reduces the risk of infection spread in the population. This may be particularly important in species where the same food is shared by the whole group. In such a situation, transmission of pathogens is particularly likely, and immune-activation-induced anorexia and social withdrawal may be highly adaptive. Obviously, adaptive immune-mediated depression-like syndrome should be transient and restricted to the time period in which the organism is sick. Indeed, many factors have been documented to tightly regulate and limit the behavioral, neural and neuroendocrine effects of cytokines, including glucocorticoids, vasopressin, and amelanocyte stimulating hormone (a-MSH) (Dantzer et al., 1996). Disruption of these regulatory factors might lead to impairment in "shutting off" the immune and neural mechanisms underlying sickness behavior, thus resulting in maladaptive depressive symptomatology. This process may underlie the Chronic Fatigue Syndrome (CFS) and the Post-Viral Fatigue Syndrome (PVFS). Both syndromes are associated with psychological changes that persist long after recuperation from viral infection (Komaroff, Fagioli, Geiger, DooHttle, Lee, Kornish, Gleit, & Guerriero, 1996). They are mainly characterized by fatigue, which reduces patients' level of everyday activity by at least 50%. Additional symptoms are low fever, sore throat, enlargement, and/or pain of the lymph nodes, headaches, depression, anxiety, confusion, bad temper, and difficulties in concentration (Klonoff, 1992; Komaroff & Buchwald, 1991). Depression occurs in 70-85% of all CFS and PVFS patients (Abbey & Garfinkel, 1991; Komaroff

300

R. Yirmiya et al.

& Buchwald, 1991). Both syndromes are also associated with aherations in immune functions, including abnormal levels of lymphocytes, hyperactive immunological responses, high levels of antibodies to Epstein-Barr Virus, and other viruses, and impaired regulation of cytokine secretion (Buchwald & Komaroff, 1991; Komaroff & Buchwald, 1998). Blood levels of Cortisol, which normally limits the extent of brain cytokine production and their behavioral impact (Goujon, Parnet, Laye, Combe, Kelley, & Dantzer, 1995; Johnson et al., 1996), are lower in these patients (Cleare, Beam, Allain, McGregor, Wessely, Murray, & O'Keane, 1995). Thus, CFS and PVFS seem to involve impairment in at least one of the mechanisms responsible for down regulating the immune system, along with its psychological effects, following recuperation from a viral infection. Other forms of depressive disorders are also associated with disruption of antiinflammatory neuromediators. For example, the levels of vasopressin and a-MSH are lower in depressed patients, particularly those with melancholia (Laruelle, Seghers, Goffinet, Bouchez, & Legros, 1990; Maes, DeJonckheere, Vandervorst, Schotte, Cosyns, Raus, & Suy, 1991), Paradoxically, acute infection occurring during severe depression may produce (in addition to exacerbated depressive symptoms) a rebound activation of the anti-inflammatory system, which could temporarily attenuate the depressive symptomatology. The effects of acute endotoxin injection on depressed patients was recently examined (Bauer, Hohagen, Gimmel, Bruns, Lis, Krieger, Ambach, Guthmann, Grunze, Fritsch-Montero, Weissbach, Ganter, Frommnerger, Riemann, & Berger, 1995); as found, proinflammatory cytokines levels and body temperature were elevated during the first 6hr following the injection, and the patients "exhibited pronounced apathy". However, 15 hr after endotoxin administration, when cytokine levels and fever subsided, the patients reported a significantly improved mood. This effect was transient and disappeared on the next day (Bauer et al., 1995). Thus, it is possible that activation of rebound "shutting off" mechanisms of immune activation in severely depressed patients can improved mood, at least transiently.

6. EFFECTS OF ANTIDEPRESSANTS ON DEPRESSION INDUCED BY IMMUNE ACTIVATION Antidepressants have been used successfully in treating depressive symptoms associated with various medical conditions (Katon & Sulhvan, 1990). Both tricyclic antidepressants (TCAs) (Schiffer & Wineman, 1990) and selective serotonin reuptake inhibitors (SSRIs) (Scott, Nussbaum, McConnell, & Brill, 1995) have been used successfully in the treatment of depression associated with multiple sclerosis, stroke (Lauritzen, Bjerg Bendsen, Vilmar, Bjerg Bendsen, Lunde, & Bech, 1994), Alzheimer's disease (Gottfries, 1997;Tueth, 1995), HIV infection (Ayuso, 1994; Rabkin, Wagner, & Rabkin, 1994), and depression induced by IFN administration (e.g., in hepatitis C or multiple sclerosis patients) (Levenson & Fallon, 1993; Mohr, Goodkin, Likosky, Gatto, Baumann, & Rudick, 1997). To further elucidate the relationship between immune activation and depression, and explore the mechanisms underlying the therapeutic action of antidepressants, we employed an experimental animal model. Specifically, we examined the effects of antidepressants on LPS- or IL-1-induced behavioral and neuroendocrine alterations in rats.

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

301

6.1. Effects of Imipramine and Fluoxetin on LPS- and IL-1-Induced Sickness Behavior and Adrenocortical Activation We first tested the effects of the tricyclic antidepressant imipramine on LPS-induced anhedonia, using the saccharin preference paradigm. Rats were habituated to a drinking deprivation schedule and to the taste of saccharin for several days before the experiment. On the first experiment day, rats were injected acutely with either imipramine or saline, immediately followed by an injection of either LPS or saline. Four hours later, animals were presented with two tubes containing either saccharin or water, and fluid consumption was measured over a priod of 20 minutes. For 3 weeks following this test, rats received only water, along with a daily injection of either imipramine or saline according to their respective group assignment during the first drinking test. On the last injection day (chronic imipramine test), the rats received either LPS or saline, according to their group assignment during the acute imipramine test, and the consumption of saccharin and water was measured 4 hours later. In the first expeimenal day, LPS significantly suppressed the consumption of saccharin, but not water, and this suppression was not attenuated by acute administration of imipramine. However, chronic imipraminetreatment, completely abolished the suppressive effect of LPS on saccharine consumption (Yirmiya, 1996). Chronic administration of imipramine (daily injection for 5 weeks) attenuated many other behavioral effects of LPS. Imipramine-treated rats exhibited facilitated recovery from LPS-induced anorexia, body-weight loss, and reduced social activity. In addition, they did not demonstrate LPS-induced suppression of locomotor and exploratory behavior in the open field test. In another experiment, we showed that acute administration of imipramine did not have such effects, i.e., there was no effect of acute imipramine on LPS-induced reduction in food consumption, body weight, social exploration, and open-field activity. The dissociation between the effects of acute and chronic imipramine treatment is important, since in clinical settings, imipramine is also effective in alleviating depression only following chronic, but not acute administration (Montgomery, 1994). In subsequent experiments we showed that chronic administration (daily injections, 5 weeks) of fluoxetin also affected some, but not all the alterations produced by LPS. Fluoxetin significantly attenuated LPS-induced reduction in food consumption and body weight, but did not affect LPS-induced decrease in social interaction and activity in the open-field test. Chronic fluoxetin treatment also attenuated LPS-induced secretion of corticosterone (Yirmiya, Weidenfeld, Pollak, Avitsur, Barak, & Ovadia, unpublished observation). In contrast to the effects of antidepressants on the responsiveness to LPS, we found no evidence for an effect of imipramine on the behavioral and neuroendocrine changes induced by IL-1. Chronic treatment with imipramine had no significant effect on IL-1-induced reduction in food consumption, body weight, social exploration, open-field activity, and corticosterone secretion. This finding indicates that the behavioral response to IL-1 in imipramine-treated rats is normal. Thus, the effects of imipramine on LPS-induced behavioral changes are at least partly mediated by changes in immune reactivity to LPS (particularly reduced production of cytokines), and not by down-stream effects on IL-1-induced sickness behavior.

R. Yirmiya et al.

302

^ • i ^

s en a.

CO _ j

^
_i

en en E

s

•o <s

1-1

ff

Vi

e

(13

5

Q B Q ^

o c

>> n

o

JD Ti

E

S

o

"3 C!

f^

ii

i?

u CD Ut

E

o E tl.

•

'

£

—

'

r

si

b

c

)L

UJ c ctj

at

u

F

;S <

Cl, c

o

.1

' w

JC

,_ en (Si

E E

S B Q ^

1

c

^ pa

3 ra u

o

F *9

^tp

FS

m td

c

o

u

^

(|iu) uojidiJunsuoQ pini-j

CL ^C

w

-^ F ^

(D

o :f c

it

o to

fl

CI

^H

c

fi

-o u

c o

«^ o

y

c

VI r/l Q-,

J O

4 «5 X) c rt 01

^ —

.

w

••-*

t> s

r:

o

"cS

JZ

s

m

_G'

r.)

•

4>

!•)

n

OS VI

O Ci

c

E

trs

1 k-

D

y

U

h _b

^ ^

o

'E*

c (fl -C u

U _] 3 y

«

o

fj

cd

C^ E •o

u

[^ 3

4=

,[-;

UJ

-* •n

1^

(|Lu) uaijdiunsuoo pmy

> i

1

u ;3

tZ E

ai)

r.,

E o u .2

^

in

fl> pa

C CJ

E pa

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

303

6.2. Mechanisms of the Effects of Antidepressants o n Immune Activation-Induced Depressive Symptoms 6.2.7. Alterations in Cytokine Production. The effects of antidepressants on immune functions, including cytokine production, have been assessed both in depressed patients, and in experimental animal models. In depressed patients, the effects of antidepressants seem to be related to the immune status of the patients at the initiation of the treatment. When depression was associated with immune activation, antidepressants suppressed immune function and cytokine secretion. For example, the increased plasma levels of IL-6 during acute depression were normalized by 8-week treatment with fluoxetin (Sluzewska et al., 1995), the increased monocyte counts in depressed patients were reduced following 6-weeks treatment with tricyclic antidepressants (Seidel et al., 1996), and the increased numbers of leukcytes and neutrophils were also reduced by antidepressant treatment (Maes, Vandoolaeghe, Van Hunsel, Bril, Demedts, Wauters, & Neels, 1997b). On the other hand, when immune functions were found to be normal, antidepressants had no effects, e.g., chronic moclobemide treatment had no effect on monocytes functions, TNFa production or IFNy levels (Landmann, Schaub, Link, & Wacker 1997). Moreover, in a study of depressed patients who exhibited immune suppression before treatment, the tricyclic antidepressant clomipramine increased the production of IL-lp, IL-2, and IL-3 (Weizman, Laor, Podliszewski, Notti, Djaldetti, & Bessler, 1994). In experimental animals, antidepressants (both TCAs and SSRIs) produce mainly immune suppression and anti-inflammatory effects (Albrecht, Helderman, Schlesser, & Rush, 1985; Audus & Gordon, 1982; Eisen, Irwin, Quay, & Livnat, 1989; Miller, Asnis, van Praag, Norin, 1986; Xiao & Eneroth, 1996). Antidepressant treatment in vivo was found to reduce immune activation in several models: it inhibited the increased acute phase response in olfactory bulbectomized rats, a useful animal model of depression (Song & Leonard, 1994), reduced IL-1 and IL-2 production in a chronic mild stress model of depression (Kubera, Symbirtsev, Basta-Kaim, Borycz, Roman, Papp, & Claesson, 1996), and inhibited immune activation in rats with experimental allergic neuritis (Zhu, Bengtsson, Mix, Thorell, Olsson, & Link, 1994). Anti-inflammatory properties were demonstrated for both fluoxetin and clomipramine in carrageenin or brewer's yeast-induced inflammation (Bianchi, Rossoni, Sacerdote, Panerai, & Berti, 1995; Bianchi, Sacerdote, & Panerai, 1994a, b). Some of the effects of antidepressants are probably mediated by a direct action on immune cells; TCAs were found to inhibit spontaneous secretion of IL-2 and IFNy form T-cells, as well as spontaneous and LPS-induced secretion of IL1 and IL-6 and TNFa from monocytes (Xia, DePierre, & Nassberger, 1996). Similarly, the antidepressant rolipram was found to suppress TNFa and (to a lesser extent) also IFNy secretion by human and rat auto-reactive T-cells (Sommer, Loschmann, Northoff, Weller, Steinbrecher, Steinbach, Lichtenfels, Meyermann, Riethmuller, Fontana, Dichgans, & Martin, 1995). These findings suggest that the effects of antidepressants on LPS-induced sickness behavior and corticosterone secretion, found in our studies, are mediated by changes in immune reactivity to LPS (particularly reduced production of cytokines). To elucidate these mechanisms we have recently examined the effects of chronic fluoxetin on the expression of TNFa and iNOS mRNA following LPS administration. Preliminary evidence suggests that LPS-induced TNFa mRNA in splenocytes was attenuated in fluoxetin-treated rats. Furthermore, LPS-induced expression of iNOS was

304

R. Yirmiya et al.

saline prozac control B ^ S B B S

P'^"" iNOS

LPS ^ S l ^ S ^^^^^3 P-actin ^ ^ ^ ^ ^ ^ ^ ^

iNOS

Figure 4. Effects of chronic treatment with fluoxetin on LPS-induced expression of iNOS and P-actin (as control) mRNA in the spleen.

also markedly attenuated (Fig. 4), suggesting that the immune-suppressive effect of chronic fluoxetin treatment on splenocytes is general, and not specific to TNFa. Following chronic treatment with either saline (Sal) or fluoxetin (Prozac, lOmg/kg injected daily for 5 weeks), rats were injected acutely with either saline or LPS (100|ig/kg). Spleens were removed 2hr post-injection and iNOS mRNA levels assessed by RT-PCR analysis. The results represent data from individual spleens, obtained during one replication of this experiment. Previous studies have clearly demonstrated that TNFa is involved in mediating the effects of various immune challenges on food consumption, body weight and HPA axis activation. Exogenous administration of TNFa produced anorexia, body weight loss, and pituitary-adrenal activation (Plata-Salaman, 1998; Sonti et al., 1996; Van der Meer, Fred Sweep, Pesman, Borm, & Hermus, 1995; Warren, Finck, Arkins, Kelley, Scamurra, Murtaugh, & Johnson, 1997). Moreover, TNFa synthesis blockers or antibodies to TNFa suppressed the anorexia, body weight loss, and adrenocortical activation produced by various immune chaflenges (Breuille, Farge, Rose,Arnel,Attaix, & Obled, 1993; Smith, & Kluger, 1993; Wohlman, Gallily, Yirmiya, & Weidenfeld, 1997). Thus, the attenuation of LPS-induced TNFa expression in fluoxetin-treated rats may explain at least some of the reduction of anorexia, body-weight loss, and adrenocortical activation in these rats. 6.2.2. Alteration in Monoaminergic Systems. In addition to their direct effects on cytokines, antidepressants may suppress the behavioral and neuroendocrine effects of LPS by modulating monoaminergic neurotransmission. Both imipramine and fluoxetin are monoamine reuptake inhibitors (imipramine inhibits the reuptake of both norepinephrine and serotonin, whereas fluoxetin is an SSRI) (Montgomery, 1994). The basis for the therapeutic effects of these drugs is not clear. Chronic, but not acute, treatment produces adaptive changes in several neural systems, particularly monoaminergic systems and the HPA axis. For example, animal studies demonstrated that chronic imipramine reduces the number and function of padrenergic and serotonergic (S-HTa) receptors (Burnet, Michelson, Smith, Gold, & Sternberg, 1994), and decreases tyrosine hydroxylase and CRH mRNA levels (Brady, Whitfield, Fox, Gold, & Herkenham, 1991). Chronic administration of fluoxetin was also found to down regulate serotonergic receptors functions (Leonard, 1994; Newman, Shapira, & Lerer, 1992). These changes may be relevant to the effects of LPS observed in our study, since LPS administration has been found to increase the turnover of norepinephrine and serotonin in several brain areas (Dunn & Welch, 1991; Dunn & Wang, 1995; Linthorst, Flachskamm, Holsboer, & Reul, 1995a; Linthorst, Flackskmann, Muller-Preuss, Holsboer, & Reul, 1995b). Although there is no direct evidence for the involvement of monoamines in mediating LPS-induced sickness behavior symptoms.

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

305

we have recently found that pretreatment with the a l adrenoreceptor antagonists prazosin and urapidil significantly attenuated IL-1-induced sickness behavior (PoUak, Avitsur, Canaan, & Yirmiya, 1998). Similarly, pretreatment with the serotonin "synthesis blocker PCPA attenuated several sickness behavior symptoms (Zubareva, Abdurasulova, Bluthe, Dantzer, & Klimenko, 1998). Thus, it is possible that following chronic treatment with antidepressants, the activation of monoaminergic systems by LPS is reduced, and this reduction may be responsible for the attenuation of the behavioral anhedonia and sickness behavior. 6.3.3. Alterations in the HPA Axis. Chronic treatment with antidepressants also alters the functioning and regulation of the HPA axis (Holsboer & Harden, 1996). Chronic, but not acute, administration of imipramine (Harden, Reul & Holsboer, 1995) or fluoxetin (Brady, Gold, Herkenham, Lynn, & Whitfield, 1992) were found to increase the levels of glucocorticoids receptors in the hippocampus (Brady et al., 1992), which normally mediate the negative feedback response to LPS-induced activation of the HPA axis (e.g., Weidenfeld & Yirmiya, 1996). In experimental animals, long term administration of fluoxetin, as well as tricyclic antidepressants, inhibits the pituitaryadrenal response to various challenges (Li, Levy, Cabrera, Brownfield, Battaglia, & Van de Kar, 1993; Reul, Stec, Soder, & Holsboer, 1993). Similarly, chronic imipramine treatment was found to normalize the hyperactive HPA axis in successfully treated depressed patients (Holsboer, Liebel, & Hofschuster, 1982). Finally, following long (8 weeks), but not short (2 weeks) treatment with Fluoxetin, CRH mRNA was decreased by 30-48% in the hypothalamic PVN (Brady et al., 1992). CRH within the brain mediates many of the behavioral and neuroendocrine effects of immune challenges (e.g., Bluthe, Crestani, Kelley, & Dantzer, 1992; Uehara, Sekiya,Takasugi,Namiki, & Arimura, 1989;Tilders et al., 1994;Turnbull & Rivier, 1995). Thus, it may be suggested that the attenuation of CRH system induced by antidepressants is also involved in the reduction of the behavioral and neuroendocrine depressive-like symptoms induced by immune activation.

7. CONCLUSIONS Accumulating evidence indicates that immune activation during various medical conditions is associated with a depressive syndrome in both humans and experimental animals. Taken together with the reports that brain cytokines influence the neurochemical systems involved in depression, these findings support the hypothesis that immune activation, via the release of peripheral and brain cytokines, might be involved in the etiology and symptomatology of "depression due to a general medical condition" and some other specific depressive syndromes (Fig. 5). Many medical conditions involve the activation of the immune system and the release of cytokines, particularly within the brain. These conditions include acute and chronic peripheral infections, the post-partum period, exposure to stressful stimuH, autoimmune diseases, such as multiple sclerosis (MS) and rheumatoid arthritis (RA), neurodegenerative diseases, such as Alzheimer's disease, stroke, and other brain traumas, and intracerebral infections with pathogens such as Borna Disease virus, herpes simplex virus (HSV), human immunodeficiency virus (HIV) and mycoplasma fermentans. All of these conditions are also associated with a high prevalence of depressive syndromes. Cytokine release within the brain can affect the neural substrate

R. Yirmiya et al.

306 Antidepressants

/ Peripheral im mum e activation and cytokine secretion

Periplieral infectons

Post parturn period

v ^

_L

^

Stress

Cenlral secretion of cytoltines (e.g , IL-1,TNF, IFN)

^

N

Alterations In neurochemical systems (e.g., NE, 5-HT, CBF)

Autoimmune l^eurodegeneratfve diseases diseases [e.g., Alzheimer) (e.g., MS. RA)

Stroke Trauma

Dspression due 10 a general medical concfiliori

frtracerebra! Infections (e.g., Soma virus. HSV, HIV, Mycoplasma)

Figure 5. Immune activation and depression due to a general medical condition: Possible mechanisms for modulation by antidepressants.

of depressive symptomatology and thus can serve as the physiological mechanism of "depression due to a general medical condition". Effective treatment of cytokineinduced depressive symptoms may be based on the ability of antidepressant drugs to interfere with the production and secretion of cytokines, either in the periphery or within the brain. In addition, antidepressants may prevent the neurochemical alterations that underlie cytokine-mediated depression. This hypothesis has direct implications for antidepressant therapy. In previous research it has been demonstrated that 1) Antidepressants can be used successfully for treating depression associated with medical conditions, as well as cytokine-induced depression in humans; 2) Antidepressants inhibit the increased immune reactivity and cytokine responses exhibited by some depressed patients; 3) Antidepressants have antiinflammatory properties in experimental animal models; and 4) Antidepressants suppress the secretion of proinflammatory cytokines following various in vitro challenges. Based on these findings and on our recent demonstration that chronic treatment with antidepressant drugs attenuated endotoxin-induced anhedonia, sickness behavior, adrenocortical activation, and expression of TNFa mRNA, we have recently suggested that some of the therapeutic effects of antidepressants can be attributed to their suppressing effects on the synthesis and/or activity of cytokines (Yirmiya, 1997). Within the brain, cytokines released by activated immune cells, glia cells or neurons have been documented to affect all of the neurochemical and neuroendocrine systems which have been implicated in depressive symptomatology. In particular, alterations in norepinephrine and serotonin neurotransmission and dysregulation of the HPA axis have been reported in various medical conditions or foUowing cytokine administration. Thus, in addition to direct effects of antidepressants on immune activation and cytokine secretion within the brain, these drugs may influence the cytokine-induced alterations in the neurochemical substrate of depression (Fig. 5). Future research should examine the effects of antidepressant drugs on immune functions and cytokine secretion, as well as the effects of cytokine synthesis-blockers and antagonists on depressive disorders associated with medical conditions.

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

307

ACKNOWLEDGMENTS Supported by grant no. 97-204 from the United States—Israel Binational Science Foundation and by a grant from the German-Israeh Foundation for Scientific research and Development.

REFERENCES Abbey, S. E. & Garfinkel, P. E. (1991). Chronic fatigue syndrome and depression: Cause, effect or covariate. Reviews of Infectious Diseases, 13, S73-83. Abramowitz, E. S., Baker, A. H., & Fleischer, S. F. (1982). Onset of depressive psychiatric crises and the menstrual cycle. American Journal of Psychiatry, 139{A), 475-478. Adler, N. E., Boyce, T., Chesney, M. A., Cohen, S., Folkman, S., Khan, R. L., & Syme, S. L. (1994). Socioeconomic status and health. American Psychologist, 49{\), 15-24. Albrecht, I , Helderman, J. H., Schlesser, M. A., & Rush, A. J. (1985). A controlled study of cellular immune function in affective disorders before and during somatic therapy. Psychiatry Research, 15, 185-193. Anderson, N. B. & Armstead, C. A. (1995). Toward understanding the association of socioeconomic status and heahh: A new challenge for the biopsychosocial approach. Psychosomatic Medicine, 57, 213-225. Anisman, H., Kokkinidis, L., & Merali, A. (1996). Interleukin-2 decreases accumbal dopamine efflux and responding for rewarding lateral hypothalamic stimulation. Brain Research, 731, 1-11. Anisman, H., Kokkinidis, L., Borowski, T, & Merali, A. (1998). Differential effects of interleukin (IL)-l, IL-2, and IL-6 on responding for rewarding lateral hypothalamic stimulation. Brain Research, 779, 177-187. Arenzana-Seisdedos, F. & Virelizer, J. L. (1983). Interferons as macrophage-activating factors. II. Enhanced secretion of interleukin-1 by lipopolysaccharide-stimulated human monocytes. European Journal of Immunology, 13, 437-440. Arvin, B., Neville, L. F, Barone, F. C , & Feuerstein, G. Z. (1996). The role of inflammation and cytokines in brain injury. Neuroscience and Biobehavioral Reviews, 20, 445-452. Aubert, A.,Vega, C , Dantzer, R., & Goodall, G. (1995). Pyrogens specifically disrupt the acquisition of a task involving cognitive processing in the rat. Brain Behavior and Immunity, 9, 129-148. Aubert, A. & Dantzer, R. (1998). Effects of lipopolysaccharide on taste reactivity in rats. Neuroimmunomodulation, 5, 3. Audus, K. L. & Gordon, M. A. (1982). Tricyclic antidepressant effects on the murine lymphocyte mitogen response. Journal of Immunopharmacology, 4, 13-27. Avitsur, R., Pollak, Y., & Yirmiya, R. (1997). Different receptor mechanisms mediate the effects of endotoxin and interleukin-1 on female sexual behavior. Brain Research, 773, 149-161. Avitsur, R., Cohen, E., & Yirmiya, R. (1998). Effects of interleukin-1 on sexual attractivity in a model of sickness behavior. Physiology and Behavior, 63, 25-30. Ayuso, J. L. (1994). Use of psychotropic drugs in patients with HIV infection. Drugs, 47(4), 599-610. Badia-Elder, N., Kiefer, S. W., & Dess, N. K. (1996). Taste reactivity in rats selectively bred for high vs. low saccharin consumption. Physiology and Behavior, 59, 749-755. Barden, N., Reul, J. M. H. M., & Holsboer, F. (1995). Do antidepressants stabilize mood through actions on the hypothalamic-pituitary-adrenocortical system? Trends in Neurosciences, 18, 6-11. Bauer, J., Hohagen, F , Gimmel, E., Bruns, F, Lis, S., Krieger, S., Ambach, W., Guthmann, A., Grunze, H., Fritsch-Montero, R., Weissbach, A., Ganter, U., Frommnerger, U., Riemann, D., & Berger, M. (1995). Induction of cytokines synthesis and fever suppresses REM sleep and improves mood in patients with major depression. Biological Psychiatry, 38, 611-621. Beers, D. R., Henkel, J. S., Kesner, R. P., & Stroop,W. G. (1995). Spatial recognition memory deficits without notable CNS pathology in rats following herpes simplex encephalitis. Journal of the Neurological Sciences, 131, 119-127. Besedovsky, H. O. & Del-Rey, A. (1996). Immune-neuro-endocrine interactions: facts and hypotheses. Endocrine Reviews, 17, 64-102. Bianchi,M., Sacerdote, P., & Panerai, A. E. (1994a). Fluoxetin reduces inflammatory edema in the rat: Involvement of the pituitary-adrenal axis. European Journal of Pharmacology, 263, 81-84. Bianchi, M., Sacerdote, P., & Panerai, A. E. (1994b). Chlomipramine differentially affects inflammatory edema and pain in the rat. Pharmacology, Biochemistry, and Behavior, 4, 1037-1040.

308

R. Yirmiya et al.

Bianchi, M., Rossoni, G., Sacerdote, P., Panerai, A. E., & Berti, F. (1995). Effects of clomipramine and fluoxetin on subcutaneous carrageenin-induced inflammation in the rat. Inflammation Research, 44, 466-469. Bluthe, R. M., Dantzer, R., & Kelley, K.W. (1991). Interleukin-1 mediates behavioral but not metaboHc effects of tumor necrosis factor alpha in mice. European Journal of Pharmacology, 209, 281-283. Bluthe, R. M., Crestani, F., Kelley, K. W., & Dantzer, R. (1992). Mechanisms of the behavioral effects of interleukin 1. Role of prostaglandins and CRF. Annals of The New York Academy of Sciences, 650, 268-275. Bluthe, R. M., Dantzer, R., & Kelley, K. W. (1997). Central mediation of the effects of interleukin-1 on social exploration and body weight in mice. Psychoneuroendocrinology, 22, 1-11. Bode, L., Zimmerman, W., Ferszt, R., Steinbach, F., & Ludwig, H. (1995). Borna disease virus genome transcribed and expressed in psychiatric patients. Nature Medicine, 1, 232-236. Bode, L., Durrwald, R., Rantam, F. A., Ferszt, R., & Ludwig, H. (1996). First isolation of infectious human Borna disease virus from patients with mood disorders. Molecular Psychiatry, I, 200-212. Borowski, T., Kokkinidis, L., Merali, A., & Anisman, H. (1997). Behavioral and central monoamine alterations induced by interleukin-1, interleukin-6, and lipopolysaccharide. Society for Neuroscience Abstracts, 23, 1508. Brady, L. S., Whitfield, H. J., Fox, R. J., Gold, P W., & Herkenham, M. (1991). Long-term antidepressant administration alters corticotropin-releasing hormone, tyrosine hydroxylase, and mineralocorticoid receptor gene expression in rat brain. Journal of Clinical Investigation, 87, 831-837. Brady, L. S., Gold, R W., Herkenham, M., Lynn, A. B., & Whitfield, H. J. (1992). The antidepressants fluoxetine, idazoxan, and phenelzine alter corticotropin-releasing hormone and tyrosine hydroxylase mRNA levels in rat brain: therapeutic implications. Brain Research, 572, 117-125. Breuille, D., Farge, M. C , Rose, F., Arnel, M., Attaix, D., & Obled, C. (1993). Pentoxifylline decreases body weight loss and muscle protein wasting characteristics of sepsis. American Journal of Physiology, 265, E660-E666. Brown, S. L., Steinberg, R. L., & van Praag, H. M. (1994). The pathogenesis of depression: reconsideration of neurotransmitter data. In J. A. Den Boer & J. M. Sitsen (Eds.), Handbook of Depression and Anxiety: A Biological Approach (pp. 529-542). New York: Marcel Dekker, Inc. Bruce, M. L., Takeuchi, D. T , & Leaf, P. J. (1991). Poverty and psychiatric status. Archives of General Psychiatry, 48, 470-474. Buchwald, D. & Komaroff, A. L. (1991). Review of laboratory findings for patients with chronic fatigue syndrome. Reviews of Infectious Diseases, 13, 12-18. Burnet, R W. J., Michelson, D., Smith, M. A., Gold, P W, & Sternberg, E. M. (1994). The effect of chronic imipramine administration on the densities of 5-HTlA and 5-HT2 receptors and the abundances of 5-HT receptor and transporter mRNA in the cortex, hippocampus, and dorsal raphe of three strains of rat. Brain Research, 638, 311-324. Cannon, J. G. & Dinarello, C. A. (1985). Increased plasma interleukin-1 activity in woman after ovulation. Science, 227, 1247-1249. Cappel, R., Gregoire, E., Thiry, L., & Sprecher, S. (1978). Antibody and cell-mediated immunity to herpes simplex virus in psychotic depression. Journal of Clinical Psychiatry, 39, 266-268. Carroll, B. J., Martin, F. I. R., & Davies, B. (1968). Resistance to suppression by dexamethasone of plasma 11hydroxycorticosterone levels in severe depressive illness. British Medical Journal, 3, 285-287. CarroU, B. J., Curtis, G. C , & Mendels, J. (1976). Neuroendocrine regulation in depression. I. Limbic system adrenocortical dysfunction. Archives of General Psychiatry, 33, 1039-1044. Cavallo, M. G., Pozzini, P., & Thorpe, R. (1993). Cytokines and autoimmunity. Clinical and Experimental Immunology, 96, 1-7. Chochinov, H. M., Wilson, K. G., Enns, M., & Lander, S. (1994). Prevalence of depression in the terminally ill: Effects of diagnostic criteria and symptom threshold judgments. American Journal of Psychiatry, 4, 537-540. Cleare, A. J., Beam, J., Allain, T, McGregor, A., Wessely, S., Murray, R. M., & O'Keane, V. (1995). Contrasting neuroendocrine responses in depression and chronic fatigue syndrome. Journal of Affective Disorders, 34, 283-289. Cohen, S. (1995). Psychological stress and susceptibility to upper respiratory infections. American Journal of Respiratory and Critical Care Medicine, 152, 53-58. Connor, T. J. & Leonard, B. E. (1998). Depression, stress and immunological activation: the role of cytokines in depressive disorders. Life Sciences, 62, 583-606. Cox, S. M., King, M. R., Casey, M. L., & Macdonald, R C. (1993). Interleukin-ip, -la, and -6 and prostaglandins

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

309

in vaginal cervical fluids of pregnant women before and during labor. Journal of Clinical Endocrinology and Metabolism, 77, 805-815. Cunningham, A. J., Murray, C. A., O'Neill, L. A. J., Lynch, M. A., & O'Connor, J. J. (1996). Interleukin-ip (ILIP) and tumour necrosis factor (TNF) inhibit long-term potentiation in the rat dentate gyrus in vitro. Neuroscience Letters, 203, 17-20. Dantzer, R., Bluthe, R. M., Aubert, A., Goodall, G., Bret-Dibat, J. L., Kent, S., Goujon, E., Laye, S., Parnet, P., & Kelley, K. W. (1996). Cytokine actions on behavior. In N. J. Rothwell (Ed.), Cytokines in the nervous system (pp. 117-144). New York: R. G. Landes Company and Chapman & Hall. Day, H. E. W. & Akil, H. (1996). Differential pattern of c-fos mRNA in rat brain following central and systemic administration of interleukin-1 beta: Implications for mechanisms of action. Neuroendocrinology, 63, 207-218. Denburg, S. D., Carbotte, R. M., & Denburg, J. A. (1997). Psychological aspects of systemic lupus erythematosus: Cognitive Function, mood, and self-report. Journal of Rheumatology, 24, 998-1003. Dess, N. K. (1993). Saccharin's aversive taste in rats: Evidence and implications. Neuroscience and Biobehavioral Reviews, 17, 359-372. Dinarello, C. A. & Wolff, S. M. (1993). The role of interleukin-1 in disease. New England Journal of Medicine, 328, 106-113. Dohrenwend, B. P. (1973). Social status and stressful life events. Journal of Personality and Social Psychology, 28, 225-235. DSM-IV, Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition). (1994). Washington, DC: American Psychiatric Association. Dunbar, P. R., Hill, J., Neale, T. J., & Mellsop, G. W. (1992). Neopterin measurement provides evidence of altered cell-mediated immunity in patients with depression, but not with schizophrenia. Psychological Medicine, 22, 1051-1057. Dunn, A. J. & Welch, J. (1991). Stress- and endotoxin-induced increases in brain tryptophan and serotonin metabohsm depend on sympathetic nervous system aciWiiy. Journal of Neurochemistry, 57, 1615-1622. Dunn, A. J. & Wang, J. (1995). Cytokine effects on CNS biogenic amines. Neuroimmunomodulation, 2, 319-328. Dunn, A. J., Chapman, Y., & Antoon, M. (1993). Endotoxin-induced behavioral changes of mice in the multicompartment chamber are distinct from those of interleukin-1. Neuroscience Research Communications, 10, 63-69. Eisen, J. N., Irwin, J., Quay, J., & Livnat, S. (1989). The effect of antidepressants on immune function in mice. Biological Psychiatry, 26, 805-817. Pent, K. & Zbinden, G. (1987). Toxicity of interferon and interleukin. Trends in Pharmacological Sciences, 8, 101-105. Foley, F W, Traugott, U , LaRocca, N. G , Smith, C. R., Perlman, K. R., Caruso, L. S., & Scheinberg, L. C. (1992). A prospective study of depression and immune dysregulation in multiple sclerosis. Archives of Neurology, 49, 238-244. Gibertini, M. (1996). IL-1 beta impairs relational but not procedural rodent learning in a water maze task. Advances in Experimental Medicine and Biology, 402, 207-217. Gibertini, M., Newton, C , Friedman, H., & Klein, T. W. (1995). Spatial learning impairment in mice with legionella pneumophila or administered exogenous interleukin-1-beta. Brain, Behavior, and Immunity, 9, 113-128. Gold, P W, Chrousos, G , Kellner, C , Post, R., Roy, A., Auerginos, P, Schulte, H., Oldfield, E., & Loriaux, D. L. (1984). Psychiatric implications of basic and clinical studies with corticotropin releasing factor. American Journal of Psychiatry, 141, 619-627. Gottfries, C. G. (1997). Recognition and management of depression in the elderly. International Clinical Psychopharmacology, 12(7), S31-S36. Goujon, E., Parnet, P, Laye, S., Combe, C , Kelley, K. W, & Dantzer, R. (1995). Stress downregulates lipopolysaccharide-induced expression of proinflammatory cytokines in the spleen, pituitary, and brain of mice. Brain, Behavior, and Immunity, 9, 292-303. Granger, D. A., Hood, K. E., Ikeda, S. C , Reed, C. L., Jones, B. C , & Block, M. L. (1997). Effects of peripheral immune activation on social behavior and adrenocortical activity in aggressive mice: genotype environment interactions. Aggressive Behavior, 23, 93-105. Greenwood, R. (1987). Residual mental disorders after herpesvirus infections. In E. Kurstak, Z. J. Lipowski, & P. V. Morozov (Eds.), Viruses, Immunity, and Mental Disorders (pp. 65-80). New York: Plenum.

310

R. Yirmlya et al.

Grill, J. J. & Norgren, R. (1978). The taste reactivity test. II. Mimetic responses to gustatory stimuli in chronic thalamic and chronic decerebrate rats. Brain Research, 143, 281-297. Grossman, C. J. (1985). Interactions between the gonadal steroids and the immune system. Science, 227, 257-261. Hall, S. & Smith, A. P. (1996a). Investigating the effects and aftereffects of naturally occurring upper respiratory tract illnesses on mood and performance. Physiology and Behavior, 59, 569-577. Hall, S. & Smith, A. P. (1996b). Behavioral effects of infectious mononucleosis. Neuropsychobiology, 33, 202-209. Hart, B. (1988). Biological basis of the behavior of sick animals. Neuroscience and Biobehavioral Reviews, 12, 123-37. Hendler, N. (1987). Infectious mononucleosis and psychiatric disorders. In E. Kurstak, Z. J. Lipowski, & P. V. Morozov (Eds.), Viruses, Immunity, and Mental Disorders (pp. 81-94). New York: Plenum. Herbert, T. B. & Cohen, S. (1993). Depression and immunity: A meta-analytic review. Psychological Bulletin, 3, 472-486. Heyser, C. J., Masliah, E., Samimi, A., Campbell, I. L., & Gold, L. H. (1997). Progressive decline in avoidance learning paralleled by inflammatory neurodegeneration in transgenic mice expressing interleukin-6 in the brain. Proceedings of the National Academy of Sciences of the United States of America, 94, 1500-1505. Hickie, I. & Lloyd, A. (1995). Are cytokines associated with neuropsychiatric syndromes in humans? International Journal of Immunopharmacology, 17, 677-683. Holsboer, F. (1995). Neuroendocrinology of mood disorders. In E E. Bloom & D. J. Kupfer (Eds.), Psychopharmacology: The Fourth Generation of Progress (pp. 957-969). New York: Raven Press. Holsboer, F. & Barden, N. (1996). Antidepressants and hypothalamic-pituitary-adrenocortical regulation. Endocrine Reviews, 17, 187-205. Holsboer, F , Liebel, R., & Hofschuster, E. (1982). Repeated dexamethasone test during depressive illness: normalization of test result compared with clinical improvement. Journal of Affective Disorders, 4, 93-101. Hutchinson, G. A., Nehall, J. E., & Simeon, D. T. (1996). Psychiatric disorders in systemic lupus erythematosus. West Indian Medical Journal, 45, 48-50. Isometsa, E.T., Aro, H. M., Henriksson, M. M., Heikkinen, M. E., & Lonnqvist, J. K. (1994). Suicide and major depression in different treatment settings. Journal of Clinical Psychiatry, 55, 523-527. Joffe, R. T , Lippert, G. P., Gray, T. A., Sawa, G., & Horvath, A. (1987). Mood disorder and multiple sclerosis. Archives of Neurology, 44, 376-378. Johnson, R. W., Propes, M. J., & Shavit, Y. (1996). Corticosterone modulates the behavioral and metabolic effects of lipopolysaccharide. American Journal of Physiology, 270, R192-R198. Katon, W. & Sullivan, M. D. (1990). Depression and chronic medical illness. Journal of Clinical Psychiatry, 51, 3-11. Katsuki, H., Nakai, S., Hirai, Y , Akaji, K., Kiso, Y, & Satoh, M. (1990). Interleukin-ip inhibits long-term potentiation in the CA3 region of mouse hippocampal slices. European Journal of Pharmacology, 191, 323-326. Kavaliers, M., Colwell, D. D., & Galea, L. A. M. (1995). Parasitic infection impairs spatial learning in mice. Animal Behaviour, 50, 223-229. Kent, S., Bluthe, R. M., Kelley, K. W., & Dantzer, R. (1992a). Sickness behavior as a new target for drug development. Trends in Pharmacological Sciences, 13, 24-28. Kent, S., Bluthe, R. M., Dantzer, R., Hardwick, A. J., Kelley, K. W., Rothwell, N. J., & Vannice, J. L. (1992b). Different receptor mechanisms mediate the pyrogenic and behavioral effects of interleukin-1. Proceedings of the National Academy of Sciences of the United States of America, 89, 9117-9120. Kiecolt-Glaser, J. K. & Glaser, R. (1991). Stress and immune function in humans. In R. Ader, D. L. Felten, & N. Cohen (Eds.), Psychoneroimmunology (pp. 849-867). San Diego: Academic Press. Klonoff, D. C. (1992). Chronic fatigue syndrome. Clinical infectious diseases, 15, 812-823. Komaroff, A. L. & Buchwald, D. S. (1991). Symptoms and signs of chronic fatigue syndrome. Reviews of Infectious Diseases, 13, 8-11. Komaroff, A. L. & Buchwald, D. S. (1998). Chronic fatigue syndrome: an update. Annual Review of Medicine, 49, 1-13. Komaroff, A. L., Fagioh, L. R., Geiger, A. M., DooHttle, T. H., Lee, J., Kornish, J., Gleit, M. A., & Guerriero, R. T. (1996). An examination of the working case definition of chronic fatigue syndrome. American Journal of Medicine, 100, 56-64. Kozak, W., Poli, V., Soszynski, D., Conn, C. A., Leon, L. R., & Kluger, M. (1996). Sickness behavior in mice

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

311

deficient in interleukin-6 during turpentine abscess and influenza pneumonitis. American Journal of Physiology, 272, R621-R630. Kozak, W., Soszynski, D., Rudolph, K., Conn, C. A., & Kluger, M. (1997). Dietary n-3 fatty acids differentially affect sickness behavior in mice during local and systemic inflammation. American Journal of Physiology, 272, R1298-R1307. Krueger, J. M. & Majde, J. A. (1994). Microbial products and cytokines in sleep and fever regulation. Critical Reviews in Immunology, 14, 355-379. Krueger, J. M., Takahashi, S., Kapas, L., Bredow, S., Roky, R., Fang, J., Floyd, R., Renegar, K. B., Guha-Thakurta, N., Novitsky, S., & Obal, F. (1995). Cytokines in sleep regulation. Advances in Neuroimmunology, 5, 171-188. Kubera, M., Symbirtsev, A., Basta-Kaim, A., Borycz, J., Roman, A., Papp, M., & Claesson, M. (1996). Effect of chronic treatment with imipramine on interleukin-1 and interleukin-2 production by splenocytes obtained from rats subjected to chronic mild stress model of depression. Polish Journal of Pharmacology, 48, 503-506. Laghrissi-Thode, E, Pollock, B. G., Szanto, K., & Reynolds, C. E (1996). Depression and suicide in medically ill patients. Current Opinion in Psychiatry, 9, 137-140. Landmann, R., Schaub, B., Link, S., & Wacker, H. R. (1997). Unaltered monocyte function in patients with major depression before and after three months of antidepressive therapy. Biological Psychiatry, 41, 675-681. Laruelle, M., Seghers, A., Goffinet, S., Bouchez, S., & Legros, J. J. (1990). Plasmatic vasopressin neurophysin in depression: Basic levels and relations with HPA axis. Biological Psychiatry, 27, 1249-1263. Lauritzen, L., Bjerg Bendsen, B., Vilmar, T., Bjerg Bendsen, E., Lunde, M., & Bech, P. (1994). Post-stroke depression: combined treatment with imipramine or desipramine and mianserin. Psychopharmacology, 114, 119-122. Leonard, B. E. (1994). Effect of antidepressants on specific neurotransmitters: Are such effects relevant to the therapeutic actions? In J. A. den Boer & J. M. Siten (Eds.), Handbook of depression and anxiety: A biological approach (pp. 379^04). New York: Marcel Dekker, Inc. Levenson, J. L. & Fallon, H. J. (1993). Fluoxetine treatment of depression caused by interferon-alpha. American Journal of Gastroenterology, 88{5), 760-761. Li, Q., Levy, A. D., Cabrera, T. M., Brownfield, M. S., Battaglia, G., & Van de Kar, L. D. (1993). Long-term fluoxetine, but not desipramine, inhibits the ACTH and oxytocin responses to the 5-HT l A agonist, 8-OH-DPAT, in male rats. Brain Research, 630, 148-156. Licinio, J., Kling, M. A., & Hauser, P. (1998). Cytokines and brain function: relevance to interferon-alphainduced mood and cognitive changes. Seminars in Oncology, 25{\), 30-38. Lieb, J. & Karmali, R. (1983). Elevated levels of prostaglandin E2 and thromboxane B2 in depression. Prostaglandins and Leukotrienes in Medicine, 10, 361-368. Lim, L., Ron, M. A., Ormerod, I. E., David, J., Miller, D. H., Logsdail, S. J., Walport, M. J., & Harding, A. E. (1988). Psychiatric and neurological manifestations in systemic lupus erythematosus. Quarterly Journal of Medicine, 66, 27-38. Linkowski, P., Mendlewicz, I , Leclercq, R., Brasseur, M., Hubain, P., Goldstein, J., Copinschi, G., & van Cauter, E. (1985). The 24-hr profile of adrenocorticotropin and Cortisol in major depressive illness. Journal of Clinical Endocrinology and Metabolism, 61, 429-438. Linnoila, M., Whorton, R., Rubinow, D. R., Cowdry, R. W., Ninan, P T., & Waters, R. N. (1983). CSF prostaglandin levels in depressed and schizophrenic patients. Archives of General Psychiatry, 40, 405-406. Linthorst, A. C. E., Flachskamm, C , Holsboer, E, & Reul, J. M. H. M. (1995a). Intraperitoneal administration of bacterial endotoxin enhances noradrenergic neurotransmission in the rat preoptic area: relationship with body temperature and hypothalamic-pituitary-adrenocortical axis activity. European Journal of Neuroscience, 7, 2418-2430. Linthorst, A. C. E., Flachskamm, C , Muller-Preuss, P, Holsboer, E, & Reul, J. M. H. M. (1995b). Effect of bacterial endotoxin and interleukin-1 beta on hippocampal serotonergic neurotransmission, behavioral activity, and free corticosterone levels: an in vivo microdialysis study. Journal of Neuroscience, 15, 2920-2934. Lycke, E., Norrby, R., & Roos, B. E. (1974). A serological study on mentally ill patients with particular reference to prevalence of herpes simplex virus infections. British Journal of Psychiatry, 124, 273. Lyketsos, C. G., Steele, C , Baker, L., Galik, E., Kopunek, S., Steinberg, M., & Warren, A. (1997). Major and minor depression in Alzheimer's disease: prevalence and impact. Journal of Neuropsychiatry and Clinical Neurosciences, 9(4), 556-561. Lynch, E. A., Dinarello, C. A., & Cannon, J. G. (1994). Gender differences in IL-la, IL-lp, and IL-1

312

R. Yirmiya et al

receptor antagonist secretion from mononuclear cells and urinary excretion. Journal of Immunology, 153, 300-306. Maes, M. (1995). Evidence for an immune response in major depression: A review and hypothesis. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 19, 11-38. Maes, M., DeJonckheere, C , Vandervorst, C , Schotte, C , Cosyns, P., Raus, J., & Suy, E. (1991). Abnormal pituitary function during melancholia: Reduced alpha melanocyte stimulating hormone secretion and increased intact ACTH non-suppression. Journal of Affective Diseases, 22, 149-157. Maes, M., Lambrechts, J., Bosmans, E., Jacobs, J., Suy, E., Vandervorst, C , De Jockheere, C , Minner, B., & Raus, J. (1992a). Evidence for systemic immune activation during depression: results of leukocyte enumeration by flow cytometry in conjunction with monoclonal antibody staining. Psychological Medicine, 22, 45-53. Maes, M., Scharpe, S., Van Grootel, L., Uyttenbroeck, W., Cooreman, W., Cosyns, P., & Suy, E. (1992b). Higher alpha-1-antitrypsin, haptoglobin, ceruloplasmin, and lower retinol binding protein plasma levels during depression: further evidence for the existence of an inflammatory response during that illness. Journal of Affective Disorders, 24, 183-192. Maes, M., Bosmans, E., Meltzer, H. Y., Scharpe, S., & Suy, E. (1993). Interleukin-lp—A Putative Mediator of HPA Axis Hyperactivity in Major Depression. American Journal of Psychiatry, 150, 1189-1193. Maes, M., Scharpe, S., Meltzer, H. Y., Okayh, G., D'Hondt, P., & Cosyns, P. (1994). Increased neopterin and interferon gamma secretion and lower L-tryptophan levels in major depression: further evidence for immune activation in severe depression. Psychiatry Research, 54, 143-160. Maes, M., Meltzer, H. Y, Bosmans, E., Bergmans, R., Vandoolaeghe, E., Ranjan, R., & Desnyder, R. (1995a). Increased plasma concentrations of interIeukin-6, soluble interIeukin-6, soluble interleukin-2, and transferrin receptor in major depression. Journal of Affective Disorders, 34, 301-309. Maes, M., Scharpe, S., Neels, H., Wauters, A., Van Gastel, A., & Cosyns, P. (1995b). Total serum protein and serum protein fractions in major depression. Journal of Affective Disorders, 34, 61-69. Maes, M., Smith, R., & Scharpe, S. (1995c). The monocyte T-lymphocyte hypothesis of major depression. Psychoneuroendocrinology, 20, 111-116. Maes, M., Bosmans, E., De Jongh, R., Kenis, G., Vandoolaeghe, E., & Neels, H. (1997a). Increased serum IL-6 and IL-1 receptor antagonist concentrations in major depression and treatment resistant depression. Cytokine, 9, 853-858. Maes, M., Vandoolaeghe, E., Van Hunsel, R, Bril, T, Demedts, P., Wauters, A., & Neels, H. (1997b). Immune disturbances in treatment-resistant depression: modulation by antidepressive treatment. Human Psychopharmacology, 12, 153-162. Magner, M. B. (1991). Psychiatric morbidity in outpatients with systemic lupus erythematosus. South African Medical Journal, 80, 291-293. Maier, S. R & Watkins, L. R. (1998). Cytokines for psychologists: Implications of bi-directional immune-tobrain communication for understanding behavior, mood, and cognition. Psychological Review, 105, 83-107. Maj, M. (1996). Depressive syndromes and symptoms in subjects with human immunodefficiency virus (HIV) infection. British Journal of Psychiatry, 168, 117-122. Marshal, P. S. (1993). Allergy and depression: A neurochemical threshold model of the relation between the illnesses. Psychological Bulletin, 113, 2 3 ^ 3 . Mathew, R. J. & Weinman, M. L. (1982). Sexual dysfunctions in depression. Archives of Sexual Behavior, 11, 323-328. McDonald, E. M., Mann, A. H., & Thomas, H. C. (1987). Interferons as mediators of psychiatric morbidity: An investigation in a trial of recombinant a-interferon in hepatitis-B carriers. Lancet, 2, 1175-1179. McGeer, P. L. & McGeer, E. G. (1995). The inflammatory response system of brain: implications for therapy of Alzheimer and other neurodegenerative diseases. Brain Research Reviews, 21, 195-218. Meijer, A., Zakay-Rones, Z., & Morag, A. (1988). Post-influenzal psychiatric disorder in adolescents. Acta Psychiatrica Scandinavica, 78, 176-181. Mendez, M. R, Martin, R. J., Smyth, K. A., & Whitehouse, P. J. (1990). Psychiatric symptoms associated with Alzheimer's disease. Journal of Neuropsychiatry and Clinical Neurosciences, 2(1), 28-33. Meyers, C. A. & Valentine, A. D. (1995). Neurologic and psychiatric adverse effects of immunological therapy. CNS Drugs, 3, 56-68. Meyers, C. A., Valentine, A. D., Wong, P. C. L., & Leeds, N. E. (1994). Reversible neurotoxicity of interleukin2 and tumor necrosis factor: Correlation of SPECT with neurophysiological testing. Journal of Neuropsychiatry and Clinical Neurosciences, 6, 285-288. Meyerson, B. J. & Lindstrom, L. H. (1973). Sexual motivation in the female rat. Acta Physiologica Scandinavica, 389, 1-80.

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

313

Miller, N. E. (1964). Some psychophysiological studies of motivation and of the behavioral effects of illness. Bulletin of the British Psychological Society, 17, 1-20. Miller, A. H., Asnis, G. M., Van Praag, H. M., & Norin, A. J. (1986). Influence of desmethylimipramine on natural killer cell activity. Psychiatry Research, 19, 9-15. Minden, S. L. & Schiffer, R. B. (1990). Affective disorders in multiple sclerosis: Review and recommendations for clinical research. Archives of Neurology, 47, 98-104. Minden, S. L., Orav, J., & Reich, P. (1987). Depression in multiple sclerosis. General Hospital Psychiatry, 9, 426-434. Mohr, D. C , Goodkin, D. E., Likosky W., Gatto, N., Baumann, K. A., & Rudick, R. A. (1997). Treatment of depression improves adherence to interferon [3 -lb therapy for multiple sclerosis. Archives of Neurology, 54{5),53l-52>2,. Montgomery, S. A. (1994). Antidepressants in long-term treatment. Annual Review of Medicine, 45, 447-457. Morag, M., Yirmiya, R., Lerer, B., & Morag, A. (in press). Influence of socioeconomic status on behavioral, emotional, and cognitive effects of rubella vaccination: A prospective, double blind sudy. Psychoneuroendocrinology. Mrak, R. E., Sheng, J. G., & Griffin, S. T. (1995). Glial cytokines in Alzheimer's disease: review and pathogenic implications. Human Pathology, 26(8), 816-823. Muller, N., Hofschuster, E., Ackenheil, M., Mempel, W., & Eckstein, R. (1993). Investigations of the cellular immunity during depression and the free interval: evidence for an immune activation in affective psychosis. Progress in Neuropsychopharmacology and Biological Psychiatry, 17, 713-730. Murphy, J. M., Olivier, D. C., Monson, R. R., Sobol, A. M., Federman, E. B., & Leighton, A. H. (1991). Depression and anxiety in relation to social status. Archives of General Psychiatry, 48, 223-229. Newman, M. E., Shapira, B., & Lerer, B. (1992). Regulation of 5-hydroxytryptamine 1A receptor function in rat hippocampus by short- and long-term administration of 5-hydroxytryptamine agonist and antidepressants. Journal of Pharmacology and Experimental Therapeutics, 260, 16-20. Niiranen, A., Laaksonen, R., livanainen, M., Mattson, K., Fakkila, M., & Cantell, K. (1988). Behavioral assessment of patients treated with alpha-interferon. Acta Psychiatrica Scandinavica, 78, 622-626. Oitzl, M. S., van Oers, H., Schobitz, B., & de Kloet, E. R. (1993). Interleukin-ip, but not interleukin-6, impairs spatial navigation learning. Brain Research, 613, 160-163. Okanoue, T., Sakamoto, S., Itoh, Y., Minami, M., Yasui, K., Sakamoto, M., Nishioji, K., Katagishi, T., Nakagawa, Y, Tada, H., Sawa, Y, Mizuno, M., Kagawa, K., & Kashima, K. (1996). Side effects of high-dose interferon therapy for chronic hepatitis C. Journal of Hepatology, 25, 283-291. Owens, M. J. & Nemeroff, C. B. (1993). The role of CRF in the pathophysiology of affective disorders: laboratory and clinical studies. In W. W. Vale (Ed.), Ciba Foundation Symposium on Corticotropin Releasing Factor, 172 (pp. 193-316). New York: John Wiley. Parker, J., Smarr, K., Anderson, S., Hewett, J., Walker, S., Bridges. A., & Caldwell, W (1992). Relationship of changes in helplessness and depression to disease activity in rheumatoid arthritis. Journal of Rheumatology, 19, 1901-1905. Parry, B. L. (1995). Mood disorders linked to the reproductive cycle in women. In F. E. Bloom & D. J. Kupfer (Eds.), Psychopharmacology: The fourth Generation of Progress (pp. 1029-1042). New York: Raven Press. Pavol, M. A., Meyers, C. A., Rexer, J. L., Valentine, A. D., Mattis, P J., & Talpaz, M. (1995). Pattern of neurobehavioral deficits associated with interferon a therapy for leukemia. Neurology, 45, 947-950. Pincus,T., Griffith, J., Pearce, S., & Isenberg, D. (1996). Prevalence of self-reported depression in patients with rheumatoid arthritis. British Journal of Rheumatology, 35, 879-883. Plata-Salaman, C. R. (1996). Anorexia during acute and chronic disease. Nutrition, 12, 69-78. Plata-Salaman, C. R. (1998). Cytokines and anorexia: a brief overview. Seminars in Oncology, 25(1), 64-72. Pollak, Y, Avitsur, R., Canaan, R., & Yirmiya, R. (1998). Central a l adrenergic receptors are involved in mediating IL-1 induced sickness behavior. NeuroImmunoModulation, 5, 38. Rabkin, J. G., Wagner, G., & Rabkin, R. (1994). Effects of setraline on mood and immune status in patients with major depression and HIV illness: an open trial. Journal of Clinical Psychiatry, 55, 433^39. Ranchor, A. V., Bouma. I , & Sanderman, R. (1996). Vulnerability and social class: Differential patterns of personality and social support over the social classes. Personality and Individual Differences, 20(2), 229-237. Renault, P E, Hoofnagle, J. H., Park, Y, Mullen, K. D., Peters, M., Jones, D. B., Rustgi, V, & Jones, E. A. (1989). Psychiatric complications of long-term interferon alpha therapy. Archives of Internal Medicine, 147, 1577-1580. Reul, J. M. H. M., Stec, I., Soder, M., & Holsboer, F. (1993). Chronic treatment of rats with antidepressant

314

R. Yirmiya et al.

amitriptyline attenuates the activity of the hypothalamic-pituitary-adrenocorticai system. Endocrinology, 133, 312-320. Rivier, C. & Rivest, S. (1993). Mechanisms mediating the effects of cytokines on neuroendocrine functions in the rat. In D. J. Chadwick, J. Marsh, & K. Ackrill (Eds.), Corticotropin-Releasing Factor (pp. 204-225). New York: John Weily & Sons. Robinson, R. G. (1997). Neuropsychiatric consequences of stroke. Annual Review of Medicine, 48, 217-229. Rodin, G. & Voshart, K. (1986). Depression in the medically ill: An overview. American Journal of Psychiatry, 143, 696-705. Romanovsky, A. A., Kulchitsky, V. A., Akulich, N. V., Koulchitsky, S. V., Simons, C. T., Sessler, D. I., & Gourine, V. N. (1996). First and second phases of biphasic fever: two sequential stages of the sickness syndrome. American Journal of Physiology, 271, R244-R253. Rothwell, N. J., Luheshi, G., & Toulmond, S. (1996). Cytokines and their receptors in the central nervous system: Physiology, pharmacology, and pathology. Pharmacology and Therapeutics, 69, 85-95. Sachar, E. J., Hellman, L., Fukushima, D. K., & Gallaghaer, T. F. (1970). Cortisol production in depressive illness: a clinical and biochemical clarification. Archives of General Psychiatry, 23, 289-298. Sadovnick, A. D., Remick, R. A., Allen, J., Swartz, E., Lee, I. M. L., Eisen, K., Farquhar, R., Hashimoto, S. A., Hooge, I , Kastrukoff, L. F., Morrison, W., Nelson, J., Ogar, J., & Paty, D. W. (1996). Depression and multiple sclerosis. Neurology, 46, 628-632. Sakic, B., Denburg, J. A., Denburg, S. D., & Szechtman, H. (1996). Blunted sensitivity to sucrose in autoimmune MRL-lpr mice: A curve shift study. Brain Research Bulletin, 41, 305-311. Schiffer, R. B. & Babigian, H. M. (1984). Behavioral disorders in multiple sclerosis, temporal lobe epilepsy, and amyotrophic lateral sclerosis: An epidemiological study. Archives of Neurolology, 41, 1067-1069. Schiffer, R. B. & Wineman, N. M. (1990). Antidepressant pharmacotherapy of depression associated with multiple sclerosis. American Journal of Psychiatry, 147, 1493-1497. Schneebaom, A. B., Singleton, J. D., & West, S. G. (1991). Association of psychiatric manifestations with antibodies to ribosoraal P proteins in systemic lupus erythematosus. American Journal of Medicine, 90, 54-62. Schobitz, B., De Kloet, E. R., & Holsboer, F. (1994). Gene expression and function of interleukin 1, interleukin 6, and tumor necrosis factor in the brain. Progress in Neurobiology, 44, 397^32. Schrott, L. M. & Crnic, L. S. (1996). Anxiety behavior, exploratory behavior, and activity in NZB X NZW Fl hybrid mice: role of genotype and autoimmune disease progression. Brain, Behavior, and Immunity, 10, 260-274. Schubert, D. S. P. & Foliart, R. H. (1993). Increased depression in multiple sclerosis patients: A meta analysis. Psychosomatics, 34, 124-130. Schwartz, J. A., Speed, N. M., Brunberg, J. A., Brewe, T. L., Brown, M., & Greden, J. F. (1993). Depression in stroke rehabilitation. Biological Psychiatry, 33, 694-699. Scott, T. F , Nussbaum, P., McConnell, H., & Brill, P. (1995). Measurement of treatment response to sertraline in depressed multiple sclerosis patients using the Carroll scale. Neurological Research, 17, 421-422. Segall, M. A. & Crnic, L. S. (1990). An animal model for the behavioral effects of interferon. Behavioral Neuroscience, 4, 612-618. Seidel, A., Arolt, V., Hunstiger, M., Rink, L., Behnisch, A., & Kirchner, H. (1995). Cytokine production and serum proteins in depression. Scandinavian Journal of Immunology, 41, 534-538. Seidel, A., Arolt, V., Hunstiger, M., Rink, L., Behnisch, A., & Kirchner, H. (1996). Major depressive disorder is associated with elevated monocyte counts. Acta Psychiatrica Scandinavica, 94, 198-204. Seilhean, D., Kobayashi, K., He, Y., Uchihara, T., Rosenblum, O., Katlama, C , Bricaire, F , Duyckaerts, C , & Hauw, J. J. (1997). Tumor necrosis factor-alpha, microglia, and astrocytes in AIDS dementia complex. Acta Neuropathologica, 93, 508-517. Sluzewska, A., Rybakowski, J., Laciak, M., Mackiewicz, A., Sobieska, M., «fe Wiktorowicz, K. (1995). Interleukin-6 serum levels in depressed patients before and after treatment with fluoxetine. Annals of the New York Academy of Sciences, 762, 474-476. Sluzewska, A., Rybakowski, J., Bosmans, E., Sobieska, M., Berghmans, R., Maes, M., & Wiktorowicz, K. (1996). Indicators of immune activation in major depression. Psychiatry Research, 64, 161-167. Smith, B. K. & Kluger, M. J. (1993). Anti-TNF-alpha Antibodies Normalized Body Temperature and Enhanced Food Intake in Tumor-Bearing Rats. American Journal of Physiology, 265, R615-R619. Smith, A. P, Tyrrell, D. A. J., Al-Nakib, W., Coyle, K. B., Donovan, C. B., Higgins, P G., & Willman, J. S. (1987).

Cytokines, "Depression Due to a General Medical Condition," and Antidepressant Drugs

315

Effects of experimentally induced respiratory virus infections and illness on psychomotor performance. Neurosychobiology, 18, 144-148. Smith, A. P., Tyrrell, D. A. J., Al-Nakib, W., Coyle, K. B., Donovan, C. B., Higgins, P. G., & Willman, J. S. (1988). The effects of experimentally induced respiratory virus infections on performance. Psychological Medicine, 18, 65-71. Smith, A. P., Tyrrell, D. A. J., & Barrow, G. I. (1992). Mood and experimentally induced respiratory virus infections and illness. Psychology and Health, 6, 205-212. Smith, A. P, Thomas, M., Brockman, P, Kent, J., & Nicholson, K. G. (1993). Effect of Influenza B Virus Infection on Human Performance. British Medical Journal, 306, 760-761. Sommer, N., Loschmann, P. A., Northoff, G. H., Weller, M., Steinbrecher, A., Steinbach, J. P., Lichtenfels, R., Meyermann, R., Riethmuller, A., Fontana, A., Dichgans, J., & Martin, R. (1995). The antidepressant rolipram suppresses cytokine production and prevents autoimmune encephalomyelitis. Nature Medicine, 1, 244-248. Song, C. & Leonard, B. E. (1994). An acute phase protein response in the olfactory bulbectomized rat: Effect of sertraline treatment. Medical Science Research, 22, 313-314. Song, C , Dinan, T, & Leonard, B. E. (1994). Changes in immunoglobulin, complement and acute phase protein levels in depressed patients, and normal controls. Journal of Affective Disorders, 30, 283-288. Sonti, G., Ilyin, S. E., & Plata-Salaman, C. R. (1996). Anorexia induced by cytokine interactions at pathophysiological concentrations. American Journal of Physiology, 270, R1394-R1402. Spadaro, F. & Dunn, A. J. (1990). Intracerebroventricular administration of interleukin-1 to mice alters investigation of stimuli in a novel environment. Brain, Behavior, and Immunity, 4, 308-322. Spriggs, D. R., Sherman, M. L., Michie, H., Arthur, K. A., Imamura, K., Wilmore, D., Frei, E., & Kufe, D. W. (1988). Recombinant human tumor necrosis factor administration as a 24-hr intravenous infusion. A phase I and pharmacologic study. Journal of National Cancer Institute, 80, 1039-1044. Suzuki, T , Funada, M., Sugano, Y., Misawa, M., Okutomi, T, Soma, G. I., & Mizuno, D. (1994). Effects of a lipopolysaccharide from pantoea agglomerans on the cocaine-induced place preference. Life Sciences, 54, PL75-PL80. Swiergiel, A. H., Smagin, G. N., & Dunn, A. J. (1996). Influenza virus infection of mice induces anorexia: Comparison with endotoxin and interleukin-1 and the effects of indomethacin. Pharmacology, Biochemistry, and Behavior, 57, 389-396. Swiergiel, A. H., Smagin, G. N., Johnson, L. J., & Dunn, A. J. (1997). The role of cytokines in the behavioral responses to endotoxin and influenza virus infection in mice: effects of acute and chronic administration of the interleukin-1-receptor antagonist (IL-lra). Brain Research, 77, 696-104. Takahashi, S., Kapas, L., Fang, J., Seyer, J. M., Wang, Y., & Krueger, J. M. (1996). An interleukin-1 receptor fragment inhibits spontaneous sleep and muramyl dipeptide-induced sleep in rabbits. American Journal of Physiology, 271, R101-R108. Tilders, F J. H., Derijk, R. H., Van Dam, A. M., Vincent, V. A. M., Schotanus, K., & Persoons, J. H. A. (1994). Activation of the Hypothalamus-Pituitary-Adrenal Axis by Bacterial Endotoxins—Routes and Intermediates. Psychoneuroendocrinology, 19, 209-232. Traskman, L., Tybring, G., Asberg, M., Bertilsson, L., & Schalling, D. (1980). Cortisol in the CSF of depressed and suicidal patients. Archives of General Psychiatry, 37,161-161. Tueth, M. J. (1995). How to manage depression and psychosis in Alzheimer's disease. Geriatrics, 50{\), 43-46, 49. Turnbull, A. V. & Rivier, C. (1995). Regulation of the HPA axis by cytokines. Brain, Behavior, and Immunity, 9, 253-275. Uehara, A., Sekiya, C , Takasugi, Y, Namiki, M., & Arimura, A. (1989). Anorexia induced by interleukin 1: involvement of corticotropin-releasing factor. American Journal of Phsyiology, 257, R613-R617. Valentine, A. D., Meyers, C. A., Kling, M. A., Richelson, E., & Hauser, P. (1998). Mood and cognitive side effects in interferon-alpha therapy. Seminars in Oncology, 25{\), 39-47. Van der Meer, M. J. M., Fred Sweep, C. G. J., Pesman, G. J., Borm, G. F , & Hermus, A. R. M. M. (1995). Synergism between IL-ip and TNF-a on the activity of the pituitary-adrenal axis and on food intake in rats. American Journal of Physiology, 268, E551-E557. Walker, L. G., Wesnes, K. P, Heys, S. D., Walker, M. B., Lolley, J., & Eremin, O. (1996). The cognitive effects of recombinant interleukin-2 (rIL-2) therapy: a controlled cUnical trial using computerised assessments. European journal of Cancer, 13, 2275-2283. Walker, L. G., Walker, M. B., Heys, S. D., Lolley, J., Wesnes, K. P, & Eremin, O. (1997). The psychological and psychiatric effects of rIL-2 therapy: a controlled clinical trial. Psychooncology, 6(4), 290-301.

316

R. Yirmiya et al

Warren, E. J., Finck, B. N., Arkins, S., Kelley, K. W., Scamurra, R. W., Murtaugh, M. P., & Johnson, R. W. (1997). Coincidental changes in behavior and plasma Cortisol in unrestrained pigs after intracerebroventricular injection of tumor necrosis factor-a. Endocrinology, 138, 2365-2371. Watkins, L. R., Maier, S. R, & Goehler, L. E. (1995a). Cytokine-to-brain communication: A review and analysis of alternative mechanisms. Life Sciences, 57, 1011-1026. Watkins, L. R., Maier, S. R, & Goehler, L. E. (1995b). Immune activation: the role of pro-inflammatory cytokines in inflammation, illness responses, and pathological pain states. Pain, 63, 289-302. Weidenfeld, J. & Yirmiya, R. (1996). Effect of bacterial endotoxin on the glucocorticoid feedback regulation of adrenocortical response to stress, Neuroimmunomodulation, 3, 352-357. Weizman, R., Laor, N., Podhszewski, E., Notti, I., Djaldetti, M., & Bessler, H. (1994). Cytokine Production in Major Depressed Patients Before and After Clomipramine Treatment. Biological Psychiatry, 35, Whitlock, R A. & Siskind, M. M. (1980). Depression as a major symptom of multiple sclerosis. Journal of Neurology, Neurosurgery and Psychiatry, 43, 861-865. Willner, P. (1994). Animal models of depression. In J. A. den Boer & J. M. Sitsen (Eds.), Handbook of depression and anxiety (pp. 291-316). New York: Marcel Dekker, Inc. Willner, P. (1997). VaHdity, reliability, and utihty of the chronic mild stress model of depression: A 10-year review and evaluation. Psychopharmacology, 134, 319-329. Wilson, E. O. (1975). Sociobiology: The new synthesis. Cambridge: Harvard University Press. Wohlman, A., Gallily, R., Yirmiya, R., & Weidenfeld, J. (1997). Characterization of the effect of Mycoplasma fermentans on the hypothalamo-pituitary-adrenal axis. Neuroendocrinology, 66, 221-228. Wong, M. L., Bongiorno, P B., Rettori, V., McCann, S. M., & Licinio, J. (1997). Interleukin (IL) Ip, IL-1 receptor antagonist, IL-10, and IL-13 gene expression in the central nervous system and anterior pituitary during systemic inflammation: Pathophysiological implications. Proceedings of the National Academy of Sciences of the United States of America, 94, 227-232. Xia, Z., DePierre, J. W, & Nassberger, L. (1996). Tricyclic antidepressants inhibit IL-6, IL-ip, and TNP-a release in human blood monocytes and IL-2 and interferon-y in T-cells. Immunobiology, 34, 27-37. Xiao, L. & Eneroth, P. (1996). Tricyclic antidepressants inhibit human natural killer cells. Toxicology and Applied Pharmacology, 137, 157-162. Yamada,T. & Yamanaka, I. (1995). Microglial localization of alpha interferon receptor in human brain tissues. Neuroscience Letters, 189, 73-76. Yehuda, R. & Nemeroff, C. B. (1994). Neuropeptide alterations in affective and anxiety disorders. In J. A. den Boer & J. M. Sitsen (Eds.), Handbook of depression and anxiaty (pp. 543-571). New York: Marcel Dekker, Inc. Yirmiya, R. (1996). Endotoxin produces a depressive-like syndrome in rats. Brain Research, 711, 163-174. Yirmiya, R. (1997). Behavioral and psychological effects of immune activation: Implications for "depression due to a general medical condition". Current Opinion in Psychiatry, 10, 470^76. Yirmiya, R., Rosen, H., Donchin, O., & Ovadia, H. (1994). Behavioral effects of lipopolysaccharide in rats: Involvement of endogenous opioids. Brain Research, 648, 80-86. Yirmiya, R., Avitsur, R., Donchin, O., & Cohen, E. (1995). Interleukin-1 inhibits sexual behavior in female but not in male rats. Brain, Behavior, and Immunity, 9, 220-33. Yirmiya, R., Barak, O., Avitsur, R., Gallily, R., & Weidenfeld, J. (1997). Intracerebral administration of Mycoplsma fermentans produces sickness behavior: role of prostaglandins. Brain Research, 749,71-81. Zacharko, R. M., Zalcman, S., Macneil, G., Andrews, M., Mendella, P. D., & Anisman, H. (1997). Differential effects of immunologic challenge on self-stimulation from the nucleus accumbens and the substantia nigra. Pharmacology, Biochemistry, and Behavior, 58, 881-886. Zhu, J., Bengtsson, B. O., Mix, E., Thorell, L. H., Olsson, T, & Link, H. (1994). Effect of monoamine reuptake inhibiting antidepressants on major histocompatibility complex expression on macrophages in normal rats and rats with experimental allergic neuritis (EAN). Immunopharmacology, 27(3), 225-244. Zorzenon, M., Colle, R., Vecchio, D., Bertoli, M., Giavedoni, A., Degrassi, A., Lavaroni, S., & Aguglia, E. (1996). Major depression, viral reactivation, and immune system. European Psychiatry, 11, 332. Zubareva, O. E., Abdurasulova, I. N., Bluthe, R. M., Dantzer, R., & Klimenko, V. M. (1998). Effects of serotonin synthesis blockade on IL-1 beta actions in the brain of rats. NeurolmmunoModulation, 5, 56.

CYTOKINES, STRESS, AND DEPRESSION Conclusions and Perspectives

Robert Dantzer, Emmanuelle E. Wollman, Ljubisa Vitkovic, and Raz Yirmiya

INTRODUCTION From the data that are reviewed in this volume, several important points emerge: (1) cytokines administered to patients and laboratory animals induce symptoms of depression, including, depressed mood, decreased interest in daily activities, anhedonia, reduced food intake, sleep disorders, hyperactivity of the HPA axis, and glucocorticoid resistance; (2) exposure to stressors can induce the expression of cytokines at the periphery and in the brain, although the exact conditions in which this occurs are still elusive; (3) depressed patients display an activation of the monocyte/macrophage arm of the immune response; (4) clinical diseases with an inflammatory component are associated with a high prevalence of depressive disorders; (5) antidepressants have anti-inflammatory properties and attenuate the behavioral effects of immune challenge. Although this convergence of data would appear at first glance to support the hypothesis that cytokines are involved in the pathophysiology of depression, there are still a number of issues that need to be clarified before such a conclusion can be accepted. The objective of this last chapter is to critically review the evidence concerning the role of cytokines in depression and present the points that are still problematic.

1. HOW CLOSE IS THE ANALOGY BETWEEN THE BRAIN EFFECTS OF CYTOKINES AND THE SYMPTOMS OF DEPRESSION? Cytokine administration to laboratory animals induces a wide range of behavioral effects that resemble symptoms of depression. Those include psychomotor retardation, reduced food intake, decreased interest in daily activities, sleep disturbances, alterations in learning and memory, and anhedonia (for examples, see the chapters by 317

318

R- Dantzer et al

Anisman & Merali; Dantzer, Aubert, Bluthe, Gheusi, Cremona, Laye, Konsman, Parnet, & Kelley; Maier, Nguyen, Deak, Milligan, & Watkins; Yirmiya, Weidenfeld, Pollak, Morag, Morag, Avitsur, Barak, Reichenberg, Cohen, Shavit, & Ovadia, in this volume). At the systems level, some behavioral effects are associated with hyperactivity of the HPA axis (Tilders, & Schmidt, this volume) and glucocorticoid resistance (Miller, Pariente, & Pearce, this volume). Although these effects are usually studied in animals acutely injected with cytokines, there is evidence that they also occur during chronic immune activation, whether the treatment is administered repeatedly or animals are chronically infected. In Hne with the effects observed in laboratory animals, cytokine therapy in patients afflicted with cancer or viral hepatitis is associated with the development of cognitive and psychiatric adverse effects. As pointed out by Meyers in her chapter, depression is one of the most frequent disorders observed in cytokine-treated cancer patients. Dysphoria, anhedonia, and helplessness in addition to fatigue, apathy, and mental slowing mainly represent depressive symptoms. Although it is not easy to separate the effects of cytokines on the mood of cancer patients from other factors that might precipitate depression in these patients, the fact that depressive symptoms spontaneously regress when the treatment is interrupted provides good evidence for a causal role of cytokines. Since cytokine therapy is still experimental, there are not many examples of controlled clinical trials with cytokines in cancer patients. However, a recent study on patients with advanced colorectal cancer revealed that compared to chemotherapy alone, immunochemotherapy resulted in reduced energy, impaired confidence, higher depressed mood and more confusion (Walker, Walker, Heys, LoUey, Wesnes, & Eremin, 1997). It is difficult to compare the effects of cytokine treatment in patients with those occurring in laboratory animals for several sets of reasons. Some of these reasons can be dealt with. They are: (1) the cytokines that are administered are different: IL-2 and IFNa are administered in clinical studies, whereas IL-1, IL-6, and TNFa are the main cytokines that are studied in experimental animals; (2) the time at which cytokines effects are observed is not the same: cytokine-induced affective and cognitive alterations are usually assessed in patients several days to a few weeks after initiation of cytokine therapy, whereas effects of cytokines in laboratory animals are studied in the hours that follow cytokine administration. It is important to note that patients treated with cytokines experience in all cases flu-like symptoms immediately after the treatment, whereas the development of psychological alterations commonly requires a few days or a few weeks. Other reasons are more difficult to be dealt with. They are: (1) large individual variability in the type and severity of specific symptoms that are reported, the reasons for which are not known; (2) patients are sick (and none of those treated with cytokines thus far, have suffered from depression as the only presenting symptom) and animals are by enlarge, healthy; (3) behaviors that have been measured and behavioral tests used in studies of humans and animals differ from each other so much so that the differences sometime preclude direct comparisons. Since the comparison between the adverse effects of cytokine therapy and the behavioral action of cytokines in animals is of little help to clarify the role of cytokines in depression, it is tempting to make recourse to analogical reasoning. If cytokines induce depression-like symptoms in animals, then cytokines are causally involved in the pathogenesis of depression in psychiatric patients. However, the analogy that is drawn between the pharmacological activity of cytokines and the clinical symptoms of

Cytokines, Stress, and Depression

319

depression does not take into account the lack of isomorphism between these two categories of effects. In laboratory animals, depression is a post hoc construct that is deduced from relatively few behavioral alterations such as the decreased exploratory activity and suppressed food intake that develop in response to acute treatment with cytokines. In contrast to experiments with animals, depression in humans is usually measured by psychiatric scales, which assess many symptoms that cannot be measured in animals. For example, 6 of the 17 symptoms of the Hamilton Rating Scale for Depression can be assessed only in humans (i.e., depressed mood, guilt, suicidal ideation, psychic anxiety, hypochondriasis, and loss of insight), 5 symptoms can be partly assessed in animals, although some aspects of these symptoms are exclusively human (e.g., the feelings of incapacity, indecision, and loss of interest in hobbies), and only 6 symptoms can be directly assessed in animals (e.g., loss of body weight). Thus, the biomedical research that is carried out on depression by the combined recourse to clinical approaches and experimental studies in animals is very problematic since the object of study is not necessarily the same. A possible way to overcome this epistemological problem is to model specific symptoms of depression in animals, and to investigate the effects of cytokines and cytokine antagonists on these more restricted components of depression. In their chapters, Anisman & Merali, as well as Maier et al, describe studies in which this approach was used to study the involvement of cytokines in two key symptoms of depression, anhedonia and learned helplessness. Anhedonia, meaning "without (an) pleasure (hedonia)", refers to the fact that depressed patients no longer enjoy activities that were previously pleasurable, such as eating, sex, social and recreational activities. Anhedonia is considered as one of the two essential features of a major depressive episode with melancholia (DSM-IV, 1994). It can be assessed in animals by a decreased rate of response to rewarding intracranial self-stimulation and by a reduced preference for sweet solutions. In their chapter, Anisman & Merali show that administration of endotoxin decreased responding of rats for a rewarding stimulation of the lateral hypothalamus. Acute administration of IL-2 also decreased intracranial self-stimulation from the lateral hypothalamus in rats and the dorso-ventral tegmentum in mice. Similar effects were observed in IL-ip-treated rats, although the effective doses were also inducing sickness. In the same manner, administration of IL-ip and TNFa reduced the consumption of chocolate milk, a highly palatable diet for mice. The same effect was observed in mice treated with endotoxin or infected with an influenza virus (Swiergel, Smagin, Johnson, & Dunn, 1997). In addition, exposure to several immune challenges, including LPS, IL-ip, and intracerebral administration of mycoplasma fermentans and the HIV glycoprotein gpl20, reduced the consumption of saccharin whereas it had little or no effect on the consumption of water (Yirmiya et al., this volume). Similar changes occurred in mice that spontaneously develop a form of systemic autoimmune lupus. During the active phase of the disease, these mice displayed a blunted sensitivity to sucrose, and this symptom was reversed by immunosuppressive treatment (Sakic, Denburg, Denburg, & Szechtman, 1996). Another aspect of depression that can be modeled in animals is learned helplessness. Learned helplessness refers to the learning of the lack of contingency between behavior and its consequences, and used to be regarded by some researchers and clinicians as one of the major causes for the cognitive, affective, and motivational aspects of depression (Seligman, 1975). Although this view has somewhat lost the favor of clinicians, it still plays an important role in psychopharmacology since it is

320

R. Dantzer et ah

at the origin of some animal models of depression. In this context, an animal model of depression is nothing more than a behavioral test that is selectively sensitive to the effects of antidepressants. In animals, learned helplessness is usually induced by exposing individual subjects to inescapable electric shocks. The effects of such exposure are compared to the effects of exposure to identical shocks that are escapable, i.e., the termination of which is contingent upon the emission of the correct response by the subject. Interestingly, inescapable shock induces anhedonic symptoms in animals, in the form of a decrease in ICSS and a reduction in preference for sucrose. In their chapter, Maier et al explore the role of brain IL-1 in learned helplessness. They first demonstrate that exposure to inescapable shock induced aspects of sickness that were very similar to those produced by infection or central administration of IL-1. These aspects included fever, which persisted several days after inescapable shock, production of acute phase proteins, and activation of macrophages, as evidenced by production of nitric oxide. A more direct evidence for a role of IL-1 in mediating the effect of inescapable shock was the observation that blockade of the effects of IL-1 in the brain antagonized the effects of inescapable shock, including interference with escape learning. The problem with this effect, however, was its lack of specificity. The same sickness-like signs as those occurring in animals exposed to escapable shock were observed in animals exposed to inescapable shock. In addition, intracerebroventricular administration of IL-1 did not mimic the effects of inescapable shock in producing learned helplessness. Although these results mean that IL-1 is necessary but not sufficient for the development of learned helplessness, they do not allow one to discard the possibility that other cytokines might act in combination with IL-1 to induce learned helplessness. Taken together, findings from animal studies reveal that cytokines might be involved in some of the most fundamental aspects underlying depression in humans, including anhedonia and learned helplessness. Although these findings appear to be convincing, at least in the case of anhedonia, there is still clearly a need for further studies. Since depressed patients have been reported to show decreased enjoyment from sweet solutions (Steiner, Lifschitz, & Perl, 1993), it may be possible to add activation a direct evaluation of changes in preference for sweet solutions to the psychiatric scales that measure anhedonia in patients who receive cytokine therapy or experience various forms of immune deficiency in patients. In addition, it would be interesting to investigate more precisely the role of anhedonia in the reported cytokineinduced decrease of libido and sexual activity in female rats (Yirmiya et al., this volume). Although the examination of the validity of the analogy between clinical symptoms of depression and brain effects of cytokines has mainly concentrated on behavioral action of these molecules, neuroendocrine and physiological changes are just as important to take into account. The fact that cytokines activate the HPA axis and that this effect is similar to the increased activity of the HPA axis observed in depressed patients, has already been alluded to. This is, however, only one aspect of the many effects of cytokines on neuroendocrine axes. Proinflammatory cytokines also enhance the secretion of vasopressin and oxytocin, whereas they decrease the activity of the gonadotropic and somatotropic axes (Turnbull & Rivier, 1996). The way these effects relate to the neuroendocrine profile of depression has received much less attention than the activation of the HPA axis.

Cytokines, Stress, and Depression

321

2. DOES STRESS INDUCE THE EXPRESSION OF CYTOKINES? The role of stress in depression cannot be reduced to a simple question of causality. TTiere are different clinical forms of depression and many types of stressors, which occur in different contexts, and do not have the same meaning for different persons. The only clinical conditions in which the causal role of psychosocial stressors is recognized in the DSM-IV classification are post-traumatic stress disorders, and adaptation disorders. The reluctance of psychiatrists to make use of the concept of stress in their practice does not prevent numerous scientists from considering the relationship between stress and depression, especially in a biological perspective. In this context, cytokines are viewed as another category of molecular factors that play a role in depression because their production and action are modulated by stress. In their chapter, Watkins, Nguyen, Lee, and Maier show that although there is some evidence for the capacity of a wide variety of physical and psychological stressors to increase the synthesis and release of proinflammatory cytokines at the periphery and in the brain, stress simply cannot be stated either to increase or to decrease cytokine production. There are many reasons for this confusion, from the diversity of stressors that have been studied to the multiple levels at which cytokine production and function are regulated. In the context of depression, however, the range of questions that can be asked concerning the relationship between stress and cytokines is more restricted since it is more important to determine whether those stressors that are assumed to precipitate depression can induce the production of cytokines than to determine the effects of stressors in general on cytokine production. The most important stressor for precipitating depression is bereavement. However, there has been no study of cytokine production in bereaved patients or in animal models of separation. Other pertinent stressors are those that are responsible for the development of learned helplessness. In the model of inescapable electric shock described previously, stressed rats displayed an increase in IL-1|3 concentrations in the pituitary, nucleus tractus solitarius, hypothalamus and hippocampus (Maier et al., this volume). However, the specificity of these changes is doubtful since rats exposed to inescapable electric shocks responded in exactly the same way as rats exposed to escapable electric shock, with the possible exception of the amygdala in which IL-ip was elevated only in rats exposed to escapable shock. When considering the relationship between stress and cytokines, it is important to remember that cytokines themselves can be regarded as stressors, at least from the viewpoint of HPA axis activation. It is known that an episode of stress can influence further response to the same or different stressors, and result in an increase (sensitization) or a decrease (habituation) of the stress response. It is therefore possible to examine how an immune stressor can alter the organism's response to a non-immune stressor. In their chapter, Tilders et al. show very convincingly that prior exposure of rats to IL-1 enhanced the pituitary-adrenal and autonomic responses to non immune stressors, including novelty, electric footshock, and the psychostimulant amphetamine. This cross sensitization developed within 1-3 weeks after the immune stressor, and was due to the reprogramming of stress-sensitive neural circuits and gene expression dependent changes in the properties of its elements. In comparison to non-immune stressors, inflammatory stimuli are particularly potent in provoking such long lasting changes in neuroendocrine circuits. Whether this phenomenon is of clinical significance, especially in elderly people who commonly develop clinical symptoms of depression that are

322

R. Dantzer et al.

correlated with concurrent infections and inflammation (Ernst, 1997), remains to be determined, but it represents an interesting issue for further research. 3. IS D E P R E S S I O N A S S O C I A T E D W I T H A C T I V A T I O N O F T H E I M M U N E SYSTEM? If cytokines are involved in the pathophysiology of depression, their production should be increased in depressed patients. The relationship between depression and immunity is certainly the topic that has been studied the most intensively in clinically oriented psychoneuroimmunology. At the beginning of this research, the expectations were quite high. In concordance with the belief that the mind influences the body, mental depression was viewed as being necessarily associated with immunosuppression. The first results were quite encouraging since assays of mitogen-induced lymphocyte proliferation and natural killer (NK) cell cytotoxic activity carried out on white blood cells confirmed that depressed patients have decreased cell mediated immunity. This decrease was more marked in patients with melancholia than in other categories of patients. Could this be one of the physiological markers of melancholia? These alterations were viewed as the consequence of the hypercortisolemia that characterizes depressed patients and, as expected, the disappearance of hypercortisolemia in response to antidepressant treatment was accompanied by normalization of immune functions. However, this first wave of optimism did not last for very long. After one decade of intense investigation of the impact of depression on immunity, it became quite clear that the results were not consistent, and that the immunosuppression, when it was present, remained within the limits of normal variation. Furthermore, at the immunological level, the concept that mental depression is necessarily associated with immune depression was challenged by findings of an increase in certain markers of immune activation. In his review of the immune changes that are observed in depressed patients, Irwin (this volume) points out that it is premature to throw the baby with the bath water. From the data available on lymphocyte functions, it is clear that depressed patients have a decreased number of lymphocytes, reduced mitogen-induced lymphocyte proliferation, and decreased NK cell activity. Some of these changes could be due to sleep disorders rather than to depression per se, and they might be associated with an increased sensitivity to certain viruses. Alternatively, a sleep disorder and/or a viral infection may be integral parts of a depressive state. For example, sleep may be affected by immune response to a viral infection (Krueger and Majde, 1994). Molecular and cellular changes may underlie neurobehavioral changes resulting in a specific depressive state. A depressive state may be an integral sum of discernable and quantifiable behaviors "out of tune" with each other. This disharmony may be better understood at one of the more basic levels of organization than the behavioral level. Concerning immune activation, there are less studies that are available, and the results are still, and not surprisingly, heterogeneous, even if it is possible to observe associations between depression and increased number of white blood cells and neutrophils, and enhanced levels of haptoglobin, a positive acute phase protein. The evidence in favor of an enhanced activation of the monocytic arm of the immune response in depressed patients has been previously summarized by Maes and his colleagues (Maes, Smith, & Scharpe, 1995) and is detailed further by Maes; Seidel, Rothermund, & Rink; and Sluzewska in their respective chapters.

Cytokines, Stress, and Depression

323

There are many methodological problems with this type of research. The most important one is the health status of patients. Normally, patients who are suffering from a concurrent organic disease should not be included in studies on the relationship between depression and immunity. However, the levels of C-reactive proteins that are displayed by some of the depressed patients characterized as having signs of immune activation would be sufficient to consider them as ill even if they do not show any clinical symptom of infection or inflammation. Another important issue is the nature of the assay that is used to assess production of cytokines. Since cytokines are not hormones, with the exception perhaps of IL-6, plasma levels of cytokines are meaningless. The capacity of white blood cells to produce cytokines in response to LPS and interferon-y is a better option, which explains why whole blood assays have been developed for the purpose of measuring cytokine production in clinical settings. However, the validation and standardization of these assays are usually rudimentary, which allows to understand why the favor is still given to more global indexes of inflammation, such as the ratio of positive to negative acute phase proteins. Whatever the case, it still remains that even if production of cytokines can be measured with adequate techniques, this is a peripheral event that does not necessarily correlate with the ability of brain cells to produce cytokines. Overactivity of the brain IL-1 system, induced by icv administration of IL-1 in rats, has been shown to be associated with increased peripheral levels of IL-6 (DeSimoni, Seroni, DeLuigi, Manfridi, Mantovani, & Ghezzi, 1990), but the way this response generalizes to other cytokines and is reflected in production of cytokines by white blood cells has not been investigated. There is clearly a need for further investigation of the relationships between the central and peripheral cytokine compartments. Independently of the site of production of cytokines, the question of specificity of the alterations in cytokine production that is observed in depressed patients still needs to be addressed properly. Elevated levels of proinflammatory cytokines (mainly IL-6 and its soluble receptor) and biochemical signs of an acute phase reaction have been evidenced in patients diagnosed with schizophrenia, mania, and post-traumatic stress disorder. However, such overlapping results are not uncommon in biological psychiatry, and they make one question the usefulness of correlating biological variables with disease entities that are defined exclusively on the basis of clinical criteria. The challenge is certainly to transcend this arbitrary categorization by identifying the biologically relevant transnosographic dimensions that are obliterated in the DSM-IV classification.

4. DO ANTIDEPRESSANTS INTERFERE WITH CYTOKINE PRODUCTION AND ACTION? The hypothesis that cytokines play a key role in the pathophysiology of depression implies that antidepressant treatments should exert their effects by interfering with cytokine production and action. Antidepressants are usually not considered as anti-inflammatory drugs even if some psychopharmacologists have proposed that they behave as central inhibitors of prostaglandins (Leonard, 1987). However, if one goes beyond such general statements to investigate in details the ability of antidepressants to affect cytokine production and action, many surprising effects can be found. The ability of antidepressants to impair the release of

324

R. Dantzer et al.

proinflammatory cytokines from activated monocytes and macrophages, act as inhibitors of chemotaxis, and enhance the expression of anti-inflammatory cytokines is reviewed by Neveu and Castanon in their chapter. Yirmiya et al (this volume) go one step further by demonstrating that chronic but not acute antidepressant treatment abrogates symptoms of LPS-induced sickness behavior in rats, specially the decreased preference for saccharin that mimics anhedonia. This effect is probably mediated by a decreased ability of monocytes and macrophages to produce proinflammatory cytokines. Although pharmacologists are prompt to jump from such positive results to pathophysiological interpretations, there is still a long way to go. The preliminary results obtained in animals should encourage more systematic studies on the effects of various classes of antidepressants and non pharmacological treatments of depression (e.g., electroshock) on the different components of sickness induced by LPS or recombinant cytokines. If these effects are confirmed, their mediating mechanisms should then to be elucidated. At the clinical level, this effort should be accompanied by investigation of the effects of antidepressant treatment on immune activation in depressed patients and the temporal relationship between these effects and mood changes.

5. ARE CYTOKINE ANTAGONISTS ABLE TO ATTENUATE THE SYMPTOMS OF DEPRESSION? If cytokines cause depression then their antagonists acting in the brain should have antidepressant activity. Besides the previously mentioned studies on the abiUty of IL-1 antagonists to block learned helplessness in certain conditions (see point 1), this possibility has received little attention so far. There is obviously room for investigating the effects of cytokine antagonists in animal models of depression and anxiety, especially since these antagonists have little or no effect on their own, besides their capacity to interfere with the production and action of proinflammatory cytokines. In such a quest, cytokine antagonists with a relatively wide spectrum of action, such as IL-10 and IL-4 which inhibit the synthesis of several proinflammatory cytokines by activated monocytes, are better suited than more specific cytokine antagonists, such as IL-lra, which interferes only with the ability of IL-la and IL-ip to bind to their receptors. At the clinical level, clinical trials on the possible effects of centrally acting anti-inflammatory compounds on depression should be undertaken, as the results of these trials are likely to play an important role in shaping the future of the field. In the meantime, it is important to note that in a pilot study designed to objectively evaluate the effects of a single dose of endotoxin on depression, administration of a low dose of LPS significantly improved mood in seven drug-free severely depressed patients on the next day following the treatment, at a time when the expression of anti-inflammatory cytokines should be expected to be maximal (Bauer, Hohagen, Gimmel, Bruns, Lis, Krieger, Ambach, Guthmann, Grunze, & FritschMontero, 1995). However, these changes in mood were transient and dissipated during the next days. These observations support the notion that depression is far too complex a disorder to be corrected by a "golden bullet" (i.e., a drug targeting a single molecule).

Cytokines, Stress, and Depression

325

6. DO THE BRAIN ACTION OF CYTOKINES AND THE SYMPTOMS OF DEPRESSION SHARE COMMON MECHANISMS? The analogy between the pharmacological effects of cytokines and symptoms of depression would be much easier to substantiate if it were possible to identify a common mediating mechanism preferably at the molecular level. There are at least two currently available candidates for this purpose, CRH and serotonin. However, before going into the details for each of these possibilities, it is important to point out that although research on the neurobiology of depression is more advanced than that on the immunology of depression, the exact biological basis of depression is still elusive. The analogy that has been drawn between brain effects of cytokines and symptoms of depression appears to be similar in its principle to the one that was proposed some time ago for CRH and depression. Like the macrophage theory of depression, the hypothesis of a causal role of CRH in the pathophysiology of depression has been built up on the convergence of two sets of data. At the experimental level, CRH infusion into the brain of laboratory rodents has been shown to induce depression-like signs, including sleep disorders, reduced appetite, decreased sexual behavior, and anxiety. At the clinical level, depressed patients are known to have an hyperactive HPA axis, as evidenced by higher levels of CRH in the cerebrospinal fluid, increased plasma Cortisol levels, and resistance to the dexamethasone suppression test. In line with the CRH hypothesis of depression, and as proposed by Miller et al in their chapter, proinflammatory cytokines would play a causal role in depression mainly because they provide another stimulatory input to the CRH system. Hyperactivity of the brain cytokine system would increase the production of CRH, which would escape the glucocorticoid feedback because of the impairment in glucocorticoid receptor translocation from the cytoplasm to the nucleus. At the same time, cytokines would increase the reactivity of the HPA axis to external stressors, by inducing the expression of vasopressin, another ACTH secretagogue, in parvocellular CRH neurons of the paraventricular nucleus of the hypothalamus (Tilders & Schmidt, this volume). This chronically increased CRH drive can by itself alter the production and action of brain cytokines (Linthorst & Reuls, this volume). However, besides the activating effects of cytokines on the HPA axis, all of this is still very speculative, and even the possibility that CRH mediates the brain effects of proinflammatory cytokines is not fully supported by the available data (Dantzer et al., this volume). Another possible mediator of the brain effects of cytokines is serotonin. The serotonin hypothesis of depression has already given rise to many controversies, and discussion of it is beyond the scope of this chapter. The synthesis of serotonin is dependent on the availability of its precursor, tryptophan. This essential amino acid enters the brain via a transport mechanism for which it competes with other neutral amino acids like tyrosine. Proinflammatory cytokines play a crucial role in the metabolism of tryptophan by enhancing the activity of indoleamine 2,3-deoxygenase (IDO). This enzyme degrades tryptophan according to the kinurenine pathway, which leads to the formation of quinolinic acid. This is not a pharmacological effect but a physiologically relevant one, as demonstrated by recent experiments on the mechanisms of immunologic tolerance of the fetus (Munn, Zhou, Attwood, Bondarev, Conway, Marshall, Brown, & Mellor, 1998). Proinflammatory cytokines produced by macrophage-like cells from the trophoblast induce IDO, which results in the decreased local availability of

326

R. Dantzer et al.

tryptophan for cytotoxic T cells. There is evidence that the enhanced expression of brain cytokines in response to systemic immune activation is accompanied by enhanced activity of IDO in microgUal cells and increased brain concentrations of quinolinic acid (Heyes, Saito, Chen, Proescholdt, Nowak, Li, Beagles, Proescholdt, Zito, Kawai, & Markai, 1997). Although cytokines have been shown to interfere with the metabolism of serotonin (Dunn et al., and Linthorst & Reuls, this volume), the role of these alterations in the brain effects of cytokines has not yet been investigated. Similarly, the effect of antidepressant treatment on cytokine-induced IDO activity is unknown.

7. P E R S P E C T I V E S The data reviewed in this volume indicate that a common molecular mechanism underlying all of depression may not exist. Although this statement may be obvious, it is not trivial. It suggests that an experimental approach to depression, as is the case with most complex disorders, must come to terms with a lack of a "final common pathway" of pathogenesis that can be corrected with one drug or a class of drugs. Depression may have to be approached at several, if not all, levels of organization of biological matter. These levels, from the bottom to the top, are: molecules, cells, cell groups, brain nuclei, circuits, organs, systems, and behavior (whole body). Activity at each level is bidirectionally integrated, from the bottom up (bottom-up) and from the top down (top-down).This "bidirectionality of causation" appears to be an essential feature of integration of biological phenomena whose principles and significance are only now beginning to be understood (Koslow, Meinecke, Lederhendler, Khachaturian, Nakamura, Karp, Vitkovic, Glanzman, & Zalcman, 1995, Vitkovic & Koslow, 1994). However, as levels of organization increase, our current knowledge of their integration decreases. This is relevant to resolving paradoxes described in this chapter and planning future experiments. Depression is now recognized as a complex mental disorder, indeed a syndrome. Symptoms are numerous and varied. Several types of depression such as manicdepressive disorder, melancholia, major depression, seasonal affective disorder, and others, have been identified and characterized on the behavioral level. These disorders are distinct and may be viewed as independent entities comprising a syndrome of depression. Moreover, some symptoms of this syndrome exist in other mental illnesses such as obsessive-compulsive disorder. Diagnostic criteria of depression have considerably improved over the last decade. However, many are subjective, a majority of which cannot be vahdated in animals and most, if not all, lack physiological bases. Thus, a gulf exists between diagnostic criteria and pathophysiology. This is one of the obstacles in understanding and treating psychopathology of depression. For example, most biological markers of depression are considered "surrogate", signifying lack of confidence in their relationship to the disorder and, therefore, therapeutic significance. This confounds biomedical research in depression. One way to move forward is to characterize the heterogeneous and complex syndrome of depression by extensive neurobehavioral testing using DSM and clinical tests as guides. In addition, diagnostic procedures for identifying specific forms of brain damage may be useful here as they have been useful in characterizing other mental disorders. A result will be a neurobehavioral testing battery. It should be applicable to humans, non-human primates and other laboratory animals such as mice. Behavioral studies with animals have provided important insights into physiological basis of

Cytokines, Stress, and Depression

327

motivated behaviors including food intake, sleep, and sex. These behaviors arise from the interaction of an organism with the external environment, are subsumed by brain-immune system interactions and are mediated by cytokines, hormones, and neurotransmitters (as described in this volume). These motivated behaviors are dysregulated in both, depression and in infections (Vitkovic & Koslow, 1994). Because these behaviors emanate from relatively well known neural and cytokine circuits, they are prime candidates for inclusion in the behavioral testing battery for depression. In addition, novel and important basic behavioral traits will likely be discovered, investigated, characterized, and incorporated into the battery. An example of such a trait is the startle response, which has distinguishable characteristics in some schizophrenics. Neurobehavioral testing may reduce the syndrome of depression into various types and reveal that a type of depression (e.g. major depression) consists of an integral sum of subtle pathologic changes occurring in a set of behaviors. Some of the affected behaviors may be motivated behaviors (e.g. food intake, sleep, etc). They are subsumed by overlapping and interacting neural circuits involving, among others, cytokine signaling. Another subtype of depression (e.g. melancholia) may also contain changes in some of the same motivated behaviors (e.g. sleep) but not others (e.g. food intake). We know that each motivated behavior is regulated by one dominant cytokine (e.g. TNF in food intake, IL-1 in sleep, etc). Other cytokines, however, may take over when needed (such as in gene knockout mutant animals. This built-in redundancy is fail-safe for an organism and perplexing for an investigator. Thus, neurobehavioral testing may lead us from the depression syndrome to types of depression, to constituent motivated behaviors, and to the lower levels of biological organization. This way, we may begin to bridge the gulf between diagnosis and pathophysiology and discover the behavioral symptoms that will unveil the pathophysiological substrates of some types of depression. At lower levels of biological organization, we have progressed in appreciating the importance of bottom-up integration for understanding complex mental disorders. "Instead of framing questions in terms of which "brain centers" or specific neurotransmitters [or cytokines] might be the root cause of a particular psychiatric illness, neuroscientists now explore how the coordinated action of neural networks, involving several brain regions and neurotransmitters, would produce the various symptoms that characterize a particular mental illness." (Koslow et al. 1995). We now need more data about cytokine changes in brains of depressed individuals. These data may be obtained by functional brain imaging during life, and by all methods at our disposal post mortem. Although studies of brain tissue from individuals who died with depression are rare, interesting information has already emerged. For example, Watson, Akil, and their collaborators found a 25% increase in pro-opiomelanocortin mRNA in anterior pituitary of individuals who died of suicide compared to individuals who died of cardiac arrest (Lopez, Palkovits, Arato, Mansur, Akil, & Watson, 1992). These data evidence chronic activation of the HPA axis at the system level and provide important validation of the principle that HPA axis is activated in clinically depressed individuals. The major remaining challenge is dealing with bi-directional causality in depression. Are changes in a motivated behavior a consequence of a particular depressive disorder (bottom-down integration), or one of the causes of the disorder (bottom-up integration), or both? Manipulating cytokine activities and measuring resulting behaviors in animals (e.g., Dantzer et al., this volume) is one way to face this challenge. Cytokines should be administered chronically, in physiological doses and by biologically relevant routes. Experimental animals should be subjected to a battery of tests

328

R. Dantzer et al.

measuring multiple behaviors known to be symptomatic of depression and having (at least partially) known neural substrates. Behaviors in the testing battery should correspond to behaviors defining a particular type of depression and should be measured similarly to the way they are measured in humans.

8. C O N C L U S I O N The intersection between cytokines and depression is a fascinating avatar of the research on brain-immune interactions. Whether this new field of research will prove important and fruitful for our understanding of the psychopathology of depression is still uncertain. In either case, it is clear that the questions addressed in this volume would not have been possible only a few years ago. The discovery of the expression of cytokines and cytokine receptors in the brain, and the demonstration of their modulation by peripheral and central immune activation have opened new avenues in neurosciences in general, and in neuropharmacology, in particular. The key role played by immune-brain interactions in mediating the adaptive response to infectious agents raises the possibility that a dysfunction of this regulatory system contributes to the pathogenesis of mental disorders (Vitkovic & Koslow, 1994). This, in turn, opens a door to therapeutics. Only five years ago, National Institute of Mental Health, National Institutes of Health, USA asked over 100 prominent scientists to assess current state of knowledge in the neurosciences and recommend areas for future development. A report that resulted from these deliberations concludes the chapter on "Neuropsychopharmacology" as follows: "A burgeoning area of research that shows promise for drug discovery stems from work done over the last few years on the neural immune axis. Neuro-immune interactions involve the HPA axis and other neural circuits in brain that represent potential targets for neuropharmacological intervention.... The involvement of interleukins and interferons in behavior and other physiological responses via brain action opens a new and important research arena: the development of pharmacological tools to shape the effects of these substances". The report enthusiastically concludes, "Such tools appear to be within our grasp" (Koslow et al. 1995). Today, the complexities of depression and neuro-immune interactions temper our enthusiasm. Although the neurobiology of depression may not yet be ready to make room for the immunology of depression, there is no doubt that previously unseen molecular and cellular players have made their entry on the stage. Their exact role in relation to the other players that are already familiar to psychiatrists still needs clarification, but their mere existence will undoubtedly shape our future views of psychopathology.

REFERENCES Bauer, J., Hohagen, F., Gimmel, E., Bruns, F., Lis, S., Krieger, S., Ambach, W., Guthmann, A., Grunze, H., & Fritsch-Montero, R. (1995). Induction of cytokine synthesis and fever suppresses REM sleep and improves mood in patients with major depression. Biological Psychiatry, 38,611-621. DeSimoni, M. G., Seroni, M., DeLuigi, A., Manfridi, A., Mantovani, A., & Ghezzi, P. (1990). Intracerebroventricular injection of IL-1 induces high circulating levels of IL-6. Journal of Experimental Medicine, 171,1773-1778. DSM-IV, Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition). (1994). Washington, DC: American Psychiatric Association. Ernst, C. (1997). Epidemiology of depression in late life. Current Opinions in Psychiatry, 10,107-112.

Cytokines, Stress, and Depression

329

Heyes, M. P., Saito, K., Chen, C. Y., Proescholdt, M. G., Nowak, T. S. Jr., Li, J., Beagles, K. E., Proescholdt, M. A., Zito, M. A., Kawai, K., & Markai, S. P. (1997). Species heterogeneity between gerbils and rats: quinoHnate production by microglia and astrocytes and accumulations in response to ischemic brain injury and systemic immune activation. Journal of Neurochemistry, 69,1519-1529. Koslow, S. H., Meinecke, D. L., Lederhendler, 1.1., Khachaturian, H., Nakamura, R. K., Karp, D., Vitkovic, L., Glanzman, D. L., & Zalcman, S., Editors (1995). The Neuroscience of Mental Health II, NIH Publication No. 95^000, Washington, DC: U.S. Government Printing Office. Krueger, J. M. & Majde, J. (1994). Microbial products and cytokines in sleep and fever regulation. Critical Reviews in Immunology 14, 335-379. Leonard, B. E. (1987). Stress, the immune system and mental illness. Stress Medicine, 3,257-258. Lopez, J. H., Palkovits M., Arato, M., Mansur A., Akil, H., & Watson, S. J. (1992). Localization and quantification of pro-opiomelanocortin mRNA and glucocorticoid mRNA in pituitaries of sicicide victims. Neuroendocrinology 56, 491-501. Maes, M., Smith, R., & Scharpe, S. (1995). The monocyte-T lymphocyte hypothesis of major depression. Psychoneuroendocrinology, 20, 111-116. Munn, D. H., Zhou, M., Attwood, J. T, Bondarev, I., Conway, S. J., Marshall, B., Brown, C , & Mellor, A. L. (1998). Prevention of allogeneic fetal rejection by tryptophan catabolism. Science, 281,1191-1193. Sakic, B., Denburg, J. A., Denburg, S. D., & Szechtman, H. (1986). Blunted sensitivity to sucrose in autoimmune MRL-lpr mice: A curve shift study. Brain Research Bulletin, 41, 305-311. Seligman, S. E. P. (1975). Helplessness. San Francisco: Freeman. Steiner, J. E., Lifschitz, L., & Perl, E. (1993). Taste and odor: reactivity in depressive disorders, a multidisciplinary approach. Perceptual and Motor Skills, 77, 1331-1346. Swiergel, A. H., Smagin, G. N., Johnson, L. J., & Dunn, A. J. (1997). The role of cytokines in the behavioral responses to endotoxin and influenza virus infection in mice: effects of acute and chronic administration of the interleukin-1 receptor antagonist (IL-lra). Brain Research, 77, 696-704. TurnbuU, A. V. & Rivier, C. (1996). Cytokine effects on neuroendocrine axes: influence of nitric oxide and carbon monoxide. In N. J. Rothwell (Ed.), Cytokines in the Nervous System (pp. 93-116). Austin: Landes. Vitkovic, L. & Koslow, S. H., Editors (1994). Neuroimmunologly and Mental Health. NIH Publication No 94-3774, Washington, D.C.: U.S. Government Printing Office. Walker, L. G., Walker, M. B., Heys, S. D., Lolley, J., Wesnes, K., & Eremin, O. (1997). The psychological and psychiatric effects of rIL-2 therapy: a controlled clinical trial. Psychooncology, 6, 290-301.

CONTRIBUTORS

Tetsuya Ando, Department of Pharmacology and Therapeutics, Louisiana State University, Shreveport, Louisiana, USA: Chapter 8 Hymie Anisman, Department of Psychology, Carleton University, Ottawa, Ontario, Canada: Chapter 12 Arnaud Aubert, Integrative Neurobiology, INRA-INSERM U394, Bordeaux, France: Chapter 6 Ronit Avitsur, Departement of Psychology, The Hebrew University, Jerusalem, Israel: Chapter 16 Ohr Barak, Departement of Psychology, The Hebrew University, Jerusalem, Israel: Chapter 16 Rose-Marie Bluthe, Integrative Neurobiology, INRA-INSERM U394, Bordeaux, France: Chapter 6 Nathalie Castanon, Integrative Neurobiology, INRA-INSERM U394, Bordeaux, France: Chapter 15 Edna Cohen, Departement of Psychology, The Hebrew University, Jerusalem, Israel: Chapter 16 Sandrine Cremona, Integrative Neurobiology, INRA-INSERM U394, Bordeaux, France: Chapter 6 Robert Dantzer, Integrative Neurobiology, INRA-INSERM U394, Bordeaux, France: Preface, Chapter 6, Conclusions and Perspectives Terrence Deak, Department of Psychology, University of Colorado at Boulder, USA: Chapter 13 Adrian J. Dunn, Department of Pharmacology and Therapeutics, Louisiana State University, Shreveport, Louisiana, USA: Chapter 8 Gilles Gheusi, Integrative Neurobiology, INRA-INSERM U394, Bordeaux, France: Chapter 6 Michael Irwin, Department of Psychiatry, University of California at San Diego, USA: Chapter 1 Keith W. Kelley, Laboratory of Immunophysiology, University of Illinois at Urbana, USA: Chapter 6 Jan-Peter Konsman, Integrative Neurobiology, INRA-INSERM U394, Bordeaux, France: Chapter 6 331

332

Contributors

Sophie Laye, Integrative Neurobiology, INRA-INSERM U394, Bordeaux, France: Chapter 6 Jacqueline E. Lee, Department of Psychology, University of Colorado at Boulder, USA: Chapter 10 Bryan E. Leonard, Department of Pharmacology, National University of Ireland, Galway, Ireland: Chapter 14 Astrid C. E. Linthorst, Max-Planck Institut fiir Psychiatric, Munich, Germany: Chapter 9 Michael Maes, University Department of Psychiatry, Anvers, Belgique: Chapter 2 Steven E Maier, Department of Psychology, University of Colorado at Boulder, USA: Chapter 10 & 13 Zul Merali, Department of Psychology, Carleton University, Ottawa, Ontario, Canada: Chapter 12 Christina A. Meyers, Department of Neuro-Oncology, University of Texas at Houston, USA: Chapter 5 Andrew H. Miller, Department of Psychiatry and Behavioral Sciences, Emory University School of Medecine, Atlanta, Georgia, USA: Chapter 7 Erin D. Milligan, Department of Psychology, University of Colorado at Boulder, USA: Chapter 13 Avraham Morag, Departement of Psychology, The Hebrew University, Jerusalem, Israel: Chapter 16 Michel Morag, Department of Psychology, The Hebrew University, Jerusalem, Israel: Chapter 16 Pierre Neveu, Integrative Neurobiology, INRA-INSERM U394, Bordeaux, France: Chapter 15 Kien T. Nguyen, Department of Psychology, University of Colorado at Boulder, USA: Chapter 10 & 13 Haim Ovadia, Departement of Psychology, The Hebrew University, Jerusalem, Israel: Chapter 16 Carmine M. Pariente, Department of Psychiatry and Behavioral Sciences, Emory University School of Medecine, Atlanta, Georgia, USA: Chapter 7 Patricia Parnet, Integrative Neurobiology, INRA-INSERM U394, Bordeaux, France: Chapter 6 Bradley D. Pearce, Department of Psychiatry and Behavioral Sciences, Emory University School of Medecine, Atlanta, Georgia, USA: Chapter 7 Yehuda Pollak, Departement of Psychology, The Hebrew University, Jerusalem, Israel: Chapter 16 Avraham Reichenberg, Departement of Psychology, The Hebrew University, Jerusalem, Israel: Chapter 16 Johannes M. H. M. Reul, Max-Planck Institut fiir Psychiatric, Munich, Germany: Chapter 9 Lothar Rink, Medizinische Universitat zu Liibeck, Germany: Chapter 3 Matthias Rothermundt, Medizinische Universitat zu Liibeck, Germany: Chapter 3 Eduard D. Schmidt, Department of Pharmacology, Vrije Universiteit, Amsterdam, The Netherlands: Chapter 11 Andreas Seidel, Medizinische Universitat zu Liibeck, Germany: Chapter 3 Yehuda Shavit, Departement of Psychology, The Hebrew University, Jerusalem, Israel: Chapter 16

Contributors

333

Anna Sluzewska,* Department of Adult Psychiatry, University of Medical Sciences, Poznan, Poland: Chapter 4 Cai Song, Department of Pharmacology, National University of Ireland, Galway, Ireland: Chapter 14 Fred J. H. Tilders, Department of Pharmacology, Vrije Universiteit, Amsterdam, The Netherlands: Chapter 11 Ljubisa Vitkovic, CNRS UPR 9023, Montpellier, France: Conclusions and Perspectives Jianping Wan, Department of Pharmacology and Therapeutics, Louisiana State University, Shreveport, Louisiana, USA: Chapter 8 Linda R. Watkins, Department of Psychology, University of Colorado at Boulder, USA: Chapter 10 & 13 Joseph Weidenfeld, Departement of Psychology, The Hebrew University, Jerusalem, Israel: Chapter 16 Emmanuelle E. WoUman, CNRS, Paris, France: Preface, Conclusions and Perspectives Raz Yirmiya, Department of Psychology, The Hebrew University, Jerusalem, Israel: Preface, Chapter 16, Conclusions and Perspectives

'• Deceased.

INDEX

Acute phase proteins, 4, 5, 26, 27, 50, 53,60, 63,110, 201,239,256,272,323 Adenosine triphosphate (ATP), 242 Adrenalectomy, 33,99-101,112,240-243 Alpha-melanotropic hormone, 243-244, 299-300 Age, 1,4,5,18,19,191,260 Alcoholism, 6,15-16 Anhedonia, 32, 76, 205-226,283, 294-296,299, 317, 319-320 Anorexia, 9,10, 32,49, 80, 85, 93-94,96-97,123124, 205, 292, 299 Anterior preoptic area: see Preoptic area of the hypothalamus, anterior Antidepressants: see Selective serotonin reuptake inhibitors; Tricyclic antidepressants Anxiety, 9,14-15, 78,114, 205-226 Astrocyte, 48, 54 ATP: see Adenosine-triphosphate Auto-antibodies, 66-67 Auto-immune disease: 35, 48, 61, 287-290, 319 Blood-brain barrier, 48, 53, 54, 87 Body weight loss, 9,10, 32, 292 Cancer, 75-81, 252,290, 318 Cholesterol, 28, 29 Orcadian rhythm, 12,14,257 Circumventricular organ, 87, 94-95 Cognitive disorders, 9,10,32,75-81, 211,285, 293 Corticotropin-releasing hormone (CRH), 34,60, 93-94,108-109,113,117,118,130,142-148, 163,184-192,204-205, 297, 325 CRH: see Corticotropin-releasing hormone Depression due to a general medical condition, 35, 283-306 Dexamethasone, 108-109, 111, 112,164,297298 Dopamine, 34, 117-125, 203-206, 208,214-215, 218-220,223-224 Dysthymia, 201

Electric shock, 33,34, 36,117-118,120,180-181, 205,236-245 Fatigue, 76, 77, 292-293, 299 Fatty acids, 28, 36, 38 Fever, 84, 91, 93, 97,124,141,145,237-238 Gender, 1, 3,4, 6-7,15,18, 35-36, 260, 288 Glucocorticoids: see Hypothalamic-pituitaryadrenal axis Glucocorticoid feedback, 99-101,113, 155-156, 298 Glucocorticoid receptors, 30, 109-114 Glucocorticoid resistance, 30,108-114 Hamilton rating scale for depression, 5, 8, 9,10, 319 Helplessness, 34,76, 236-237, 317, 319-320 Hippocampus, 30, 91, 132-139,208, 223-224, 241242 Hospitalization, 8, 9, 260 Hypothalamic-pituitary-adrenal axis, 25, 30, 31,32, 34,38,49, 59,62, 64-65, 84, 91, 92-93,108114,123, 141,144-145,183-192, 204-205, 214,218, 223-226, 253, 297, 301, 304-305, 321-322 Hypothalamus, 30, 33, 48,91,117-125, 241-242; See also Paraventricular nucleus; Preoptic area of the hypothalamus, anterior Interferon-alpha (IFN-a), 32,48, 49,75-79,122, 255, 268, 288, 318 Interferon-gamma (IFN-y), 27, 30, 34, 35,37,48,52, 271 Interleukin-1 (IL-1), 27,29, 30, 33, 34, 36, 37, 48,49, 51, 59, 84, 91,96-98,110-112,119-121,135137,139,144,161-168, 215-222, 235-246, 255,318, 327 receptor, 30,48,84, 91-92, 97,101,166,167-168 receptor antagonist, 34, 35, 36, 37, 51, 64, 84, 9698, 122-123, 135-136,166-167, 244-245, 324 Interleukin-2 (IL-2), 11,12, 27, 29, 30, 34, 36, 48, 49, 52, 77, 110, 121, 135,139, 210-215, 318 soluble receptor, 4, 26, 52, 60, 61, 63, 64, 271 335

Index

336 Interleukin-4 (IL-4), 110,324 InterIeukin-6 (IL-6), 4,26,27,29,30,32, 33, 34, 37, 48, 51, 59,63, 67, 84, 91, 98,110-112,121, 123,144, 271-272, 318, 323 soluble receptor, 4, 51, 63,323 Interleukin-10 (IL-10), 34, 35,37, 52,271, 324 Intracranial self-stimulation, 86, 206, 208, 213, 214, 217, 295-296,319 In vivo microdialysis, 131-132, 210, 220-221 Iron, 29 Lipopolysaccharide (LPS), 32, 33, 91, 96-98, 118-119,132-134,137-139,141-142,144, 154,205,207-210, 268-269, 292-305 Lymphocyte proliferation, 1, 3,4, 6, 7, 8,15,17, 34, 49, 61,144, 255, 260, 269-270, 322 Lymphocyte subsets, 2, 3,12,18, 270 Lymphokine-activated killer cell activity, 11,13 Macrophage, 33,48,255 Major depression, 14, 32, 33, 38, 59, 63-68,108-110, 142,201,271 MelanchoHa, 9,201, 291,327 Meta-analysis, 1,2, 4, 5,9,25 Microglia, 48, 53, 54, 255 Monocyte, 2, 25, 34,50, 51,60 Motivation, 85-87, 201 Natural killer (NK) cell, 1, 3, 6, 7, 8,10,11,12,15, 16,17, 20,49, 50, 260, 270, 322 Neuropsychology, 75-79 Neurotoxicity, 77 Neutrophil, 25 Nitric oxide, 92, 94, 239-240, 304 Noradrenahne, 34,117-125,137-139,141, 203, 205, 208-209, 214-215, 218-220, 223-226, 304-305 Novelty, 181, 205, 211, 221 Nucleus accumbens, 214, 218, 220, 223-224 Nucleus tractus solitarius, 94-95,118,140, 241,243, 246 Olfactory bulbectomy, 26, 36,65 Pain, 293,299 Panic disorder, 15 Paraventricular nucleus, 120,123,139,184-186, 208, 223-224 Physical exercise, 154,157-158,159-160

Preoptic area of the hypothalamus, anterior, 137-141 Prostaglandins, 25,26, 52, 60,61, 87, 94,134-135, 260 Psychomotor retardation, 9, 32,49, 76,292-293, 298 Psychosocial stressors, 5, 7, 20, 33,35,38,67,158, 160, 259 Restraint, 33,100,118,154,157,159,160, 237, 245 Selective serotonin reuptake inhibitors (SSRI), 36, 37,38,51, 79, 80,113, 273-277,300-305, 323-324 Sensitization, 177-192, 205, 223-226 Serotonin, 25,30, 31,34, 37, 61, 66-67,117-125, 132-137,141-142,145, 203-205, 214-215, 218-220, 223-225, 259,304-305, 325 Sexual behaviour, 296 Sickness behaviour, 32, 60, 83-102,, 110,130,141142,145-148,200-201,221,235-236, 283, 291,301-302,317 Sleep disorders, 6,9,10-14, 32,49,292,322 Smoking, 5, 6, 16-17 Social exploration, 32, 85,96,293 Spatial learning, 211-212,293 Stress: see Electric shock; Psychosocial stressors; Restraint Stressor controllability, 237,245 Substance P, 154-156, 159 Sympathetic nervous system, 155-156,181, 253 Thl/Th2 balance, 52, 257 Thyroid gland, 30, 31, 32,38 Treatment-resistant depression, 62-68 Tricyclic antidepressants, 36, 37, 38,51, 52, 62, 79, 80,113, 269, 273-277, 300-305, 323-324 Tryptophan,31,32,118-122, 325-326 Tumor necrosis factor-alpha (TNF-a), 30, 34,48, 59, 84, 91,122,135,137,144, 222-226, 304, 318,327 Vagus nerve, 89-91 Vasopressin, 98-99, 184-189, 205, 299-300 Viral infection, 18,20, 35, 48,67-68,109,118, 254255, 268,285, 289, 291, 299, 322 White blood cells, 5,16, 25,255 Whole blood assay of cytokines, 37, 50, 323

Cytokines: Stress and Immunity

Read more

Cytokines Stress and Immunity

Read more

Stress, the Brain and Depression

Read more

Stress, the Brain and Depression

Read more

Stress, the Brain and Depression

Read more

Cytokines and Autoimmune Diseases

Read more

Cytokines and Disease

Read more

Stress, Anxiety, Depression: A Practical Workbook

Read more

Mechanical Stretch and Cytokines

Read more

Cytokines and the Brain

Read more

Proinflammatory Cytokines

Read more

Cytokines and the Abdominal Surgeon

Read more

Cytokines (Growth Factors & Cytokines in Health & Disease)

Read more

Cytokines and Chemokines in Autoimmune Disease

Read more

Stress

Read more

Depression And Bipolar Disorder

Read more

Stress

Read more

Depression (Diseases and Disorders)

Read more

Depression and Physical Illness

Read more

Stress

Read more

Cytokines and Chemokines in Infectious Diseases Handbook

Read more

Cytokines and Chemokines in Infectious Diseases Handbook

Read more

Cytokines and the CNS, Second Edition

Read more

Depression and Heart Disease

Read more

Depression (Diseases and Disorders)

Read more

Anxiety, Depression, and Emotion

Read more

Structural and Stress Analysis

Read more

Stress and Cardiovascular Disease

Read more

Growth Factors & Cytokines in Health & Disease, Vol. 2A: Cytokines

Read more

Adversity, Stress, and Psychopathology

Read more

Recommend Documents

Cytokines: Stress and Immunity

CYTOKINES Stress and Immunity edited by Nicholas P. Plotnikoff Robert E. Faith Anthony J. Murgo Robert A. Good CRC P...

Cytokines Stress and Immunity

Stress, the Brain and Depression

This page intentionally left blank Stress, the Brain and Depression Can stress cause depression? Studies generally p...

Stress, the Brain and Depression

Stress, the Brain and Depression

This page intentionally left blank Stress, the Brain and Depression Can stress cause depression? Studies generally p...

Cytokines and Autoimmune Diseases

CYTOKINES AND AUTOIMMUNE DISEASES Edited by Vijay K. Kuchroo, DVM, PhD Nora Sarvetnick, PhD David A. Hafler, MD Lindsa...

Cytokines and Disease

Cytokines and Disease Marc Feldmann* and Fionula M. Brennan Cytokine and Cellular Immunology Division, Kennedy Institute...

Stress, Anxiety, Depression: A Practical Workbook

title: author: publisher: isbn10 | asin: print isbn13: ebook isbn13: language: subject publication date: lcc: ddc: su...

Mechanical Stretch and Cytokines

Mechanical Stretch and Cytokines Mechanosensitivity in Cells and Tissues Volume 5 Series Editors A. Kamkin Departmen...

Cytokines and the Brain

Cytokines and the Brain Neuroimmune Biology, Volume 6 Neuroimmune Biology Series Editors I. Berczi, A. Szentivanyi A...