Current Topics in Cellular Regulation, Volume 35 (Current Topics in Cellular Regulation)

CURRENT TOPICS IN

Cellular Regulation Volume 35

This Page Intentionally Left Blank

CURRENT TOPICS IN

Cellular Regulation edited by Earl R. Stadtman

National Institutes of Health Bethesda, Maryland

P. Boon Chock

National Institutes of Health Bethesda, Maryland

Volume 35

ACADEMIC PRESS San Diego New York Boston London Sydney Tokyo Toronto

This book is printed on acid-free paper. @ Copyright © 1997 by ACADEMIC PRESS All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the Publisher. The appearance of the code at the bottom of the first page of a chapter in this book indicates the Publisher's consent that copies of the chapter may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, Massachusetts 01923) for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-1997 chapters are as shown on the title pages. If no fee code appears on the title page, the copy fee is the same as for current chapters. 0070-2137/97 $25.00 Academic Press 15 East 26th Street, 15th floor. New York, New York 10010 http://www.apnet.com Academic Press Limited 24-28 Oval Road, London NWl 7DX, UK http://www.hbuk.co.uk/ap/ International Standard Serial Number: 0070-2137 International Standard Book Number: 0-12-152835-9 PRINTED IN THE UNITED STATES OF AMERICA 97 98 99 00 01 02 EB 9 8 7 6 5 4 3 2

Contents

Regulation of Iron Metabolism in Eukaryotes TRACEY ROUAULT AND RICHARD KLAUSNER I. Need for Regulation of Iron Metabolism II. Translational Regulation of the Iron Sequestration Protein Ferritin III. Expression of TfR Regulated by Intracellular Iron Levels IV. IREs as Binding Sites for IRP V. Steric Hindrance Model to Account for Function of IRPl in Regulation VI. Iron-Sulfur Cluster of IRPl as Key to Sensing of the Iron Levels ... VII. Residues in Active Site Cleft Essential in IRE Binding VIIL Role of IRPl in Vivo IX. Assembly and Disassembly of Iron-Sulfur Cluster of IRPl X. IRP2 as Ubiquitous IRE Binding Protein Regulated by Degradation When Iron Levels Are High XL Cysteines Indispensable to Function of Degradation Domain XII. IRPs Expressed in Cells: Redundant Function or Undefined Unique Roles XIII. Differences in IRE Binding Sites of IRPl and IRP2 Permit IRPs to Bind Unique Targets XTV. Role for IRE Mutations in Human Disease XV. Speculations on Evolution of IRPs XVI. Other G6nes That May Be Regulated by IRE-IRP Regulatory System XVII. Summary and Conclusions References

1 2 3 3 4 5 6 8 8 9 10 11 12 13 13 14 16 16

Structure, Mechanism, and Specificity of Protein-Tyrosine Phosphatases ZHONG-YIN ZHANG I. Introduction II. Structural Characterization

21 25

VI

CONTENTS

III. Catalytic Mechanism IV. Substrate Specificity V. Conclusion and Perspective References

30 48 62 63

Regulation of Fas-Mediated Apoptosis ROBERTA A. GOTTLIEB AND BERNARD M . BABIOR I. II. III. IV.

Introduction Regulatory Factors Regulation of Fas-Mediated Apoptosis Conclusion References

69 72 86 92 92

Aging and Regulation of Apoptosis HuBER R. W A R N E R

I. II. III. IV. V. VI. VII.

Introduction Genes Involved in Apoptosis Cell Senescence and Apoptosis Immunological Aging and Apoptosis Caloric Restriction and Apoptosis Neurodegenerative Disease and Apoptosis Summary References

107 108 110 112 114 115 117 119

Gene Regulation by Reactive Oxygen Species FiLiBERTO CiMiNO, F R A N C A E S P O S I T O , R O S A R I O A M M E N D O L A , A N D

TOMMASO RUSSO I. Introduction II. Sensitivity of Transcription Factors to Intracellular Redox Changes III. Effects of Redox Changes on Regulation of Gene Expression IV. ROS and Extracellular Signal Transduction V. Concluding Remarks References

123 124 134 139 143 144

Regulation of N F - K B and Disease Control: Identification of a Novel Serine Kinase and Thioredoxin as Effectors for Signal Transduction Pathway for N F - K B Activation TAKASHI OKAMOTO, SHINSAKU SAKURADA, JIAN-PING YANG, AND JOCELYN P . M E R I N

I. Introduction II. Transcription Factor NF-KB and Its Activation Pathways

149 150

CONTENTS III. Involvement of N F - K B in Disease Processes IV. Screening Strategy for Anti-NF-zcB Compounds V. Summary References

Vll 154 157 157 158

Regulation of Bacterial Responses to Oxidative Stress JUDAH L . ROSNER AND GiSELA S T O R Z I. Introduction II. Regulators of Escherichia coli Responses to Oxidative Stress III. Oxidative Stress Responses in Salmonella, Haemophilus, Mycobacterium, and Bacillus IV. Concluding Remarks References

163 164 169 174 175

Mechanism and Regulation of Bone Resorption by Osteoclasts NOBUHIKO K A T U N U M A

I. Historical Background and Future Prospects II. Suppression of Bone Resorption by Cathepsin L Family Inhibitors III. Functional Differentiation of Osteoclasts and Macrophages with Respect to Cathepsins L and B IV. Mechanism and Regulation of Procathepsin L Secretion from Osteoclasts V. Cathepsin L Secreted from Osteoclasts as Precursor Form and Processed by Cysteine Proteinase(s) in Bone Lacunae VI. Inhibitory Mechanisms of H*-ATPase, Carbonic Anhydrase II, and Monensin on Pit-Forming Assay VII. Suppression of Bone Resorption by Cathepsin L Family Inhibitors in Vivo VIII. Possible Strategies for New Drug Design to Protect Bone Resorption References INDEX

179 180 183 184 187 188 189 190 191 193

This Page Intentionally Left Blank

CURRENT TOPICS IN CELLULAR REGULATION, VOLUME 35

Regulation of Iron Metabolism in Eukaiyotes TRACEY ROUAULT RICHARD KLAUSNER

Cell Biology and Metabolism Branch National Institutes of Child Health and Human Disease National Institutes of Health Bethesda, Maryland 20892

I. Need for Regulation of Iron Metabolism Iron is indispensable to the function of many proteins in prokaryotes and eukaryotes, including enzymes t h a t are involved in such fundamental processes as respiration, photosynthesis, and nitrogen fixation. Features t h a t account for the frequent use of iron in prosthetic groups include the ability of iron to facilitate oxidation-reduction reactions and the flexible coordination chemistry of iron. Because iron is required for vital functions, systems t h a t are devoted to the efficient solubilization and uptake of iron have evolved in cells. In addition to solving the problem of ensuring t h a t supplies of iron are sufficient, cells must also provide defenses against iron-related toxicity. Free iron can participate in reactions that produce toxic by-products, the most well known of which is the Fenton reaction in which reduced metal species, such as ferrous iron, react with hydrogen peroxide to form superoxide anion and hydroxyl radical. Hydrogen peroxide is formed in cells as a spontaneous and potentially toxic by-product of respiration, and its rapid breakdown into water and oxygen is enzymatically facilitated by cellular catalases to protect against formation of toxic oxygen species. The highly reactive hydroxyl radical formed in the Fenton reaction can abstract electrons from target molecules in the cell, including DNA, proteins, and lipids, thereby producing unstable free radicals and irreversible damage in target molecules (1,2). The potential toxicity of these reactions may explain why proteins t h a t are devoted to intracellular iron sequestration are maintained in many cell types. Through regulation of the processes of both iron uptake and sequestration, eukaryotic cells can maintain tight control over amounts of free iron and avoid iron-related toxicity. In this chapter, we will discuss regulation of iron metabolism in eukaryotic cells, focusing on regulation of the iron sequestration pro-

Z.

TRACEY ROUAULT AND RICHARD KLAUSNER

tein, ferritin, and regulation of the iron uptake receptor, the transferrin receptor (TfR). The expression levels of both ferritin and the TfR are regulated by two proteins known as iron regulatory proteins 1 and 2 (IRPl and IRP2), formerly known as iron-responsive element (IRE) binding proteins (IRE-BPs) (3,4), as iron regulatory factor (IRF) (5), or as ferritin repressor protein (FRF) (6). Discussion of the mechanisms of regulation of these proteins will provide an overview of mechanisms of posttranscriptional gene regulation, as four different types of such regulation are involved at different steps in iron regulation. Ferritin is translationally regulated, whereas the TfR is regulated through regulation of mRNA degradation (7-9). IRPs sense iron levels directly, either through changes in the status of an associated prosthetic group of IRPl or through an iron-dependent change in the rate of degradation of IRP2. Coordination of uptake and sequestration to the needs of the cell is accomplished by the regulatory proteins IRPl and IRP2. II. TranslatJonal Regulation of the Iron Sequestration Protein Ferritin Ferritin is a multimeric protein composed of 24 subunits which coassemble to form a hollow sphere (10). Within the sphere, iron is precipitated as iron hydroxyphosphate. Subunits are of two types: the H chain, which is heavier (approximately 20 kDa) and is the predominant form in heart, and the L chain, which is lighter (approximately 18 kDa) and is the predominant form in liver cells. The functional H chain gene in humans is located on chromosome 11, whereas the functional L chain gene in humans is located on chromosome 19 (11). The 5'UTRs of ferritin transcripts contain cis-acting elements that mediate the translational regulation of ferritin. When cells are iron-replete, mRNAs encoding ferritin subunits are freely translated, whereas when cells are depleted of iron, the translation of these transcripts is repressed. Shorter sequences within the 5'UTRs of ferritins of H and L chains of numerous species responsible for mediating the translational regulation of these transcripts were first identified by deletional analyses of the 5'UTRs (12,13). The functional motifs within 5'UTRs form RNA stem-loop motifs known as iron-responsive elements, or IREs (14) (see Fig. 1). Phylogenetic comparisons, mutational analyses, and comparisons among functional IREs in different species and between H and L chains have permitted definition of the features of a consensus IRE. An IRE consists of a base-paired stem interrupted by an unpaired cytosine five base pairs removed from a 6-membered loop. The sequence of the loop is

REGULATION OF IRON METABOLISM IN EUKARYOTES

6

almost always CAGUG(X), where X can be any base but G (7). The fact that the IRE is sufficient to mediate translational regulation was established by studies of chimeric transcripts. When IREs were positioned in the 5'UTR of reporter genes, the transcripts became translationally regulated by iron (13,14), confirming that the IRE could transfer iron-mediated translational regulation to unrelated mRNAs expressed in mammalian cells.

III. Expression of TfR Regulated by Intracellular Iron Levels When cells are iron-replete, levels of TfR decrease, whereas in cells treated with iron chelators, synthesis and expression of TfRs are high. These changes in levels of protein expression refiect changes in the levels of the mRNA for TfR (7-9). In iron-depleted cells the mRNA for TfR is relatively stable, whereas in iron-replete cells the mRNA is rapidly degraded. Five IREs are found in the 3'UTR of the TfR transcript, and when the IRE-containing portion of the 3'UTR is ligated to reporter genes, the mRNA half-life of those genes becomes regulated by iron levels (5,15). Structural studies using oligonucleotides and RNase mapping have confirmed that IRE structures are present in the 3'UTR of the TfR (16). When IREs are mutated, the TfR transcript is rapidly degraded regardless of the iron status of the cell. There is a rapid turnover determinant in the 3'UTR between IREs which is required for mRNA degradation (17).

IV. IREs as Binding Sites for IRP Gel retardation assays were the initial means by which cytosolic regulatory proteins that bound to IREs were identified (18,19). Lysates were shown to contain an IRE binding protein that bound to IREs with high affinity (K^ = 10-50 pM) and specificity (20,21). The gel retardation assay was used to monitor purification of the IRE binding protein, which was accomplished using RNA affinity chromatography (22), traditional column chromatography (23,24), or a combination of the two approaches (25). A second assay used to monitor purification of the protein binding to IREs was based on repression of translation of ferritin mRNA in vitro, and the trans-acting protein was therefore referred to as the ferritin repressor protein (FRP) (6). In gel retardation assays performed on lysates of cells, IRE binding activity accurately reflects the iron status of the cells prior to lysis. When cells are irondepleted, IRE binding activity is high, whereas in cells that are ironreplete, binding activity is low (20).

4

TRACEY ROUAULT AND RICHARD KLAUSNER

Cloning of the IRE binding protein (3,23,24,26), now referred to as IRPl, proved to be unexpectedly informative as the sequence of IRPl turned out to be almost 30% identical to that of porcine mitochondrial aconitase (27). Mitochondrial aconitase had been the subject of extensive biochemical studies (28,29) and the crystal structure was known (30,31). The enzymatic active site residues of mitochondrial aconitase had been identified and studied by site-directed mutagenesis (32), and though overall sequence identity between mitochondrial aconitase and IRPl was only 28%, all of the active site residues between the two proteins were identical (27). Mossbauer spectroscopy and biochemical studies had established that mitochondrial aconitase contained an ironsulfur cluster (33,34) which was cubane (35). Of the four irons in the [4Fe-4S] cluster, three were shown to be directly bound to cysteines of the peptide, whereas the fourth had potentially open coordination sites and could bind directly to the enzymatic substrates citrate, isocitrate, and cis-aconitate (31). An analogous cluster was found in IRPl in ironreplete cells, and IRPl was revealed to be the source of aconitase activity in the cytosol of eukaryotes (36).

V. Steric Hindrance Model to Account for Function of IRP1 in Regulation When the IRP is bound to IREs in the 5'UTR of transcripts, the protein interferes with translational initiation and prevents active translation of the mRNA transcripts of ferritin and erythocyte aminolevulinic acid synthase (eALAS, 5-aminolevulinate synthase), the enzyme which is the rate-limiting step in the biosynthesis of heme (3740). When IRP is bound to IREs in the 3'UTR of the transferrin receptor, binding prevents cleavage of the TfR mRNA at an endonucleolytic cleavage site, most likely by inhibiting access of RNases to the site (17). In both cases, it is likely that binding of IRP creates a steric block. For IREs that are located in the 5'UTR, proximity to the 5' cap is required for function, and there is a loss of translational regulation when the IRE is located more than 67 nucleotides 3' of the cap site (41). The assembly of the 43S subunit of the ribosome at the 5' cap is inhibited (42). The mechanism is likely steric blockage of translational initiation as binding of other unrelated proteins to RNA motifs in the 5'UTR can also inhibit translation (43). Similarly, it appears that IRPs bound to IREs in the 3'UTR may prevent access of an endonuclease to a preferred cleavage site (17). Thus, binding by IRPs simultaneously prevents synthesis of ferritin and eALAS and indirectly increases synthesis of the TfR by increasing the amount of TfR transcript (see Fig. 1).

REGULATION OF IRON METABOLISM IN EUKARYOTES Ferritin mRNA

One IRE in 5'UTR AAAAAA

40S c ribosomal subunit

AAAAAA

60S Q 7

IRE is occupied by IRP inhibiting translation initiation.

ribosomal subunit

"^^ '® ""occupied allowing polysome formation and ferritin synthesis.

Five IREs in 3'UTR

TfR mRNA Protein Coding

ftffl^

AAAAAA

ft

One or more IREs are occupied by IRP protecting mRNA from degradation.

ft

/^ AAAAAA

IREs are unoccupied rendering mRNA susceptible to endonuclease which may be rate-determining step in mRNA degradation.

FIG. 1. A model for IRP-mediated regulation of expression of ferritin and TfR. Binding of IRP to the IRE at the 5' end of ferritin mRNA prevents initiation of translation in iron-depleted cells, whereas translation proceeds in the iron-replete cell. Binding of IRP(s) to the IREs in the 3'UTR of the TfR protects an endonucleolytic cleavage site, which is readily cleaved in iron-replete cells in which IRP does not bind IREs.

VI. Iron-Sulfur Cluster of IRP1 as Key to Sensing of Iron Levels Amounts of IRP 1 that are detectable immunologically do not change significantly when cells are made iron-replete or are iron-depleted (44,45). However, changes in intracellular iron status alter the status of the iron-sulfur cluster and the function of the protein. In cells that are depleted of iron, IRPl is found predominantly as an apoprotein.

b

TRACEY ROUAULT AND RICHARD KLAUSNER

devoid of an iron-sulfur cluster (46-48). The apoprotein form of IRPl binds with high affinity to IREs, whereas the [3Fe-4S] form of the cluster does not bind IREs (46). Studies with recombinant IRPl have shown that the protein functions either as an active cytosolic aconitase containing a [4Fe-4S] cluster in cells that are iron-replete or as a high affinity IRE binding protein in cells that are iron-depleted (47-49). The two functions of IRPl, as cytosolic aconitase (holoprotein) or as the IRE binding protein (apoprotein), are mutually exclusive properties (7,50). When the three cysteines that ligate the iron-sulfur cluster of IRPl are individually mutagenized to serine, the recombinant protein expressed in cells constitutively binds IREs in gel retardation assays of lysates, regardless of the iron status of the cell prior to lysis (51-53). The results of mutagenesis of the cysteines, along with the correlation of aconitase activity with the iron-replete status, supports the view that the iron sulfur cluster is itself the determinant of function. When cells are iron-replete, the holoprotein, which does not bind IREs, is present, whereas when cells are depleted of iron or are incapable of assembling an iron-sulfur cluster because of mutations of cluster ligands, IRPl is found as apoprotein which binds with high affinity to IREs (46,54). VII. Residues in Active Site Cleft Essential in IRE Binding In the crystal structure of mitochondrial aconitase, an active site cleft separates the fourth domain of the protein from domains 1-3. The fourth domain is connected to the first three domains by a hinge-linker peptide which is not highly structured and which would be expected to exhibit considerable fiexibility. Residues that are analogous to the known active site residues of mitochondrial aconitase have been shown to be important in RNA binding. Ultraviolet (UV) cross-linking studies of IRE and protein have identified both short peptide fragments that contain active site residues (55) and longer peptide fragments from the presumed active site cleft of the protein (56,57). Arginine residues from both sides of the putative active site cleft have been shown to be important in high affinity IRE binding (53). These arginine residues align with known enzymatic active site residues of mitochondrial aconitase and are thought to contribute to the enzymatic active site of IRPl. When these arginines are mutagenized to glutamine, aconitase activity of IRPl is eliminated (53). Although the residues identified as being important in IRE binding are believed to be located in the putative active site cleft of IRPl, the dimensions of the cleft, as revealed in the crystal structure of

REGULATION OF IRON METABOLISM IN EUKARYOTES

7

mitochondrial aconitase, are such that an RNA stem-loop would be physically excluded from this area (see Fig. 2) unless the dimensions of the active site cleft were changed. The fact that IRPl functions as a cytosolic aconitase that is just as efficacious as mitochondrial aconitase supports the view that the geometry of the active site is very similar in the two proteins (36). However, there is no information about the geometry of the IRE binding site in the apoprotein, as crystals of the IRE bound to protein have not been obtained. It is likely that the flexible hinge-linker permits movement of the fourth domain of IRPl with respect to the remainder of the protein and that potential binding surfaces that were previously not accessible become contacts for IREs (see Fig. 2). The fact that the RNA binding site overlaps with the enzymatic active site explains why the two activities are mutually

flexible hingelinker structure of mitochondrial aconitase FIG. 2. A model of mitochondrial aconitase showing the flexible hinge-linker which connects domains 1-3 to domain 4. Residues analogous to those that have been implicated in IRE binding are illustrated in black space-fill form. Residues 100 and 101 of IRP were identified as important to IRE binding in UV cross-linking, and the residues that align with those residues are shown in the backbone alignment of IRPl. Residues R780, R541, and R536 of IRPl were identified as important in IRE binding by IRPl and IRP2 (see text), and the corresponding residues of mitochondrial aconitase are labeled in the ribbon diagram of mitochondrial aconitase. The IRE binding affinity of R780Q to IRPl was decreased 10,000-fold; R541Q was decreased 1000-fold; R536Q was decreased 100-fold. DlOO HlOl was contained within cross-linked peptide.

8

TRACEY ROUAULT AND RICHARD KLAUSNER

exclusive properties of the protein. In mitochondrial aconitase, the substrate binds to residues on either side of the active site cleft that separates domain 4 from the first three domains (31), a feature which would be expected to contribute to the apposition of domain 4 to domains 1-3 in the holoprotein. Although the active site residues identified as important in IRE binding by UV cross-linking and mutagenesis are present in mitochondrial aconitase, mitochondrial aconitase does not bind IREs (58). The aconitase of Escherichia coli is highly related to IRPl, with a 60% sequence identity between the proteins and identity between active site residues (59), and yet there is also no indication that the E. coli aconitase can bind IREs. Clearly, although some important observations have been made on residues that contribute to the IRE binding site, much remains to be learned about the IRE binding site, as it is not clear why neither mitochondrial aconitase nor E. coli aconitase can bind IREs. There is one circumstance in which it appears that holoprotein can bind IREs: when the holoprotein is treated with high concentrations of reducing agents, including 2% 2-mercaptoethanol (v/v), high affinity binding of IREs is induced (60). The cluster does not dissociate completely from the protein, as high concentrations of 2-mercaptoethanol do not convert the protein to the IRE binding form permanently, and the protein ceases to bind IREs as soon as the 2-mercaptoethanol is removed by chromatography. In this situation, it is possible that the cluster is sufficiently altered by the treatment with the 2-mercaptoethanol that local conformation is changed, permitting access to residues that would normally be inaccessible for IRE binding. VIII. Role of IRP1 In Vivo When the form of IRPl which contains cysteine-to-serine mutations is expressed in cells, regulation of ferritin and TfR is impaired. Cells are unable to derepress ferritin translation when iron-replete and are similarly unable to decrease biosynthesis of the TfR in iron-replete cells (61). These studies of the effect of the mutagenized protein on intracellular iron metabolism support the hypothesis that the ironsulfur cluster of IRPl permits sensing of intracellular iron levels and determines regulatory function in vivo. IX. Assembly and Disassembly of Iron-Sulfur Cluster of IRP1 In the regulation of IRPl, it is important to understand the mechanisms of assembly and disassembly of iron-sulfur clusters, as it is

REGULATION OF IRON METABOLISM IN EUKARYOTES

9

apparent that the iron-sulfur cluster is the key determinant of whether the protein will bind to IREs. The iron-sulfur clusters of many proteins are known to be degraded by oxidants (61a), and the key to stability of these prosthetic groups appears to be determined by whether the cluster is positioned in the protein such that it is accessible to oxidants and solvent. In the case of IRPl, the cluster is positioned in a hydrophilic cleft in the protein where it is readily accessible to oxidants and solvent. The iron-sulfur cluster of IRPl is destabilized by exposure to nitric oxide (62-64) or to hydrogen peroxide (65,66). In addition, the ironsulfur cluster of E. coli aconitase, which is 60% identical in primary sequence to IRPl, is known to be rapidly destabilized by superoxide anion (67), and other hydrolases ofE. coli that contain [4Fe-4S] clusters are similarly sensitive to superoxide (68). Thus, it is quite possible that the iron-sulfur cluster of IRPl is relatively easily degraded under conditions of aerobic growth, whereas the protein itself may have a relatively long half-life. In a setting in which the iron-sulfur cluster is turned over much faster than the protein itself, it is possible to imagine how the iron-sulfur cluster could allow the protein to serve as an iron sensor: the cluster would be reassembled when sufficient iron and sulfur were present with the aid of cluster assembly enzymes, but if iron levels were low, the cluster would not be reassembled and apoprotein, the form which represses translation of ferritin and increases biosynthesis of the TfR, would accumulate (61a). X. IRP2 as Ubiquitous IRE Binding Protein Regulated by Degradation When Iron Levels Are High At the time of cloning of IRPl, the cDNA encoding a second, highly related protein, IRP2, was cloned (3). The protein was 58% identical in primary amino acid sequence, and recombinant protein expressed from the sequence was found to bind consensus IREs with high affinity similar to IRPl (4). However, there were several noteworthy differences between IRPl and IRP2. Unlike IRPl, the predicted active site residues of IRP2 were not identical to those of mitochondrial aconitase; in particular, the serine proposed to function as the catalytic base in the aconitase reaction was noted to be a glutamine at the analogous position in IRP2. In mitochondrial aconitase, mutation of the active site residue results in a loss of aconitase activity (32), and there is no detectable aconitase activity associated with IRP2 (69,70). In addition to the lack of aconitase activity, IRP2 differs from IRPl in the mode of regulation by iron. Unlike IRPl, which remains present in the iron-replete cell, although the IRE binding activity is lost, IRP2 is physically absent in iron-replete cells (69,70). Experiments have

10

TRACEY ROUAULT AND RICHARD KLAUSNER

shown t h a t the rates of synthesis of IRP2 are equal in iron-replete and iron-depleted cells but t h a t the rates of degradation vary markedly (70,71). In the iron-replete cell, IRP2 is degraded within minutes to hours, whereas in the iron-depleted cell, IRP2 is relatively stable, with a half-life of 9-12 h r (45,72,73). One notable difference between the two proteins is t h a t IRP2 contains an insertion of 73 amino acids relative to I R P l , and the amino acid insertion is encoded by a separate genomic exon. When the amino acids corresponding to the unique exon are excised, the remaining protein is still able to bind IREs with high affinity and is therefore not misfolded as a result of the deletion. However, rapid degradation in the presence of iron is no longer seen. When the IRP2-specific exon is ligated to the analogous position in an expression construct of I R P l , the chimeric I R P l acquires the ability to be rapidly degraded in the presence of iron (72). Thus, the IRP2-specific exon is an iron-dependent degradation domain, capable of conferring the rapid degradation phenotype on a recipient protein t h a t is otherwise stable in iron-replete cells.

XI. Cysteines Indispensable to Function of Degradation Domain The degradation domain of IRP2 has an unusual distribution of amino acids, containing 5 cysteines, 10 prolines, and 8 glycines. When 3 of the cysteines are mutagenized simultaneously to serines, the rapid degradation in iron-replete cells is no longer seen (72). It is quite likely t h a t the cysteines directly bind iron, either singly or as part of an ironsulfur cluster. Although there is still no direct proof t h a t iron is bound by the degradation domain, it is attractive to hypothesize t h a t direct binding of iron to the degradation domain leads to a change in the protein t h a t results in targeting for degradation. One possibility is t h a t a conformational change driven by the binding of iron results in surface exposure of an epitope t h a t is recognized as a degradation target. Another is t h a t ferrous iron bound but not fully coordinated by the protein can react with hydrogen peroxide to release hydroxyl radicals t h a t cause damage in the vicinity of the iron-binding site. A similar process, known as metal-catalyzed oxidation, has been described for the degradation of the E, coli protein glutamine synthetase (2) ( g l u t a m a t e ammonia ligase), and is believed to account for turnover of other proteins. IRP2 is degraded by the proteasome, as evidenced by studies which show t h a t inhibitors of proteases, including MG132 and lactacystin, inhibit the degradation of IRP2 (71,72). It is not known whether IRP2

REGULATION OF IRON METABOLISM IN EUKARYOTES

11

is ubiquitinated, and little is known about the process by which IRP2 is targeted to the proteasome.

XII. IRPs Expressed in Cells: Redundant Function or Undefined Unique Roles In cells that have been examined to date, both IRPl and IRP2 are expressed, although the ratios of the two proteins vary among cell types. In most cells with the exception of brain tissue, IRPl is far more abundant than IRP2, as judged by Northern analyses (70) and by gel retardation assays (74) in which the complexes arising from binding of each protein are separable. Because both proteins bind to consensus IREs with equally high affinity, the advantages of expressing both IRPs in individual cells are not obvious. When transcripts containing IREs are translated in in vitro translation systems, IRPl and IRP2 are just as efficacious as translational repressors (75). One obvious difference between the two proteins is that IRPl functions as a cytosolic aconitase in iron-replete cells, whereas IRP2 does not have aconitase activity and is rapidly degraded in iron-replete cells. It is not clear what role cytosolic aconitase has in cytosolic metabolism. Citrate and isocitrate are both present in the cytosol, and citrate gives rise to precursors of fatty acid synthesis, whereas isocitrate is a precursor of glutamate and glutamine synthesis. When cytosolic aconitase is absent in cells, the interconversion of citrate and isocitrate can be performed in mitochondria by mitochondria aconitase, and precursors and products can cross the mitochondrial membrane. There are several approaches that can aid in the determination of the role of each IRP in the cell. They include expression of mutant forms of the protein, as in the case of IRPl, in which a cysteinyl cluster ligand has been mutagenized, or "knockout" of the function of the two IRPs, either individually or simultaneously. When the constitutive IRE binding form of IRPl is expressed in cells, derepression of ferritin synthesis does not take place in iron-replete cells and regulation of the TfR is similarly impaired (61), indicating that IRPl can mediate regulation of these transcripts in vivo. When IRPl function is eliminated in mice through homologous recombination, there is no apparent impairment in regulation of iron metabolism in the mice, and there is no apparent phenotype associated with the loss of cytosolic aconitase activity (K. Iwai, R. D. Klausner, and T. A. Renault, unpublished observations). Elimination of IRP2 activity by homologous recombination should help to establish whether one IRP has a predominant role in iron regulation or whether each IRP serves as a backup for the other.

12

TRACEY ROUAULT AND RICHARD KLAUSNER

Regulation of protein function is commonly accomplished through degradation of the unwanted activity, as exemplified by regulation of IRP2. Degradation ensures that the unwanted activity cannot be inappropriately activated under certain circumstances, leading to misregulation. In the case of IRPl, it appears that binding activity is activated not only by iron deprivation but also by oxidative stress (see Section IX). In fact, the sensitivity to oxidants may be the means by which the cluster is normally turned over, opening the possibility for sensing of iron levels by the apoprotein. It is quite possible that an important role of IRPl will turn out to be its role as a sensor of oxidative stress, in addition to its role in regulation of iron metabolism.

XIII. Differences in IRE Binding Sites of IRP1 and IRP2 Permit IRPs to Bind Unique Targets Although both IRPl and IRP2 bind IREs with high affinity, each IRP has, in addition, a ligand identified which is bound by one of the two IRPs but not the other. These ligands were identified by performing SELEX procedures on IRPl (76,77) and IRP2 (77). SELEX, which stands for selective evolution of ligands by exponential enrichment (78), is a procedure in which the protein selects from a group of randomized RNA ligands those ligands which are bound with the highest affinity. The results of the studies performed with IRPl and IRP2 have shown that each IRP has, in addition to the ability to bind the consensus IRE, specificity for a ligand that is not bound by the other IRP. In IRPl, the alternative ligand for IRPl has the sequence UAGUA(X) in the loop (76,77). Since both the consensus IRE and the U1A5 alternative ligand have the potential to form base pairs between the first and fifth nucleotides of the loop, it is possible that base pairing in the loop is present in the bound IRE. Such issues could be resolved by cocrystal structures. IRP2 binds a ligand with the sequence GGGAG(X) where the sixth nucleotide can be any one of the four nucleotides. This sequence is bound with lower affinity than the consensus IRE and does not appear to be capable of mediating translational regulation in an in vitro translation assay (77). The GGGAG(X) alternative ligand is bound poorly by IRPl, an observation which emphasizes that the IRE binding sites are different, although they are similar in their ability to bind the consensus IRE. Although endogenous transcripts that contain these alternative ligands have not been found, it is possible that there is a set of endogenous transcripts that is bound by one IRP but not the other. This feature could be valuable if, for instance, IRPl was serving as a sensor of oxidative stress in cells. Under these circumstances.

REGULATION OF IRON METABOLISM IN EUKARYOTES

13

expression of genes that would exacerbate the toxicity of oxidative stress could be translationally repressed by IRPl, and gene products involved in mitigation of oxidative stress could contain alternative ligands in the 3' end which might lead to increased half-life of the mRNA and increased expression. Differences in the RNA binding sites of the two IRPs revealed by differences in ligand specificities are further underscored by sitedirected mutagenesis studies. Whereas mutation of a single arginine of the enzymatic active site of IRPl results in substantial loss of IRE binding affinity, mutation of the analogous arginine in IRP2 has no measurable impact on IRE binding affinity. However, when this mutation is combined with mutations in other active site arginine residues, binding affinity is profoundly decreased, revealing that although similar residues contribute to binding, their relative importance between the two proteins differs (77). XIV. Role for IRE Mutations in Human Disease A new genetic disorder has been described in humans: the hereditary hyperferritinemia-cataract syndrome. Affected individuals have a combination of elevated serum ferritin and congenital bilateral nuclear cataracts (79,80). Mutations have been identified in the IRE of the ferritin L chain; in one case, the mutation changes the sequence of the IRE loop from CAGUGU to CGGUGU, whereas in the other case, the mutation is also in the loop but the sequence is CACUCU. In both cases, the mutations to the loop result in loss of IRP binding, as would be expected on the basis of previous mutagenesis studies (76,81), which have shown that the sequence CAGUG is required for high affinity binding. In a SELEX study of both IRPl and IRP2, the two nucleotides of the consensus sequence that were never substituted in high affinity binding forms were the unpaired bulge C of the stem and the first G of the loop (77), which is mutated in one of the affected families (79). The pathophysiology of hyperferritinemia is clear, but the reason for an association with formation of cataracts is unclear. XV. Speculations on Evolution of IRPs Iron-sulfur clusters may have been one of the earliest types of protein prosthetic groups in widespread use. When early life forms began to evolve, the atmosphere of the Earth was anaerobic, and iron was probably highly abundant and soluble. Geothermal vents may have led to high concentrations of hydrogen sulfide in certain locales. Iron-sulfur

14

TRACEY ROUAULT AND RICHARD KLAUSNER

clusters of various stoichiometries, including cubane [4Fe-4S] clusters, have been shown to form spontaneously (82) when a reducing atmosphere and high concentrations of iron and sulfide are found (83). Favorable synthetic conditions may have been present in certain locales early in the history of the Earth, and this could explain why iron-sulfur clusters are found as prosthetic groups in many enzymes central to metabolism, including several enzymes of the Krebs cycle, a central metabolic pathway in which precursors for other metabolic pathways are synthesized. The ancestral protein to I R P l was probably the aconitase ofE. coli, which is 60% identical to I R P l (59), a homology which is striking and which far exceeds the homology between the mammalian mitochondrial and cytosolic aconitases. Much experimental evidence supports the view t h a t iron-sulfur clusters are quite labile in the presence of oxidants (61a). If the iron-sulfur cluster of aconitase is accessible to solvent and oxidants in solution, it could be degraded and subsequently regenerated by iron-sulfur cluster assembly processes in the cell. However, it seems apparent t h a t such regeneration of the cluster could take place only in the presence of sufficient iron and sulfur, and t h a t apoprotein would accumulate in iron-depleted cells. On the basis of the relationship of I R P l to mitochondrial aconitase, we have proposed t h a t the loss of the iron-sulfur cluster permits motion in the region of a flexible hinge-linker of the fourth domain with respect to domains 1-3 and thereby permits access of large molecules to a portion of the protein which is normally accessible only to solvent. Under these conditions, binding of endogenous mRNAs to sequence spaces specific to the apoprotein could occur. Because the binding site would be available only in iron-depleted cells, it would be reasonable that such a site could be incorporated into regulatory pathways.

XVI. Other Genes That May Be Regulated by IRE-IRP Regulatory System Other pathways t h a t could benefit from regulation by an IRE binding protein could include pathways in which proteins require incorporation of iron for function. In the interests of economy, synthesis of such proteins could be repressed during times of relative iron depletion, since the product of synthesis would not fulfill its physiological role without iron. An example of such economy exists with the example of the eALAS protein, the enzyme proposed to be the rate-limiting step in the biosynthesis of heme. Heme is the product of incorporation of iron into protoporphyrin IX, and it is not only nonfunctional in the

REGULATION OF IRON METABOLISM IN EUKARYOTES

15

absence of iron, but there is significant toxicity associated with the accumulation of heme precursors. An IRE present in the 5'UTR of eALAS prevents active biosynthesis of ALAS in erythrocyte cells t h a t are iron-depleted, and the resulting economies to the cell are clear (39,40). It is interesting to note t h a t IREs are contained in two enzymes of the Krebs cycle, succinate dehydrogenase subunit b (SDHb) ofDrosophila melanogaster (84) and porcine mitochondrial aconitase (38,85). The transcripts of four different mammalian species—human, porcine, bovine, and mouse—contain identically conserved IREs in the 5'UTR of mitochondrial aconitase, and this conservation among species in the 5'UTR, which is normally quite divergent among species, implies a potentially important role in regulation of this protein and in physiology. In fact, it appears t h a t this IRE is sufficient to mediate translational regulation in in vitro translation systems; furthermore, there are demonstrable differences in total levels of mitochondrial aconitase in tissues of animals on high versus low iron diets (86). In the case of the IRE found in succinate dehydrogenase (SDHb) of Drosophila, the IRE is sufficient to mediate translational regulation of a chimeric gene expressed in mammalian cells and to shift the distribution of SDHb mRNA from polysomes (actively translated) to mRNPs (translationally repressed) when the Drosophila cells are treated with iron chelators (87). Interestingly, an IRE is not found in h u m a n SDHb mRNA (88), but the h u m a n mitochondrial aconitase gene contains a consensus IRE (86). It is interesting to note t h a t these are the two enzymes of the Krebs cycle t h a t require iron-sulfur prosthetic groups. Iron-sulfur clusters are also found in many of the proteins of the mitochondrial electron transport chain. If synthesis of an iron-requiring enzyme of the Krebs cycle is repressed by lack of iron, this could lead to impairment of the function of the entire cycle. Thus, it is possible t h a t the Krebs cycle itself is regulated by iron levels, and therefore t h a t fundamental features of respiration and energy storage are regulated by iron. If the Krebs cycle were made less efficient, oxygen levels would be expected to rise, since oxygen would no longer be as rapidly consumed. The increases in oxygen tension could lead to more rapid oxidative disassembly of the clusters of remaining cytosolic aconitase, which would then lead to increased availability of iron for other cellular processes. If this were true, then the fates of iron and oxygen would be intertwined under conditions of iron excess, in which the reaction of iron with oxygen promotes toxicity, and iron depletion, in which an increase in oxygen tension could lead to an increase in levels of chelatable iron, and thus produce a self-correction in the system.

16

TRACEY ROUAULT AND RICHARD KLAUSNER

XVII. Summary and Conclusions Iron metabolism is regulated in cells to ensure that iron supplies are adequate and nontoxic. The expression of iron metabolism is regulated primarily by posttranscriptional mechanisms. Ferritin, eALAS, SDHb of Drosophila, and mammalian mitochondrial aconitase are translationally regulated. The TfR is regulated at the level of mRNA stability. Iron regulatory proteins are regulated either by assembly or by disassembly of an iron-sulfur cluster (IRPl) or by rapid degradation in the presence of iron (IRP2). The list of targets for IRP-mediated regulation is growing longer, and a range of possibilities for versatile regulation exists, as each IRP can bind to unique targets that differ from the consensus IRE. The reactivity of iron with oxygen and the creation of toxic by-products may be the evolutionary stimulus that produced this system of tight posttranscriptional gene regulation. REFERENCES 1. Imlay, J. A., and Linn, S. (1988). Science 240, 1302-1319. 2. Stadtman, E. R. (1993). Annu. Rev. Biochem. 62, 797-821. 3. Rouault, T. A., Tang, C. K , Kaptain, S., Burgess, W. H., Haile, D. J., Samaniego, F., McBride, O. W., Harford, J. B., and Klausner, R. D. (1990). Proc. Natl. Acad. Sci. U.S.A. 87, 7958-7962. 4. Rouault, T. A., Haile, D. H., Downey, W. E., Philpott, C. C , Tang, C , Samaniego, F., Chin, J., Paul, L, Orloff, D., Harford, J. B., and Klausner, R. D. (1992). BioMetals 5, 131-140. 5. MuUner, E. W., Neupert, B., and Kuhn, L. C. (1989). Cell (Cambridge, Mass.) 58, 373-382. 6. Brown, P. H., Daniels-McQueen, S., Walden, W. E., Patino, M. M., Gaffield, L., Bielser, D., and Thach, R. E. (1989). J. Biol. Chem. 264, 13383-13386. 7. Klausner, R. D., Rouault, T. A., and Harford, J. B. (1993). Cell {Cambridge, Mass.) 72,19-28. 8. Melefors, O., and Hentze, M. W. (1993). Blood Rev. 7, 251-258. 9. Kuhn, L. C. (1994). Bailliere's Clin. Haematol. 7, 763-785. 10. Theil, E. C. (1987). Annu. Rev. Biochem. 56, 289-315. 11. Worwood, M., Brook, J. D., Cragg, S. J., Hellkuhl, B., Jones, B. M., Perara, P., Roberts, S. H., and Shaw, D. (1985). J. Hum. Genet. 69, 159-164. 12. Hentze, M. W., Rouault, T. A., Caughman, S. W., Dancis, A., Harford, J. B., and Klausner, R. D. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 6730-6734. 13. Aziz, N., and Munro, H. N. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 8478-8482. 14. Hentze, M. W., Caughman, S. W., Rouault, T. A., Barriocanal, J. G., Dancis, A., Harford, J. B., and Klausner, R. D. (1987). Science 238, 1570-1573. 15. Casey, J. L., Koeller, D. M., Ramin, V. C , Klausner, R. D., and Harford, J. B. (1989). EMBO J. 8, 3693-3699. 16. Horowitz, J. A., and Harford, J. B. (1992). New Biol. 4, 330-338. 17. Binder, R., Horowitz, J. A., BasiHon, J. P., Koeller, D. M., Klausner, R. D., and Harford, J. B. (1994). EMBO J. 13, 1969-1980. 18. Leibold, E. A., and Munro, H. N. (1988). Proc. Natl. Acad. Sci. U.S.A. 85,2171-2175.

REGULATION OF IRON METABOLISM IN EUKARYOTES

17

19. Rouault, T. A., Hentze, M. W., Caughman, S. W., Harford, J. B., and Klausner, R. D. (1988). Science 241, 1207-1210. 20. Haile, D. J., Hentze, M. W., Rouault, T. A., Harford, J. B., and Klausner, R. D. (1989). Mol. Cell Biol. 9, 5055-5061. 21. Barton, H. A., Eisenstein, R. S., Bomford, A., and Munro, H. N. (1990). J. Biol. Chem. 265, 7000-7008. 22. Rouault, T. A., Hentze, M. W., Haile, D. J., Harford, J. B., and Klausner, R. D. (1989). Proc. Natl. Acad. Sci. U.S.A. 86, 5768-5772. 23. Patino, M. M., and Walden, W. E. (1992). J. Biol. Chem. 267, 19011-19016. 24. Yu, ¥., Radisky, E., and Leibold, E. A. (1992). J. Biol. Chem. 267, 19005-19010. 25. Neupert, B., Thompson, N. A., Meyer, C , and Kuhn, L. C. (1990). Nucleic Acids Res. 18, 51-55. 26. Hirling, H., Emery-Goodman, A., Thompson, N., Neupert, B., Seiser, C , and Kuhn, L. C. (1992). Nucleic Acids Res. 20, 33-39. 27. Rouault, T. A., Stout, C. D., Kaptain, S., Harford, J. B., and Klausner, R. D. Cell (Cambridge, Mass.) 64, 881-883. 28. Beinert, H., and Kennedy, M. C. (1989). Eur. J. Biochem. 186, 5 - 1 5 . 29. Beinert, H. (1990). FASEB J. 4, 2483-2491. 30. Robbins, A. H., and Stout, C. D. (1989). Proteins 5, 289-312. 31. Lauble, H., Kennedy, M. C , Beinert, H., and Stout, C. D. (1992). Biochemistry 3 1 , 2735-2748. 32. Zheng, L., Kennedy, M. C , Beinert, H., and Zalkin, H. (1992). J. Biol. Chem. 267, 7895-7903. 33. Kent, T. A., Huynh, B. H., and Munck, E. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 6574-6576. 34. Kennedy, M. C , Emptage, M. H., Dreyer, J. L., and Beinert, H. (1983). J. Biol. Chem. 258, 11098-11105. 35. Werst, M. M., Kennedy, M. C , Houseman, A. L., Beinert, H., and Hoffman, B. M. (1990). Biochemistry 29, 10533-10540. 36. Kennedy, M. C , Mende-Mueller, L., Blondin, G. A., and Beinert, H. (1992). Proc. Natl. Acad. Sci. U.S.A. 89, 11730-11734. 37. Cox, T. C , Bawden, M. J., Martin, A., and May, B. K. (1991). EMBO J. 10,1891-1902. 38. Dandekar, T., Stripecke, R., Gray, N. K., Goossen, B., Constable, A., Johansson, H. E., and Hentze, M. W. (1991). EMBO J. 10, 1903-1909. 39. Melefors, O., Goossen, B., Johansson, H. E., Stripecke, R., Gray, N. K., and Hentze, M. W. (1993). J. Biol. Chem. 268, 5974-5978. 40. Bhasker, C. R., Burgiel, G., Neupert, B., Emery-Goodman, A., Kuhn, L. C , and May, B. K. (1993). J. Biol. Chem. 268, 12699-12705. 41. Goossen, B., and Hentze, M. W. (1992). Mol. Cell. Biol. 12, 1959-1966. 42. Gray, N. K , and Hentze, M. W. (1994). EMBO J. 13, 3882-3891. 43. Stripecke, R., Oliveira, C. C , McCarthy, J. E., and Hentze, M. W. (1994). Mol. Cell. Biol. 14, 5898-5909. 44. Tang, C. K , Chin, J., Harford, J. B., Klausner, R. D., and Rouault, T. A. (1992). J. Biol. Chem. 267, 24466-24470. 45. Henderson, B. R., and Kuhn, L. C. (1995). J. Biol. Chem. 270, 20509-20515. 46. Haile, D. J., Rouault, T. A., Harford, J. B., Kennedy, M. C , Blondin, G. A., Beinert, H., and Klausner, R. D. (1992). Proc. Natl. Acad. Sci. U.S.A. 89, 11735-11739. 47. Emery-Goodman, A., Hirling, H., Scarpellino, L., Henderson, B., and Kuhn, L. (1993). Nucleic Acids Res. 2 1 , 1457-1461.

18

TRACEY ROUAULT AND RICHARD KLAUSNER

48. Gray, N. K , Quick, S., Goossen, B., Constable, A., Hirling, H., Kuhn, L. C , and Hentze, M. W. (1993). Eur. J. Biochem. 218, 657-667. 49. Haile, D. J., Rouault, T. A., Tang, C. K., Chin, J., Harford, J. B., and Klausner, R. D. (1992). Proc. Natl. Acad. Sci. U.S.A. 89, 7536-7540. 50. Klausner, R. D., and Rouault, T. A. (1993). Mol. Biol. Cell 4, 1-5. 51. Philpott, C. C , Haile, D. J., Rouault, T. A., and Klausner, R. D. (1993). J. Biol. Chem. 268, 17655-17658. 52. Hirling, H., Henderson, B. R., and Kuhn, L. C. (1994). EMBO J. 13, 4 5 3 - 4 6 1 . 53. Philpott, C. C , Klausner, R. D., and Rouault, T. A., Proc. Natl. Acad. Sci. U.S.A. 91, 7321-7325. 54. Basilion, J. P., Kennedy, M. C , Beinert, H., Massinople, C. M. Klausner, R. D., and Rouault, T. A. (1994). Arch. Biochem. Biophys. 311, 517-522. 55. Basilion, J. P., Rouault, T. A., Massinople, C. M., Klausner, R. D., and Burgess, W. H. (1994). Proc. Natl. Acad. Sci. U.S.A. 9 1 , 574-578. 56. Swenson, G. R., and Walden, W. E. (1994). Nucleic Acids Res. 22, 2627-2633. 57. Neupert, B., Menotti, E., and Kuhn, L. C. (1995). Nucleic Acids Res. 23, 2579-2583. 58. Kaptain, S., Downey, W. E., Tang, C , Philpott, C. C , Haile, D. J., Orloff, D. G., Harford, J. B., Rouault, T. A., and Klausner, R. D. (1991). Proc. Natl. Acad. Sci. U.S.A. 88, 10109-10113. 59. Prodromou, C , Artymiuk, P. J., and Guest, J. R. (1992). Eur. J. Biochem. 204, 599-609. 60. Hentze, M. W., Rouault, T. A., Harford, J. B., and Klausner, R. D. (1989). Science 244, 357-359. 61. DeRusso, P. A., Philpott, C. C , Iwai, K., Mostowski, H. S., Klausner, R. D., and Rouault, T. A. (1995). J. Biol. Chem. 270, 15451-15454. 61a. Rouault, T. A., and Klausner, R. D. (1996). Trends Biochem. Sci. 21, 174-177. 62. Weiss, G., Goossen, B., Doppler, W., Fuchs, D., Pantopoulos, K., Werner-Felmayer, G., Wachter, H., and Hentze, M. W. (1993). EMBO J. 12, 3651-3657. 63. Drapier, J. C , Hirling, H., Wietzerbin, J., Kaldy, P., and Kuhn, L. C. (1993). EMBO J. 12, 3643-3649. 64. Pantopoulos, K., and Hentze, M. W. (1995). Proc. Natl. Acad. Sci. U.S.A. 92, 1 2 6 7 1271. 65. Pantopoulos, K., and Hentze, M. W. (1995). EMBO J. 14, 2917-2924. 66. Martins, E. A., Robalinho, R. L., and Meneghini, R. (1995). Arch. Biochem. Biophys. 316, 128-134. 67. Gardner, P. R., and Fridovich, I. (1991). J. Biol. Chem. 266, 1478-1483. 68. Flint, D. H., Tuminello, J. F., and Emptage, M. H. (1993). J. Biol. Chem. 2 6 8 , 2 2 3 6 9 22376. 69. Guo, B., Yu, Y., and Leibold, E. A. (1994). J. Biol. Chem. 269, 24252-24260. 70. Samaniego, F., Chin, J., Iwai, K , Rouault, T. A., and Klausner, R. D. (1994). J. Biol. Chem. 269, 30904-30910. 71. Guo, B., Phillips, J. D;, Yu, Y., and Leibold, E. A. (1995). J. Biol. Chem. 270, 2 1 6 4 5 21651. 72. Iwai, K , Klausner, R. D., and Rouault, T. A. (1995). EMBO J. 14, 5350-5357. 73. Pantopoulos, K., Gray, N. K , and Hentze, M. W. (1995). RNA 1, 155-163. 74. Henderson, B. R., Seiser, C , and Kuhn, L. C. (1993). J. Biol. Chem. 268,27327-27334. 75. Kim, H. Y., Klausner, R. D., and Rouault, T. A. (1995). J. Biol. Chem. 270,4983-4986. 76. Henderson, B. R., Menotti, E., Bonnard, C , and Kuhn, L. C. (1994). J. Biol. Chem. 269, 17481-17489.

REGULATION OF IRON METABOLISM IN EUKARYOTES

19

77. Butt, J., Kim, H., Basilion, J. P., Cohen, S., Iwai, K., Philpott, C. C, Altschul, S., Klausner, R. D., and Rouault, T. A. (1996). Proc. Natl Acad. Sci. U.S.A. 93, 43454349. 78. Tuerk, C, and Gold, L. (1990). Science 249, 505-510. 79. Girelli, D., Corrocher, R., Bisceglia, L., Olivieri, O., De Franceschi, L., Zelante, L., and Gasparini, P. (1995). Blood 86, 4050-4053. 80. Beaumont, C., Leneuve, P., Devaux, I., Scoazec, J. Y., Berthier, M., Loiseau, M. N., Grandchamp, B., and Bonneau, D. (1995). Nat. Genet. 11, 444-446. 81. Jaffrey, S. R., Haile, D. J., Klausner, R. D., and Harford, J. B. (1993). Nucleic Acids Res. 21, 4627-4631. 82. Berg, J. M., and Holm, R. H. (1981). In "Metal Ions in Biology" (T. G. Spiro, ed.), pp. 1-66. Wiley, New York. 83. Maden, B. E. H. (1995). Trends Biochem. Sci. 20, 337-341. 84. Au, H. C., and Scheffler, I. E. (1994). Gene 149, 261-265. 85. Zheng, L., Andrews, P. C., Hermodson, M. A., Dixon, J. E., and Zalkin, H. (1990). J. Biol. Chem. 265, 2814-2821. 86. Kim, H. Y., LaVaute, T., Iwai, K, Klausner, R. D., Rouault, T. A. (1996). J. Biol. Chem. 271, 24226-24230. 87. Kohler, S. A., Henderson, B. R., and Kuhn, L. C. (1995). J. Biol. Chem. 270, 3078130786. 88. Au, H. C., Ream-Robinson, D., Bellew, L. A., Broomfield, P. L., Saghbini, M., and Scheffler, I. E. (1995). Gene 159, 249-253.

This Page Intentionally Left Blank

CURRENT TOPICS IN CELLULAR REGULATION, VOLUME 35

Structure, Mechanism, and Specificity of Protein-Tyrosine Phosphatases ZHONG-YIN ZHANG

Department ofMolecular Pharmacology Albert Einstein College of Medicine Bronx, New York 10461

I. Introduction Protein tyrosine phosphorylation and dephosphorylation are fundamental cellular signaling mechanisms t h a t control cell growth and differentiation, mitogenesis, cell cycle, metabolism, gene transcription, cytoskeletal integrity, neuronal development, and the immune response (Yarden and Ullrich, 1988; Bishop, 1991; Cantley et al, 1991; Hunter, 1995). Hundreds of protein kinases and protein phosphatases and their substrates are integrated v^ithin an elaborate signal transducing network. The defective or inappropriate operation of this network is at the root of such widespread diseases as cancers, diabetes, and autoimmune disorders. Consequently, the characterization of the individual components and the delineation of the circuitry of this regulatory network have emerged as one of the most active fields in biomedical research. In vivo, the level of tyrosine phosphorylation in a given protein is regulated by the opposing actions of protein-tyrosine kinases (PTKs, EC 2.7.1.112) and protein-tyrosine phosphatases (FTPases, EC 3.1.3.48). PTKs are enzymes t h a t catalyze the transfer of the y-phosphate of ATP to the 4-hydroxyl of tyrosyl residues within specific protein/peptide substrates. PTPases are hydrolytic enzymes t h a t remove phosphate from the phosphorylated tyrosine residue(s). The sequence context surrounding the target tyrosine plays a key role in determining its recognition by PTKs and PTPases. Comprehension of the physiological roles of protein tyrosine phosphorylation, and its potential as a mechanism for reversible modulation of protein function and cell physiology, must necessarily encompass the characterization of PTPases in addition to the PTKs. The structure and function of PTKs have been extensively studied (Hunter, 1987, 1991). However, only recently has attention been focused on PTPases. Several reviews of PTPases have appeared 21

Copyright © 1997 by Academic Press. All rights of reproduction in any form reserved. 0070-2137/97 $25.00

22

ZHONG-YIN ZHANG

(Fischer et al, 1991; Charbonneau and Tonks, 1992; Brautigan, 1992; Walton and Dixon, 1993; Zhang and Dixon, 1994; Barford et al, 1995). This chapter will focus primarily on the structure, mechanism, and substrate specificity of the PTPases. A. Classification of PTPases

The PTPases constitute a growing family of enzymes that rival the PTKs in terms of structural diversity and complexity. Unlike protein kinases, where tyrosine-specific and serine/threonine-specific kinases share sequence identity, the PTPases show no sequence similarity with serine/threonine phosphatases or with the broad-specificity phosphatases such as acid or alkaline phosphatases. Collectively, PTPases can be broadly categorized into two groups: receptor-like and intracellular PTPases (Fig. 1). The receptor-like PTPases, exemplified by the leukocyte phosphatase, CD45, generally have an extracellular domain, a single transmembrane region, and one or two C3^oplasmic PTPase domains. The intracellular PTPases, exemplified by PTPIB and the Yersinia PTPase, contain a single catalytic domain and various amino- or carboxyl-terminal extensions, including SH2 domains that may have targeting or regulatory functions. Although many PTPases are proteins with more than 400 amino acids, their catalytic domains are usually contained within a span of 250 residues referred to as the PTPase domain. This domain is the only structural element that has amino acid sequence identity among all PTPases from bacteria to mammals (Zhang e^ al., 1994a). The unique feature that defines the whole PTPase family is the active site sequence (HA^)C(X)5R(S/T) called the PTPase signature motif in the catalytic domain (Zhang et al., 1994d). Interestingly, the PTPase signature motif can also be found in the structures of two additional groups of enzymes that can bring about phosphomonoester hydrolysis, namely, the VHl(VaccmJa open reading frame iifi)-like dual-specificity phosphatases (Guan et al., 1991) and the low molecular weight (18-kDa) PTPases (Zhang and Van Etten, 1990; Cirri et al, 1993) (Fig. 1). The low molecular weight (Mr) and the VHl-like phosphatases display little amino acid sequence identity with the classical PTPases. The only similarities among these three groups of phosphatases are the relative placement of the essential cysteine and arginine residues in the active sites that constitute the PTPase signature motif (HA^)C(X)5R(S/T). Unlike the PTPases, which show substrate specificity strictly restricted to phosphotyrosyl proteins (Sparks and Brautigan, 1985; Tonks et al., 1988; Guan and Dixon, 1990), the dual-specificity phosphatases and the low M^ phosphatases are unusual in that they can utilize protein substrates containing phosphotyrosine, phosphoserine, and phosphothreonine (Gautier et al.,

23

PROTEIN-TYEOSINE PHOSPHATASES

Protein—Tyrosine

I

Phosphatases

-Il"- "I

I^I

Yop51 PTP1B SHPTP CD45 PTPa

^ ^1^^^ ^1.^^^ ^ ^1^^^ ^ 1 ^ ^ ^ ^

Dual— Specificity

Phosphatases VHR VH1 MKP1 cdc25

Stpl

WvWWWWN

•

PTPase Signature (H/V)C(X)5R(S/T)

^

PTPase

W

Dual—Specificity Phosphatase Catalytic Domain

I I N- or

Catalytic

C-terminal

Motif: Domain

Extension

• M ^ ^

SH2

Domain

Transmembrane Low

Mr

Fibronectin

Region

Phosphatase Type

III

Domain

FIG. 1. Structural features of receptor-like and intracellular phosphatases containing the PTPase signature motif: C(x)5R(S/T).

1991; Guan et al, 1991; Ishibashi et al, 1992; Zhang et al, 1995a). However, despite the variation in the primary structures and the differences in the active site substrate specificities, the PTPases, the dualspecificity phosphatases, and the low Mr phosphatases utilize a common mechanism to effect catalysis (see Section III). B. Biological Functions of PTPases

The functional role of PTPases in cellular signaling processes is just beginning to be appreciated (Sun and Tonks, 1994; Hunter, 1995). Because deregulated PTKs, such as src, Ick, and neu, can function as

24

ZHONG-YIN ZHANG

dominant oncogenes, it has been assumed that at least some PTPases function as products of tumor suppressor genes. Indeed, mutations in SH-PTPl lead to severe immune dysfunction, giving rise to the motheaten phenotype in mice (Shultz et al., 1993). Thus, SH-PTPl may be an important negative regulator of cytokine signaling; its loss results in sustained tyrosine phosphorylation, with consequent enhanced proliferation (KlingmuUer et al., 1995). The Fas-associated phosphatase (FAP-1) inhibits Fas-generated signals that lead to apoptosis (Sato et al,, 1995). The gene for the receptor-like PTPy has been proposed to be a tumor suppressor gene located on chromosome 3p21, a segment frequently altered in renal and lung carcinomas (LaForgia et al., 1991; Wary et al., 1993). In addition, expression of PTPIB has been shown to block transformation mediated by neu (Brown-Shimer et al., 1992) and to partially revert transformation by src (Woodford-Thomas et al., 1992). On the other hand, there is mounting evidence that some PTPases potentiate rather than antagonize the action of PTKs. This behavior enhances mitogenic signaling, leading to cellular transformation. Thus the receptor PTPase CD45, through its capacity to dephosphorylate and activate the src family of PTKs, is essential for initiating downstream signaling processes in response to stimulation of T- and Bcell receptors (Pingel and Thomas, 1989). SH-PTP2 and its Drosophila homolog corkscrew are positive mediators of growth factor signaling (Perkins e^ aZ., 1992;Noguchie^aZ., 1994). The cell cycle regulator cdc25 dephosphorylates Tyr-15 of cdc2, thereby activating the cdc2-cyclin B complex, which, in turn, promotes mitosis (Gould and Nurse, 1989; Millar and Russell, 1992). Strikingly, ectopic expression of PTPce produces a transformed phenotype in rat embryonic fibroblasts (Zheng et al, 1992). Indeed, the plant oncogene rolB, from Agrobacterium rhizogenes, has been shown to encode a protein that displays PTPase activity (Filippini et al., 1996). The importance of PTPases in cellular physiology is further emphasized by the fact that they are often targets for microbial or viral intervention. For instance, the pathogenic bacterium Yersinia encodes a PTPase essential for its virulence (Guan and Dixon, 1990; Bliska et al., 1991), and the Vaccinia virus encodes a dualspecificity phosphatase, VHl (Guan et al., 1991), that is essential for viral transcription and infectivity (Liu et al., 1995). Since the discovery of the Vaccinia VHl phosphatase, a number of additional dual-specificity phosphatases have been identified. The mammalian dual-specificity protein phosphatases have surfaced as key regulators of mitogenic signaling pathways as well as the cell cycle itself(SuneifaZ., 1993; Arocae^ aZ., 1995; Poon and Hunter, 1995;Keyse, 1995). The low M^ PTPases, whose biological function is unknown, were

PROTEIN-TYROSINE PHOSPHATASES

25

previously thought to exist only in mammalian species. Genetic studies of fission yeast (Schizosaccharomyces pombe) with temperature-sensitive mutations of cdc25 (i.e., cdc25-22) have identified a gene that, when overexpressed, rescues cdc25-22 (Mondesert et aL, 1994). This gene, named stpl^ (small tyrosine phosphatase), encodes a protein that is highly similar (42% identical) to the mammalian low M^ PTPases. The low Mr PTPases have been shown to dephosphorylate both phosphotyrosyl and phosphoseryl/threonyl protein substrates (Zhang et al, 1995a). These findings suggest t h a t low My. PTPases may represent a subgroup of dual-specificity phosphatases t h a t may be involved in cell cycle control.

II. Structural Characterization A. PTPases Structures The structures of the PTPase catalytic domains of P T P I B (residues 1-321; Barford etal, 1994) and Yersinia PTPase (residues 163-468; Stuckey et al, 1994) have been determined at 2.8 A and 2.5 A resolution, respectively. Although the amino acid sequence of the Yersinia PTPase is only —20% identical to t h a t of PTPIB, it is clear t h a t the two structures share a common secondary structural scaffold, with close similarity in tertiary structure. P T P I B is composed of a single domain, with 8 a helices and 12 j8 strands. Similarly, the Yersinia PTPase is a single domain protein with 5 a helices and 8 j8 strands (Fig. 2). The central feature of the P T P I B and Yersinia PTPase tertiary folds is a highly twisted, mixed /3 sheet fianked by a helices on both sides. The PTPase active site is located within a crevice on the molecular surface and is contained on (32,138, (33, and pi in Yersinia PTPase (^83, i812, j84, and j811 in PTPIB), as well as on the associated helices and loops t h a t form the core of the PTPase catalytic region (Fig. 2). Residues of the PTPase signature motif in the Yersinia PTPase are located on the loop structure between the ^ t u r n at the COOH terminal of (38 (residues 403-406) and the first t u r n of helix a5 (407-410). The same motif in PTPIB (residues 214-222) is located between the COOH terminus of j812 and the first t u r n of Q;4. This loop is termed the PTPase phosphate binding loop or P-loop (Stuckey et al, 1994). The P-loop includes the invariant residue Cys-403 in the Yersinia PTPase, which is located at the base of the catalytic pocket. The pocket is —10 A deep and —13 A wide and is surrounded by five polypeptide loops or turns (al-j81, a6-a7, j82-a2, a3-i84, and ^4-^87) and helix a 3 . Similarly, the catalytic cleft of PTPIB is surrounded by four loops (cel-jSl, /33-ce2, f311-a3, and a5-a6), j85, and j86. The majority of the invariant amino acid residues conserved

26

ZHONG-YIN ZHANG

FIG. 2. Ribbon diagram of the Yersinia PTPase structure. Reprinted with permission from Nature [J. A. Stuckey et al, Nature (London) 370, 571-575 (1994)]. Copyright 1994 Macmillan Magazines Limited.

in all PTPases from bacteria to mammals (Zhang and Dixon, 1994) are located in and around the enzyme active site (Barford et al., 1994; Stuckey et al, 1994). This suggests that all PTPases are likely to recognize phosphotyrosine in a similar way and have similar catalytic mechanisms. As will be discussed below, some of these residues are important for active site integrity, while others are involved in catalysis. Tungstate is a competitive inhibitor of both the Yersinia PTPase and PTPIB. The crystal structures of both the Yersinia PTPase- (Stuckey et al, 1994) and the PTPlB-tungstate complexes (Barford e^ al, 1994) have also been solved and provide information about the interactions between the enzyme and the phosphate moiety of the substrate. The structures of the PTPase-tungstate complexes show that the tungstate oxygen atom(s) ion pair with the positively charged Arg-409 and Arg221 of the Yersinia PTPase and PTPIB, respectively. The Yersinia PTPase-tungstate structure indicates that there are two hydrogen

PROTEIN-TYROSINE PHOSPHATASES

27

bonds between the guanidinium group of Arg-409 and two of the tungstate oxygen atoms, a and b (Fig. 3). The tungstate oxygen atoms denoted as a, b, and c are hydrogen bonded to the NH amides of the peptide backbone making up the P-loop. The tungstate oxygen atom denoted as d is projecting out of the active site pocket and interacting with the side chain of the invariant Gln-446 in Yersinia PTPase and Gln-262 in PTPIB. This oxygen atom most hkely corresponds to the oxygen atom present in the substrate scissile phosphate ester bond (Stuckey et al, 1994). Thus, the side chain of R409 acts together with the amides of the P-loop to Ugand the tungstate. The invariant Cys residue (Cys-403 in Yersinia PTPase and Cys-215 in PTPIB) has been shown to be essential for PTPase activity and formation of a covalent phosphoenzyme intermediate. The PTPaseWO4 complexes reveal that the Sy atom of Cys-403 in the Yersinia PTPase is poised 3.6 A from the W atom, while the Sy atom of Cys215 in PTPIB is poised 3.1 A from the W atom. In addition, the sulfur atom is directly opposite the scissile bond oxygen d, and is in a position

FIG. 3. The phosphate binding loop with tungstate bound. Only the peptide backbone from residues 403-409 of the Yersinia PTPase is shown. Tungstate is located in the center of the loop structure. The side chain of Cys-403 corresponds to the active site nucleophile which directly attacks the phosphorus atom in a substrate. The tungstate oxygen atoms, a and b, form two hydrogen bonds with the guanidinium group of Arg409. The tungstate oxygen atoms a, b, and c, also form multiple hydrogen bonds with the NH amides of the peptide backbone making up the phosphate binding loop. The tungstate oxygen atom d is projecting out of the active site and is equivalent to the oxygen present in the scissile ester bond. Reprinted with permission from Z.-Y. Zhang etal, Biochemistry 33,15266-15270 (1994). Copyright 1994 American Chemical Society.

28

ZHONG-YIN ZHANG

for an SN2 nucleophilic attack on a substrate's phosphorus atom (Fig. 3). This is consistent with its role as a nucleophile in the catal3rtic mechanism, as will be discussed below. The apparent thiol pifa of the active site Cys-403 was found to be 4.7 (Zhang and Dixon, 1993). This suggests that Cys-403 exists as a thiolate anion at physiological pH. The crystal structure also provides an explanation for the lowering of the pifa of the invariant Cys residue thiol group from 8.5 in aqueous solution to 4.7 in the PTPase active site. Unlike the cysteine proteinases, which stabilize an active site thiolate anion via an ion pair with a protonated histidine, the PTPases stabilize the active site thiolate by an extensive network of hydrogen bonds radiating out from the Ploop (Barford et aL, 1994; Stuckey et al., 1994). The side chain of the invariant His-402 in the Yersinia PTPase is not involved in ion pair interaction with the side chain of Cys-403 but participates in an array of H bonds that includes the carbonyl of Cys-403, the side chain of His402, the Tyr-301 hydroxyl, and finally, the His-270 side chain. A similar network of H bonds including His-214 is also present in the PTPIB structure. Substitution of the invariant histidine would be expected to disrupt this H-bond network and elevate the apparent pi^a of Cys-403. Indeed, H402N and H402A mutations in the Yersinia PTPase display apparent thiol pifa of 6.0 and 7.4, respectively (Zhang and Dixon, 1993). A second array of H bonds from the carbonyls of residues Ala-405 and Gly-406 in Yersinia VTVsiSe (Ala-217 and Gly-218 in PTPIB) mediate interaction with the buried guanidinium group of Arg-440 (Arg-257 in PTPIB) to stabilize the thiolate anion or bound tungstate. Finally, the helix dipole of ab in Yersinia PTPase (a4 in PTPIB) is positioned in such a way that it will also contribute to the stability of the anion. B. Structures of VHR and Low Mr Phosphatase

The VHl-like dual-specificity phosphatases and the low M^ phosphatases display little amino acid sequence identity with the classical PTPases. The only similarities among these three groups of phosphatases are the relative placements of the essential cysteine and arginine residues in the active sites that constitute the PTPase signature motif (HA/^)C(X)5R(S/T) (Fig. 4). While the P-loop structure appears to be conserved in PTPases from bacteria to mammals, it is striking to find that the P-loop of the PTPase also has a structure similar to that of the active site loop of VHR (Yuvaniyama et al, 1996). VHR (for VHlrelated; Ishibashi e^ aZ., 1992), which is believed to be responsible for activation of cdk-cyclin complex(es) at some stage of the cell cycle (Aroca et aZ./1995), has emerged as the prototype for the dual-specificity phosphatases. Not only has this particular phosphatase been overex-

29

PROTEIN-TYROSINE PHOSPHATASES

Sequence

Phosphatase Yop51 PTPIB CD45

V I H V V H V V H

VHl VHR PAC-1

L V H L V H L V H

Low M^

I

R A G V G S A G I G S A G V G |A A G V N R E G Y S Q A G I S H

FTP

L F V

L G N I C

human cdc25

F V H

E F S S E laG P

FIG. 4. The PTPase signature motif. Reprinted with permission from Z.-Y. Zhang et al, Biochemistry 34, 16389-16396 (1995). Copyright 1995 American Chemical Society.

pressed and purified to homogeneity but, in addition, it has been the subject of several detailed enzymological investigations (Zhou et al,, 1994; Denu et al, 1995a,b; Zhang et al, 1995b). The structure of VHR reveals a general fold that occurs in the Yersinia PTPase and the PTPIB structures, and many of the secondary structural elements of PTPases are also present in VHR. Although VHR is smaller than PTPases, the VHR structure retains the same starting and ending secondary structure elements as the PTPase catalytic domains, suggesting that the VHR structure may define a minimal core structure for both dualspecificity phosphatases and the PTPases (Yuvaniyama et al, 1996). In addition to the general lack of sequence identity between the low Mr phosphatases and the PTPases, the low M^ enzymes are also relatively smaller and contain a PTPase signature motif close to the NH2 terminus of the protein, whereas the signature motif occurs closer to the COOH terminus of the PTPases and the dual-specificity phosphatases. The bovine low M^ PTPase has distinct topologies compared with those of the PTPIB, the Yersinia PTPase, and the VHR (Su et al, 1994; M. Zhang et al, 1994). However, it is interesting that residues of the PTPase signature motif (12-19 in the low M^ PTPase, 403-410 in the Yersinia PTPase, 214-222 in PTPIB, and 123-131 in VHR) form a similar loop structure, termed the phosphate-binding loop, between the ^ turn at the COOH terminus of a j8 strand and the first turn of an a helix (Su et al, 1994; M. Zhang e^ al, 1994; Barford et al, 1994; Stuckey et al, 1994; Yuvaniyama et al, 1996). Furthermore, a structural comparison between the bovine low M^ phosphatase and PTPIB reveals that in addition to the confined similarity in the active sites, the central

30

ZHONG-YIN ZHANG

parallel (3 sheet of low Mj. phosphatase superimposes closely with the four central ^8 strands of PTPIB (Barford et al, 1995). Thus, these different phosphatase structures are striking examples of convergent evolution achieving highly similar active site clefts, and the similarities in the conserved active site motifs may suggest a common mechanism to bring about phosphate monoester hydrolysis in these otherwise very differernt molecules. Indeed, the invariant Cys residue has been shown to be essential for phosphatase activity and formation of a covalent cysteinyl phosphoenzyme intermediate (Zhang, 1990; Guan and Dixon, 1991; Wo et al, 1992; Cho et al, 1992; Zhou e^ al, 1994), whereas the invariant Arg residue in the signature motif has been shown to play an important role in substrate binding and transition state stabilization (Zhang et al, 1994d). These phosphatase structures, taken together with earlier biochemical and more recent mutational studies of active site residues, provide a unique opportunity to identify common mechanistic features associated with this novel family of biological catalysts. In addition, the availability of the structures combined with the ability to mutate specific residues provides an opportunity to test these mechanistic hypotheses and identify key residues for catalysis and substrate/inhibitor recognition. III. Catalytic Mechanism A. An Overview The mechanism of phosphate monoester hydrolysis [Eq. (1)] has been the subject of many investigations (CuUis, 1987; Frey, 1989; Thatcher and Kluger, 1989), ROPOi" + H2O ^ ROH + HOPOi"

(1)

because the hydrolysis of phosphate monoesters is linked to energy metabolism, to metabolic transformation and regulation, and, more recently, to a wide variety of signal transduction pathways. Hydrolysis of phosphate monoesters is a thermodynamically favorable process (AG^ < -2.15 kcal/mol), but in the absence of enzymes, phosphate monoesters are almost kinetically inert (Westheimer, 1987). The means by which their reactivity is enhanced by PTPases presents both an old and a new problem in mechanistic enzymology. A central question concerning the mechanism of phosphatases is whether substitution at phosphorus proceeds by a single-displacement or double-displacement pathway. Single displacement involves the direct transfer of a phosphoryl group from the monoester to water. In

PROTEIN-TYROSINE PHOSPHATASES

31

the double-displacement mechanism, the phosphoryl group is first transferred to a nucleophilic group of the enzyme, forming a phosphoenzyme intermediate which is then hydrolyzed by water. The kinetic and catalytic mechanisms of both alkaline (Kim and Wyckoff, 1991; Coleman, 1992) and acid phosphatases (Van Etten, 1982) have been demonstrated to involve a phosphoserine (Schwartz and Lipmann, 1961) and a phosphohistidine intermediate (Van Etten and Hickey, 1977), respectively. A carboxyl phosphate intermediate is involved in a sarcoplasmic reticulum ATPase-catalyzed reaction (Degani and Boyer, 1973). Detailed studies of the catalytic mechanism of PTPases are beginning to appear in the literature. B. Novel Thiol Phosphate Covalent Intermediate

The PTPase-catalyzed reaction involves a phosphoenzyme intermediate which can be trapped by addition of a denaturant soon after mixing the enzyme with a ^^P-labeled substrate (Zhang, 1990; Guan and Dixon, 1991; Pot et al, 1991; Wo et aL, 1992; Cho et al, 1992). This suggests that the PTPase-catalyzed hydrolytic reaction is nucleophilic in nature, and involves both the formation and the breakdown of a phosphoenzyme intermediate. An analysis of the chemical stability of the trapped phosphoenzyme intermediate in the PTPase-catalyzed reaction suggests that it has the characteristics of a previously undescribed thiophosphate linkage (-S-POi~) (Zhang, 1990; Guan and Dixon, 1991; Wo et al., 1992). Site-directed mutagenesis experiments show that the invariant Cys residue (e.g., Cys-403 in the Yersinia PTPase and Cys-215 in PTPl) is required for PTPase activity (StreuH et al, 1990; Guan and Dixon, 1990; Gautier et al, 1991; Cirri et al, 1993; Zhou et al, 1994). Replacement of the catalytically essential Cys-215 residue in PTPl with Ser destroys its ability to form a phosphoenzyme intermediate, suggesting that the intermediate is a phosphocysteine (Guan and Dixon, 1991; Cirri et al, 1993; Zhou et al, 1994). In addition, the inactivation of PTPases, the dual specificity, and the low Mj. phosphatases by iodoacetate selectively modifies the active site Cys residues of these phosphatases (Camici et al., 1989; Pot and Dixon, 1992; Zhang and Dixon, 1993; Zhou et al., 1994). The participation of a cysteine residue in the phosphoenzyme intermediate is further supported by ^^P NMR analysis of the trapped intermediate in the low M^ PTPase (Wo et al., 1992) and the receptor-like PTPase, LAR (leukocyte antigenrelated PTPase), catalyzed reaction (Cho et al, 1992). The ^^P NMR chemical shift of the phosphoenzyme intermediate is typical of a thiophosphate bond. Moreover, the formation and decay of the intermediate have been shown to be kinetically competent by fast quench technique

32

ZHONG-YIN ZHANG

(Cho et al., 1992). All of the above experiments point to the existence of a covalent cysteinylphosphate enzyme intermediate on the kinetic pathway. The involvement of a phosphocysteine intermediate in phosphatasecatalyzed hydrolysis reactions is novel. A thiophosphate linkage in proteins has been reported only in bacterial thioredoxin (Pigiet and Conley, 1978) and mannitol carrier-specific transporter enzyme II (Pas et al., 1991). The biological significance of thiophosphate in thioredoxin is not clear. A phosphocysteine intermediate proposed for the mannitolcarried specific transporter enzyme in bacteria may be required for phosphate transfer from phosphoenolpyruvate to mannitol. Thiophosphate esters have been suggested as "models" for intermediates in enzymatic phosphoryl transfers (Walsh, 1952), but evidence for their existence in enzymatic systems has been weak until now. Because the energy of the P-S bond (45-50 kcal/mol) is considerably less than that of the P - 0 bond (95-100 kcal/mol), P - S bond cleavage is much more facile than P - 0 cleavage (Bruice and Benkovic, 1966). Thus, it is not unreasonable for PTPases to utilize a thiophosphate as a covalent enzyme intermediate in catalysis. Previous work suggests that nonenzymatic thiophosphate ester hydrolysis exhibits many of the characteristics of its oxygen counterparts (Herr and Koshland, 1957; Dittmer et al, 1963; Bruice and Benkovic, 1966; Milstien and Fife, 1967). PTPases provide a unique system in which to examine the enzyme-catalyzed phosphoryl transfer where a cysteine nucleophile is directly involved. C. Invariant Arginine in PTPase Signature [\/lotif

As discussed above, the invariant Cys present in the PTPase signature motif, which includes the active site sequence (HyV)C(X)5R(S/T) (where X can be any amino acid; Fig. 4), is absolutely required for catalysis and is directly involved in phosphoenzyme formation. The importance of the invariant Arg residue in the PTPase signature motif (Arg-409 in Yersinia PTPase) has also been recognized. Mutations of the invariant Arg residue in the PTPase signature motif resulted in complete loss of enzymatic activity for two receptor-like PTPases, LAR and CD45 (Streuh et al, 1990; Johnson et al, 1992) and the low M, phosphatase from bovine liver (Cirri et al, 1993). However, from such studies, it is not clear whether the Arg residue is essential for PTPase folding and structure or catalysis, or both. The extraordinary reactivity of the Yersmia PTPase (Zhang et al, 1992) has made it possible to examine the effect of mutagenesis on the invariant Arg residue within the active site. The higher intrinsic activity of Yersinia PTPase has made it possible to measure ^cat and K^ values for both the R409A and

PROTEIN-TYROSINE PHOSPHATASES

33

R409K site-directed mutants to address the functional significance of Arg-409 in terms of substrate binding and catalysis (Zhang et al., 1994d). A 8200-fold decrease in ^cat and a 26-fold increase in K^ were observed for the R409A mutant. Interestingly, the R409K mutant displayed a ^cat value identical to that of R409A, and the apparent K^ value for pNPP was only 1.9-fold higher than that of the wild-type enzyme. The R409A mutation decreases the arsenate binding affinity by 47-fold, while R409K decreases the arsenate binding affinity by 18fold. These results suggest that Arg-409 plays a critical role in phosphosubstrate binding. The most important function of Arg-409 may be transition state stabilization, which is reflected by the dramatic reduction in both ^cat/^m and ^cat values for the Arg-409 mutants. Tungstate is a competitive inhibitor of the Yersinia PTPase, with a binding constant of 61 IJLM at pH 7.0 and ionic strength of 0.15 M. The three-dimensional structure of the Yersinia PTPase-tungstate complex (Stuckey et al., 1994) shows that tungstate oxygen atom(s) ion pair with the positively charged guanidinium group of Arg-409 (Fig. 3). Tungstate oxygen atoms labeled a and b form hydrogen bonds with the Ns and NT/ of Arg-409. The tungstate oxygen atom denoted as c is hydrogen bonded to the NH amides of the peptide backbone making up the phosphate-binding loop. The tungstate oxygen atom denoted as d is projecting out of the active site pocket and most likely corresponds to the oxygen atom present in the scissile substrate phosphate ester bond. Figure 3 shows that the thiolate anion of Cys-403 is positioned at the base of the active site and would likely attack the phosphate ester, forming a trigonal bipyramidal transition state. The side chain of Arg-409 is in an ideal position for effective stabilization of the transition state (Fig. 3). Cleavage of the phosphoester linkage would then produce the Cys-403 thiophosphate intermediate with release of phenol/alcohol. Hydrolysis of the phosphoenzyme intermediate would most likely require water to attack in the same axial position from which the phenol/ alcohol departed. The attack of water on the phosphoenzyme intermediate would again form a trigonal bipyramidal transition which can be similarly stabilized by Arg-409. The subsequent cleavage of the thiophosphate bond would regenerate the thiolate anion of Cys-403 and yield the other product of the reaction, inorganic phosphate. Studies have shown that a guanidinium group (present in an arginine) is ideally suited for interaction with phosphate by virtue of its planar structure and its ability to form multiple hydrogen bonds with the phosphate moiety (Cotton et al., 1973). The ability of the guanidinium group to form a coplanar bidentate complex with two equatorial oxygen atoms present on the phosphate during catalysis provides a

34

ZHONG-YIN ZHANG

plausible mechanism for stabilization of the trigonal bipyramidal transition state(s). The geometry associated with the amino group of the alternate cationic Lys side chain would not be expected to be able to form a coplanar bidentate complex with the trigonal bipyramidal transition state, which may explain why the R409K mutant showed no improvement in ^cat when compared to R409A. The fact that a Lys residue at position 409 can partially replace the Arg residue in terms of substrate and inhibitor binding, while at the same time being unable to substitute for the Arg in catalysis, suggests that the transition state(s) likely employ the unique structural properties of the guanidinium side chain of Arg-409. It is most likely that the conserved Arg residue in the PTPase active site is geometrically positioned in such a manner that it interacts more favorably with the transition state than with the ground state. D. Conserved Serine/Threonine in PTPase Signature IViotif

In addition to the essential Cys and Arg residues, a conserved Ser or Thr can be found in the PTPase signature motif immediately after the invariant Arg residue (Fig. 4, Zhang et al., 1995c). The dual-specificity phosphatase Cdc25 is the only one that lacks a hydroxyl group at this position. Interestingly, Cdc25 is several orders of magnitude less reactive than other PTPases (Dumphy and Kumagai, 1991; Zhang et al, 1992). In the bovine low M, PTPase structure (Su et al, 1994; M. Zhang, et al., 1994), as well as in the Yersinia PTPase (Stuckey et al., 1994) and the human PTPIB (Barford et al, 1994) structures, the hydroxyl group of the conserved Ser/Thr is approximately 3 A to the Sy of the active site Cys residue (Fig. 5), making a reasonable good S-HO hydrogen bond (Gregoret et al, (1991). Because the PTPases, the dual-specificity phosphatases, and the low Mj. phosphatases effect catalysis through a covalent thiophosphate enzyme intermediate, the catalyzed reaction must be composed of at least two chemical steps, i.e., formation and breakdown of the phosphoenzyme intermediate. The phosphoryl group in the substrate is first transferred to the nucleophilic active site thiolate group of the enzyme to form the phosphoenzyme intermediate (^2)? which is then hydrolyzed by water (^3) (Scheme I). The kinetic scheme is composed of substrate binding, followed by two chemical steps. ki

E + ROPOf^

1 E • ROPOi

< E-P ROH

SCHEME 1

ko

- E + Pi

PROTEIN-TYROSINE PHOSPHATASES

35

FIG. 5. Active site conformation of the PTPase signature motif corresponding to residues Cys-12 to Ser-19 in the bovine low M, PTPase (M. Zhang et al, 1994). The hydrogen bond between the sulfur atom of Cys-12 and the hydroxyl group of Ser-19 is highlighted. Reprinted with permission from Y. Zhao and Z.-Y. Zhang, Biochemistry 35,11797-11804 (1996). Copyright 1996 American Chemical Society.

phosphorylation {k^ and dephosphorylation (^3), where E is the enzyme, ROPO3 the substrate, E-ROPOi the enzyme-substrate Michaehs complex, E-P the phosphoenz3mae intermediate, ROH the phenol, and Pi inorganic phosphate. If the net rate of intermediate breakdown is slower than that of intermediate formation, one would predict a "burst" ofp-nitrophenol production using p-nitrophenyl phosphate (pNPP) as a substrate. Burst kinetics has been demonstrated with the Yersinia PTPase (Zhang e^aZ., 1995c), rat PTPl (Zhang, 1995a), the dual-specificity phosphatase VHR (Zhang et al, 1995b), and the low M^ phosphatases (Zhang and Van Etten, 1991; Zhang et al, 1995a). This has permitted the determination of individual rate constants directly associated with the formation {k^ and breakdown (^3) of the phosphoenzyme intermediate. Thus, an evaluation of the burst kinetics combined with the technique of site-directed mutagenesis should allow one to ascertain specific contributions of active site residues to the individual steps of the phosphatasecatalyzed reaction. It was shown that under most conditions, the decomposition of the cysteinylphosphate enzyme intermediate (^3) is the rate-limiting step for the overall phosphatase-catalyzed hydrolysis. Supporting evidence includes that up to 27-74% of PTPase can be trapped as a covalent adduct using ^^P-labeled p-nitrophenyl phosphate (Zhang, 1990; Guan and Dixon, 1991; Wo et al, 1992) and that such an intermedi-

36

ZHONG-YIN ZHANG

ate can be prepared in sufficient amount for ^^P NMR analysis (Wo et al, 1992; Cho e^aZ., 1992). Site-directed mutagenesis and pre-steady-state stopped-flow kinetic experiments have demonstrated that the conserved hydroxyl group in the PTPase signature motif plays a critical role in efficient E - P hydrolysis in the Yersinia PTPase (Zhang et al., 1995b), the dual-specificity phosphatase, VHR (Denu and Dixon, 1995), and the low Mr phosphatase-catalyzed reaction (Zhao and Zhang, 1996). It appears that the elimination of the hydroxyl group in the conserved Ser/Thr has only a modest effect on k^, whereas its major impact seems to be primarily reflected in ^3. For example, in the reaction catalyzed by the low M^ phosphatase (Stpl) from the fission yeast S. pombe (Mondesert et al, 1994), elimination of the hydroxyl group at Ser-18 decreases the rate of E - P formation (^2) and breakdown (ks) by 4.3- and 35.7-fold, respectively (Zhao and Zhang, 1996). Similarly, substitution of the corresponding Thr residue by an Ala in the Yersinia PTPase resulted in a 2.4- and 30.4-fold reduction in ^2 and k^, respectively, at pH 5.8 (Zhang et al., 1995c). Figure 6 shows the differential effects of Ala substitution at residue Thr-410 of the Yersinia PTPase on ^2 and kz at pH 6.0 (Zhang et al, 1995c). Thus, results from the Yersinia PTPase, the dual-specificity 0.25

E

c o ^

0.2

0.15

0

o c

CO

0.1

o

<

0.05

h

0.02

0.04

0.06

0.08

0.1

Time (s) FIG. 6. Burst kinetics observed with the wild-type (WT) and the T410A mutant Yersinia PTPase. The wild-type enzyme concentration was 41.4 IJLM, and that of T410A was 45 ixM. The pNPP concentration was 20 and 4 mM for the wild-type and T410A Yersinia PTPase, respectively. Each stopped-flow trace was an average of at least six individual experiments. The solid line represents a theoretical fit of the data to the equation [pnitrophenolate] = At -^ B{1 - e~^*). Reprinted with permission from Z.-Y. Zhang et al., Biochemistry 34, 16389-16396 (1995). Copyright 1995 American Chemical Society.

PROTEIN-TYROSINE PHOSPHATASES

37

phosphatase VHR, and the low M^. phosphatase Stpl suggest that the main function of the conserved hydroxyl group in the PTPase signature motif is to faciUtate the breakdown of the phosphoenzyme intermediate. How this is accompUshed by the hydroxyl side and why disruption of the hydroxyl-thiolate interaction causes differential effects on the two chemical steps will be discussed in conjunction with the description of the nature of the transition state in the PTPase-catalyzed reaction. E. General Acid-Base Catalysis In addition to nucleophilic catalysis and transition state stabilization, it appears that all three groups of phosphatases also utilize general acid-base to facilitate catalytic turnover. By site-directed mutagenesis and pH dependency kinetic analysis, Asp-356 in the Yersinia PTPase was first demonstrated to be the general acid in the PTPase-catalyzed reaction (Zhang et al., 1994a). Detailed kinetic analyses of the Yersinia PTPase and the rat PTPl-catalyzed reaction have been performed in order to elucidate further the mechanism of phosphate monoester hydrolyses by this family of enzymes (Zhang et al., 1994c; Zhang, 1995a). The hydrolysis of phosphotyrosine-containing peptides as well aspNPP by PTPases display a bell-shaped profile when values of ^cat are plotted vs. pH. The apparent pi^a values obtained in the pH vs rate profile suggest that acidic residues are responsible for both the ascending and descending limbs of the profile. Site-directed mutagenesis was used to probe the function of several conserved acidic residues (Zhang et al., 1994a). Detailed kinetic analysis of each of these mutant PTPases shows that mutations at either residue 290 or 356 in the Yersinia PTPase led to a marked reduction in catal3^ic activity and an alteration in the pH dependence for catalysis. Replacement of Glu-290 with Gin caused a reduction of 130-fold in ^cat? whereas replacement of Asp-356 withAsn caused a reduction of 2.4 X 10^-fold in ^cat- Elimination of the carboxyl group at residue 290 (E290Q) caused the disappearance of the acidic limb of the pH profile, whereas elimination of the carboxyl group at residue 356 (D356N) caused the disappearance of the basic limb of the pH profile. Furthermore, when both carboxyl side chains at residues 290 and 356 were eliminated, the ^cat value was reduced by 5.4 X 10^-fold and the double mutant-catalyzed phosphomonoester hydrolysis was completely pH independent. These results have led to the conclusion that residues Asp-356 and Glu-290 function as general acid-general base catalysts in the Yersinia PTPase-catalyzed hydrolysis. Interestingly, these two residues are the only acidic residues that are invariant in all PTPases (Zhang et al,, 1994a), suggesting that a common catalytic mechanism is utilized by this family of enzymes.

38

ZHONG-YIN ZHANG

In the Yersinia PTPase, Asp-356 is found on a flexible loop which undergoes a major conformational change on binding of tungstate (Stuckey et al, 1994). In the unliganded Yersinia PTPase structure, Asp-356 is greater than 10 A from the phosphate-binding site. In the tungstate-complexed structure, the surface loop (residues 351-360) that contains the Asp-356 residue has moved like a "flap" to cover the active site (Stuckey et aL, 1994). The C« of Asp-356 itself moves 6 A toward the active site on tungstate binding, positioning its carboxylate 3.5 A away from the tungstate oxygen (d in Fig. 3) which is structurally homologous to the scissile oxygen of a phosphotyrosine substrate. This is consistent with the role of the invariant Asp-356 donating a proton to the tyrosine leaving group during the flrst catalytic step. Similarly, binding of either phosphortyrosine or phosphotyrosine-containing peptide to the Cys-215 to Ser mutant mammalian PTPase, PTPIB, causes a conformational change of an equivalent surface loop that brings the corresponding Asp-181 into the catalytic site {^metal, 1995). Substitution of the equivalent Asp-181 in PTPl by an Asn residue (Hengge et aL, 1995) results in a 300-fold reduction in ^cat and a loss of the basic limb of the bell-shaped pH-rate profile (L. Wu and Z.-Y. Zhang, unpublished results). In the structure of the PTPlB(C215S)-substrate complex, Asp-181 forms a network of hydrogen bonds to the phenolic oxygen of phosphotyrosine and a buried water molecule (Jia et aL, 1995). Thus Asp-181 in PTPl may also function as a general acid to donate a proton either directly or through the water molecule to the phenolic oxygen. The conserved Glu-290, originally thought of as a general base, is located on another loop and its side chain extends into the active site, where it forms an important bidentate ionic interaction with the side chain of the invariant Arg-409 in the phosphate-binding loop. Similar interaction is also noted in the PTPIB structure (Barford et aL, 1994). This suggests that the carboxyl group of Glu-290 may be required to position Arg-409 for effective substrate binding and transition state stabilization. The critical role of Glu-290 in orienting Arg-409 explains why it has features of an apparent general base. The bidentate hydrogen bonding interaction between Arg-409 and Glu-290 would be disrupted once the carboxylate is protonated. When an approach identical to the one demonstrated for the Yersinia PTPase (Zhang etaL, 1994a) was used, the involvement of an aspartic acid as a general acid was also confirmed in the dual-specificity phosphatase VHR-catalyzed reaction (Denu et aL, 1995b). Interestingly, although Asp-92 in VHR plays a role similar to that of the aspartic acids in the Yersinia PTPase and PTPIB (Asp-356 and Asp-181, respectively), it is located on a different side of the active site and may

PROTEIN-TYROSINE PHOSPHATASES

39

protonate the bridging oxygen in the phosphate ester substrate from a different angle than that observed in the PTPases (Yuvaniyama et aU 1996). It is noteworthy that in the low M^. phosphatase, the active site arginine is also held in position by an aspartic acid (M. Zhang et al., 1994). Although the low Mr phosphatases share no discernible sequence identity with any of the PTPases except for the conserved cysteine and arginine in the PTPase signature motif, the crystal structures of the two bovine M^ phosphatases have also allowed identification of the possible candidate for the proton donor (Su et ah, 1994; M. Zhang et aL, 1994). Both structures reveal that Asp-129, located on a loop adjacent to the phosphate-binding loop but opposite the nucleophilic cysteine, is pointed toward the bound sulfate and phosphate, respectively. In one structure, solved at pH 5.5 with a bound sulfate at the active site (Su et aL, 1994), the side chain of Asp-129 is shown to form a hydrogen bond with one oxygen of the sulfate anion. The distance between the carboxylate oxygen and the sulfate oxygen is 2.7 A. This suggests that the side chain of Asp-129 is protonated in this condition. In the other structure, determined at pH 7.5 with a phosphate anion bound at the active site (M. Zhang et al, (1994), the side chain of Asp-129 is 3.66 A away from the phosphate. This increase in distance could be caused by charge repulsion between the phosphate and the carboxylate, since at pH 7.5 the carboxyl group of Asp-129 is likely deprotonated. Because the oxygen atom in the bound oxyanion that interacts with Asp-129 is structurally homologous to the scissile oxygen of a phosphate monoester substrate, one role proposed for Asp-129 is that it donates a proton to the leaving group during phosphoenzyme formation (Su et al., 1994; M. Zhang et al., 1994). Subsequent mutational studies provided evidence for the importance of Asp-129 in catalysis, but the detailed mechanism and the specific step(s) that are effected by Asp-129 remained controversial due to the inherent limitations of steady-state kinetics (Taddei et al, 1994; Z. Zhang et aL, 1994). In both cases, Asp-129 was changed to an Ala residue. However, the conclusions of the two papers differ drastically. For example, in one study, Asp-129 is concluded to be the proton donor to the leaving group in the phosphorylation step, and its mutation to alanine is believed to result in a change in the rate-limiting step of the catalysis from dephosphorylation (^3) to phosphorylation (^2) (Z. Zhang et al., 1994). In the other study, Asp-129 is suggested to be involved in both steps of the catalytic mechanism, and the rate-limiting step of the D129A mutant is reported to still correspond to ^3 (Taddei et al., 1994).

40

ZHONG-YIN ZHANG

The detailed mechanism and the specific step(s) that are effected by the Asp residue have been revealed by an investigation that utilizes a combination of techniques including site-directed mutagenesis, as well as pre-steady-state and steady-state kinetic analysis (Wu and Zhang, 1996). Asp-128 in the yeast enzyme Stpl has been replaced by a Glu, an Asn, and an Ala. The ^cat for the hydrolysis of pNPP decreases by factors of 6.7,400, and 650 for the mutants D128E, D128N, and D128A, respectively. An evaluation of the burst kinetics demonstrates that Asp-128 plays a role in both phosphoenzyme intermediate formation (^2) and breakdown (^3). Thus, substitution at Asp-128 by a Glu, an Ala, or an Asn reduces ^2 by 17-, 7480-, and 11,900-fold and k^ by 6.2, 380 and 40-fold, respectively. The greater effect on k^ than k^ is consistent with a dissociative transition state for the low M^ PTPasecatalyzed reaction (see Section III,F). Taken together, these results are consistent with Asp-128 acting as a general acid to donate a proton to the phenolate leaving group in the phosphorylation step (^2)? and the same carboxylate side chain acting as a general base to activate a nucleophilic water molecule in the dephosphorylation step (A3). F, Nature of Transition State Reaction of phosphate monoesters with water or other nucleophiles [Eq. (1)] resulting in hydrolysis or transfer of the phosphoryl (PO3) group can, in principle, occur via one of two limiting mechanisms analogous to those for substitution at tetrahedral sp^ carbon (Cox and Ramsay, 1964; Benkovic and Schray, 1978): dissociative [SNl-like, Eq. (2)], or associative [SN2-like, Eq. (3)]. In the SNI mode, dissociation of the leaving group 0 II RO-p-0~ 1

o R O - P - O" I

„^ H20

^ROH+

^^ \

o I

II P

^RO-P-OH / \

0 H2O

0 II ^HO-p-0~ I

(2)

o ^ROH + H O - P - O I

(3)

produces a highly electron-deficient tricoordinated phosphorous species, "monomeric metaphosphate," as an intermediate (Butcher and Westheimer, 1955; Westheimer, 1981). In the SN2 mode, phosphorous compounds have the further potential compared with carbon compounds of reacting via a stable pentacoordinated phosphor ane interme-

PROTEIN-TYROSINE PHOSPHATASES

41

diate (Benkovic and Schray, 1978). There is much evidence to support the proposal that the monoanion and dianion of phosphate monoesters hydrolyze via a dissociative mechanism (Butcher and Westheimer, 1955; Bunton et al, 1967; Kirby and VarvogUs, 1967, 1968; Gorenstein et ah, 1977). However, there is no direct evidence for the existence of a metaphosphate intermediate in aqueous solutions (Jencks, 1980; Herschlag and Jencks, 1986). A free metaphosphate has only been implicated in a solvent such as ^er^butanol (Friedman ej^ al., 1988). In fact, all of the available data in aqueous solution can be interpreted by a concerted SN2 mechanism involving an unsymmetrically "exploded" metaphosphate-like transition state in which bond formation to the incoming nucleophile is minimal and bond breaking between phosphorus and the leaving group is substantial (Bourne and Williams, 1984; Skoog and Jencks, 1984; Jencks et al, 1986; Herschlag and Jencks, 1986, 1989). Secondary ^^O kinetic isotope effect in the nonbridging positions of the phosphate group (Weiss et al, 1986; Cleland and Hengge, 1995) and primary ^^O kinetic isotope effect in the leaving group oxygen (Gorenstein et aL, 1977; Hengge et al., 1994) are consistent with a dissociative transition state for nonenzymatic hydrolysis of phosphate monoesters. As discussed above, nonenzymatic nucleophilic displacement reactions on phosphate monoesters are believed to involve an exploded metaphosphate-like dissociative transition state (Cleland and Hengge, 1995). A prerequisite for the understanding of PTPase-catalyzed phosphoryl transfer reactions is the elucidation of the nature of the enzymatic transition state. What is the preferred mechanistic route for PTPase-catalyzed phosphoryl transfer reactions? Do PTPases catalyze phosphoryl transfers via a dissociative or an associative mechanism? Different strategies may be utilized by a PTPase to accelerate a reaction, depending on the nature of the transition state. An associative mechanism had been favored for enzyme-catalyzed phosphoryl transfer reactions, since an enzyme could stabilize the increased negative charge on the phosphoryl moiety in the transition state (Hasset et al, 1982; Mildvan and Fry, 1987; CuUis, 1987). It has been argued that general acid catalysis to accelerate the departure of the leaving group should be enzymatically important regardless of the mechanism of phosphorylation (i.e., associative or dissociative) and that there should be little requirement for general base catalysis in the dissociative mechanism (Benkovic and Schray, 1978). In order to maximize the catal3^ic power from proton transfer to a general base at the active site, it is conceivable that the enzyme may select a some-

42

ZHONG-YIN ZHANG

what more associative transition state with more bond formation to the entering nucleophile (Herschlag and Jencks, 1989). In order to determine the nature of the transition state for the PTPase-catalyzed reaction, heavy-atom kinetic isotope effects have been measured (Hengge et aL, 1995). The transition-state structure of the Yersinia PTPase and the PTPl-catalyzed hydrolysis of pNPP has been probed by measuring the secondary ^^O isotope effect for the phosphoryl oxygen atoms, the primary ^^O isotope effect in the leaving group oxygen atom, as well as the secondary ^^N isotope effect in the nitrogen atom of the leaving group. All isotope effects were measured by the competitive method using an isotope ratio mass spectrometer (O'Leary and Marlier, 1979), which is by far the most accurate method for the measurement of the comparatively small magnitudes of heavy-atom isotope effects. Because the competitive method measures effects on kcJK^, the isotope effects reported are those on the first part of the mechanism through the first irreversible step, shown as ^2 in Scheme I. It is demonstrated that for both the Yersinia PTPase and rat PTPl, (1) P - 0 bond cleavage is rate-limiting for the k^JK^^ portion of the mechanism and (2) the proton from the general acid is largely transferred to the bridge oxygen in the transition state. Furthermore, comparisons of the heavy-atom kinetic isotope effects on the PTPase-catalyzed pNPP hydrolysis with the solution data suggest that the transition state for the phosphorylation step (^2) of the PTPase reaction is highly dissociative in character, as is the case for the nonenzymatic reaction (Hengge e^ al, 1995). Similar results have also been obtained with VHR (Hengge et al, 1996) and Stpl (A. C. Hengge and Z.-Y. Zhang, unpublished result). A dissociative transition state for the phosphorylation of PTPase is suggested in Fig. 7. The P - 0 bond to the leaving group is largely broken, proton transfer to the leaving group oxygen is correspondingly advanced such that the departing phenol has no charge, and the central phosphoryl group resembles metaphosphate in structure. To probe the transition state of the dephosphorylation step, the linear free energy relationship or Br0nsted correlation has been applied to study the effect of changing pi^a on the second-order rate constants for the reaction of j8-substituted ethanols with the phosphoenzyme intermediate (E-P) (Zhao and Zhang, 1996). This is based on the fact that E - P can partition its reaction with water to give the hydrolysis product and with alcohols to produce the alkyl phosphates. Scheme II illustrates the partitioning of E - P in the presence of alcohol ROH, in which ks i=k3 [H2Q]) is the rate of hydrolysis and k^ [ROH] is the rate of E - P reacting with ROH to form the phosphorylated alcohol

43

PROTEIN-TYROSINE PHOSPHATASES

E-P Breakdown

E-P Formation

E-S

+

p T D a C A

E-S-PO3

RO-PO3

p T p M A M

,AAAA/>yV\AArU\/>/VVV\AAA/VVVV» A S D

+

HoO

.AAAAA/VWAAAAyVWAAAAAAAA/'

ASP

H ...p___.

-0 —

C y s — S-

R

p.--. O

H

H

O

O

Ser/Thr

Ser/Thr

I

-0^

O

I

I

I

I

FIG. 7. Suggested transition state structures for the phosphorylation and dephosphorylation steps in the PTPase-catalyzed reaction. In the E - P formation step, the proton is shown predominantly on the side chain the SerAThr residue, while in the E - P breakdown step, the proton is shown predominantly on the attacking water molecule, as expected from the nature of the transition states. Reprinted with permission from Y. Zhao and Z.-Y. Zhang, Biochemistry 35,11797-11804 (1996). Copyright 1996 American Chemical Society.

(ROPO3 ). The j8nu: parameter, obtained from structure-reactivity correlations of reactivity as a function of the pi^a of the attacking nucleophile, may be viewed as an empirical index of the fraction of charge transferred to the nucleophile and correspondingly may reflect the degree of bond formation between the nucleophile and the phosphorus in the transition state (Jencks and Gilchrist, 1968). Thus, information about the transition state of E - P dephosphorylation can be obtained by studyks [ H 2 O ]

K E + ArOPOi

:^EArOPOr

E-P ArOH SCHEME II

V

E + Ropor ^4 [ ROH ]

44

ZHONG-YIN ZHANG

ing the selective influence of a nucleophilic acceptor on the partitioning of E - P in the Stpl-catalyzedpNPP hydrolysis. A Br0nsted plot of log ^4 against the ipK^ of the attacking alcohol gives a slope of ^Snu: = 0.14 for the wild-type Stpl-catalyzed reaction (Fig. 8). Thus the nucleophilic reactivity of alcohols toward E - P is not significantly influenced by the basicity of the nucleophiles. In nonenzymatic solution reaction, the most striking characteristic of the reactions of nucleophilic reagents with phosphate monoesters is the very low sensitivity of the rates of the reactions to the basicity of the attacking nucleophilic reagent. For example, the /3nu: value is 0.13 for pNPP dianion with amine nucleophile (Kirby and Jencks, 1965). The iSnu: value for the reaction of the mixed anhydride acetyl phosphate with alcohols was found to be in the range of 0.10 to 0.14 (Herschlag and Jencks, 1989). A ^Snu: value of 0.07 was obtained from studies of reactions of phosphoanhydrides with alcohols (Admiraal and Herschlag, 1995). These are consistent with the prevailing view in the literature that in non-enzyme-catalyzed solution reactions, phosphate monoesters hydrolyze via a dissociative mechanism involving an unsymmetrically exploded metaphosphate-like transition state in which bond formation to the incoming nucleophile is minimal and bond breaking between phosphorus and the leaving group is substantial (Benkovic and Schray, 1978; Thatcher and Kluger, 1989; Cleland and Hengge, 1995).

O

FIG. 8. Br0nsted plots of the second-order rate constant for the alcoholysis of E - P as a function of the basicity of the alcohol in the Stpl (•) and Stpl/S18A (•) catalyzed reactions. Reprinted with permission from Y. Zhao and Z.-Y. Zhang, Biochemistry 35, 11797-11804 (1996). Copyright 1996 American Chemical Society.

PROTEIN-TYROSINE PHOSPHATASES

45

The weak dependence of the second-order rate constant (k^ on basicity (jSnu: = 0.14) for the reactions of alcohols with E - P suggests that the transition state for the dephosphorylation of E - P in the Stplcatalyzed reaction is also highly dissociative (Fig. 7). On the other hand, the fact that phosphoryl transfer is dependent on nucleophile basicity also indicates that the entering nucleophile is a required participant in the reaction pathway and argues against the existence of a free metaphosphate intermediate in the PTPase-catalyzed reaction. It is thus concluded that for both of the chemical steps (i.e., E - P formation and breakdown) in the PTPase-catalyzed reaction, the transition states are highly dissociative and similar to those in non-enzyme-catalyzed solution reactions. The fact that the transition states for E - P formation and hydrolysis are both dissociative is in accord with mutational and kinetic observations. An important strategy to use in promoting a dissociative mechanism is to stabilize the buildup of negative charge on the leaving group (Benkovic and Schray, 1978; Herschlag and Jencks, 1990). The dissociative transition state for E - P formation in the PTPase reaction is stabilized by an active site Asp residue which facilitates the departure of the leaving phenoxide (Hengge et aL, 1995; Zhang et al., 1995b; Wu and Zhang, 1996). For example, it is interesting to note that, in general, substitutions at the conserved Asp residue have a more profound effect on the step leading to intermediate formation compared to its decomposition (Wu and Zhang, 1996). This is consistent with a dissociative transition state: more help is needed to facilitate the departure of the leaving group in the phosphoenzyme formation step (JZQ) than to activate nucleophilic water in the phosphoenzyme hydrolysis step (^3). Because the transition state for the E - P dephosphorylation step is also dissociative, charge stabilization on the leaving thiolate will be important since P-S bond breaking is substantial in the transition state (Zhang et al, 1995c; Zhao and Zhang, 1996). This charge stabilization can be effected by a hydrogen bond between the sulfur atom of the active site Cys residue and the conserved hydroxyl group in the PTPase signature motif. The loss of such an interaction would make the phosphocysteine more resistant to breakdown. Thus, a more dramatic decrease in the rate of E - P hydrolysis is observed when the hydroxyl group is eliminated (Zhang et al, 1995c; Denu and Dixon, 1995; Zhao and Zhang, 1996). In essence, the conserved hydroxyl group in the PTPase signature motif provides the driving force for the thiophosphate enzyme intermediate to go through a dissociative pathway. Because there is very little bond formation between the phosphorus and the attacking nucleophile in a dissociative transition state, minimal activation is

46

ZHONG-YIN ZHANG

required for the nucleophilic attack in the E - P formation step. Thus, it is understandable that ehmination of the hydroxyl group PTPase signature motif has a minimal effect on the phosphorylation step, because little nucleophilic activation is needed for the reaction. One would also predict that elimination of the active site hydroxyl group would make the dephosphorylation reaction less dissociative. Indeed, this has been experimentally verified: the Br0nsted plot of the second-order rate constant for the alcoholysis (k^ of E - P as a function of the basicity of the alcohol in the Stpl/S18A-catalyzed reaction is shown in Fig. 8, which gives a slope of j8nu: = 0.26. In fact, it is noteworthy that the jSnu: value for the SISA-catalyzed reaction is close to the values for reactions of oxyanions with phosphate triesters, which range from 0.30 to 0.48, depending on the leaving groups (Khan and Kirby, 1970). Because phosphate triesters react with associative transition states (Benkovic and Schray, 1978; Cleland and Hengge, 1995), it is thus conceivable that the transition state for the S18A-catalyzed reaction contains substantial associative character. G. Mechanism for PTPase-Catalyzed Phosphate Monoester Hydrolysis

The structure of the phosphate-binding loop, the invariant Cys and Arg residues, and the essential Ser/Thr residue in the Yersinia PTPase, PTPIB, the low Mr phosphatase, and the VHR have been conserved from bacteria to mammals (Barford et al., 1994; Stuckey et aL, 1994; Su et al, 1994; M. Zhang et al, 1994; Yuvaniyama et al, 1996). It is likely that PTPases, dual-specific phosphatases, and low M^ phosphatases utilize similar structural features, the phosphate-binding loop and an essential Asp residue, within their active sites and employ a common strategy for phosphate monoester hydrolysis. In the common catalytic strategy (Fig. 9), the Cys residue acts as the nucleophile to attack the phosphate ester, forming a thiophosphate intermediate, and the invariant Arg residue plays a role in both substrate recognition and transition state stabilization. The step leading to cysteinylphosphate formation is facilitated by the protonation of the ester oxygen atom in the leaving group, which is accomplished by the conserved Asp residue acting as a general acid. This is required for stabilization of the trigonal bipyramidal transition state toward loss of the oxygen in the leaving group rather than the sulfur of the cysteinyl nucleophile. After formation of the phosphoenzyme intermediate, the dephosphorylation event would occur by attack of water that approaches from the just vacated leaving group side on the phosphoenzyme intermediate, with subsequent release of inorganic phosphate. The same Asp residue functions

47

PROTEIN-TYROSINE PHOSPHATASES

Step 2

PTPase +

HO, ^o

FIG. 9. A unified chemical mechanism for the reaction catalyzed by the PTPases, the dual-specificity phosphatases, and the low Mj. phosphatases.

as a general base in the second step by activating a water molecule for the hydrolysis of the phosphoenzyme intermediate. As part of the phosphate-binding loop, the conserved Thr/Ser residue immediately following the invariant Arg residue facilitates the breakdown of the phosphoenzyme intermediate. It is particularly interesting that nature appears to have utilized this strategy on more than one occasion to hydrolyze phosphate monoesters (e.g., PTPases and low Mr phospha-

48

ZHONG-YIN ZHANG

tase) and that this strategy differs markedly from those employed by alkaline, acid, or Ser/Thr protein phosphatases. IV. Substrate Specificity A. An Overview Although considerable progress has been made in the identification and characterization of new PTPases, relatively little is known about the physiological function of these enzymes. Further understanding of the specific functional roles of PTPases in cellular signaling requires detailed investigation of PTPase substrate specificity. Such knowledge is also crucial for the design and development of novel PTPase inhibitors containing elaborate functionality. A central question in the field is how PTPases distinguish the diversity of substrates they encounter in the cell. There have been relatively few biochemical analyses of the mechanisms that govern PTPase substrate specificity. Enzyme activities of PTPases with artificial substrates are at least three orders of magnitude higher than those of PTKs (Fischer et ah, 1991; Tonks et al.j 1988). In addition, it is clear from the existence of multiple PTPases that more than one PTPase may reside in a single cell. This implies that substrate specificity is likely to be controlled by factors in addition to the amino acid sequence surrounding the site of dephosphorylation. Current data suggest that despite the moderate PTPase substrate specificity observed in vitro, the recognition of PTPases with their physiological phosphorylated substrates in intact cells is rather specific. Evidence suggests that both catalytic and noncatalytic regions of PTPases are important for phosphotyrosyl substrate recognition. In some cases, noncatalytic domains localize the PTPases to specific intracellular compartments in which the effective local concentration of a substrate is high. For example, the COOH-terminal extension of the PTPIB tyrosine phosphatase has been shown to be necessary and sufficient for targeting the enzyme to the cytoplasmic side of the endoplasmic reticulum (Frangioni et al., 1992). The presence of various targeting domains in PTPases also provides a mechanism for directing PTPases to particular subcellular locations. This may be important in both defining and restricting the substrate specificity of the PTPases. In other cases, noncatalytic segments of PTPases play a role in modulating enzyme activity in an allosteric fashion. Occupancy of both SH2 domains of the PTPase SHPTP2 (also known as Syp, PTPID, or PTP-2C) stimulates phosphatase activity (Sugimoto et al, 1994; Pei et al, 1994; Pluskey et al, 1995). Thus, cellular SH-PTP2 activity and specificity are regulated by the interaction of the enzyme with phosphotyrosine-containing proteins

PROTEIN-TYROSINE PHOSPHATASES

49

that bind to its SH2 domains. Other factors which are also hkely to be important in regulating PTPase activity include ligand binding and covalent modifications. Potential ligands for the extracellular domains of several receptor-like PTPases have been identified (Brady-Kalnay et al, 1993; Gebbink et al, 1993; Sap et al, 1994; Poles et al, 1995). However, it has not been demonstrated Whether ligand binding can cause significant change in intrinsic PTPase activity. Similarly, covalent modification (i.e., phosphorylation) has been demonstrated with the PTPases (Feng et al, 1993; Vogel e^ al, 1993; Flint et al, 1993), although the change in phosphorylation state is often accompanied only by small changes in PTPase activity. However, in addition to targeting and localization domains, PTPase substrate specificity must be mediated by intrinsic substrate specificities of the active site, as well as by structural features in the vicinity of the phosphorylated tyrosine residue. B. Amino Acid Sequence Specificity

Using synthetic phosphotyrosine-containing peptides that correspond to natural phosphorylation sites in proteins, several groups have demonstrated that PTPases display a range ofk^JKr^ values for these relatively short peptide substrates (Madden et al., 1991; Fallen et al., 1991; Cho et al, 1991, 1993; Ramachandran et al, 1992; Zhang et al, 1993a,b, 1994b; Hippen et al, 1993; Ruzzene et al, 1993; Harder et al, 1994; Dechert et al, 1995). In fact, the k^JK^ values for some of the peptide substrates approach the efficiency limited by diffusion events, suggesting that short optimal phosphopeptide sequences may contain all the information that is needed for in vivo PTPase recognition (Zhang et al, 1993a,b). Although phosphotyrosine itself binds to PTPases, it does so very weakly. It has been shown that amino acid residues fianking the pTyr moiety are important for efficient PTPase catalysis (Table I) (Zhang et

TABLE I Yersinia P T P A S E SUBSTRATE SEQUENCE SPECIFICITY

Substrate

^eat (s"^)

K^ (fiM)

pNPP pTyr DADEpYLIPQQG DADEpY-NH2 DADEpYL

89 164 1310 1400 1380

1700 9580 59 291 100

kJK^

ifjiM-' s'')

0.052 0.017 22 4.8 14

50

ZHONG-YIN ZHANG

al., 1993a,b, 1994b). For example, the Yersinia PTPase displays a k^J K^ value for the undecapeptide, DADEpYLIPQQG, t h a t is 1300-fold higher t h a n t h a t of pTyr itself (Table I). However, it is also clear t h a t the presence of phosphotyrosine within the peptides is crucial for substrate recognition because unphosphorylated peptides do not bind to the phosphatase (Zhang et al., 1993a; Ruzzene et al., 1993). It has also been shown t h a t the replacement of phosphoserine for phosphotyrosine within a suitable peptide substrate gives rise to a totally inert derivative (Ruzzene et al, 1993). The ability of FTPases to distinguish phosphotyrosine-containing proteins from those with phosphoserine or phosphothreonine may result from the depth of the catalytic "pocket" since the smaller phosphoserine and phosphothreonine side chains would not reach to the phosphate binding pocket (Barford et al., 1994; Stuckey et al., 1994). Because neither phosphotyrosine nor the peptide alone is sufficient to confer specificity and high affinity binding, the recognition of a phosphotyrosine-containing peptide must involve recognition of both phosphotyrosine and residues within the peptide backbone. The contribution of individual amino acid side chains to binding and catalysis was ascertained utilizing a strategy in which each amino acid within a specific peptide was sequentially substituted for by an Ala residue (Ala scan) (Zhang et al., 1993b). It was found t h a t a number of FTPases prefer acidic residues NH2 terminal to phosphotyrosine (Zhang et al., 1993a,b; Hippen et al., 1993; Ruzzene et al., 1993). It was further revealed t h a t a minimum of six amino acid residues are required for the most efficient FTFase binding and catalysis (Zhang et al., 1994b). These include pTyr, four amino acid residues NH2-terminal and one amino acid residue COOH-terminal to the pTyr. Indeed, the hexapeptide Asp-Ala-Asp-Glu-pTyr-Leu-NH2 is an excellent substrate for both F T F l and the Yersinia FTFase t h a t exhibits a kJK^ of 2.24 X 10^ M'^ s-i and 1.41 X 10^ M'^ s ' ^ respectively (Table I; Zhang et al, 1994b). It is noteworthy t h a t SH2 domains also recognize linear sequences surrounding phosphorylated tyrosines. Unlike FTFases, the binding specificity of SH2 domains appears to be dictated by amino acids immediately COOH-terminal to phosphotyrosine. The crystal structure of the src SH2 domain (Waksman et aL, 1993) complexed with a phosphotyrosine-containing peptide reveals t h a t only five to seven residues make direct contact with the protein, with most extensive interactions centered on phosphotyrosine, the + 1 and the + 3 positions. The crystal structures of FTFases (Barford et aL, 1994; Stuckey et al., 1994) reveal t h a t residues around the phosphotyrosine can be positioned to interact with residues on the surface surrounding the catalytic cleft. Several positively charged residues (R228, R295, R404, R409,

PROTEIN-TYROSINE PHOSPHATASES

51

R437, and K447 in Yersinia PTPase, and K36, R45, R47, K116, K120, R221, and R254 in PTPIB) at the opening of the catalytic pocket may also contribute to a positive electrostatic field that could explain the preference of the enzymes for acidic residues that are commonly found adjacent to physiological sites of tyrosine phosphorylation. The crystal structure of the active site Cys-215 to the Ser mutant of PTPIB complexed with Asp-Ala-Asp-Glu-pTyr-Leu-NH2 reveals specific interactions between acidic side chains of the substrate and basic residues of the enzyme (Jia et al, 1995). These results suggest that the sequence surrounding the pTyr residue plays a key role in determining its recognition by PTPases. Interestingly, kinetic studies have also demonstrated that PTPl can hydrolyze a variety of peptide substrates differing in sequence and length with almost equal k^JK^ values (Zhang et al, 1993a,b, 1994b). For example, the peptide substrates Neu(546-556) (Asp-Asn-Leu-TyrpTyr-Trp-Asp-Gln-Asn-Ser-Ser) and p60src(523-531) (Thr-Glu-ProGln-pTyr-Gln-Pro-Gly-Glu) display kinetic parameters similar to those of EGFR(988-998) (Asp-Ala-Asp-Glu-pTyr-Leu-Ile-Pro-Gln-Gln-Gly) (Table II). These results cannot be easily explained by the observed interactions in the crystal structure of PTPIB complexed with AspAla-Asp-Glu-pTyr-Leu-NH2 and suggest that additional determinants for peptide substrate recognition by PTPl must exist. In order to investigate further the molecular mechanism of substrate recognition, photoaffinity labeling experiments on the native form of PTPl have been carried out to identify region(s) of the enzyme involved in peptide substrate binding (Zhang et al., 1996). The photoactive probe used was a phosphotyrosyl peptide derived from EGFR(988-998) in which the Glu residue immediately NH2-terminal to the phosphotyrosine was replaced with the amino acid p-benzoylphenylalanine (Bpa)

TABLE II KINETIC CONSTANTS FOR HYDROLYSIS OF PHOSPHOTYROSINE AND PHOSPHOTYROSINE-CONTAINING PEPTIDES BY P T P l

Substrate

^cat (S ^)

pY (phosphotyrosine) EGFR(988-998), DADEpYLIPQQG DADEpYL-NH2 Neu(546-556), DNLYpYWDQNSS p60src(523-531), TEPQpYQPGE DADBpapYLIPQQG

44.1 75.7 71.8 55.7 78.9 69.3

K^ (fJiM) 4930 2.63 3.20 3.40 6.16 3.30

10-^ X ^eat/i^m (M-l S-l) 8.94 X 10-^ 2.88 2.24 1.64 1.28 2.10

52

ZHONG-YIN ZHANG

to generate DADBpapYLIPQQG. This photoactive amino acid, hke other benzophenone-type labels, will cross-link efficiently to a wide range of binding sites in proteins (Williams and Coleman, 1982; Miller, 1991; Dorman and Prestwich, 1994). Kinetic parameters /^cat and if^ for PTPl-catalyzed hydrolysis of DADBpapYLIPQQG are similar to those for the peptide EGFR(988-998) (Table II) and place it among the best substrates for PTPl. The Bpa-containing peptide cross-linked to a helical region in the NH2 terminus of PTPl in two separate experiments. The a helix modified by the photoaffinity label [designated a!2J in Barford et al. (1994)] lies near the active site cleft of PTPl (Fig. 10). However, this helix is situated on the opposite side of the cleft from Arg-47, the residue in the Cys215 ^^ Ser mutant of PTPIB that makes contact with the bound peptide Asp-Ala-Asp-Glu-pTyr-Leu-NH2 at the - 1 and - 2 positions in the cocrystal structure (Jia et al,, 1995). No additional interactions between helix a2' and the bound peptide are present in the cocrystal structure. Photo-cross-Unking of DADBPapYLIPQQG to helix a2' implies that the peptide is bound to the enzyme active site in a different conformation than that observed for Asp-Ala-Asp-Glu-pTyr-Leu-amide in the crystal structure. Because the photoaffinity label-containing phosphopeptide acts as a good substrate for the enzyme, it must consequently

Helix a2'

FIG. 10. Crystal structure of PTPIB complexed with tungstate. A ribbon diagram of the PTPl structure (Barford et al, 1994) was created using the program MOLSCRIPT (Krauhs, 1991). Tungstate ion is shown in CPK format. The positions of residues Arg-47 and ne-23 (in heHx a2') are indicated. From Z.-Y. Zhang et al, J. Biol Chem. 271,5386-5392 (1996). Copyright 1996 The American Society for Biochemistry and Molecular Biology.

PROTEIN-TYROSINE PHOSPHATASES

53

be bound in the active site in a conformation that is compatible with PTPl catalytic function. Replacement of the negative charge at position - 1 of Asp-Ala-Asp-Glu-pTyr-Leu-NH2 by a bulky hydrophobic Bpa residue may in fact reduce the interaction with the side chain of Arg-47 to such an extent that the mode of binding seen in the crystal structure is no longer energetically favored. Thus, interaction with helix a2' may be important in substrates of PTPl which do not contain an acidic residue NH2-terminal to tyrosine. Although rat PTPl and its human homolog PTPIB are localized to the cytoplasmic side of endoplasmic reticulum (Grangioni et ah, 1992; Woodford-Thomas et ah, 1992), which may provide a level of regulation, in vitro they are not very specific PTPases because they dephosphorylate a wide variety of substrates. As shown in the three-dimensional structure of the catalytic domain human PTPIB (Barford et al., 1994), the protein surface surrounding the catalytic cleft is relatively open and consists of a number of depressions and protrusions. This may allow numerous modes of peptide recognition and is consistent with the ability of PTPl to hydrolyze a wide variety of phosphotyrosinecontaining substrates with nearly equal efficiency. It is reasonable to suggest, based on both kinetic and structural studies, that binding interactions between PTPl and the side chain of tyrosine, the phosphate group, the main-chain nitrogens of phosphotyrosine and the +1 residue are conserved for all peptide substrates. The orientations of individual peptide/protein substrates may be different from each other and may be dictated by specific interactions between amino acid side chains in the vicinity of phosphotyrosine and residues near the enzyme active site cleft. These observations have implications for the design and development of PTPase inhibitors; they indicate that a systematic investigation of amino acid specificity at sites in close proximity to the phosphotyrosine may reveal sequences that are preferentially recognized by PTPases. C. Active Site Substrate Specificity

It is generally accepted that PTPases exhibit strict substrate specificity toward phosphotyrosine-containing proteins/peptides (Sparks and Brautigan, 1985; Tonks et al, 1988; Guan and Dixon, 1990). Studies using synthetic phosphopeptides have demonstrated that PTPases display amino acid sequence sensitivity surrounding the phosphotyrosine. Little is known about the substrate specificity inherent within the PTPase active site, namely, the molecular features that enable PTPases to favor aryl over alkyl phosphates. This is due largely to our incomplete understanding of the scope and limitation of the active site substrate

54

ZHONG-YIN ZHANG

specificity of PTPases, namely, the range of molecular moieties that can be readily accommodated and processed by the catalytic apparatus of this family of enzymes. It has been demonstrated that in addition to aryl phosphates, both the Yersinia PTPase and the rat PTPl can dephosphorylate alkyl phosphates including pyridoxal phosphate, glycerophosphate, phosphoserine, and phosphothreonine (Zhang, 1995b). Furthermore, it has been found that the human dual-specificity phosphatase, VHR, exhibits a rather distinct active site specificity from those of the tyrosine-specific PTPases, catalyzing the hydrolysis of aromatic as well as aliphatic phosphate monoesters with similar efficiency (Zhang et aL, 1995b). These observations provide exciting new opportunities for mechanistic investigations as well as PTPase inhibitor design, since a new dimension of aliphatic functionalities, in addition to the aromatic moieties, can can be incorporated into a peptide template. A chemical strategy has been devised that fuses peptidic and nonpeptidic components into compounds that have proven to be extraordinarily useful in probing the active site specificity of protein kinases (Kwon et al., 1993; Lee et al., 1994). With the use of this strategy, a structurally diverse array of phosphorylated peptide-aminoalcohol fusion compounds (Glu-Glu-Glu-Glu-NH-R-OPOi) has been constructed. These compounds have been used to assess the active site substrate specificity of the Yersinia PTPase (Dunn et al., 1996). In contrast to the currently held belief that pTyr is absolutely essential for PTPase recognition of protein- and peptide-based substrates, the Yersinia PTPase was found to catalyze the hydrolysis of a wide variety of both aromatic and aliphatic phosphates fused to an active site-directed peptide (Tables III and IV). Yersinia PTPase activity is sensitive to stereochemistry at the acarbon of the pTyr residue. (Glu)4-p(L)Tyr-NH2 (1) exhibits a 10-fold larger kcJKj^ value than (Glu)4-p(D)Tyr-NH2 (2). Surprisingly, the achiral phosphoi^yramme-substituted peptide 3, which lacks the a-carboxamide moiety present in pTyr, is a 5.6-fold more efficient substrate (in terms of KJK^ than its naturally occurring chiral counterpart 1 (Table III, structures 3-5) The peptide-free derivative 4 behaves as a typical aryl phosphate, with ^cat (315 s~^) and Xm (3.2 mM) values that are nearly identical to those exhibited by p-nitrophenyl phosphate [^cat (345 s-i) and i^^ (2.6 mM)] (Zhang et al, 1994c). These results reveal two significant facets of the substrate specificity of the Yersinia PTPase. First, in the absence of amino acid residues on the carboxy-terminal side of pTyr, substituents at the a position of the phosphoresidue interfere with PTPase activity. Second, a peptide moiety can dramatically enhance the substrate efficacy of a simple aromatic phosphate such as

PROTEIN-TYROSINE PHOSPHATASES

55

TABLE III Yersinia ACTIVE SITE AKOMATIC/BENZYLIC PHOSPHATE SUBSTRATE SPECIFICITY

Structure number

Substrate

^cat (1 X 10^ s-^)

KJ^mM)

1.2 ± 0.1

0.05 ± 0.02

22 ± 7

4

0.32 ± 0.03

3.2 ± 0.5

0.098 ± 0.016

5

0.50 ± 0.06

0.42 ± 0.17

1.2 ± 0.5

KJK,^ {^iM-^ s-^)

0-P03^-

(Glu)4-HN 0-P03^-

•"HaN

0-P0.2" (Glu)4-HN

phosphotyramine (a 230-fold enhancement in k^JK^ of 3 versus 4). In addition to the influence exerted by stereochemistry, the distance between the peptide backbone and the aromatic phosphate moiety was also found to control the efficacy of PTPase-catalyzed hydrolysis. When the distance between the peptide unit and the aryl phosphate group is shortened by a single methylene unit (i.e., compound 5), a 19-fold reduction in k^JK^ occurs relative to that observed with 3. Barford proposed that PTPase specificity for pTyr-containing peptides probably results from the depth of the active site cleft since the smaller phosphoserine and phosphothreonine side chains should be unable to reach the phosphate binding site (Barford et aL, 1994). Because simple alkyl phosphates can be processed by PTPases (although not as efficiently as aryl phosphates) (Zhang, 1995b), and because the presence of a peptide enhances substrate reactivity, a peptide-linked alkyl phosphate should be expected to make a reasonable PTPase substrate. This should especially be the case if the distance between the peptide backbone and the phosphate moiety corresponds to that present in a peptide-linked phosphorylated Tyr. Such an optimal distance should be approximately six CH2 units. A series of peptide-based phosphorylated aliphatic alcohols (Table IV, structures 6-12) were prepared in order to explore the validity of this supposition. The h^JK^ for the Yersinia PTPase-catalyzed hydrolysis of (Glu)4-NH-(CH2)2-OPOi" (6) is 0.20 mM~^ s"^ (Table IV). Although this value is significantly lower than the value of all of the aromatic phosphates illustrated in Table

56

ZHONG-YIN ZHANG TABLE IV Yersinia ACTIVE SITE ALIPHATIC PHOSPHATE SUBSTRATE SPECIFICITY

Substrate

Structure number

(Glu)4-NH-(CH2)2-OP032(Glu)4-NH-(CH2)3-OP032(Glu)4-NH-(CH2)4-OP032(Glu)4-NH-(CH2)5-OP032(Glu)4-NH-(CH2)6-OP032(Glu)4-NH-(CH2)7-OP032(Glu)4-NH-(CH2)8-OP032-

6 7 8 9 10 11 12

^cat (S ^)

0.32 3.2 4.2 3.6 11 25 2.9

± ± ± ± ± ± ±

0.03 0.3 0.5 0.2 2 1 0.4

K^imM) 1.6 4.1 4.4 1.9 2.1 1.8 3.5

± ± ± ± ± ± ±

0.2 0.4 0.7 0.2 0.6 0.1 0.8

KJK^ 0.20 0.78 0.95 1.9 5.2 14 0.83

(mM-1 s-i) ± ± ± ± ± ± ±

0.03 0.11 0.20 0.2 1.8 1 0.22

I, peptide 6 is a considerably better substrate than its nonpeptidic counterpart, 0-phosphorylethanolamine (Zhang, 1995b). It was estimated that compound 6, which lacks the positively charged j8-substituent and contains an appended active site-directed peptide, is hydrolyzed approximately 770-fold more efficiently than 0-phosphorylethanolamine (Dunn et al., 1996), thereby demonstrating that interactions between substrate and enzyme, which are removed from the site of bond cleavage and formation, can be beneficial for catalysis. As the number of methylene groups n in (Glu)4-NH-(CH2)„-OP03 increases from 2 to 5 (compounds 6-9), an overall 10-fold increase in k^JK^ is apparent. However, this improvement accelerates rapidly ain = 6 and peaks at 71 = 7. The 26-fold enhancement in KJKrr, for 10 (compared to 6) and the 70-fold enhancement in substrate efficacy for 11 are due to substantial improvements in the ^cat parameter. However, the catalytic efficiency of compound 12 (n = 8) is more than an order of magnitude less than that of its zi = 7 counterpart. These results are consistent with the hypothesis that the optimal distance between the phosphate moiety and the peptide backbone corresponds to the length of a Tyr side chain. As noted above, PTPases have been shown to utilize such nonpeptidic species as glycerophosphate and pyridoxal phosphate as substrates (Zhang, 1995b; Zhang et al, 1995b). Small aryl phosphates, such as pNPP, serve as resonably efficient PTPase substrates as well (Zhang, 1995a; Zhang et al, 1994c). In general, these species exhibit K^ values in the mM range. By systematic variation of the substitution patterns on the phenyl phosphate framework, several extraordinarily potent low-molecular-weight nonpeptidic PTPase substrates that exhibit Michaelis constants in the low fjuM range for PTPl and the dualspecificity phosphatase VHR have since been identified (Montserat et

PROTEIN-TYEOSINE PHOSPHATASES

57

al., 1996; Chen et al., 1996) and (in several instances) distinguish between PTPl and VHR. A few of these compounds (13-16) are shown in Table V. It is striking that some of these low M^ substrates display kinetic parameters that are comparable or better than those of the best pTyr-containing peptide substrates. Considering their relatively modest structural framework, the substrate efficacy of these compounds is unprecedented. In general, negatively charged substituents para and hydrophobic substituents meta and para to the hydrolyzable phosphate moiety are responsible for enhanced substrate efficacy. Although amino acid residues surrounding the catalytic site are highly conserved among PTPases, even at this relatively early stage in our understanding of the enzymology of PTPases, it has become evident that significant differences exist among PTPases. For example, although phenolphthalein monophosphate 15 serves as a substrate for both the VHR and PTPl enzymes, the K^ exhibited by these enzymes toward this compound differs dramatically (69 ^iM for VHR and 2.2 mM for PTPl). In addition, the K^ values associated with PTPase substrates tend to correlate well with Ki values of structurally analogous competitive inhibitors. It has also been found that phenolphthalein diphosphate 16 displays K^ values of 24 ^M toward VHR, 360 iiM toward PTPl, and 2.43 mM toward PTPa. These results are surprising in view of the structural information on phosphotyrosine recognition gained from the structure of the PTPlB-pTyr complex (Jia et al., 1995). Invariant nonpolar residues in the PTPIB catalytic domain (Tyr-46, Val-49, Phe-182, Gln-262) form the binding site for the phenyl ring of phosphotyrosine. The phosphoryl group in phosphotyrosine is surrounded by residues corresponding to the PTPase signature motif. This suggests that the mechanism for phosphotyrosine recognition among all PTPases is similar. In contrast, the above kinetic results indicate that significant differences exist among these PTPases in terms of their ability to recognize low M^. compounds. These results demonstrate the promise of this approach in identifying unique Mr compounds that are specific for different PTPase isoenzymes. D. Inhibitor Development

Selective and potent PTPase inhibitors may not only serve as useful probes to help define the physiological function of PTPase, but may ultimately constitute a novel, and potentially valuable, family of agents for the treatment of cancers, autoimmune diseases, and diabetes. However, in spite of the potential value of PTPase inhibitors for the study of signal transduction pathways and for therapeutic intervention, relatively little has been reported on the development of such agents. Vana-

P^

^y-"

t^°

^

I

+ 1 +1 O ^ CO (N CO

d d +1 +1 (N CO id ^

o +1 +1

CO lO ^

oa CX) CD

+1 +1

o o X CO

+1 +1

CO CO CO O iH CO

+1 +1

d d

CO rA

00 O

+ 1 +1

o o GO (N

CO r-{

d d +1 +1 a o (N rA

PROTEIN-TYROSINE PHOSPHATASES

59

date is the reagent of choice when studying the biological role of PTPases (Swarup et al., 1982). Because vanadate exerts its inhibitory effect on many phosphatases, specificity is always a major concern. Another compound with considerable utility as a PTPase inhibitor is phenylarsine oxide (Garcia-Morales et al., 1990). Phenylarsine oxide is also known to be a powerful inactivating agent for other enzymes that have spatially close thiols (Brown et al,, 1987), so its effect on PTPases would not be expected to occur in a fairly specific fashion. The toxicity associated with these compounds and their nonspecific nature toward phosphatases make them less attractive reagents for biological studies. Phosphonic acid derivatives are nonhydrolyzable phosphate mimetics that have found wide use in biological systems (Blackburn, 1981; Engel, 1983). They are isosteric with parent phosphates and yet are resistant to the action of phosphatases. For example, benzylphosphonic acid was shown to be a competitive inhibitor for the low M^ PTPase with a Ki value of 4.6 mM (Zhang and Van Etten, 1990). This value is higher than that of inorganic phosphate, which is 2 mM. In order to build tighter-binding inhibitors, additional molecular features have to be incorporated into the phosphonate functionality. Because amino acids flanking the phosphotyrosyl residue (pTyr) contribute to high affinity substrate binding (Zhang et al., 1994b), one current approach toward the design of potent and selective PTPase inhibitors relies on the incorporation of a nonhydrolyzable analog of pTyr into specific optimal phosphopeptide templates. This is because pTyr itself is absolutely essential for PTPase recognition of peptide/protein-based substrates. For example, PTPases do not bind tyrosine-bearing peptides that lack the phosphate or phosphate-mimicking functionality (Zhang et al., 1993a). 0-Methylation of pTyr in a peptide suppresses its ability to bind PTPase (Ruzzene et al., 1993). Furthermore, the recognition pocket for phosphotyrosine represents the dominant driving force for peptide binding since phosphotyrosine contributes about 53% of the peptide solvent-accessible surface area (Jia et al., 1995). This implies that a tyrosine moiety, in conjunction with the negatively charged phosphate group, is crucial for PTPase recognition. Based on these structural considerations, several nonhydrolyzable analogs of pTyr have been prepared and inserted into PTPase-targeted peptides. Phosphonomethylphenylalanine (Pmp) is a phosphonate-based surrogate of pTyr in which the phosphate ester oxygen has been replaced by a methylene unit (Fig. 11). Indeed, Pmp-containing peptides have been shown to be effective, reversible inhibitors of PTPases (Chatterjee et al., 1992; Zhang et al, 1994b). Peptides containing sulfotyrosyl (Liotta

60

ZHONG-YIN ZHANG

COO

O^

coo

coo

"O

Phosphotyrosine

Phosphonomethyl Phenylalanine (Pmp)

Phosphonodifluoromethyl Phenylalanine (F2Pmp)

coo

s=zo Sulfotyrosine

Malonyltyrosine

Cinnamic Acid

FIG. 11. Structures of phosphot3n:'osine and the nonhydrolyzable phosphotjnrosyl analogs.

et al, 1994), malonyl (Kole et al., 1995a), and cinnamic acid derivatives (Moran et aL, 1995) have also been shown to be effective inhibitors of PTPases (Fig. 11). The mechanism-based phosphatase inactivator, 4(fiuoromethyl)phenylphosphate (FMPP) (Myers and Widlanski, 1993; Wang et al,, 1994), is another molecular motif that can be built into a peptide. FMPP is a suicide inhibitor which, on the action of phosphatase, generates a reactive quinone methide in the active site. This species can then be attacked by nucleophiles at the enzyme active site, resulting in the inactivation of the enzyme. Surprisingly, substitutions

PROTEIN-TYROSINE PHOSPHATASES

61

by two fluorines at the methylene position in Pmp (F2Pmp, Fig. 11) (Burke et al., 1993) bring about a three orders of magnitude improvement in the inhibitory potency toward PTPIB (Burke et al., 1994; Chen et al.j 1995). The higher PTPase affinity of the difluorophosphonate moiety was further demonstrated by small non-peptide-based aryl difluorophosphonates which also exhibit good PTPase inhibition (Kole et al, 1995b). Further kinetic analysis suggest that the F2Pmp-containing peptide is superior to the Pmp-containing peptide possibly because the two fluorine atoms in F2Pmp can restore or enhance the hydrogen-bonding interactions normally existing between the phenolic oxygen in pTyr and side chains in the active site of PTPl (Chen et al., 1995). This hypothesis is supported by biochemical and structural data. PTPases utilize a general acid to facilitate the departure of the phenolic leaving group by protonation of the bridging oxygen. Indeed, the structure of PTPIB (with Cys-215 to Ser substitution) complexed with a peptide substrate (Jia et al., 1995) reveals that the phenolic oxygen of pTyr forms a network of hydrogen bonds with the side chain of the corresponding general acid Asp-181 and a buried water molecule. Thus, it is conceivable that the fluorine atoms in the F2Pmp-containing peptide interact directly with Asp-181, the bound water molecule, and/or other active site residues. The average sequence identity among mammalian PTPase catalytic domains is about 40%, and most of the conserved amino acid residues are located in and around the active site where catalysis occurs (Barford et al., 1994; Stuckey et al., 1994). Consequently, it has been assumed that much of what we learn about one PTPase should prove applicable to other PTPase family members as well. In contrast, comparative analysis of the inhibition of PTPl, PTPa, and LAR suggests that there is remarkable variability in the degree of inhibition of PTPases by the F2Pmp peptides (Chen et al, 1995). Thus, Ac-Asp-Ala-Asp-Glu(F2Pmp)-Leu-NH2 is a potent inhibitor for PTPl {K^ = 0.18 ± 0.02 ^M) but a significantly poorer one for LAR {Ki = 376 ± 43 IJLM). Similarly, Ac-Asp-(F2Pmp)-Val-Pro-Met-Leu-NH2 is a potent inhibitor for PTPl {Ki = 0.12 + 0.01 ixM) but a much poorer one for PTPa {K^ = 465 ± 130 fjiM). Although the general fold of the polypeptide backbone is conserved among PTPases, unique features at the active sites of individual PTPases may provide a basis for understanding the differences in substrate/inhibitor specificity. It appears that significant differences exist within the active sites of various PTPases and that selective, tight-binding PTPase inhibitors can be developed.

62

ZHONG-YIN ZHANG

V. Conclusion and Perspective PTPases are encoded by a family of genes that have been shown to play pivotal roles in cell growth and signal transduction. The study of PTPases has increased dramatically over the last few years in terms of discovering new PTPases, identifying their biological functions, and analyzing their mechanism of catalysis and substrate specificity. It would appear that the activity of PTPases will have to be tightly regulated, since multiple PTPases are likely to be present in the same cell. The combination of catal3^ic domains with a variety of noncatal3d:ic and functional motifs will provide additional modes of regulating PTPase substrate specificity. As noted earlier, an important aspect of the regulation of PTPase activity appears to be their localization. By limiting the access of phosphorylated proteins for each PTPase, de facto substrate specificity is created. However, results described in this review demonstrate that substrate specificity of PTPases can also be determined at the active site and at the primary structural level. Systematic variation of peptide substrates in terms of size and amino acid sequence will undoubtedly shed light on the question of substrate specificity in the primary structure context. A combinatorial phosphopeptide library may be useful to facilitate the search for the optimal amino acid sequence for PTPase substrate recognition. Similar approaches have been successfully applied to study protein kinase and SH2 domain specificity (Songyang et al., 1993, 1994). The problem of substrate specificity can also be tackled by studying the structure-function relationship of PTPases using site-directed mutagenesis. A deeper understanding of PTPase substrate specificity will speed up the identification of physiological substrates and the design of specific inhibitors. The findings that appropriately functionalized aromatic phosphates are hydrolyzed by the tyrosine-specific phosphatase PTPl and the dual-specificity phosphatase VHR as efficiently as the very best peptide substrates reported for these enzymes (Montserat et al., 1996; Chen et al., 1996) provide good starting points for the development of low Mj. nonpeptidic PTPase inhibitors. A comparative analysis of the active site specificities of PTPases may reveal key differences in the ability of individual enzymes to tolerate specific structural motifs. Clearly, any observed differences should prove useful in the design of PTPase-specific inhibitors. The latter would be of decided benefit in helping to define the role of PTPases in cellular signaling. Finally, as the three-dimensional structures of the PTPases containing bound substrates/inhibitors become available, they will reveal the intimate details of key molecular interactions between active site residues and inhibitor functionality. These

PROTEIN-TYROSINE PHOSPHATASES

63

studies will set the stage for a detailed structure-function analysis of the mechanism of molecular recognition of PTPases. The structural information will also be utilized to guide site-directed mutagenesis of PTPases in order to ascertain the role of critical residues in substrate recognition and catalysis. As our understanding of the substrate specificity and the factors t h a t control this specificity increases, efforts in the area of rational design of inhibitors t h a t are targeted for specific PTPases should become possible. The existence of a covalent phosphoenzyme intermediate on the kinetic pathway of the PTPase-catalyzed reaction is firmly established. The detailed mechanism which PTPases utilize to facilitate the hydrolysis of phosphate monoesters is beginning to emerge. In addition, the nature of the transition states of the PTPase-catalyzed reaction has been elucidated. Site-directed mutagenesis has proven to be a powerful tool in elucidating the mechanism of catalysis when this approach is coupled with traditional chemical modification studies. The three-dimensional structures of several PTPase catalytic domains are available, which will greatly enhance our understanding of the PTPase catalytic mechanism. Because all PTPases share a catalytic domain with considerable amino acid sequence identity, indepth kinetic and structural analysis of a limited number of PTPases will most likely yield insightful mechanistic information t h a t may be applicable to the rest of the family. The challenge for the future will be (1) to determine the molecular basis for PTPases catalysis, substrate recognition, and regulation of activity and (2) to identify PTPase cellular substrates and establish the role of these enzymes in signal transduction. Such knowledge is highly desirable for understanding the regulation of the function of PTPases in normal and transformed cells. ACKNOWLEDGMENT

This work was supported by a grant from the National Institutes of Health CA69202. REFERENCES

Admiraal, S. J., and Herschlag, D. (1995). Chem. Biol 2, 729-739. Aroca, P., Bottaro, D. P., Ishibashi, T., Aaronson, S. A., and Santos, E. (1995). J. Biol. Chem. 270, 14229-14234. Barford, D., Flint, A. J., and Tonks, N. K. (1994). Science 263, 1397-1404. Barford, D., Jia, Z., and Tonks, N. K. (1995). Nat. Struct. Biol. 2, 1043-1053. Benkovic, S. J., and Schray, K. J. (1978). In "Transition States of Biochemical Processes" (R. D. Gandour and R. L. Schowen, eds.), pp. 493-527. Plenum, New York. Bishop, J. M. (1991). Cell {Cambridge, Mass.) 64, 235-248. Blackburn, G. M. (1981). Chem. Ind. (London), pp. 134-138.

64

ZHONG-YIN ZHANG

Bliska, J. B., Guan, K. L., Dixon, J. E., and Falkow, S. (1991). Proc. Natl Acad. Set U.S.A. 8 8 , 1 1 8 7 - 1 1 9 1 . Bourne, N., and Williams, A. (1984). J. Am. Chem. Soc. 106, 7591-7596. Brady-Kalnay, S. M., Flint, A. J., and Tonks, N. K. (1993). J. Cell Biol. 122, 961-972. Brautigan, D. L. (1992). Biochim. Biophys. Acta 114, 63-77. Brown, S. B., Turner, R. J., Roche, R. S., and Stevenson, K. J. (1987). Biochemistry 26, 863-871. Brown-Shimer, S., Johnson, K. A., Hill, D. E., and Bruskin, A. M. (1992). Cancer Res. 52, 478-482. Bruice, T. C., and Benkovic, S. J. (1966). "Bioorganic Mechanisms," Vol. 2, pp. 1-103. Benjamin, New York. Bunton, C. A., Fendler, E. L., Humeres, E., and Yang, K.-U. (1967). J. Org. Chem. 32, 2806-2811. Burke, T. R., Jr., Smyth, M., Nomizu, M., Otaka, A., and Roller, P. P. (1993). J. Org. Chem. 58, 1336-1340. Burke, T. R., Jr., Kole, H. K , and Roller, P. P. (1994). Biochem. Biophys. Res. Commun. 204, 129-134. Butcher, W. W., and Westheimer, F. H. (1955). J. Am. Chem. Soc. 77, 2420-2424. Camici, G., Manao, G., Cappugi, G., Modesti, A., Stefani, M., and Ramponi, G. (1989). J. Biol. Chem. 264, 2560-2567. Cantley, L. C., Auger, K. R., Carpenter, C., Duckworth, B., Graziani, A., Kapeller, R., and Soltoff, S. (1991). Cell (Cambridge, Mass.) 64, 281-302. Charbonneau, H., and Tonks, N. K. (1992). Annu. Rev. Cell Biol. 8, 463-493. Chatterjee, S., Goldstein, B. J., Csermely, P., and Shoelson, S. E. (1992). Pept, Proc. Am. Pept. Symp., 12th, pp. 553-555. Chen, L., Wu, L., Otaka, A., Smyth, M. S., Roller, P. P., Burke, T. R., den Hertog, J., and Zhang, Z.-Y. (1995). Biochem. Biophys. Res. Commun. 216, 976-984. Chen, L., Montserat, J., Lawrence, D. S., andZhang, Z.-Y. (1996). Biochemistry 35,9349-9354. Cho, H., Ramer, S. E., Itoh, M., Winkler, D. G., Kitas, E., Bannwarth, W., Burn, P., Saito, H., and Walsh, C. T. (1991). Biochemistry 30, 6210-6216. Cho, H., Krishnaraj, R., Kitas, E., Bannwarth, W., Walsh, C. T., and Anderson, K. S. (1992). J. Am. Chem. Soc. 114, 7296-7298. Cho, H., Krishnaraj, R., Itoh, M., Kitas, E., Bannwarth, W., Saito, H., and Walsh, C. T. (1993). Protein Sci. 2, 977-984. Cirri, P., Chiarugi, P., Camici, G., Manao, G., Raugei, G., Cappugi, G., and Ramponi, G. (1993). Eur. J. Biochem. 214, 647-657. Cleland, W. W., and Hengge, A. C. (1995). FASEB J. 9, 1585-1594. Coleman, J. E. (1992). Annu. Rev. Biophys. Biomol. Struct. 2 1 , 441-483. Cotton, F. A., Hazen, E. E., Jr., Day, V. W., Larsen, S., Norman, J. G., Jr., Wong, S. T. K , and Johnson, K. H. (1973). J. Am. Chem. Soc. 95, 2367-2369. Cox, J. R., and Ramsay, O. B. (1964). Chem. Rev. 64, 317-352. CuUis, P. M. (1987). In "Enzyme Mechanisms" (M. L. Page, ed.), pp. 179-220. Royal Society of Chemistry, London. Dechert, U., Affolter, M., Harder, K. W., Mathews, J., Owen, P., Clark-Lewis, I., Thomas, M. L., Aebersold, R., and Jirik, F. R. (1995). Eur. J. Biochem. 231, 673-681. Degani, C , and Boyer, P. D. (1973). J. Biol. Chem. 248, 8222-8226. Denu, J. M., and Dixon, J. E. (1995). Proc. Natl. Acad. Sci. U.S.A. 92, 5910-5914. Denu, J. M., Zhou, G., Wu, L., Zhao, R., Yuvaniyama, J., Saper, M. A., and Dixon, J. E. (1995a). J. Biol. Chem. 270, 3796-3803. Denu, J. M., Zhou, G., Guo, Y., and Dixon, J. E. (1995b). Biochemistry 34, 3396-3403.

PROTEIN-TYKOSINE PHOSPHATASES

65

Dittmer, D. C , Ramsay, 0 . B., and Spalding, R. E. (1963). J. Org. Chem. 28,1273-1278. Dorman, G., and Prestwich, G. D. (1994). Biochemistry 33, 5661-5673. Dumphy, W. G., and Kumagai, A. (1991). Cell {Cambridge, Mass.) 67, 189-196. Dunn, D., Chen, L., Lawrence, D. S., and Zhang, Z.-Y. (1996). J. Biol. Chem. 271,168-173. Engel, R. (1983). In "The Role of Phosphonates in Living Systems" (R. L. Hilderbrand, ed.), pp. 97-138. CRC Press, Boca Raton, FL. Feng, G.-S., Hui, C.-C., and Pawson, T. (1993). Science 259, 1607-1611. Filippini, F., Rossi, V., Marin, O., Trovato, M., Costantino, P., Downey, P. M., and Terzi, M. (1996). Nature (London) 379, 499-500. Fischer, E. H., Charbonneau, H., and Tonks, N. K. (1991). Science 253, 401-406. Flint, A. J., Gebbink, M. F., Franza, B. R., Jr., Hill, D. E., and Tonks, N. K. (1993). EMBO J. 12, 1937-1946. Frangioni, J. V., Beahm, P. H., Shifrin, V., Jost, C. A., and Neel, B. G. (1992). Cell (Cambridge, Mass.) 68, 545-560. Frey, P. A. (1989). Adv. Enzymol. 62, 119-201. Friedman, J. M., Freeman, S., and Knowles, J. R. (1988). J. Am. Chem. Soc. 110,1268-1275. Garcia-Morales, P., Minami, V., Luong, E., Klausner, R. D., and Samelson, L. E. (1990). Proc. Natl. Acad. Sci. U.S.A. 87, 9255-9259. Gautier, J., Soloman, M. J., Booher, R. N., Bazan, J. F., and Kirschner, M. W. (1991). Cell (Cambridge, Mass.) 67, 197-211. Gebbink, M. F., Zondag, G. C., Wubbolts, R. W., Beijersbergen, R. L., Van Etten, L, and Moolenaar, W. H. (1993). J. Biol. Chem. 268, 16101-16104. Gorenstein, D. G., Lee, Y.-G., and Kar, D. (1977). J. Am. Chem. Soc. 99, 2264-2267. Gould, K. L., and Nurse, P. (1989). Nature (London) 342, 39-45. Gregoret, L. M., Rader, S. D., Fletterick, R. J., and Cohen, F. E. (1991). Proteins: Struct, Funct, Gene. 9, 99-107. Guan, K. L., and Dixon, J. E. (1990). Science 249, 553-556. Guan, K. L., and Dixon, J. E. (1991). J. Biol. Chem. 266, 17026-17030. Guan, K. L., Broyles, S., and Dixon, J. E. (1991). Nature (London) 350, 359-361. Harder, K. W., Owen, P., Wong, L. K., Aebersold, R., Clark-Lewis, L, and Jirik, F. R. (1994). Biochem. J. 298, 395-401. Hasset, A., Blattler, W., and Knowles, J. R. (1982). Biochemistry 21, 6335-6340. Hengge, A. C , Edens, W. A., and Elsing, H. (1994). J. Am. Chem. Soc. 116, 5045-5049. Hengge, A. C , Sowa, G., Wu, L., and Zhang, Z.-Y. (1995). Biochemistry 34,13982-13987. Hengge, A. C , Denu, J. M., and Dixon, J. E. (1996). Biochemistry 35, 7084-7092. Herr, E. B., Jr., and Koshland, D. E., Jr. (1957). Biochim. Biophys. Acta 25, 219-220. Herschlag, D., and Jencks, W. P. (1989). J. Am. Chem. Soc. I l l , 7579-7586. Herschlag, D., and Jencks, W. P. (1990). Biochemistry 29, 5172-5179. Hippen, K. L., Jakes, S., Richards, J., Jena, B. P., Beck, B. L., Tabatabai, L. B., and Ingebritsen, T. S. (1993). Biochemistry 32, 12405-12412. Hunter, T. (1987). Cell (Cambridge, Mass.) 50, 823-829. Hunter, T. (1991). In "Methods in Enzymology" (T. Hunter and B. M. Sefton eds.). Vol. 200, pp. 3-37. Academic Press, San Diego, CA. Hunter, T. (1995). Cell (Cambridge, Mass.) 80, 225-236. Ishibashi, T., Bottara, D. P., Chan, A., Miki, T., and Aaronson, S. A. (1992). Proc. Natl. Acad. Sci. U.S.A. 89, 12170-12174. Jencks, W. P. (1980). Ace. Chem. Res. 13, 161-169. Jencks, W. P., and Gilchrist, M. (1968). J. Am. Chem. Soc. 90, 2622-2637. Jencks, W. P., Haber, M. T., Herschlag, D., and Nazaretian, K. L. (1986). J.Am. Chem. Soc. 108, 479-483.

66

ZHONG-YIN ZHANG

Jia, Z., Barford, D., Flint, A. J., and Tonks, N. K. (1995). Science 268, 1754-1758. Johnson, P., Ostergaard, H. L., Wasden, C , and Trowbridge, I. S. (1992). J. Biol. Chem. 267, 8035-8041. Keyse, S. M. (1995). Biochim. Biophys. Acta 1265, 152-160. Khan, S. A., and Kirby, A. J. (1970). J. Chem. Soc. B, pp. 1172-1182. Kim, E. E., and Wyckoff, H. W. (1991). J. Mol. Biol. 218, 449-464. Kirby, A. J., and Jencks, W. P. (1965). J. Am. Chem. Soc. 87, 3209-3216. Kirby, A. J., and Varvoghs, A. G. (1967). J. Am. Chem. Soc. 89, 415-423. Kirby, A. J., and Varvoghs, A. G. (1968). J. Chem. Soc. B, pp. 135-141. KlingmuUer, U., Lorenz, U., Cantley, L. C , Neel, B. G., and Lodish, H. F. (1995). {Cambridge, Mass.) 80, 729-738. Kole, H. K , Akamatsu, M., Ye, B., Yan, X., Barford, D., Roller, P. P., and Burke, T. R., J r . (1995a). Biochem. Biophys. Res. Commun. 209, 817-822. Kole,H.K,Smyth,M.S.,Russ,P.L.,andBurke,T.R.,Jr.(1995b).SJocAem. J. 311,1025-1031. Krauhs, P. J. (1991). J. Appl. Crystallogr. 24, 946-950. Kwon, Y.-G., Mendelow, M., Srinivasan, J., Lee, T. R., Pluskey, S., Salerno, A., and Lawrence, D. S. (1993). J. Biol. Chem. 268, 10713-10716. LaForgia, S., Morse, B., Levy, J., Barnea, G., Ganniz-Weston, L. A., Harris, C. C., Drabkin, H., Patterson, D., Croce, C. M., Schlessinger, J., and Huebner, K. (1991). Proc. Natl. Acad. Sci. U.S.A. 88, 5036-5040. Lee, T. R., Niu, J., and Lawrence, D. S. (1994). Biochemistry 33, 4245-4250. Liotta, A. S., Kole, H. K., Fales, H. M., Roth, J., and Bernier, M. (1994). J. Biol. Chem. 269, 22996-23001. Liu, K., Lemon, B., and Traktman, P. (1995). J. Virol. 69, 7823-7834. Madden, J. A., Bird, M. L, Man, Y., Raven, T., and Myles, D. D. (1991). Anal. Biochem. 199, 210-215. Mildvan, A. S., and Fry, D. C. (1987). Adv. Enzymol. 59, 241-313. Millar, J. B. A., and RusseU, P. (1992). Cell (Cambridge, Mass.) 68, 407-410. Miller, W. T. (1991). In "Methods in Enzymology" (T. Hunter and B. M. Sefton, eds.), Vol. 200, pp. 500-508. Milstien, S., and Fife, T. H. (1967). J. Am. Chem. Soc. 89, 5820-5826. Mondesert, O., Moreno, S., and Russell, P. (1994). J. Biol. Chem. 269, 27996-27999. Montserat, J., Chen, L., Lawrence, D. S., and Zhang, Z.-Y. (1996). J. Biol. Chem. 271, 7868-7872. Moran, E. J., Sarshar, S., Cargill, J. F., Shahbaz, M. M., Lio, A., Mjalh, A. M. M., and Armstrong, R. W. (1995). J. Am. Chem. Soc. 117, 10787-10788. Myers, J. K , and Widlanski, T. S. (1993). Science 262, 1451-1453. Noguchi, T., Matozaki, T., Horita, K , Fujioka, Y., and Kasuga, M. (1994). Mol. Cell. Biol. 14, 6674-6682. O'Leary, M. H., and Marlier, J. F. (1979). J. Am. Chem. Soc. 101, 3300-3306. Fallen, C. J., Lai, D. S. Y., Chia, H. P., Boulet, L, and Tong, P. H. (1991). Biochem. J. 276, 315-323. Pas, H. H., Meyer, G. H., Kruizinga, W. H., Tamminga, K. S., van Weeghel, R. P., and Robillard, G. T. (1991). J. Biol. Chem. 266, 6690-6692. Pei, D., Lorenz, U., KlingmuUer, U., Neel, B. G., and Walsh, C. T. (1994). Biochemistry 33, 15483-15493. Peles, E., Nativ, M., Campbell, P. L., Sakurai, T., Martinez, R., Lev, S., Clary, D. 0., Schilling, J., Barnea, G., Plowman, G. D., Grumet, M., and Schlessinger, J. (1995). Cell (Cambridge, Mass.) 82, 251-260. Perkins, L. A., Larsen, L, and Perrimon, N (1992). Cell (Cambridge, Mass.) 7 0 , 2 2 5 - 2 3 6 . Pigiet, v., and Conley, R. R. (1978). J. Biol. Chem. 253, 1910-1920.

PROTEIN-TYROSINE PHOSPHATASES

67

Pingel, J. T., and Thomas, M. L. (1989). Cell {Cambridge, Mass.) 58, 1055-1065. Pluskey, S., Wandless, T. J., Walsh, C. T., and Shoelson, S. E. (1995). J. Biol Chem. 270, 2897-2900. Poon, R. Y. C , and Hunter, T. (1995). Science 270, 9 0 - 9 3 . Pot, D. A., and Dixon, J. E. (1992). J. Biol Chem. 267, 140-143. Pot, D. A., Woodford, T. A., Remboutsika, E., Haun, R. S., and Dixon, J. E. (1991). J. Biol. Chem. 266, 19688-19696. Ramachandran, C , Aebersold, R., Tonks, N. K., and Pot, D. A. (1992). Biochemistry 31, 4232-4238. Ruzzene, M., Donella-Deana, A., Marin, O., Perich, J. W., Ruzza, P., Borin, G., Calderan, A., and Pinna, L. A. (1993). Eur. J. Biochem. 211, 289-295. Sap, J., Jiang, Y.-P., Friedlander, D., Grumet, M., and Schlessinger, J. (1994). Mol. Cell. Biol. 14, 19-25. Sato, T., Irie, S., Kitada, S., and Reed, J. G. (1995). Science 268, 411-415. Schwartz, J. H., and Lipmann, F. (1961). Proc. Natl. Acad. Sci. U.S.A. 47, 1996-2005. Shultz, L. D., Schweitzer, P. A., Rajan, T. V., Yi, T., Ihle, J. N., Matthews, R. J., Thomas, M. L., and Beier, D. R. (1993). Cell (Cambridge, Mass.) 73, 1445-1454. Skoog, M. T., and Jencks, W. P. (1984). J. Am. Chem. Soc. 106, 7597-7606. Songyang, Z., Shoelson, S. E., Chaudhuri, M., Gish, G., Pawson, T. Haser, W., King, F., Roberts, T., Ratnofsky, S., Lechleider, R. L., Neel, B. G., Birge, R. B., Fajardo, J. E., Chou, M. M., Hanafusa, H., Schaffhausen, B., and Cantley, L. C. (1993). Cell (Cambridge, Mass.) 72, 767-778. Songyang, Z., Blechner, S., Hoagland, H., Hoekstra, M. F., Piwnica-Worms, H., and Cantley, L. C. (1994). Curr. Biol. 4, 973-982. Sparks, J. W., and Brautigan, D. L. (1985). J. Biol. Chem. 260, 2042-2045. Streuh, M., Krueger, N. X., Thai, T., Tang, M., and Saito, H. (1990). EMBO J. 9, 2 3 9 9 2407. Stuckey, J. A., Fauman, E. B., Schubert, H. L., Zhang, Z.-Y., Dixon, J. E., and Saper, M. A. (1994). Nature (London) 370, 571-575. Su, X.-D., Taddei, N., Stefani, M., Ramponi, G., and Nordlund, P. (1994). Nature (London) 370, 575-578. Sugimoto, S., Wandless, T., Shoelson, S. E., Neel, B. G., and Walsh, C. T. (1994). J. Biol. Chem. 269, 13614-13622. Sun, H., and Tonks, N. K. (1994). Trends Biochem. Sci. 19, 480-485. Sun, H., Charles, C. H., Lau, L. F., and Tonks, N. K. (1993). Cell (Cambridge, Mass.) 75, 487-493. Swamp, G., Spreeg, K. V., Cohen, S., and Garbers, D. L. (1982). J. Biol. Chem. 257,7298-7301. Taddei, N., Chiarugi, P., Cirri, P., Fiaschi, T., Stefani, M., Camici, G., Giovanni, R., and Ramponi, G. (1994). FEBS Lett. 350, 328-332. Thatcher, G. R. J., and Kluger, R. (1989). Adv. Phys. Org. Chem. 25, 99-265. Tonks, N. K , Diltz, C. D., and Fischer, E. H. (1988). J. Biol. Chem. 263, 6731-6737. Van Etten, R. L. (1982). Ann. N Y. Acad. Sci. 390, 2 7 - 5 1 . Van Etten, R. L., and Hickey, M. E. (1977). Arch. Biochem. Biophys. 183, 250-259. Vogel, W., Lammers, R., Huang, J., and Ullrich, A. (1993). Science 259, 1611-1614. Waksman, G., Shoelson, S. E., Pant, N., Cowburn, D., and Kuriyan, J. (1993). Cell (Cambridge, Mass.) 72, 779-790. Walsh, E. O'F. (1952). Nature (London) 169, 546. Walton, K. M., and Dixon, J. E. (1993). Annu. Rev. Biochem. 62, 101-120. Wang, Q., Dechert, U., Jirik, F., and Withers, S. G. (1994). Biochem. Biophys. Res. Commun. 200, 577-583.

68

ZHONG-YIN ZHANG

Wary, K. K , Lou, Z., Buchberg, A. M., Siracusa, L. D., Druck, T., LaForgia, S., and Huebner, K (1993). Cancer Res. 53, 1498-1502. Weiss, P. M., Knight, W. B., and Cleland, W. W. (1986). J. Am. Chem. Soc. 108,2761-2762. Westheimer, F. H. (1981). Chem. Rev. 81, 313-326. Westheimer, F. H. (1987). Science 235, 1173. Williams, N., and Coleman, P. S. (1982). J. Biol. Chem. 257, 2834-2841. Wo, Y.-Y. P., Zhou, M.-M., Stevis, P., Davis, J. P., Zhang, Z.-Y., and Van Etten, R. L. (1992). Biochemistry 3 1 , 1712-1721. Woodford-Thomas, T. A., Rhodes, J. D., and Dixon, J. E. (1992). J. Cell Biol. 117,401-404. Wu, L., and Zhang, Z.-Y., (1996). Biochemistry 35, 5426-5434. Yarden, Y., and Ullrich, A. (1988). Annu. Rev. Biochem. 57, 443-478. Yuvaniyama, J., Denu, J. M., Dixon, J. E., and Saper, M. A. (1996). Science 272, 1 3 2 8 1331. Zhang, M., Van Etten, R. L., and Stauffacher, C. V. (1994). Biochemistry 33,11097-11105. Zhang, Z., Harms, E., and Van Etten, R. L. (1994). J. Biol. Chem. 269, 25947-25950. Zhang, Z.-Y. (1990). Ph.D. Thesis, Purdue University, West Lafayette, IN. Zhang, Z.-Y. (1995a). J. Biol. Chem. 270, 11199-11204. Zhang, Z.-Y. (1995b). J. Biol. Chem. 270, 16052-16055. Zhang, Z.-Y., and Dixon, J. E. (1993). Biochemistry 32, 9340-9345. Zhang, Z.-Y., and Dixon, J. E. (1994). Adv. Enzymol. 68, 1-36. Zhang, Z.-Y., and Van Etten, R. L. (1990). Arch. Biochem. Biophys. 282, 39-49. Zhang, Z.-Y., and Van Etten, R. L. (1991). J. Biol. Chem. 266, 1516-1525. Zhang, Z.-Y., Clemens, J. C , Schubert, H. L., Stuckey, J. A., Fischer, M. W. F., Hume, D. M., Saper, M. A., and Dixon, J. E. (1992). J. Biol. Chem. 267, 23759-23766. Zhang, Z.-Y., Maclean, D., Thieme-Sefler, A. M., Roeske, R., and Dixon, J. E. (1993a). Anal Biochem. 2 1 1 , 7-15. Zhang, Z.-Y., Thieme-Sefler, A. M., Maclean, D., McNamara, D. J., Dobrusin, E. M., Sawyer, T. K , and Dixon, J. E. (1993b). Proc. Natl. Acad. Sci. U.S.A. 90, 4446-4450. Zhang, Z.-Y., Wang, Y., and Dixon, J. E. (1994a). Proc. Natl. Acad. Sci. U.S.A. 9 1 , 1 6 2 4 1627. Zhang, Z.-Y., Maclean, D., McNamara, D. J., Dobrusin, E. M., Sawyer, T. K , and Dixon, J. E. (1994b). Biochemistry 3 3 , 2285-2290. Zhang, Z.-Y., Malachowski, W. P., Van Etten, R. L., and Dixon, J. E. (1994c). J. Biol. Chem. 269, 8140-8145. Zhang, Z.-Y., Wang, Y., Wu, L., Fauman, E., Stuckey, J. A., Schubert, H. L., Saper, M. A., and Dixon, J. E. (1994d). Biochemistry 33, 15266-15270. Zhang, Z.-Y., Zhou, G., Denu, J. M., Wu, L., Tang, X., Mondesert, O., Russell, P., Butch, E., and Guan, K.-L. (1995a). Biochemistry 34, 10560-10568. Zhang, Z.-Y., Wu, L., and Chen, L. (1995b). Biochemistry 34, 16088-16096. Zhang, Z.-Y., Palfey, B. A., Wu, L., and Zhao, Y. (1995c). Biochemistry 34,16389-16396. Zhang, Z.-Y., Walsh, A. B., Wu, L., McNamara, D. J., Dobrusin, E. M., and Miller, W. T. (1996). J. Biol. Chem. 271, 5386-5392. Zhao, Y., and Zhang, Z.-Y. (1996). Biochemistry 35, 11797-11804. Zheng, Z. M., Wang, Y., and Fallen, C. J. (1992). Nature (London) 359, 336-339. Zhou, Z., Denu, J. M., Wu, L., and Dixon, J. E. (1994). J. Biol. Chem. 269, 28084-28090.

CURRENT TOPICS IN CELLULAR REGULATION, VOLUME 35

Regulation of Fas-Mediated! Apoptosis ROBERTA A. GOTTLIEB BERNARD M . BABIOR Department of Molecular and Experimental Medicine The Scripps Research Institute La Jolla, California 92037

I. introduction Apoptosis, originally a botanical term referring to the shedding of leaves from deciduous trees, was first applied by Wyllie in 1980 to a process he identified whereby cells kill themselves according to a stereotyped program (203). Many events take place during the execution of this program. Nuclear DNA is degraded into fragments «:*200 bp long, a length roughly equal to the length of the DNA wrapped around the histones in a nucleosome; on agarose gel electrophoresis, these fragments appear as a ladder (Fig. 1)—the so-called nucleosomal ladder of apoptosis. DNA degradation is accompanied by the cleavage of important structural proteins in the nucleus, including lamin (132). The cytoplasm acidifies, and major alterations take place in energy metabolism, including the abolition of the transmembrane A/XH+ in mitochondria and the conversion of NAD^ to poly(ADP-ribose) and nicotinamide, alterations that culminate in the complete loss of ATP from the apoptotic cell. Tissue transglutaminase is activated, possibly by the fall in cytoplasmic pH, leading to the extensive cross-linking of proteins in both the nucleus and the cytoplasm. Changes in the cell surface, including an increase in surface phosphatidylserine (158) and integrins (43), mark the cell for engulfment by mononuclear phagocytes, the garbage cans of multicellular organisms. Poorly defined cytoskeletal changes result in the fragmentation of the nucleus and the shedding of cell substance as "apoptotic bodies" that start as blebs but then separate from the dying cell to float free in the extracellular medium (Fig. 2). The end result of this process is that the apoptotic cell is eliminated without releasing its contents into the tissues. Apoptosis is a general process that is required for the attainment of many biological goals. It is indispensable in embryological growth and development, the most thoroughly studied example being its role in 69

Copyright © 1997 by Academic Press. All rights of reproduction in any form reserved. 0070-2137/97 $25.00

70

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

FIG. 1. Oligonucleosomal fragmentation of DNA from apoptotic cells. During apoptosis in many cell types, DNA is degraded first into 300- and 50-kb pair fragments and subsequently into oligonucleosomal fragments, the "nucleosomal ladder" of apoptosis.

the development of the roundworm Caenorhadbitis elegans (79). Virally infected cells (167) and cells containing severely damaged DNA (101) are eliminated through apoptosis. The atrophy of hormone-responsive tissues such as the prostate that takes place when hormone concentrations fall with aging is mediated by apoptosis (104). Immunological tolerance depends in part on the death through apoptosis of self-reactive lymphocytes (148). These are just a few examples of essential biological functions that are mediated through apoptosis. Apoptosis is precipitated by a wide variety of stimuli, some causing nonspecific damage to affected cells and others functioning as physiological regulators. Stimuli that induce apoptosis by inflicting nonspecific damage include ultraviolet radiation (52), oxidizing agents (123), and agents used for chemotherapy (36), all of which cause DNA strand breaks; reperfusion injury, a form of damage that takes place during the first few seconds after the blood supply is restored to ischemic tissues (56); and inhibitors of macromolecule synthesis such as cycloheximide and actinomycin D (120). Examples of physiological regulators include growth factors and extracellular matrix, whose presence can delay or prevent apoptosis in cells that would otherwise die sponta-

REGULATION OF FAS-MEDIATED APOPTOSIS

71

FIG. 2. Characteristic morphological changes in apoptotic cells. Jurkat cells exposed to anti-Fas IgM demonstrate membrane blebbing, volume loss, chromatin condensation, and nuclear fragmentation. An apoptotic body is indicated by the arrow.

neously (16,49,50,198); cytokines such as tumor necrosis factor (TNF), which precipitates apoptosis in cells displaying the TNF receptor (114); perforin and granzyme B, which induce apoptosis when implanted by cytotoxic T lymphocytes into a membrane of a target cell (the "kiss of death") (10); steroid hormones, which precipitate apoptosis in lymphocytes (204); and the Fas receptor-Fas ligand system, whose function is reviewed below. The responses of cells undergoing apoptosis are highly variable. The degradation of DNA into nucleosome-sized fragments is regarded as the hallmark of apoptosis, but cell death by apoptosis can occur without the appearance of a nucleosomal ladder—indeed, without a nucleus (164). This finding suggests that what is called apoptosis is really an aggregate of several independent processes, and that the final end point of apoptosis—namely, the elimination of a cell without release of its contents—can be attained by means of suitable subsets of these processes. In addition, various cells respond differently to agents that

72

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

activate apoptosis. For instance, cycloheximide precipitates apoptosis in some cell types (120) but prevents it in others (21). As another example, corticosteroids induce apoptosis only in lymphocytes (204). Rather than attempt to survey all aspects of the regulation of apoptosis in all the cells in which it has been studied, a task well beyond the scope of this chapter, we will focus on the regulation of apoptosis as initiated by a single agency, although one of particular physiological importance: the occupation of the Fas receptor.

II. Regulatory Factors From the wilderness of molecular features that characterize apoptosis in different cell types, certain groups of factors have emerged that appear to be specific regulators of apoptosis. These include Bcl-2 and its relatives; certain proteases; and ceramide, a lipid. A. The Bcl-2 Family

1. BcL-2 The Bcl-2 family (see Table I) is a group of structurally related proteins that participate in the regulation of apoptosis, some as inhibitors and others as activators of this process. Bcl-2, the first member of this family to be discovered, is an antagonist of apoptosis. It initially appeared in the guise of an oncogene located at the t(14;18) translocaTABLEI BcL-2 FAMILY"^

Name Bcl-2 Ced-9 Bax BC1-XL

Bcl-xs Bad Bak Bag-1 BHRF-1

GenBank Accession No.

% Similarity

% Identity

Activity

M14745 L26546 L22473 Z23115 Z23116 L37296 U23765 U17162 A22899

50 52 59 51 43 56 37 41

19 28 43 31 19 26 14 24

Life Life Death Life Death Death Death Life Life

"" Bcl-2 family members participate in regulating apoptosis. Comparisons of homology (similarity, %, and identity, %) are made to Bcl-2. Higher homology occurs within the self-associating domains BHl and BH2. The biological activity of the gene product is denoted as favoring survival (life) or promoting apoptosis (death). Accession numbers are listed for reference.

REGULATION OF FAS-MEDIATED APOPTOSIS

73

tion characteristic of follicular lymphomas (66,77) and was shortly thereafter shown to extend the life of B lymphocytes in which it was expressed (35) without augmenting proliferation. It has now been found to occur in a large variety of cells (110), where it acts to prevent apoptosis initiated by many stimuli, but not all (59,70,174). Its most important physiological role has been revealed by studies with mice whose Bcl-2 genes have been experimentally altered. In transgenic mice in which Bcl-2 was overexpressed, the only obvious phenotypic alteration was a great expansion of the lymphoid follicles in the spleen and nodes (166). These expanded follicles contained mature B cells that express immunoglobulin M (IgM) and IgD on their surfaces. The phenotype of the Bcl-2 knockout mouse, however, showed more widespread alterations, with polycystic kidney disease, gray hair, and massive involution of the thymus and spleen leading to early postnatal death (191). It is clear from these phenotypes that Bcl-2 is an important regulator of the death of cells in the lymphoid lineage. There is some evidence that Bcl-2 prevents apoptosis by acting as an antioxidant (78,97), but this question remains unanswered (89,169). As to the properties of the molecule itself, Bcl-2 is a 26-kDa protein found in the outer membranes of mitochondria, the endoplasmic reticulum, and the nuclear membrane. It is anchored in these membranes by means of a C-terminal transmembrane segment (residues 212-235), but a membrane location is not essential for its antiapoptotic activity, since soluble forms of Bcl-2 created by eliminating the transmembrane segment are still active, although not as active as the membraneassociated protein (78,133). Two domains, known as BHl (residues 136-155) and BH2 (residues 187-202), are necessary for Bcl-2 to express its antiapoptotic activity. Bcl-2 forms a homodimer, binding to itself through a head-tail interaction between residues near the N terminus of one molecule and residues in the C-terminal half of the other (68,160). The BHl and BH2 domains, however, are not involved in homodimer formation. 2.

BAX

The second member of the Bcl-2 family to be discovered was Bax. This protein was found during a yeast two-hybrid search for proteins that interact with Bcl-2 (143). Bax shows a high degree of homology to Bcl-2, especially in the important BHl and BH2 domains that are necessary for the antiapoptotic effect of Bcl-2 (Fig. 3). Nevertheless, Bax opposes the action of Bcl-2, accelerating rather than inhibiting apoptosis.

74

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

Bel-2

MAHAGRTGYD NREIVMKYIH YKLSQRGYEW DAGDVGAAPP GAAPAPGIFS

Bax

MDGSGE

QPRGGGPTSS EQIMKTGALL

SQPGHTPHTA ASRDPVARTS PLQTPAAPGA AAGPALSPVP PWHLTLRQA LQGFIQDRAG

RMGGEAP ELALDPVPQD ASTKKLSEC-

-LKRI

I GDDFSRRYRR DFAEMSRQLH LTPFTARGRF ATWEELFRD G-VNWGRIVAF GDELDSN--M ELQRMIAAVD TDS--PREVF FRVAADMFSD GNFNWGRWAL II FEFGGVMCVE SVNREMSPLV DNIALWMTEY LNRHLHTWIQ DNGGWDAFVE FYFASKLVLK ALCTKVPELI RTIMGWTLDF LRERLLGWIQ DQGGWDGLLS LYGPSMRPLF DFSWLSLKTL LSLALVGACI TLGAYLGHK YFGTP--

--TWQTVTIF VAGVLTAS-- -LTIWKKMG

FIG. 3. Sequence similarity in BHl and BH2 domains of Bcl-2 and Bax. Domains BHl (I) and BH2 (II) are indicated in boldface type. Sequence identity is denoted by (:), and similarity is denoted by (.).

Bax occurs in three forms that are produced by alternative sphcing of the Bax message. The active form of Bax is Bax-alpha (a), a 21-kDa protein with a single transmembrane domain. Other forms are Baxbeta (/3) and Bax-gamma (y), soluble proteins whose function is obscure. The distribution of Bax in the tissues is very broad, even more extensive than that of Bcl-2 (109). It is of interest that within a given cell type (e.g., colonic epithelium), cells destined to undergo apoptosis contain levels of Bax that are higher than average, whereas their levels of Bcl2 are lower than average. Little is known about how Bax-a promotes apoptosis. One aspect of its function, however, is well understood. Bax itself has been reported to form a homodimer (68,160,213), and it is thought that the Bax dimer is the lethal form of the protein. Bcl-2 disrupts the Bax dimer by displacing one of its Bax chains, forming a Bax • Bcl-2 dimer that no longer promotes apoptosis. A high Bcl-2/Bax ratio in a cell would therefore be expected to suppress apoptosis by favoring the formation of the innocuous Bax • Bcl-2 dimer; conversely, a low Bcl-2/Bax ratio would promote apoptosis (106). It is through its BHl and BH2 regions that Bcl-2 binds to Bax. Missense mutations in the BHl domain (G145A, G145E) or the BH2 domain (W188A) eliminate the ability of Bcl-2 to bind to Bax and, at the same time, abolish the activity of Bcl-2 (210). Similarly, a Bcl-2

REGULATION OF FAS-MEDIATED APOPTOSIS

75

molecule in which both BH domains have been deleted [A(138154,188-196)] is unable to bind Bax, although dimerization is completely normal (68). These findings represent the strongest evidence t h a t the antiapoptotic activity of Bcl-2 depends on its ability to bind to Bax. Heterodimer formation, however, must be only part of the story, because the Bcl-2 deletion mutant A l l - 3 3 was able to bind to Bax without eliminating its lethal function (68). 3. Bcl-x A number of other members of the Bcl-2 family have been reported (Table I). Of these, the most extensively studied have been the two forms of Bcl-x. The bcl-x gene was first isolated from chickens by means of its homology with the bcl-2 gene. The chicken gene was then used as a probe to show t h a t humans express two forms of Bcl-x: BC1-XL, a 26-kDa protein, and Bcl-Xs, a 19-kDa protein t h a t is made by splicing out the portion of the Bcl-x message t h a t encodes residues 126-188 of BC1-XL, excising B H l and half of BH2 from the protein. Like Bcl-2, both forms of Bcl-x contain C-terminal transmembrane segments and are located largely in the mitochondria (45). The function of BC1-XL was demonstrated by experiments showing t h a t a cell line dependent on interleukin 3 (IL-3) was able to survive the withdrawal of the growth factor when transfected with BC1-XL. This result showed t h a t BC1-XL, like Bcl-2, opposed apoptosis. The same cells were neither protected from nor sensitized to the withdrawal of IL-3 when transfected with Bclxs alone; but when Bcl-xs was cotransfected with Bcl-2, the protection against IL-3 withdrawal normally afforded by Bcl-2 was eliminated. Bcl-xs, then, promoted apoptosis, not through an action of its own but by opposing the action of Bcl-2 (13). Like Bcl-2, Bcl-x is expressed widely, but the pattern of expression differs from t h a t of Bcl-2. In particular, Bcl-x is prominently expressed in neurons and hematopoietic tissue (13,55,110) and in certain types of lymphocytes (45). The physiological significance of these expression patterns is indicated by the abnormalities seen in Bcl-x knockout mice. These mice showed extensive apoptosis of neurons and hematopoietic cells, as well as some abnormalities affecting lymphocytes (127), a picture very different from t h a t seen in Bcl-2 knockouts. Interactions of BC1-XL and -Xs with themselves and other members of the Bcl-2 family have been studied using the yeast two-hybrid system. Results show t h a t BC1-XL interacts with itself and with Bcl-2, Bax, and Bcl-xs, binding to both the N-terminal and C-terminal halves of Bcl-2 (160). Bcl-xs interacts with Bcl-2, an interaction dependent on residues 2 2 - 7 4 of Bcl-Xs, on the B H l domain, and on the C-terminal hydrophobic

76

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

region of Bcl-2 (160). The dependency on the BHl domain was shown by the finding that Bcl-Xs failed to bind to the G145E mutant of Bcl2. It is of interest that this same mutant also failed to take up Bax (see above), suggesting that Bcl-Xs and Bax bind to the same region of Bcl-2 and that Bcl-Xs may work by competing with Bax for Bcl-2, preventing the formation of the Bax • Bcl-2 dimer and the resulting neutralization of Bax. 4. OTHER PROTEINS FUNCTIONALLY RELATED TO BCL-2

The Bcl-2 family contains other antiapoptotic proteins besides those discussed above (Table I). One of these is encoded by ced-9, a gene found in the roundworm Caenorhahditis elegans. During the development of C. elegans, 131 cells out of a total of 1090 die by apoptosis. By using Nomarski optics to screen several thousand ethylmethane sulfonatetreated worms, Horvitz and associates identified a number of genes that were involved in the death of these cells. Some of these genes were necessary for normal apoptosis, while others prevented apoptosis in cells normally destined to die during development. Among the latter was the gene ced-9 (73), which was found to encode a protein homologous to Bcl-2 (75). Gain-of-function mutations of ced-9 prevented normal developmental apoptosis, while loss-of-function mutations led to widespread apoptosis culminating in the death of the embryonic worm (73,74). The effects of the loss-of-function mutations could be corrected by the overexpression of bcl-2 in a transgenic worm (75). Related proteins in mammalian cells include Bak, Bad, Mcl-l, and BAG-1. Two of these proteins, Mcl-l and BAG-1, prevent apoptosis (153,180). Mcl-l is homologous to Bcl-2, interacts with Bcl-2, and, like Bcl-2, is localized in the mitochondria, although it differs from Bcl-2 in its distribution among other subcellular compartments (107,207). Like Bcl-2, Mcl-l is found in a wide variety of tissues, but within a given tissue its pattern of distribution tends to differ from that of Bcl2 (107,108). The disparate patterns of distribution of Bcl-2 and Mcl-l suggest that these two proteins play different biological roles. BAG-1 resembles Bcl-2 and Mcl-l in opposing apoptosis, but it has no structural homology to Bcl-2 (180). Of particular interest, cells that overexpress both BAG-1 and Bcl-2 are much more resistant to apoptosis than are cells overexpressing either one of those proteins alone. As to the other two proteins, Bak, found in a large number of tissues, promotes apoptosis in cells exposed to a death stimulus (102). Bad also promotes apoptosis, dimerizing with BC1-XL and blocking its antiapoptotic effect, possibly by displacing Bax from a BC1-XL • Bax complex (206). Certain viral proteins affect apoptosis in infected cells. BHRFl,

77

REGULATION OF FAS-MEDIATED APOPTOSIS

an Epstein-Barr virus protein, is homologous to the carboxyl portions of several members of the Bcl-2 family, has a subcellular distribution similar to t h a t of Bcl-2, and inhibits apoptosis in cells in which it (i.e., BHRFl) is overexpressed (38,72). 5. CONTROL OF EXPRESSION OF THE BCL-2 FAMILY

The control of apoptosis depends to a great extent on the level of expression of Bcl-2 family protein in the cell of interest. There is an extensive literature on this subject, a few examples from which are given in Table II. The most general conclusion t h a t can be drawn is TABLE II REGULATION OF EXPRESSION OF BCL-2 FAMILY MEMBERS"

Setting

Stimulus

Response

Outcome

Reference 117

Bcell

IL-1 CD40

Bcl-2 t

Thymocytes

Interaction with self-MHC Aging

Bcl-2 t

sig cross-linking Addition of c-kit ligand Withdrawal of IL-3 Expression of activated Ras

Baxt Bcl-2 t

Memory cells maintained Positive selection Peripheral deletion Apoptosis Survival

Bcl-2 i

Apoptosis

Bcl-2 t , Bcl-

Survival without IL-3

Bcl-2 i

Apoptosis

9

Bcl-2 i Bcl-2 i

Apoptosis Apoptosis

20 165

wt p53 expression

Bcl-2 i , Baxt

Apoptosis

165

IL-6 addition

BCI-XLT

Survival

119

UV exposure Menstrual cycle

Bcl-2 i Bcl-2 varies

Apoptosis Apoptosis inverse

52 54, 157

CD4% CD4- CD8cells BL-41 B cell line NK cells Mast cells IL-3-dependent stem cell line HL-60 cells

HL-60 cells M l myeloid leukemia cells M l myeloid leukemia cells IL-6-dependent myeloma line Rat epidermis Breast and endometrium

TNF-a, ceramide, ionizing radiation Differentiation TGF-/3 addition

Bcl-2 i

117 182 6 124 103 23

"^ Expression of Bcl-2 relatives changes in response to certain signals (or withdrawal of signals), leading to a change in the balance of apoptotic and antiapoptotic factors. The outcome determines life or death.

78

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

that the expression of the various members of the Bcl-2 family is highly variable, depending critically on the type of cell, the state of development of the organism that contains the cell, and the exogenous factors (proteins, hormones, etc.) to which the cell has been exposed. B. Proteases Certain specific proteases (Table III) appear to be important in initiating the events of apoptosis. Many of these are cysteine proteases; others are serine proteases. Most of these proteases act to cause the suicide of the cell that has produced them, but a few act as murderers, killing cells into which they have been injected by cytotoxic T lymphocytes. In view of the fact that only a few special enzymes have been identified as so-called death proteases, it is of considerable interest that apoptosis (or a process very much like apoptosis) can be initiated by the delivery into a cell of an arbitrary protease such as trypsin, chymotrypsin, or proteinase K (199). The death proteases are characterized by the unusual feature that they are specific for aspartyl-X bonds. 1. CED-3

The gene ced-3 is another of the C elegans genes that affects apoptosis. In contrast to ced-9, whose inactivation results in the death of cells normally destined to survive, the inactivation of ced-3 prevents the death of the cells that are normally lost during C elegans edevelopment (41). The ced-3 gene was isolated by positional cloning (212) and TABLE III IGE/CED-3 FAMILY OF DEATH PROTEASES''

Cleaves Protease

Ref.

IL-1

PARP

Inhibitors

% Identity to hICE

Ced-3 ICE Yama/CPP32 PrICE NEDD2/ICH-1 TX/ICH-2 MCH2 ICErei-II, -III

212 24 188 115 112, 195 46, 95 48 128

IL-1 IL-1 No No

No Poorly Yes Yes

p35, CrmA, YVAD p35 CrmA, YVAD P35, CrmA, DEVD YVAD p35, CrmA p35 DEVD

29 100 31

No Yes No

28 53 29 53,50

"^ Ever-expanding family of proteases participate in apoptosis. Their activities with respect to prototypic substrates such as prointerleukin-lj8 and poly(ADP ribose) polymerase (PARP) are incompletely characterized, as is their susceptibility to inhibitors.

REGULATION OF FAS-MEDIATED APOPTOSIS

79

was found to encode a protein with a predicted mass of 57 kDa that showed considerable homology to the mammalian IL-lj8 converting enzyme (ICE), which generates IL-lj8 from its inactive precursor. This homology, plus the presence in the Ced-3 protein of the pentapeptide QACRG, which is found in ICE and several related proteins, strongly suggested that, like ICE, Ced-3 was a protease, a conjecture that has now been proven experimentally (205). 2. ICE Known for some time as the enzyme that activates IL-lj8, ICE was cloned on the basis of this activity from neutrophils (24) and monocytes (189). Crystallography showed that ICE is a tetramer composed of two identical heterodimers, each containing a lOK and a 20K subunit (plO and p20, respectively) (194). The protease is synthesized as a 45K proenzyme from which the active enzyme is released by limited proteolysis. Active ICE is a cysteine protease with the very unusual feature, shared by several death proteases, that it cleaves proteins at Asp-X bonds. For example, the best-characterized of the ICE-mediated proteolytic reactions, the release of IL-lj8 from its precursor, involves the cleavage of the Asp^^^-Ala bond in the sequence YVHDA (80). Five different forms of ICE are produced by alternative splicing of its mRNA (3). Of these five isoforms, ICE-ce (the full-length protein), ICE-/3 and ICE-y caused apoptosis of cells in which they were overexpressed, while ICE-5 and ICE-e showed antiapoptotic activity. ICE-y is missing most of its propeptide (amino acids 20-112 are deleted), leading to the speculation that this isoform may be constitutively active, and in fact may be responsible for the activation of ICE-a during the initiation of the apoptosis program. It was the homology between Ced-3 and ICE that suggested a probable proteolytic function for Ced-3. Conversely, the same homology suggested a probable death protease function for ICE. This has now been demonstrated in several experimental systems. Overexpression of wildtype ICE in Rat-1 fibroblasts resulted in the death of the cells by apoptosis, while overexpression of an inactive ICE mutant had no such effect (125). Overexpression of ICE was also found to accelerate apoptosis induced by ligation of the Fas receptor (118). Apoptosis in all these cases was prevented or delayed by coexpression of CrmA, a 38-kDa vaccinia virus serpin that specifically inhibits several death proteases (51,118,125,187), presumably by combining with them to form an inactive complex* and prevents apoptosis under a variety * Formation of an inactive ICE-CrmA complex has been shown experimentally (18).

80

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

of circumstances (51,187,195). In addition, apoptosis induced by the ligation of receptors or by certain chemotherapeutic agents was prevented by Ac-YVAD chloroketone and by Z-Asp-carbinol dichlorobenzoate, low molecular weight inhibitors of ICE (51,122). Experiments with mice in which the ice gene has been inactivated, however, show t h a t there must be other mammalian death proteases besides ICE (111,116). Unlike normal mice, the ICEless mice produce little if any ILla or IL-lj8 when treated with endotoxin and are resistant to endotoxin shock. Consistent with these findings, monocytes from the ICEless mice failed to secrete IL-la or IL-lj8 after stimulation with endotoxin. Furthermore, although their thymocytes were resistant to Fasmediated apoptosis, the thymocytes died by apoptosis after exposure to irradiation or corticosteroids, and the ICEless mice themselves developed in a normal fashion. 3. OTHER DEATH PROTEASES

Several other mammalian death proteases have been identified. (a) ICH-li (195). Ich-1 is a gene t h a t was cloned on the basis of its resemblance to ced 3 and ice, Ich-1 encodes 2 proteins: I C H - 1 L , a 48kDa protein homologous to both subunits of ICE as well as to Ced-3, and ICH-ls, a 34-kDa protein t h a t is synthesized from an alternatively spliced ich-1 message. Overexpression of I C H - 1 L induces apoptosis, while overexpression of ICH-ls suppresses apoptosis. NEDD-2 is the murine homolog of ICH-1, initially identified by subtraction cloning from developing brain (112). (b) prICE (protease like ICE) {115), In a cell-free system, prICE activity results in the appearance of a DNA ladder typical of apoptosis. As to possible substrates, prICE can cleave nuclear poly(ADPribose)polymerase, a 116-kDa enzyme t h a t is activated by DNA strand breaks to add long chains of poly(ADP-ribose) to nearby proteins such as histones (201). Poly(ADP-ribose)polymerase is frequently, but not invariably, cleaved during the execution of the apoptosis program (99,100). Cleavage by prICE removes the 31-kDa DNA-binding domain from the enzyme by cutting an Asp-Ala bond (115), with preservation of residual catalytic activity t h a t is no longer upregulated by fragmented DNA. In contrast to ICE, prICE has no effect on pro-IL-ljS. Conversely, ICE readily cleaves pro-IL-lj8 but is only weakly active against poly(ADP-ribose)polymerase, cleaving it only at high enzyme concentrations (62). PrICE has not yet been cloned and may be identical to one of the already-cloned proteases listed here.

REGULATION OF FAS-MEDIATED APOPTOSIS

81

(c) TX (46) (also known as Ich-2 [95]). TX, a protein «50% identical to ICE, appears on the basis of homology to be a cysteine protease that cleaves at Asp-X bonds. The expressed protein was confirmed to be a protease, cleaving itself but not acting on pro-IL-lj8. COS cells in which TX was overexpressed underwent apoptosis. TX is also able to cleave poly(ADP-ribose)polymerase (62). (d) ICErei proteases. Homology cloning has identified additional members of the ICE/Ced-3 family, including ICEreJI and ICEreJH (128). (e) Others. Other proteases implicated in apoptosis include an enzyme that cleaves the 70-kDa protein of the Ul small nuclear ribonucleoprotein (SNURP) (23); a 24-kDa protease whose activity is increased by an order of magnitude in U937 cells undergoing apoptosis (202); a serine protease isolated from apoptotic Jurkat cells that induces the formation of a DNA ladder in isolated thymocyte nuclei (162); and several serine proteases, one of which cleaves Asp-X bonds, from the granules of cytotoxic lymphocytes (168). Granzyme A is able to process prointerleukin-lj8 to its active form by cleaving at Arg-120, a finding of unclear significance to the process of apoptosis induction (82). More interesting is the finding that granzyme B is able to generate the active form of Yama (see below), and initiate the death protease cascade directly (37). The proteasome is an ATP-dependent multisubunit protease that targets polyubiquinated proteins for degradation. The proteasome may also participate in apoptosis (61). In the tobacco hornworm Manduca sexta, there are major increases in the expression of several of the subunits of the proteasomes, as well as increases in proteasomal activity and coordinated induction of the ubiquitin conjugation pathway, in the period just before eclosion (emergence of the adult hawkmoth), an event during which there is extensive apoptosis of abdominal muscles (39,64). Additional evidence supporting a role for the proteasome in Fasmediated signaling is provided by the report by Wright et al. (202a) that UBC-FAP, the mammalian homolog of UBC9, an ubiquitin-conjugating enzyme, is associated with the cytosolic portion of Fas, and that mutations which abrogate Fas death signaling eliminate binding of UBC9. However, the relationship of this association to propagation of the death signaling cascade is, as yet, unclear. 4. YAMA

Of the mammalian apoptosis-associated proteases described to date, the one whose properties may be most consistent with its active partici-

82

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

pation in the cell death program is the cysteine protease known as Yama (originally called CPP32j8 and also known as apopain) (47,188). Named after the Hindu god of death, Yama is a heterodimer composed of 11-kDa and 20-kDa subunits t h a t is generated by limited proteolysis of an inactive 32-kDa precursor, probably by cleavage at an Asp-X bond. Within the p20 subunit is the diagnostic motif QACRG, which contains the active site cysteine and is also found in the death proteases Ced-3 and ICE. Three criteria that have been proposed by Tewari et al. (188) to define the death protease: (1) similarity with other apoptosis-associated proteases. Yama is homologous to the death proteases Ced-3 and ICE, contains the unique death protease motif QACRG, and cleaves proteins specifically at Asp-X bonds; (2) inactivation by CrmA. Yama was inactivated when incubated with CrmA and coimmunoprecipitated with CrmA after inactivation; neither occurred when Yama was incubated with the loss-of-function m u t a n t CrmA T291R (188); and (3) cleavage of poly(ADP-ribose) polymerase. Like prICE, Yama cleaves poly(ADPribose) to an 85-kDa polypeptide. Yama fulfills these three criteria, and accordingly has been put forward as perhaps the true death protease. Two recent papers identified a ICE/CED-3-like protease (FLICE) associated with FADD/Mort-1, as part of the Fas and TNF-R complexes involved in death signaling. Although it is yet not clear how receptor activation leads to activation of this protease (which has a prodomain and structure similar to other ICE/CED-3 family members), it is clear t h a t this protease, once activated, can cleave PARP to its signature fragments. Interestingly, instead of the usual highly conserved QACRG motif, FLICE contains the pentapeptide QACQG. 5. INHIBITORS

Two inhibitors of ICE family proteases have been identified to date, and these potently block apoptosis in a variety of systems. The first, CrmA, is a gene product of the cowpox virus and interacts with Yama but not ICE. The second is the baculovirus protein known as p35 (33). This protein, produced by the nuclear polyhedrosis virus Autographa californica and localized chiefly in the cytoplasm, was found to block apoptosis induced by a number of stimuli in cultured SF21 Spodoptera frugiperda (fall armyworm) cells, mammalian neural cells, and transgenic C. elegans and Drosophila (21,71,76,151,178). Homologs of p35 have been found in other insect viruses (12,96,147). The p35 protein is cleaved by several of the death proteases but then dissociates from the protease slowly or not at all, thereby inhibiting the protease. In in

REGULATION OF FAS-MEDIATED APOPTOSIS

83

vitro assays, p35 has been shown to inhibit ICE, Yama/CPP32, ICHl/Nedd-2, and ICH--2/TX but not granzyme B (18). 6. SUBSTRATES

Proteolysis during apoptosis appears to be widespread. Specific substrates of these death proteases include pro-IL-1/3, poly(ADP-ribose) polymerase, and the 70-kDa protein of Ul (already mentioned) (186). In addition, apoptosis was shown to be accompanied by the cleavage of ce-fodrin (121). PITSLRE jS-l, a 58-kDa protein kinase, is cleaved into a more active 50-kDa form in cells undergoing apoptosis (113). Other studies have identified many other proteins that are cleaved during apoptosis (155,193). In particular, Voelkel-Johnson et al. (193) showed that lamin B, topoisomerase I, histone HI, protein kinase Cj81, and phospholipase A2 were all cleaved during tumor necrosis factorinduced apoptosis (Table IV). Among these proteolytic events, it is not clear which if any are essential components of the apoptosis program. There is even some evidence for the existence of one or more protease-independent routes to apoptosis (159). With regard to Fas-mediated apoptosis, the preponderance of the evidence favors a death protease-dependent program, but as will be discussed below, there are experiments suggesting that Fasmediated apoptosis may occur without the participation of a protease. Even when an apoptotic event clearly depends on a protease, it is difficult to know which protease is responsible unless a gene knockout or double knockout experiment proves to be informative. Analysis of the internal cleavage sites of the proteases and the one well-characterized substrate, poly(ADP-ribose) polymerase, leads to a limited generalization about a possible consensus sequence for other putative substrates of the ICE family proteases. Assigning the aspartic acid residue as PO, then the preceding four amino acids tend to include one to three acidic residues and one to three hydrophobic residues. The first residue following the Asp is restricted to alanine, glycine, or serine. Three documented exceptions occur: M^^^ in Yama, Q^^^ in NEDD2, and N^^^ in ICE. These generalizations allow one to inspect the sequences in proteins reported to be cleaved during apoptosis and predict the possible cleavage sites, assuming the responsible protease is an ICEfamily protease with substrate specificity resembling the proteases themselves (Table IV). The other interesting feature of the known cleavage sites is that two cleavage sites are present in almost every known substrate. Considering that the tetrameric structure contains two catalytic domains, one can envision a substrate protein wrapped around the protease, with two sites being clipped on the same molecule.

84

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR TABLE IV SUBSTRATES FOR ICE-FAMILY PROTEASES"

Protein

ICE protease

Yama/CPP32i8 Yama/CPP32)8 MCH2 NEDD2/Ich-1 NEDD2/Ich-1 TX/Ich-2 ICE ICE ICE ICE CED-3 CED-3 ICErelll ICErellll pro-Il-li8 pro-IL-lj8 PITSLREiS Lamin B Ul SNURP PARP

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No No

Sequence around cleavage site GIETDSGVDD GVDDDMAC ?

WRGVDQQDK CEESDAGKE ?

FPVLDSVGVS EFEDDAIKKA LNMQDQGV PQAVDNPAMP FPVLDSVDG WDNRDGPLF NLEEDAVYKTH NLEADSVCKIH FFEADGPKQ AYVHDAPVRSLNCT ? ? ? ?

" Substrates for ICE-family proteases include the proteases themselves, which are activated by limited proteolysis. In every case examined so far, the protease cleaves at the bond D-X, where X is generally a small amino acid residue, most often G or A. In the majority of cases, additional acidic residues lie close to the cleavage site. The aspartate at the cleavage site is shown in boldface type.

The distance between the cleaved Asp residues is a minimum of 6 residues (in Yama) to 19 residues (in ICE). This predicts that other substrates may be cleaved at a second, nearby site and may explain why the sizes of the fragments observed do not always add up to the size of the intact protein. C. Ceramide Ceramide is a neutral lipid consisting of a long-chain fatty acid attached by an amide linkage to sphingosine, a long-chain amino alcohol (see Diagram I). It is produced chiefly by the hydrolysis of sphingomyelin, an abundant phospholipid in which the hydroxyl group of ceramide is attached to the hydroxyl group of choline through a phosphodiester linkage. Ceramide is similar in structure to diacylglycerol, another neutral lipid that contains two long hydrocarbon chains. Diacylglycerol

REGULATION OF PAS-MEDIATED APOPTOSIS

85

Ceramide

Diacylglycerol

CH2OH

DIAGRAM I. Structure of ceramide and its antagonist, diacylglycerol. Ceramide generated by the activity of either acidic or neutral sphingomyelinase is thought to act as a second messenger in signals leading to apoptosis. In contrast, structurally similar diacylglycerol participates in signaling pathways that oppose apopotosis. Adapted from Hannun and Obeid (69).

is well known as a regulator of biological processes that involve protein kinase C. It is therefore not altogether surprising that ceramide also functions as a biological regulator. The participation of ceramide in apoptosis has been demonstrated in many systems (58,69). Studies with WEHI-231, a murine B-cell lymphoma line, have shown that these cells accumulated ceramide in response to several inducers of apoptosis, including anti-IgM (149); that treatment of these cells with ceramide induced both apoptosis and oxidant production, and that both responses were blocked by antioxidants (and by BC1-XL) (44); and that a subline of WEHI-231 that is resistant to apoptosis induced by anti-IgM produced abnormally small amounts of ceramide (57). Treatment with exogenous ceramide induced apoptosis in HL-60 and U937 cells (90,91,93). However, the concentration of exogenous ceramide needed to induce apoptosis is much higher than the amount of ceramide generated in response to a physiological signal. Using a water-soluble ceramide (C2-ceramide, in which the acyl group is acetate) and related compounds (11,37), it was shown that only ceramides with the correct stereochemical configuration were able to induce apoptosis, further evidence in support of the physiological nature of ceramide-induced apoptosis. During apoptosis, ceramide is made available by the activation of a sphingomyelinase, either acidic (32) or neutral (65). Depending on the experimental system, it has been

86

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

postulated to work through the activation of the guanine nucleotidebinding protein Ras (105) or of certain ceramide-dependent serine/ threonine-specific protein kinases (65)—leading to phosphorylation of Rafl, with subsequent phosphorylation of MEKl (MAP kinase or ERK kinase (209)—or by inhibiting the expression of Bcl-2 (25) or after hydrolysis, sphingosine serving as the true apoptosis-inducing agent in HL-60 under U937 cells (141,142). In inducing apoptosis, ceramide may meet with opposition. Diacylglycerol (see Diagram I) is an antagonist of ceramide-induced apoptosis in several systems, suggesting a yin-yang relationship between these two closely related lipids (65,90-92). Most surprising of all is the report that ceramide itself may be an antagonist of apoptosis, since it appears to prevent the death of sympathetic neuron primary cultures deprived of nerve growth factor (83,105). III. Regulation of Fas-Mediated Apoptosis Fas (also known as Fas antigen, APO-1, and CD95) and Fas ligand are the two physiological components of an extensively studied system that activates the apoptosis program. Fas is a member of the TNF receptor superfamily, a family that also includes the TNF receptor, the nerve growth factor receptor, the B-cell antigen CD40, and several others. When the Fas molecules on a cell are occupied, either by the Fas ligand or by a suitable anti-Fas antibody, a signal is transmitted instructing the cell to commit suicide. Whether the cell obeys this death signal depends on factors that regulate the response of the cell to the signal. The regulation of apoptosis is immensely complex, varying in fundamental ways depending on the cell type and the means by which apoptosis is induced. For instance, L929 cells transfected with Fas cDNA express both Fas and the closely related TNF receptor (TNFR). Apoptosis was induced in these cells by occupation of either of the two receptors. Even though the transfected L929 cells were killed by the occupation of either of these two very similar receptors, killing via TNFR was quite different from kiUing through Fas. TNFR killed the cells by means of oxygen radicals released by deranged mitochondria, with the participation of poly(ADP-ribose) polymerase and phospholipase A2, the latter activated by an influx of Ca^^. Furthermore, the dying cells showed none of the characteristic morphological features of apoptosis. In contrast, killing by Fas was independent of mitochondrial metabolism, phospholipase, and poly(ADP-ribose) polymerase and was associated with the development of typical apoptotic morphology in the

REGULATION OF FAS-MEDIATED APOPTOSIS

87

dying cells (163). In contrast to the results with L929 cells, occupation of either Fas or TNFR led to phospholipase A2 activation as well as apoptosis in a breast carcinoma line MCF7 that was transfected with a Fas cDNA and expressed both receptors (88). In a series of tumor lines that express both Fas and TNFR, some are susceptible to killing through Fas but not TNFR, others through TNFR but not Fas (200). Indeed, under certain circumstances, the engagement of Fas even induces T cells and fibroblasts to proliferate (1). Both macrophages and endothelial cells display Fas on their surfaces, but only macrophages die when Fas is occupied (154). As another example. Fas-bearing thymocytes can be driven into apoptosis by exposure to glucocorticoids or by occupation of Fas; the former process is inhibited but the latter is stimulated when protein synthesis is blocked (139). Furthermore, socalled double-positive T lymphocytes, which have not yet committed themselves to the CD4+ (helper) or CD8^ (cytotoxic) T cell line and are positive for both markers, i.e., (CD4^CD8^), are very susceptible to killing via Fas, while single-positive T cells (CD4+ or CD8^) are resistant to Fas-induced apoptosis even though Fas is as abundant on the surface of single-positive cells as it is on double-positive cells (139). These examples give some sense of the quagmire confronting investigators seeking to understand the regulation of apoptosis. Fasdependent or otherwise. The field is very new, however, and even though a unifying hypothesis concerning the regulation of apoptosis is not yet on the horizon, it is useful to review the information currently available on Fas-dependent apoptosis as a point of entry into this morass, a topic as bewildering as it is important. A. Fas and Fas Ligand

Fas was first cloned in 1991 by Nagata's group using as a probe an anti-Fas monoclonal antibody that was found to induce apoptosis in susceptible cells (86). An antigen known as APO-1, purified and cloned slightly later by Oehm et al., was shown by sequencing to be identical to Fas (138). These studies showed that Fas was a 48-kDa glycoprotein 335 amino acids in length with (1) a cysteine-rich extracellular portion consisting of 4 homologous «40 residue domains arranged approximately in tandem; (2) a single transmembrane domain lying approximately between residues 174 and 190; and (3) a so-called death domain consisting of residues 228 to 297 (34,53,185), mutations which abolish the ability of Fas to initiate apoptosis (84). The death domain associates with itself, and the apoptosis signal appears to arise from this selfassociation (15). Fas is also synthesized in soluble forms that arise through alternative splicing of the Fas message (22). The Fas gene

88

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

contains nine exons spread over 25 kb of chromosome (10q24.1 (7,81). Exon 6 encodes the transmembrane region and exon 9 the death domain. Several proteins that interact with the cytosohc portion of Fas have been described, as summarized in Cleveland and Ihle (34). FADD (27), also known as MORT-1 (14), and RIP (172) interact with the death domain of Fas, possess death domains of their own, and induce apoptosis when overexpressed. Apoptosis in response to FADD is blocked by CrmA, indicating that cell death in response to this protein involves a death protease. RIP contains, in addition to a death domain, a region homologous to protein kinases, while the unnamed protein contains a region that promotes self-association. A third protein, FAP-1, interacts with a region at the extreme C terminus of Fas that opposes the action of the death domain (161). FAP-1, a protein tyrosine phosphatase, is expressed most strongly in cells that are relatively resistant to apoptosis. Most recently, a Fas-associated protein (p 55.11) was identified by yeast two-hybrid screening that bears sequence homology to a subunit of the 268 proteasome (15). Perhaps this will provide a clue as to how the ICE-family proteases first become activated. In experimental systems. Fas-mediated apoptosis is usually induced by receptor clustering mediated by an anti-Fas IgM antibody. The physiological inducer, however, is the Fas ligand (FasL). FasL, a member of the TNF family, is a 45-kDa transmembrane glycoprotein containing 281 amino acids. It was first purified and then cloned from a C3i:otoxic T-lymphocyte hybridoma (176,179). FasL is encoded by an 8-kb gene containing four exons that is found on chromosome lq23. Expression of Fas is fairly widespread, including liver, heart (183), lung, and ovary (150), in addition to neutrophils, monoc5^es (87), and T cells and B cells at certain stages of their development (19,87,140). Because of its wide distribution. Fas-mediated apoptosis can be induced in vivo in other tissues if the proper stimulus is used. For example, injection of an anti-Fas antibody into the peritoneal cavities of mice caused rapid apoptosis of their thymocytes, as expected (139), but also resulted in the death of the animals due to apoptosis involving the hepatocytes (140). The expression of FasL is somewhat more limited. It occurs on lymphoid tissue and is upregulated on activated T cells, which can also secrete a soluble 26-kDa form of the ligand (177,184). It is also found at a relatively low concentration in lung and small intestine and at a somewhat higher concentration in testis, which synthesizes a slightly shorter form of the protein. Expression of FasL by Sertoli cells of the testis has been shown to be responsible for the basis of immune privilege in that organ (8). Similarly, expression of FasL in

REGULATION OF FAS-MEDIATED APOPTOSIS

89

the eye induces apoptosis in infiltrating lymphocytes and thereby confers immune privilege (60). In terms of normal physiology, Fas and FasL are chiefly involved with T-lymphocyte function. One function of the Fas-FasL system has to do with the development of immunological tolerance by T cells. Occupancy of the T-cell receptors of a set of lymphocytes that are CD4^CD8^ and are therefore of intermediate maturity upregulates Fas and induces expression of FasL, with subsequent destruction of the activated lymphocytes by Fas-mediated apoptosis (2,17,94,136,139, 192,211). The antigen by far most likely to occupy the receptor on such a cell is a self-antigen, so this sequence of events provides a mechanism to eliminate autoreactive T lymphocytes. More mature lymphoc3^es, which still display Fas but are positive only for CD4 or CDS, are resistant to Fas-mediated killing. Supporting a major role for Fas in the deletion of autoreactive lymphocytes is the finding that mice that are homozygous for the Ipr or the gld mutation, which inactivate Fas or FasL, respectively, accumulate large quantities of T cells (particularly CD4"CD8~ T cells), developing lymphadenopathy and an autoimmune disease resemblinglupus erythematosus (30,126,129,130,170,171,173). The Fas-FasL system is also involved in the destruction of targets by cytotoxic T lymphocytes and NK cells, both of which express FasL and kill cells that display Fas on their surfaces (5,156). Killing via the Fas-FasL system may be important in the host defense against viral diseases such as viral hepatitis (140). Sometimes the participation of the Fas-FasL system in antiviral defense constitutes too much of a good thing, as for example in the case of acquired immunodeficiency syndrome (AIDS), in which the depletion of CD4^ T cells results at least in part from sensitization of infected T cells to Fas-mediated killing owing to upregulation of Fas by the viral transcription factor HIV-1 Tat (98,197). B. Regulation of Fas-Mediated Apoptosis

Like numerous other biochemical processes. Fas-mediated apoptosis is regulated by protein phosphorylation. In cells susceptible to Fasmediated killing, several proteins are phosphorylated on tyrosine residues within a minute after ligation of Fas. The phosphorylation of these proteins, some of which are themselves phosphorylation-dependent kinases, is transient, lasting less than 30 min (40). This transient phosphorylation of tyrosines must be relevant to Fas-induced apoptosis, however, because inhibitors of tyrosine kinases will delay apoptosis in these cells (40). The effects of tyrosine kinase inhibitors can be partly explained by the finding that with cytotoxic T cells, at least, the upregu-

90

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

lation of FasL that occurs on stimulation with antigen or with a protein kinase C activator is blocked by inhibitors of tyrosine kinases (4). The protein kinase PITSLRE was produced and activated (by proteolysis) under conditions suggesting t h a t it could participate in the initiation of apoptosis (113). Further evidence supporting protein phosphorylation as an important positive regulator of Fas-mediated apoptosis are the results of Sato et al. showing t h a t FAP-1, a protein-tyrosine phosphatase t h a t associates with the intracellular C terminus of Fas, inhibited Fas-mediated apoptosis (161); the demonstration by Takizawa et al. t h a t Fas expression was increased in cells infected with influenza virus by the interferon-mediated activation of the double-stranded RNA-dependent kinase (181); and the finding of Haldar et al. t h a t the antiapoptosis protein Bcl-2 was inactivated by phosphorylation on serine (66). Dephosphorylation, however, may also be associated with apoptosis, and phosphorylation may delay or prevent it. When the Tcell receptor on cytotoxic T cells was occupied by antigen, causing the upregulation of FasL, phosphotyrosines on the receptor subunits themselves were partly dephosphorylated (144). The upregulation of FasL under these circumstances was blocked not only by tyrosine kinase inhibitors, as mentioned above, but also by cyclosporine, an inhibitor of the calcium-dependent protein-tyrosine phosphatase calcineurin (4). In CEM-6 cells, initiation of apoptosis requires a specific phosphatase, the "hematopoietic cell protein tyrosine phosphatase," and inhibition of this phosphatase by pervanadate prevents apoptosis in these cells (175). Fas-mediated killing of cultured hepatocytes is aggravated by inhibitors of protein kinase C, suggesting t h a t certain protein(s), when phosphorylated on serines and/or threonines, are able to protect these cells against apoptosis (132). Finally, the phosphorylation of the antioncoprotein Rb by the cyclin E/Cd2 kinase prevents apoptosis by allowing cells to pass through the Gi/S checkpoint and begin to proliferate (26). The actions of protooncogenes and protoantioncogenes also affect Fas-mediated apoptosis. The antiapoptotic effect of phosphorylated Rb was discussed in the preceding paragraph. The transcription factor p53, however, tends to promote apoptosis by inducing the production of Fas (146) and by upregulating p21-WAFl/CIPl, which blocks the passage of cells through the Gi/S checkpoint by inhibiting cyclin-dependent kinases (26). Apoptosis induced in T-cell hybridomas by ligation of the T-cell receptor, a process mediated through the Fas-FasL system (208), is blocked by antisense oligonucleotides against Q-myc and by a dominant negative mutation of Max, the protein t h a t binds to Myc to form the active Myc • Max transcription factor (58,121). Apoptosis in

REGULATION OF FAS-MEDIATED APOPTOSIS

91

this system was also blocked by Y-abl, which encodes a tyrosine kinase whose production is regulated in a defective manner (58). The participation of the low molecular weight guanine nucleotide-binding protein Ras in Fas-mediated apoptosis is suggested by the finding that in certain lymphoid lines Ras is activated by ligation of Fas, and that apoptosis in this system is partly blocked by inhibition of Ras (63). As discussed above, Bcl-2 and related proteins, certain Asp-X-specific proteases, and ceramide appear to be specific regulators of apoptosis. Like other factors that affect apoptosis, the effects of these specific regulators vary from cell to cell, even though the method of inducing apoptosis—i.e., ligation of F a s t i s the same in each case. Thus, Bcl-2 effectively inhibits Fas-induced apoptosis in Jurkat cells, IL-3dependent FDC-Pl cells, and the WR19L lymphoma line (85) but has no effect on T-cell receptor-induced apoptosis of T-cell hybridomas or on Fas-mediated target cell killing by cytotoxic T lymphocytes (28,58). Proteases participate extensively but not universally in Fas-mediated apoptosis. For example, thymocytes from mice in which ICE had been knocked out were resistant to Fas-mediated apoptosis (but not to apoptosis in response to steroids or ionizing radiation) (111), and Fasmediated apoptosis in many systems is blocked by antiproteases, including the very specific vaccinia virus serpin CrmA (29,42,118,122, 131,162,187), but Fas-dependent apoptosis of CD4+ or CD8+ lymphoblasts is not inhibited by antiserine or anticysteine proteases (159). The participation of ceramide in Fas-mediated apoptosis is suggested by the finding that ceramide has been reported to appear in cells in which Fas is occupied, and that exposure of L cells to a Fas ligand activates an acidic sphingomyelinase (32,63). Other factors are involved in the regulation of Fas-mediated apoptosis. In cultured hepatocytes, short-lived proteins that protect against apoptosis seem to be produced, because Fas-mediated apoptosis in these cells is aggravated by cycloheximide and actinomycin D (134). In the B-cell line FMO, Fas ligation induces a rise in calcium concentration that is necessary for DNA ladder formation and cell fragmentation, because calcium chelators prevent these changes in cells in which Fas is occupied (145). Apoptosis of T-cell hybridomas in which the T-cell receptor has been occupied is prevented by retinoic acid and glucocorticoids, probably because these agents prevent the upregulation of FasL on the hybridoma cells (208). Apoptosis in gliomas is regulated by cytokines via their ability to regulate the expression of Fas (196). The foregoing show the complexity of the mechanisms that regulate Fasmediated apoptosis.

92

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

IV. Conclusion Features of many of the regulatory mechanisms of biological systems are becoming clear, at least in outline. New information about receptors, G proteins, and phospholipases has dispelled some of the mists surrounding signal transduction. An understanding of cyclins and related proteins has yielded important insights into how cell division is regulated. Discoveries of homeoboxes and master genes have even shed some light on the mysteries of embryogenesis and development. Apoptosis, however, remains largely unmapped—a riddle wrapped in a mystery inside an enigma (31). Many novel events occur in cells undergoing apoptosis—chromatin degradation, acidification, alterations of surface lipids, proteolysis, cell fragmentation, and others. Which of these are essential for apoptosis and which are epiphenomena? How are the death proteases activated, and what are their substrates? What is Bcl2 doing? What is the role of ceramide? Is there a single mechanism underlying apoptosis, or are cell types undergoing apoptosis like unhappy families, each unhappy in its own way (190)? Penty of work remains before these questions can begin to be answered. ACKNOWLEDGMENT Supported in part by USPHS Grants AG-13501 and AI-24227 and by a grant from Sandoz. R.A.G. is the recipient of Grant K08 AI-01345. REFERENCES 1. Aggarwal, B. B., Singh, S., LaPushin, R., and Totpal, K. (1995). Fas antigen signals prohferation of normal human diploid fibroblast and its mechanism is different from tumor necrosis factor receptor. FEBS Lett. 364, 5-8. 2. Alderson, M. R., Tough, T. W., Davis-Smith, T., Braddy, S., Falk, B., Schooley, K. A., Goodwin, R. G., Smith, C. A., Ramsdell, F., and Lynch, D. H. (1995). Fas ligand mediates activation-induced cell death in human T lymphocytes. J. Exp. Med. 181, 71-77. 3. Alnemri, E. S., Fernandes-Alnemri, T., and Litwack, G. (1995). Cloning and expression of four novel isoforms of human interleukin-1 beta converting enz5mae with different apoptotic activities. J. Biol. Chem 270, 4312-4317. 4. Anel, A., Buferne, M., Boyer, C, Schmitt-Verhulst, A. M., and Golstein, P. (1994). T cell receptor-induced Fas ligand expression in cytotoxic T lymphocyte clones is blocked by protein tyrosine kinase inhibitors and cyclosporin A. Eur. J. Immunol. 24, 2469-2476. 5. Arase, H., Arase, N., and Saito, T. (1995). Fas-mediated cjrtotoxicity by freshly isolated natural killer cells. J. Exp. Med. 181, 1235-1238. 6. Bargou, R. C, Bommert, K, Weinmann, P., Daniel, P. T., Wagener, C, Mapara, M. Y. and Dorken, B. (1995). Induction of Bax-alpha precedes apoptosis in a human B lymphoma cell line: Potential role of the bcl-2 gene family in surface IgM-mediated apoptosis. Eur. J. Immunol. 25, 770-775. 7. Behrmann, I., Walczak, H., and Krammer, P. H. (1994). Structure of the human APO-1 gene. Eur. J. Immunol. 24, 3057-3062.

REGULATION OF FAS-MEDIATED APOPTOSIS

93

8. Belgrau, D., Gold, D., Selawry, H., Moore J, J., Franzusoff, A., and Duke, R. C. (1995). A role for CD95 ligand in preventing graft rejection. Nature (London) 377, 630-632. 9. Benito, A., Grillot, D., Nunez, G., and Fernandez-Luna, J. L. (1995). Regulation and function of Bcl-2 during differentiation-induced cell death in HL-60 promyelocytic cells. Am J Pathol. 146, 481-490. 10. Berke, G. (1995). The CTL's kiss of death (review). Cell (Cambridge, Mass.) 8 1 , 9 - 1 2 . 11. Bielawska, A., Crane, H. M., Liotta, D., Obeid, L. M., and Hannun, Y. A. (1993). Selectivity of ceramide-mediated biology. Lack of activity of erjrthro-dihydroceramide. J. Biol. Chem. 268, 26226-26232. 12. Birnbaum, M. J., Clem, R. J., and Miller, L. K. (1994). An apoptosis-inhibiting gene from a nuclear polyhedrosis virus encoding a polypeptide with Cys/His sequence motifs. J. Virol. 68, 2521-2528. 13. Boise, L. H., Gonzalez-Garcia, M., Postema, C. E. et al. (1993). bcl-x, a 6c/-2-related gene t h a t functions as a dominant regulator of apoptotic cell death. Cell (Cambridge, Mass.) 74, 597-608. 14. Boldin, M. P., Varfolomeev, E. E., Pancer, Z., Mett, I. L., Camonis, J. H., and Wallach, D. (1995). A novel protein t h a t interacts with the death domain of Fas/ A P O l contains a sequence motif related to the death domain. J. Biol. Chem. 270, 7795-7779. 15. Boldin, M. P., Mett, I. L., and Wallach, D. (1995). A protein related to a proteasomal subunit binds to the intracellular domain of the p55 TNF receptor upstream to its "death domain." FEBS Lett. 367, 39-44. 15a. Boldin, M. P., Goncharov, T. M. Goltsev, Y. V., and Wallach, D. (1996). Involvement of MACH, a novel MORTl/FADD-interacting protease, in Fas/APO-1- and TNF receptor-induced cell death. Cell (Cambridge, Mass.) 85, 803-815. 16. Boudreau, N., Sympson, C. J., Werb, Z., and Bissell, M. J. (1995). Suppression of ICE and apoptosis in mammary epithelial cells by extracellular matrix. Science 267,891-893. 17. Brunner, T., Mogil, R. J., LaFace, D., Yoo, N. J., Mahboubi, A., Echeverri, F., Martin, S. J., Force, W. R., Lynch, D. H., Ware, C. F., and Green, D. R. (1995). Cell-autonomous Fas (CD95)/Fas-ligand interaction mediates activation-induced apoptosis in T-cell hybridomas (see comments). Nature (London) 373, 441-444. 18. Bump, N. J., Hackett, M., Hugunin, M., Seshagiri, S., Brady, K , Chen, P., Ferenz, C , Franklin, S., Ghayur, T., Li, P., Licari, P., Mankovich, J., Shi, L., Greenberg, A. H., Miller, L. K , and Wong, W. W. (1995). Inhibition of ICE family proteases by baculovirus antiapoptotic protein p35. Science 269, 1885-1888. 19. Carsetti, R., Kohler, G., and Lamers, M. C. (1995). Transitional B cells are the target of negative selection in the B cell compartment. J. Exp. Med. 181,2129-2140. 20. Carson, W. E., Haldar, S., Baiocchi, R. A., Croce, C. M., and Caligiuri, M. A. (1994). The c-kit ligand suppresses apoptosis of h u m a n natural killer cells through the upregulation of bcl-2. Proc. Natl. Acad. Sci. U.S.A. 9 1 , 7553-7557. 21. Cartier, J. L., Hersberger, P. A., and Friesen, P. D. (1994). Suppression of apoptosis in insect cells stably transfected with baculovirus p35: Dominant interference by N-terminal sequences p35(l-76). J. Virol. 68, 7728-7737. 22. Cascino, I., Fiucci, G., Papoff, G., and Ruberti, G. (1995). Three functional soluble forms of the h u m a n apoptosis-inducing Fas molecule are produced by alternative splicing. J. Immunol. 154, 2706-2713. 23. Casciola-Rosen, L. A., Miller, D. K , Anhalt, G. J., and Rosen, A. (1994). Specific cleavage of the 70-kDa protein component of the U l small nuclear ribonucleoprotein

94

24. 25.

26. 27.

28.

29.

30.

31. 32.

33. 34. 35. 36. 37.

38.

39.

40.

41.

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR is a characteristic biochemical feature of apoptotic cell death. J. Biol. Chem. 269, 30757-30760. Ceretti, D. P., and Kozlosky, C. J. (1992) Molecular cloning of the interleukin-1 beta converting enzyme. Science 256, 97-100. Chen, M., Quintans, J., Fuks, Z., Thompson, C , Kufe, D. W., and Weichselbaum, R. R. (1995). Suppression of Bcl-2 messenger RNA production may mediate apoptosis after ionizing radiation, tumor necrosis factor alpha, and ceramide. Cancer Res. 55, 991-994. Chiarugi, V., MagneUi, L., Cinelli, M., and Basi, G. (1994). Apoptosis and the cell cycle (review). Cell. Mol. Biol. Res. 40, 603-612. Chinnaiyan, A. M., O'Rourke, K , Tewari, M., and Dixit, V. M. (1995). FADD, a novel death domain-containing protein, interacts with the death domain of Fas and initiates apoptosis. Cell (Cambridge, Mass.) 8 1 , 505-512. Chiu, V. K., Walsh, C. M., Liu, C. C , Reed, J. C., and Clark, W. R. (1995). Bcl-2 blocks degranulation but not Fas-based cell-mediated cytotoxicity. J. Immunol. 154,2023-2032. Chow, S. C , Weis, M., Kass, G. E., Holmstrom, T. H., Eriksson, J. E., and Orrenius, S. (1995). Involvement of multiple proteases during Fas-mediated apoptosis in T lymphocytes. FEBS Lett. 364, 134-138. Chu, J. L., Ramos, P., Rosendorff, A., Nikolic-Zugic, E., Lacy, E., Matsuzawa, A., and Elkon, K. B. (1995). Massive upregulation of the Fas ligand in Ipr and gld mice: Implications for Fas regulation and the graft-versus-host disease-like wasting syndrome. J. Exp. Med. 181, 393-398. Churchill, W. L. S. (1939). Speech, Oct. 1. Cifone, M. G., De Maria, R., RoncaioH, P., Rippo, M. R., Azuma, M., Lanier, L. L., Santoni, A., and Testi, R. (1994). Apoptotic signaling through CD95 (Fas/Apo-1) activates an acidic sphingomyelinase. J. Exp. Med. 180, 1547-1552. Clem, R. J., and Miller, L. K. (1994). Control of programmed cell death by the baculovirus genes p 5 5 and iap. Mol. Cell. Biol. 14, 5212-5222. Cleveland, J. L., Ihle, J. N. (1995). Contenders in FasL/TNF death signaling. Cell (Cambridge, Mass.) 8 1 , 479-482. Cory, S. (1995). Regulation of lymphoc3rte survival by the bcl-2 gene family (review). Annu. Rev. Immunol. 13, 513-543. D'Amico, A. V., and McKenna, W. G. (1994). Apoptosis and a re-investigation of the biologic basis for cancer therapy. Radiother. Oncol. 33, 3-10. Darmon, A. J., Nicholson, D. W., and Bleackley, R. C. (1995). Activation of the apoptotic protease CPP32 by cytotoxic T-cell-derived granzyme B. Nature (London) 377, 446-448. Dawson, C. W., Eliopoulos, A. G., Dawson, J., and Young, L. S. (1995). B H R F l , a viral homologue of the Bcl-2 oncogene, disturbs epithelial cell differentiation. Oncogene 10, 69-77. Dawson, S. P., Arnold, J. E., Mayer, N. J., Reynolds, S. E., Billett, M. A., Gordon, C , CoUeaux, L., Kloetzel, P. M., Tanaka, K , and Mayer, R. J., (1995). Developmental changes of the 26 S proteasome in abdominal intersegmental muscles oiManduca sexta during programmed cell death. J. Biol. Chem. 270, 1850-1858. Eischen, C. M., Dick, C. J., and Leibson, P. J., (1994). Tyrosine kinase activation provides an early and requisite signal for Fas-induced apoptosis. J. Immunol. 153, 1947-1954. Ellis, H. M., and Horvitz, J. R. (1986). Genetic control of programmed cell death in the nematode C. elegans. Cell (Cambridge, Mass.) 44, 817-829.

REGULATION OF FAS-MEDIATED APOPTOSIS

95

42. Enari, M, Hug, H., and Nagata, S. (1995). Involvement of an ICE-like protease in Fas-mediated apoptosis. Nature {London) 375, 7 8 - 8 1 . 43. Fadok, V. A., Savill, J. S., Haslett, C, Bratton, D. L., Doherty, D. E., Campbell, P. A., and Henson, P. M. (1992). Different populations of macrophages use either the vitronectin receptor or the phosphatidylserine receptor to recognize and remove apoptotic cells. J. Immunol. 149, 4029-4035. 44. Fang, W., Rivard, J. J., Ganser, J. A., LeBien, T. W., Nath, K. A., Mueller, D. L., and Behrens, T. W. (1995). BC1-XL rescues WEHI 231 B lymphocytes from oxidantmediated death following diverse apoptotic stimuli J. Immunol. 155, 66-75. 45. Fang, W., Rivard, J. J., Mueller, D., and Behrens, T. W. (1994). Cloning and molecular characterization of mouse Bcl-x in B and T lymphoc5d:es. J. Immunol. 1 5 3 , 4 3 8 8 4398. 46. Faucheu, C , Diu, A., Chang, A. W., Blanchet, A. M., Miossec,C, Herve, F., CollardDutilleul, v., Gu, Y., Aldape, R. A., Lippke, J. K., et al. (1995). A novel h u m a n protease similar to the interleukin-1 beta converting enzyme induces apoptosis in transfected cells. EMBO J. 14, 1914-1922. 47. Fernandes-Alnemri, T., Litwack, G., and Alnemri, E. S. (1994). CPP32, a novel h u m a n apoptotic protein with homology to Caenorhabditis elegans cell death protein Ced-3 and mammalian interleukin-1 beta-converting enzyme. J. Biol. Chem. 269,30761-30764. 48. Fernandes-Alnemri, T., Litwack, G., and Alnemri, E. S. (1995). Mch2, a new member of the apoptotic Ced.3/Ice cysteine protease gene family. Cancer Res. 55,2737-2742. 49. Franceschi, C. (1989). Cell proliferation, cell death and dying (review). Aging 1, 3-15. 50. Frisch, S. M., and Francis, H. (1994). Disruption of epithelial cell-matrix interactions induces apoptosis J. Cell Biol. 124, 619-626. 51. Gagliardini, V., Fernandez, P. A., Lee, R. K , Drexler, H. C , Rotello, R. J., Fishman, M. C , and Yuan, J. (1994). Prevention of vertebrate neuronal death by the crmA gene. Science 263, 826-828. 52. Gillardon, F., Eschenfelder, C , Uhlmann, E., Hartschuh, W., and Zimmermann, M. (1994). Differential regulation oic-fos, fosB, c-junJunB, bcl-2 and bax expression in r a t skin following single or chronic ultraviolet irradiation and in vivo modulation by antisense oligodeoxynucleotide superfusion. Oncogene 9, 3219-3225. 53. Golstein, P., Marguet, D., and Depraetere, V. (1995). Homology between reaper and the cell death domains of Fas and T N F R l (letter). Cell (Cambridge, Mass.) 81, 185-186. 54. Compel, A., Sabourin, J. C , Martin, A., Yaneva, H., Audouin, J., Decroix, Y., and Poitout, P. (1994). Bcl-2 expression in normal endometrium during the menstrual cycle. Am. J. Pathol. 144, 1195-1202. 55. Gonzalez-Garcia, M., Garcia, I., Ding, L., O'Shea, S., Boise, L. H., Thompson, C. B., and Nunez, G. (1995). Bcl-x is expressed in embryonic and postnatal neural tissues and functions to prevent neuronal cell death. Proc. Natl. Acad. Sci. U.S.A. 92,4304-4308. 56. Gottlieb, R. A., Burleson, K. O., Kloner, R. A., Babior, B. M., and Engler, R. L. (1994). Reperfusion injury induces apoptosis in rabbit cardiomyocytes. J. Clin. Invest. 94, 1621-1628. 57. Gottschalk, A. R., McShan, C. L., Kilkus, J., Dawson, G., and Quintans, J. (1995). Resistance to anti-IgM-induced apoptosis in a WEHI-231 subline is due to insufficient production of ceramide. Eur. J. Immunol. 25, 1032-1038.

96

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

58. Gottschalk, A. R., and Quintans, J. (1995). Apoptosis in B lymphocytes: The WEHI231 perspective (review). Immunol. Cell Biol. 73, 8-16. 59. Green, D. R., Mahboubi, A, Nishioka, W., Oja, S., Echeverri, F., Shi, Y., Glynn, J., Yang, Y., Ashwell, J., and Bissonnette, R. (1994). Promotion and inhibition of activation-induced apoptosis in T-cell hybridomas by oncogenes and related signals (review). Immunol. Rev. 142, 321-342. 60. Griffith, T. S., Brunner, T., Fletcher, S. M., Green, D. R., and Ferguson, T. A. (1995). Fas ligand-induced apoptosis as a mechanism of immune privilege. Science 270, 1189-1192. 61. Grimm, L., and Osborne, B. (1995). Proteasome inhibition in thymocyte apoptosis. Abstr. Meet. Programmed Cell Death, 1995. Cold Spring Harbor Laboratory, New York, p. 61. 62. Gu, Y., Sarnecki, C., Aldape, R. A., Livingston, D. J., and Su M. S. (1995). Cleavage of poly(ATP-ribose)polymerase by interleukin-1 beta converting enzyme and its homologs TX and Nedd-2. J. Biol. Chem. 270, 18715-18718. 63. Gulbins, E., Bissonnette, R., Mahboubi, A., Martin, S., Nishioka, W., Brunner, T., Baier, G., Baier-Bitterlich, G., B3rrd, C , Lang, F.etal. (1995). FAS-induced apoptosis is mediated via a ceramide-initiated RAS signaling pathway. Immunity 2 , 3 4 1 - 3 5 1 . 64. Haas, A. L., Baboshina, O., Williams, B. M., and Schwartz, L. M. (1995). Coordinated induction of the ubiquitin conjugation pathway accompanies the developmentally programmed death of insect skeletal muscle. J. Biol. Chem. 270, 9407-9412. 65. Haimovitz-Friedman, A., Kan, C. C , Ehleiter, D., Persaud, R. S., McLoughlin, M., Fuks, Z., and Kolesnick, R. N. (1994). Ionizing radiation acts on cellular membranes to generate ceramide and initiate apoptosis. J. Exp. Med. 180, 525-535. 66. Haldar, S., Beatty, C , Tsujimoto, Y., and Croce, C. M. (1989). The bcl-2 gene encodes a novel G protein. Nature (London) 342, 195-198. 67. Haldar, S., Jena, N., and Croce, C. M. (1995). Inactivation of Bcl-2 by phosphorylation. Proc. Natl. Acad. Sci. U.S.A. 92, 4507-4511. 68. Hanada, M., Aime-Sempe, C , Sato, T., and Reed, J. C. (1995). Structure-function analysis of Bcl-2 protein. Identification of conserved domains important for homodimerization with Bcl-2 and heterodimerization with Bax. J. Biol. Chem. 270,11962— 11969. 69. Hannun, Y. A., and Obeid, L. M. (1995). Ceramide: An intracellular signal for apoptosis (review). Trends Biochem. Sci. 20, 73-77. 70. Hawkins, C. J., and Vaux, D. L. (1994). Analysis of the role of Bcl-2 in apoptosis (review). Immunol. Rev. 142, 127-139. 71. Hay, B. A., Wolff, T., and Rubin, G. M. (1994). Expression of baculovirus p35 prevents cell death in Drosophila. Development (Cambridge, UK) 120, 2121-2129. 72. Henderson, S., Huen, D., Rowe, M., Dawson, C , Johnson, G., and Rickinson, A. (1993). Epstein-Barr virus-coded B H R F l protein, a viral homologue of Bcl-2, protects h u m a n B cells from programmed cell death. Proc. Natl. Acad. Sci. U.S.A. 90, 8479-8483. 73. Hengartner, M. O., Ellis, R. E., and Horvitz, H. R. (1992). Caenorhabditis elegans gene ced-9 protects cells from programmed cell death. Nature (London) 356, 494-499. 74. Hengartner, M. O., and Horvitz, H. R. (1994). Activation of C. elegans cell death protein CED-9 by an amino-acid substitution in a domain conserved in Bcl-2. Nature (London) 369, 318-320. 75. Hengartner, M. O., and Horvitz, H. R. (1994). C. elegans cell survival gene ced-9 encodes a functional homolog of the mammalian proto-oncogene bcl-2. Cell (Cambridge, Mass.) 76, 665-676.

REGULATION OF FAS-MEDIATED APOPTOSIS

97

76. Hershberger, P. A., LaCount, D. J., and Friesen, P. D. (1994). The apoptotic suppressor p35 is required early during baculovirus replication and is targeted to the cytosol of infected cells. J. Virol. 68, 3467-3477. 77. Hockenbery, D., Nunez, G., Milliman, C., Schreiber, R. D., and Korsmeyer, S. J. (1990). Bcl-2 is an inner mitochondrial membrane protein t h a t blocks programmed cell death. Nature (London) 348, 334-336. 78. Hockenbery, D. M., Oltvai, Z. N., Yin, X.-M., Milliman, C. L., and Korsmeyer, S. J. (1993). Bcl-2 functions in an antioxidant pathway to prevent apoptosis. Cell (Cambridge, Mass.) 75, 241-251. 79. Horvitz, H. R. (1994). The ins and outs of programmed cell death during C. elegans development (review). Philos. Trans. R. Soc. London, B Ser. 345, 243-246. 80. Howard, A. D., Kostura, M. J., Thornberry, N., Ding, G. J. F., Limjuco, G., Weidner, J., Salley, J. P., Hogquist, K. A., Chaplin, D. D., Mumford, R. A., Schmidt, J. A., and Tocci, M. J. (1991). IL-1 converting enzyme requires aspartic acid residues for processing of the IL-1/3 precursor at two distinct sites and does not cleave 31-kDa IL-IQ;. J. Immunol. 147, 2964-2969. 81. Inazawa, J., Itoh, N., Abe, T., and Nagata, S. (1992). Assignment of the h u m a n Fas antigen gene (Fas) to 10q24.1. Genomics 14, 821-822. 82. Irmler, M., Hertig, S., MacDonald, H. R., Sadoul, R., Becherer, J. D., Proudfoot, A., Solari, R., and Tschopp, J. (1995). Granzyme A is an interleukin 1 beta-converting enzyme. J. Exp. Med. 181, 1917-1922. 83. Ito, A., and Horigome, K. (1995). Ceramide prevents neuronal programmed cell death induced by nerve growth factor deprivation. J. Neurochem. 65, 463-466. 84. Itoh, N., and Nagata, S. (1993). A novel protein domain required for apoptosis. Mutational analysis of h u m a n Fas antigen. J. Biol. Chem. 268, 10932-10937. 85. Itoh, N., Tsujimoto, Y., and Nagata, S. (1993). Effect of bcl-2 on Fas antigenmediated cell death. J. Immunol. 151, 621-627. 86. Itoh, N., Yonehara, S., Ishii, A., Yonehara, M., Mizushima, S., Sameshima, M., Hase, A., Seto, Y., and Nagata, S. (1991). The polypeptide encoded by the cDNA for h u m a n cell surface antigen Fas can mediate apoptosis. Cell (Cambridge, Mass.) 66, 233-243. 87. Iwai, K., Miyawaki, T., Takizawa, T., Konno, A., Ohta, K , Yachie, A., Seki, H. and Taniguchi, N. (1994). Differential expression of bcl-2 and susceptibility to anti-Fasmediated cell death in peripheral blood lymphocj^es, monocytes, and neutrophils. Blood 84, 1201-1208. 88. Jaattela, M., Benedict, M., Tewari, M., Shayman, J. A., and Dixit, V. M. (1995). Bcl-x and Bcl-2 inhibit TNF and Fas-induced apoptosis and activation of phospholipase A2 in breast carcinoma cells. Oncogene 10, 2297-2305. 89. Jacobson, M. D., and Raff, M. C. (1995). Programmed cell death and Bcl-2 protection in very low oxygen. Nature (London) 374, 814-816. 90. Jarvis, W. D., Fornari, F. A., Jr., Browning, J. L., Gewirtz, D. A., Kolesnick, R. N., and Grant, S. (1994). Attenuation of ceramide-induced apoptosis by diglyceride in h u m a n myeloid leukemia cells. J. Biol. Chem. 269, 31685-31692. 91. Jarvis, W. D., Kolesnick, R. N., Fornari, F. A., Traylor, R. S., Gewirtz, D. A., and Grant, S. (1994). Induction of apoptotic DNA damage and cell death by activation of the sphinogmyelin pathway. Proc. Natl. Acad. Sci. U.S.A. 91, 73-77. 92. Jayadev, S., Liu, B., Bielawska, A. E., Lee, J. Y., Nazaire, F., Pushkareva, M. Yu., Obeid, L. M., and Hannun, Y. A. (1995). Role for ceramide in cell cycle arrest. J. Biol. Chem. 270, 2047-2052.

98

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

93. Ji, L., Zhang, G., Uematsu, S., Akahori, Y., and Hirabayashi, Y. (1995). Induction of apoptotic DNA fragmentation and cell death by natural ceramide. FEBS Lett. 358, 211-214. 94. J u , S. T., Panka, D. J., Cui, H., Ettinger, R. el-Khatib, M. Sherr, D. H. Stanger, B. Z., and Marshak-Rothstein, A. (1995). Fas(CD95)/FasL interactions required for programmed cell death after T-cell activation (see comments). Nature (London) 373, 444-448. 95. Kamens, J., Paskind, M., Hugunin, M., Talanian, R. V., Allen, H., Banach, D., Bump, N., Hackett, M., Johnston, C. G., Li, P. et al. (1995). Identification and characterization of ICH-2, a novel member of the interleukin-1 beta-converting enzyme family of cysteine proteases. J. Biol. Chem. 270, 15250-15256. 96. Kamita, S. G., Mijima, K., and Maeda, S. (1993). Identification and characterization of the p35 gene of Bombyx mori nuclear polyhedrosis virus t h a t prevents virusinduced apoptosis. J. Virol. 67, 455-463. 97. Kane, D. J., Sarafian, T. A., Anton, R., Hahn, H., Gralla, E. B., Vallentine, J. S., Ord, T., and Bredesen, D. E. (1993). Bcl-2 inhibition of neural death: Decreased generation of reactive oxygen species. Science 262, 1274-1277. 98. Katsikis, P. D., Wunderlich, E. S., Smith, C. A., Herzenberg, L. A., and Herzenberg, L. A. (1995). Fas antigen stimulation induces marked apoptosis of T lymphocytes in h u m a n immunodeficiency virus-infected individuals. J. Exp. Med. 181, 2029-2036. 99. Kaufmann, S. H., Desnoyers, S., Ottaviano, Y., Davidson, N. E., and Poirier, G. G. (1993). Specific proteolytic cleavage of poly(ADP-ribose)polymerase: An early marker of chemotherapy-induced apoptosis. Cancer Res. 53, 3976-3985. 100. Kaufmann, S. H. (1989). Induction of endonucleolytic DNA cleavage in h u m a n acute myelogenous leukemia cells by etoposide, camptothecin, and other cytotoxic anticancer drugs: A cautionary note. Cancer Res. 49, 5870-5878. 101. Kerr, J. F., Winterford, C. M., and Harmon, B. V. (1994). Apoptosis. Its significance in cancer and cancer therapy (review). Cancer (Philadelphia) 73, 2013-2026. 102. Kiefer, M. C., Brauer, M. J., Powers, V. C., Wu, J. J., Umansky, S. R., Tomei, L. D., and Barr, P. J. (1995). Modulation of apoptosis by the widely distributed Bcl-2 homologue Bak. Nature (London) 374, 736-739. 103. Kinoshita, T., Yokota, T., Arai, K , and Miyajima, A. (1995). Regulation of Bcl-2 expression by oncogenic Ras protein in hematopoietic cells. Oncogene 10, 2 2 0 7 2212. 104. Kiplesund, K. M., Halgunset, J., Fjosne, H. E., and Sunde, A. (1988). Light microscopic morphometric analysis of castration effects in the different lobes of the r a t prostate. Prostate 13, 221-232. 105. Kolesnick, R., and Fuks, Z. (1995). Ceramide: A signal for apoptosis or mitogenesis? (comment) (review). J. Exp. Med. 181, 1949-1952. 106. Korsmeyer, S. J., Shutter, J. R., Veis, D. J., Merry, D. E., and Oltvai, Z. N. (1993). Bcl-2/Bax: A rheostat t h a t regulates an anti-oxidant pathway and cell death (review). Semin. Cancer Biol. 4, 327-332. 107. Krajewski, S., Bodrug, S., Gascoyne, R., Berean, K , Krajewska, M., and Reed, J. C. (1994). Immunohistochemical analysis of Mcl-l and Bcl-2 proteins in normal and neoplastic lymph nodes. Am. J. Pathol. 145, 515-525. 108. Krajewski, S., Bodrug, S., Krajewska, M., Shabaik, A., Gascoyne, R., Berean, K., and Reed, J. C. (1995). Immunohistochemical analysis of Mcl-l protein in h u m a n tissues. Differential regulation of Mcl-l and Bcl-2 protein production suggests a unique role for Mcl-l in control of programmed cell death in vivo. Am. J. Pathol. 146, 1309-1319.

REGULATION OF FAS-MEDIATED APOPTOSIS

99

109. Krajewski, S., Krajewska, M., Shabaik, A., Miyashita, T., Wang, H. G., and Reed, J. C. (1994). Immunohistochemical determination of in vivo distribution of Bax, a dominant inhibitor of Bcl-2. Am. J. Pathol. 145, 1323-1336. 110. Krajewski, S., Krajewska, M., Shabaik, A., Wang, H. G., Irie, S., Fong, L., and Reed, J. C. (1994). Immunohistochemical analysis of in vivo patterns of Bcl-x expression. Cancer Res. 54, 5501-5507. 111. Kuida, K., Lippke, J. A., Ku, G., Harding, M. W., Livingston, D. J., Su, M. S., and Flavell, R. A. (1995). Altered cytokine export and apoptosis in mice deficient in interleukin-1 beta converting enzyme. Science 267, 2000-2003. 112. Kumar, S., Kinoshita, M., Noda, M., Copeland, N. G., and Jenkins, N. A. (1994). Induction of apoptosis by the mouse Nedd2 gene, which encodes a protein similar to the product of the Caenorhabditis elegans cell death gene ced-3 and the mammalian IL-1 beta-converting enzyme. Genes Dev. 8, 1613-1626. 113. Lahti, J. M., Xiang, J., Heath, L. S., Campana, D., and Kidd, V. J. (1995). PITSLRE protein kinase activity is associated with apoptosis. Mol. Cell. Biol. 15, 1-11. 114. Laster, S. M., Wood, J. G., and Gooding, L. R. (1988). Tumor necrosis factor can induce both apoptotic and necrotic forms of cell death. J. Immunol. 141,2629-2634. 115. Lazebnik, Y. A., Kaufmann, S. H., Desnoyers, S., Poirier, G. G., and Earnshaw, W. C. (1994). Cleavage of poly(ADP-ribose)polymerase by a proteinase with properties like ICE. Nature (London) 371, 346-347. 116. Li, P., Allen, H., Banerjee, S., Franklin, S., Herzog, L., Johnston, C , McDowell, J., Paskind, M., Rodman, L., Salfeld, J. et al. (1995). Mice deficient in IL-1 betaconverting enzjmtie are defective in production of mature IL-1 beta and resistant to endotoxic shock. Cell (Cambridge, Mass.) 80, 401-411. 117. Linette, G. P., and Korsmeyer, S. J. (1994). Differentiation and cell death: Lessons from the immune system. Curr. Opin. Curr. Biol 6, 809-815. 118. Los, M., Van de Craen, M., Penning, L. C , Schenk, H., Westendorp, M., Baeuerle, P. A., Droge, W., Krammer, P. H., Fiers, W., and Schulze-Osthoff, K. (1995). Requirement of an ICE/CED-3 protease for Fas/APO-1-mediated apoptosis. Nature (London) 375, 8 1 - 8 3 . 119. Loten, J., and Sachs, L. (1995). Regulation of bcl-2, bcl-xi and bax in the control of apoptosis by hematopoietic cytokines and dexamethasone. Cell Growth Differ. 6,647-653. 120. Martin, S. J., Lennon, S. V., Bonham, A. M., and Cotter, T. G. (1990). Induction of apoptosis (programmed cell death) in h u m a n leukemic HL-60 cells by inhibition of RNA or protein synthesis. J. Immunol. 145, 1859-1867. 121. Martin, S. J., O'Brien, G. A., Nishioka, W. K , McGahon, A. J., Mahboubi, A., Saido, T. C , and Green, D. R. (1995). Proteolysis of fodrin (non-erythroid spectrin) during apoptosis. J. Biol. Chem. 270, 6425-6428. 122. Mashima, T., Naito, M., Kataoka, S., Kawai, H., and Tsuruo, T., (1995). Aspartatebased inhibitor of interleukin-1 beta-converting enzyme prevents antitumor agentinduced apoptosis in h u m a n myeloid leukemia U937 cells. Biochem. Biophys. Res. Commun. 209, 907-915. 123. McConkey, D. J., Hartzell, P., Nicotera, P., Wyllie, A. H., and Orrhenius, S. (1988). Stimulation of endogenous endonuclease activity in hepatocytes exposed to oxidative stress. Toxicol Lett. 42, 123-130. 124. Mekori, Y. A., Oh, C. K., and Metcalfe, D. D. (1995). The role of c-kit and its ligand, stem cell factor, in mast cell apoptosis (review). Int. Arch. Allergy Immunol. 107, 136-138.

100

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

125. Miura, M., Zhu, H., Rotello, R., Hartwieg, E. A., and Yuan, J. (1993). Induction of apoptosis in fibroblasts by IL-1 beta-converting enz5rme, a mammalian homolog of the C. elegans cell death gene ced-3. Cell (Cambridge, Mass.) 75, 653-660. 126. Mixter, P. F., Russell, J. Q., and Budd, R. C. (1994). Delayed kinetics of T lymphocyte anergy and deletion in Ipr mice J. Autoimmun. 7, 697-710. 127. Motoyama, N., Wang, F., Roth, K. A., Sawa, H., Nakayama, K , Negishi I., Senju, S., Zhang, Q., Fujii, S., and Loh, D. Y. (1995). Massive cell death of immature hematopoietic cells and neurons in Bcl-x-deficient mice. Science 267, 1506-1510. 128. Munday, N. A., Vaillancourt, J. P., Ali, A., Casano, F. J., Miller, D. K , Molineaux, S. M., Yamin, T.-T., Yu, V. L., and Nicholson, D. W. (1995). Molecular cloning and pro-apoptotic activity of ICEreill and ICEreilll, members of the ICE/CED-3 family of cysteine proteases. J. Biol. Chem. 270, 15870-15876. 128a. Muzio, M., Chinnaiyan, A. M., Kischkel, F. C , O'Rourke, K., Shevchenko, A., Ni, J., Scaffidi, C , Bretz, J. D., Zhang, M., Gentz, R., Mann, M., Krammer, P. H., Peter, M. E., and Dixit, V. M. (1996). FLICE, a novel FADD-homologous ICE/CED-3-like protease, is recruited to the CD95 (Fas/APO-1) death-inducing signaling complex. Cell 85, 817-827. 129. Nagata, S., and Suda, T. (1995). Fas and Fas ligand: Ipr andgld mutations (review). Immunol. Today 16, 3 9 - 4 3 . 130. Nagata, S. (1994). Mutations in the Fas antigen gene in Ipr mice (review). Semin. Immunol. 6, 3 - 8 . 131. Nakajima, H., Golstein, P., and Henkart, P. A. (1995). The target cell nucleus is not required for cell-mediated granzyme- or Fas-based cytotoxicity. J. Exp. Med. 181, 1905-1909. 132. Neamati, N., Fernandez, A., Wright, S., Kiefer, J., and McConkey, D. J. (1995). Degradation of lamin B1 precedes oligonucleosomal DNA fragmentation in apoptotic thymoc3^es and isolated th5rmocyte nuclei. J. Immunol. 154, 3788-3795. 133. Nguyen, M., Branton, P. E., Walton, P. A., Oltvai, Z. N., Korsmeyer, S. J., and Shore, G. C. (1994). Role of membrane anchor domain of Bcl-2 in suppression of apoptosis caused by ElB-defective adenovirus. J. Biol. Chem. 269, 16521-16524. 134. Ni, R., Tomita, Y., Matsuda, K , Ichihara, A., Ishimura, K , Ogasawara, J., and Nagata, S. (1994). Fas-mediated apoptosis in primary cultured mouse hepatocytes. Exp. Cell Res. 215, 332-337. 135. Nicholson, D. W., Ali, A., Thornberry, N. A., Vaillancourt, J. P., Ding, C. K , Gallant, M., Gareau, Y., Griffin, P. R., Labelle, M., Lazebnik, Y. A. et al. (1995). Identification and inhibition of the ICE/CED-3 protease necessary for mammalian apoptosis (see comments). Nature (London) 376, 3 7 - 4 3 . 136. Nishimura, Y., Ishii, A., Kobayashi, Y., Yamasaki, Y., and Yonehara, S. (1995). Expression and function of mouse Fas antigen on immature and mature T cells. J. Immunol. 154, 4395-4403. 137. Obeid, L. M., Linardic, C. M., Karolak, L. A., a n d H a n n u n , Y. A. (1993). Programmed cell death induced by ceramide. Science 259, 1769-1771. 138. Oehm, A., Behrmann, L, Falk, W., Pawlita, M., Maier, G., Klas, C , Li-Weber, M., Richards, S., Dhein, J., Trauth, B. C , Ponstingl, H., and Krammer, P. H. (1992). Purification and molecular cloning of the APO-1 cell surface antigen, a member of the tumor necrosis factor/nerve growth factor receptor superfamily. Sequence identity with the Fas antigen J. Biol. Chem. 267, 10709-10715. 139. Ogasawara, J., Suda, T., and Nagata, S. (1995). Selective apoptosis of CD4 + CD8 + thymoc3^es by the anti-Fas antibody. J. Exp. Med. 181, 485-491.

REGULATION OF FAS-MEDIATED APOPTOSIS

101

140. Ogasawara, J., Watanabe-Fukunaga, R., Adachi, M., Matsuzawa, A., Kasugai, T., Kitamura, Y., Itoh, N., Suda, T., and Nagata, S. (1993). Lethal effect of the antiFas antibody in mice. Nature {London) 364, 806-809. 141. Ohta, H., Sweeney, E. A., Masamune, A., Yatomi, Y., Hakomori, S. and Igarashi, Y. (1995). Induction of apoptosis by sphingosine in h u m a n leukemic HL-60 cells: A possible endogenous modulator of apoptotic DNA fragmentation occurring during phorbol ester-induced differentiation. Cancer Res. 55, 691-697. 142. Ohta, H., Yatomi, Y. Sweeney, E. A., Hakomori, S., and Igarashi, Y. (1994). A possible role of sphingosine in induction of apoptosis by tumor necrosis factor-alpha in h u m a n neutrophils. FEES Lett. 355, 267-270. 143. Oltvai, Z. N., Milliman, C. L., and Korsmeyer, S. J. (1993). Bcl-2 heterodimerizes in vivo with a conserved homolog, Bax, that accelerates programmed cell death. Cell {Cambridge, Mass.) 74, 609-619. 144. Orchansky, P. L., and Teh, H. S. (1994). Activation-induced cell death in proliferating T cells is associated with altered tyrosine phosphorylation of TCR/CD3 subunits. J. Immunol. 153, 615-622. 145. Oshimi, Y., and Miyazaki, S. (1995). Fas antigen-mediated DNA fragmentation and apoptotic morphologic changes are regulated by elevated cytosolic Ca2+ level. J. Immunol. 154, 599-609. 146. Owen-Schaub, L. B., Zhang, W., Cusack, J. C , Angelo, L. S., Santee, S. M., Fujiwara, T., Roth, J. A., Deisseroth, A. B., Zhang, W. W., Kruzel, E., and Radinsky, R. (1995). Wild-type h u m a n p53 and a temperature-sensitive m u t a n t induce Fas/APO-1 expression. Mol. Cell Biol. 15, 3032-3040. 147. Passarelli, A. L., and Miller, L. K. (1994). Identification and transcriptional regulation of the baculovirus lef-6 gene. J. Virol. 68, 4458-4467. 148. Pircher, H., MuUer, K. P., Kyewski, B. A., and Hengartner, H. (1992). Thymocytes can tolerize thymocytes by clonal deletion in vitro. Int. Immunol. 4, 1065-1069. 149. Quintans, J., Kilkus, J., McShan, C. L., Gottschalk, A. R., and Dawson, G. (1994). Ceramide mediates the apoptotic response of W E H I 2 3 1 cells to anti-immunoglobulin, corticosteroids and irradiation. Biochem. Biophys. Res. Commun. 202,710-714. 150. Quirk, S. M., Cowan, R. G., Joshi, S. G., and Henrikson, K. P. (1995). Fas antigenmediated apoptosis in h u m a n granulosa/luteal cells. Biol. Reprod. 52, 279-287. 151. Rabizadeh, S., LaCount, D. J., Friesen, P. D., and Bredesen, D. E. (1993). Expression of the baculovirus p35 gene inhibits mammalian neural cell death. J. Neurochem. 61, 2318-2321. 152. Ray, C. A., Black, R. A., Kronheim, S. R., Greenstreet, T. A., Sleath, P. R., and Salvesen, G. S. (1992). Pickup DJ. Viral inhibition of inflammation: Cowpox virus encodes an inhibitor of the interleukin-lj8-converting enz5ane. Cell {Cambridge, Mass.) 69, 597-604. 153. Reynolds, J. E., Yang, T., Qian, L., Jenkinson, J. D., Zhou, P., Eastman, A., and Craig, R. W. (1994). Mcl-l, a member of the Bcl-2 family, delays apoptosis induced by c-Myc overexpression in Chinese hamster ovary cells. Cancer Res. 54,6348-6352. 154. Richardson, B. C , Lalwani, N. D., Johnson, K. J., and Marks, R. M. (1994). Fas ligation triggers apoptosis in macrophages but not endothelial cells. Eur. J. Immunol. 24, 2640-2645. 155. Robaye, B., Doskeland, A. P., Suarez-Huerta, N., Doskeland, S. O., and Dumont, J. E. (1994). Apoptotic cell death analyzed at the molecular level by two-dimensional gel electrophoresis. Electrophoresis 15, 503-510. 156. Rothstein, T. L., Wang, J. K , Panka, D. J., Foote, L. C , Wang, Z., Stanger, B., Cui, H., Ju, S. T., and Marshak-Rothstein, A. (1995). Protection against Fas-depen-

102

157.

158.

159. 160.

161. 162. 163.

164. 165.

166.

167. 168.

169.

170.

171.

172.

173.

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR dent Thl-mediated apoptosis by antigen receptor engagement in B cells. Nature {London) 374, 163-165. Sabourin, J. C , Martin, A., Baruch, J., True, J. B., Gompel, A., and Poitout, P. (1994). Bcl-2 expression in normal breast tissue during the menstrual cycle. Int. J. Cancer 59, 1-6. Sambrano, G. R., and Steinberg, D. (1995). Recognition of oxidatively damaged and apoptotic cells by an oxidized low density lipoprotein receptor on mouse peritoneal macrophages: Role of membrane phosphatidylserine. Proc. Natl. Acad. Sci. U.S.A. 92, 1396-1400. Sarin, A., Nakajima, H., and Henkart, P. A. (1995). A protease-dependent TCRinduced death pathway in mature lymphocytes. J. Immunol. 154, 5806-5812. Sato, T., Hanada, M., Bodrug, S. et al. (1994). Interactions among members of the Bcl-2 protein family analyzed with a yeast two-hybrid system. Proc. Natl. Acad. Sci. U.S.A. 9 1 , 9238-9242. Sato, T., Irie, S., Kitada, S., and Reed, J. C. (1995). FAP-1: A protein tyrosine phosphatase that associates with Fas. Science 268, 411-415. Schlegel, J., Peters, L, and Orrenius, S. (1994). Isolation and partial characterization of a protease involved in Fas-induced apoptosis. FEBS Lett. 364, 139-42. Schulze-Osthoff, K, Krammer, P. H., and Droge, W. (1994). Divergent signalling via APO-1/Fas and the TNF receptor, two homologous molecules involved in physiological death. EMBO.J 13, 4587-4596. Schulze-Osthoff, K , WalczakH., Droge, W., and Krammer, P. H. (1994). Cell nucleus and DNA fragmentation are not required for apoptosis J. Cell Biol. 127, 15-20. Selvakumaran, M., Lin H. K , Miyashita, T., Wang, H. G., Krajewski, S. Reed, J. C., Hoffman, B., and Liebermann, D. (1994). Immediate early up-regulation of bax expression by p53 but not TGF beta 1: A paradigm for distinct apoptotic pathways. Oncogene 9, 1791-1798. Sentman, C. S., Shutter, J. R., Hockenberg, D., Kanagawa, O., and Korsmeyer, S. J. (1991). Bcl-2 inhibits multiple forms of apoptosis but not negative selection in thymocytes. Cell {Cambridge, Mass.) 67, 879-888. Shen, Y., and Shenk, T. E. (1995). Viruses and apoptosis (review). Curr. Opin. Genet. Dev. 5, 105-111. Shi, L., Kam, C. M., Powers, J. C., Aebersold, R., and Greenberg, A. H. (1992). Purification of three C3^otoxic lymphocyte granule serine proteases t h a t induce apoptosis through distinct substrate and target cell interactions. J. Exp. Med. 176, 1521-1529. Shimizu, S., Eguchi, Y., Kosaka, H., Kamiike W., Matsuda, H., and Tsujimoto, Y. (1995). Prevention of hypoxia-induced cell death by Bcl-2 and BC1-XL. Nature {London) S74, 811-813. Singer, G. G., Carrera, A. C., Marshak-Rothstein, A., Martinez, C , and Abbas, A. K. (1994). Apoptosis, Fas and systemic autoimmunity: The MRL-lpr/lpr model (review). Curr. Opin. Immunol. 6, 913-920. Sobel, E. S., Kakkanaiah, V. N., Rapoport, R. G, Eisenberg, R. A., and Cohen, P. L. (1995). The abnormal /pr double-negative T cell fails to proliferate in vivo. Clin. Immunol. Immunopathol. 74, 177-184. Stanger, B. Z., Leder, P., Lee, T. H., Kim E., and Seed, B. (1995). A novel protein containing a death domain t h a t interacts with Fas/APO-1 (CD95) in yeast and causes cell death. Cell {Cambridge, Mass.) 81, 513523. Steinberg, A. D. (1994). MRL-lpr/lpr disease: Theories meet Fas (review). Semin. Immunol. 6, 55-69.

REGULATION OF FAS-MEDIATED APOPTOSIS

103

174. Steller, H. (1995). Mechanisms and genes of cellular suicide. Science 267, 1 4 4 5 1449. 175. Su, X., Zhou, T., Wang, Z., Yang, P., Jope, R. S., and Mountz, J. D. (1995). Defective expression of hematopoietic cell protein tyrosine phosphatase (HCP) in lymphoid cell blocks Fas-mediated apoptosis. Immunity 2, 353-362. 176. Suda, T., and Nagata, S. (1994). Purification and characterization of the Fas-ligand t h a t induces apoptosis. J. Exp. Med. 179, 873-879. 177. Suda, T., Okazaki, T., Naito, Y., Yokota, T., Arai, N., Ozaki, S., Nakao, K. and Nagata, S. (1995). Expression of the Fas ligand in cells of T cell lineage. J. Immunol. 154, 3806-3813. 178. Sugimoto, A., Friesen, P. D., and Rothman, J. H. (1994). Baculovirus p35 prevents developmentally programmed cell death and rescues a ced-9 m u t a n t in the nematode Caenorhabditis elegans. EMBO J. 13, 2023-2028. 179. Takahashi, T., Tanaka, M., Inazawa, J., Abe, T., Suda, T., and Nagata, S. (1994). H u m a n Fas ligand: Gene structure, chromosomal location and species specificity. Int. Immunol. 6, 1567-1574. 180. Takayama, S., Sato, T., Krajewski, S., Kochel, K., Irie, S., Millan, J. A., and Reed, J. C. (1995). Cloning functional analysis of BAG-1: A novel Bcl-2-binding protein with anti-cell death activity. Cell (Cambridge, Mass.) 80, 279-284. 181. Takizawa, T., Fukuda, R., Miyawaki, T., Ghashi, K., and Nakanishi, Y. (1995). Activation of the apoptotic Fas antigen-encoding gene upon infiuenza virus infection involving spontaneously produced beta-interferon. Virology 209, 288-296. 182. Tamura, A., and Yui, K. (1995). Age-dependent reduction of Bcl-2 expression in peripheral T cells of Ipr and gld m u t a n t mice. J. Immunol. 155, 499-507. 183. Tanaka, M., Ito, H., Adachi, S., Akimoto, H., Nishikawa, T., Kasajima, T., Marumo, F., and Hiroe, M. (1994). Hypoxia induces apoptosis with enhanced expression of Fas antigen messenger RNA in cultured neonatal r a t cardiomyoc5i:es. Circ. Res. 75, 426-433. 184. Tanaka, M., Suda, T., Takahashi, T., and Nagata, S., (1995). Expression of the functional soluble form of h u m a n fas ligand in activated lymphocytes. EMBO J. 14, 1129-1135. 185. Tartaglia, L. A., Ayres, T. M., Wong, G. H., and Goeddel, D. V. (1993). A novel domain within the 55 kD TNF receptor signals cell death. Cell (Cambridge, Mass.) 74, 845-853. 186. Tewari, M., Beidler, D. R., and Dixit, V. M. (1995). CrmA-inhibitable cleavage of the 70kDa protein component of the U l small nuclear ribonucleoprotein during Fas- and tumor necrosis factor induced apoptosis. J. Biol. Chem. 270,18738-18741. 187. Tewari, M., and Dixit, V. M. (1995). Fas- and tumor necrosis factor-induced apoptosis is inhibited by the poxvirus crmA gene product. J. Biol. Chem. 270,3255-3260. 188. Tewari, M., Quan, L. T., O'Rourke, K , Desnoyers, S., Zeng, Z., Beidler, D. R., Poirier, G. G., Salvesen, G. S., and Dixit, V. M. (1995). Yama/CPP32 beta, a mammalian homolog of CED-3, is a CrmA-inhibitable protease t h a t cleaves the death substrate poly (ADP-ribose)polymerase. Cell (Cambridge, Mass.) 8 1 , 801-809. 189. Thornberry, N. A., and Bull, H. G. (1992). The novel heterodimeric cysteine protease is required for interleukin-1 beta processing in monocytes. Nature (London) 356, 768-744. 190. Tolstoy, L. N. (1912). "Anna Karenina." Alfred Knopf, New York. 191. Veis, D. J., Sorenson, C. M., Shutter, J. R., and Korsmeyer, S. J. (1993). Bcl-2deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell (Cambridge, Mass.) 75, 229-240.

104

ROBERTA A. GOTTLIEB AND BERNARD M. BABIOR

192. Vignaux, F., Vivier, E., Malissen, B., Depraetere, V., Nagata, S., and Golstein, P. (1995). TCR/CD3 coupling to Fas-based cytotoxicity. J. Exp. Med. 181, 781-786. 193. Voelkel-Johnson, C , Entingh, A. J., Wold, W. S., Gooding, L. R., and Laster, S. M. (1995). Activation of intracellular proteases is an early event in TNF-induced apoptosis. J. Immunol. 154, 1707-1716. 194. Walker, N. P., Talanian, R. V., Brady, K. D., Dang, L. C., Bump, N. J., Ferenz, C. R., Franklin S., Ghayur, T., Hackett, M. C., Hammill, L. D. et al. (1994). Crystal structure of the cysteine protease interleukin-1 beta-converting enzyme: A (p20/ pl0)2 homodimer. Cell {Cambridge, Mass.) 78, 343-352. 195. Wang, L., Miura, M., Bergeron, L., Zhu, H., and Yuan, J. (1994). Ich-1, an Ice/ ce(i-5-related gene, encodes both positive and negative regulators of programmed cell death. Cell (Cambridge, Mass.) 78, 739-750. 196. Weller, M., Frei, K., Groscurth, P., Krammer, P. H., Yonekawa, Y., and Fontana, A. (1994). Anti-Fas/APO-1 antibody-mediated apoptosis of cultured h u m a n glioma cells. Induction and modulation of sensitivity by cytokines. J. Clin. Invest. 94, 954-964. 197. Westendorp, M. O., Frank, R., Ochsendauer, C., Strieker, K., Dhein J., Walczak, H., Debatin, K. M., and Krammer, P. H. (1995). Sensitization of T cells to CD95mediated apoptosis by HIV-1 Tat and gp 120. Nature (London) 375, 497-500. 198. WiUiams, G. T., Smith, C. A., Spooncer, E., Dexter, T. M., and Taylor, D. R. (1990). Haemopoietic colony stimulating factors promote cell survival by suppressing apoptosis. Nature (London) 343, 76-79. 199. Williams, M. S., and Henkart, P. A. (1994). Apoptotic cell death induced by intracellular proteolysis. J. Immunol. 153, 4247-4255. 200. Wong, G. H., and Goeddel, D. V. (1994). Fas antigen and p55 TNF receptor signal apoptosis through distinct pathways. J. Immunol, 152, 1751-1753. 201. Wong, M., and Smulson, M. (1984). A relationship between nuclear poly(ADP ribosylation) and acetylation posttranslational modifications. 2. Histone studies. Biochemistry 23, 3272-3730. 202. Wright, S.C., Wei, Q. S., Zhong, J., Zheng, H., Kinder, D. H., and Larrick, J. W. (1994). Purification of a 24-kD protease from apoptotic tumor cells t h a t activates DNA fragmentation. J. Exp. Med. 180, 2113-2123. 202a. Wright, D. A., Futcher, B., Ghosh, P., and Geha, R. S. (1996). Association of h u m a n Fas (CD95) with a ubiquitin conjugating enzyme (UBC-FAP). J. Biol. Chem. 27, 31037-31043. 203. WyUie, A. H., Kerr, J. F., and Currie, A. R. (1980) Cell death: The significance of apoptosis (review). Int. Rev. Cytol. 68, 251-306. 204. Wyllie, A. H., Morris, R. G., Smith, A. L., and Dunop, D. (1984). Chromatin cleavage in apoptosis: Association with condensed chromatin morphology and dependence on macromolecular synthesis. J. Pathol. 142, 66-77. 205. Xue, D., and Horvitz, H. R. (1995). Inhibition of the Caenorhabditis elegans celldeath protease CED-3 by a CED-3 cleavage site in baculovirus p35 protein. Nature (London) 377, 248-251. 206. Yang, E., Zha, J., Jockel, J., Boise, L. H., Thompson, C. B., and Korsmeyer, S. J. (1995). Bad, a heterodimeric partner for BC1-XL and Bcl-2, displaces Bax and promotes cell death. Cell (Cambridge, Mass.) 80, 285-291. 207. Yang, T., Kozopas, K. M., and Craig, R. W. (1995). The intracellular distribution and pattern of expression of MCI-1 overlap with, but are not identical to, those of Bcl-2. J. Cell Biol. 128, 1173-1184.

REGULATION OF FAS-MEDIATED APOPTOSIS

105

208. Yang, Y., Mercep, M., Ware, C. F., and Ashwell, J. D. (1995). Fas and activationinduced Fas ligand mediate apoptosis of T cell hybridomas: Inhibition of Fas ligand expression by retinoic acid and glucocorticoids. J. Exp. Med. 181, 1673-1682. 209. Yao, B., Zhang, Y., Delikat, S., Mathias, S., Basu, S., and Kolesnick, R. (1995). Phosphorylation of Raf by ceramide-activated protein kinase. Nature (London) 378, 307-310. 210. Yin, X-M., Oltvai, Z. N., and Korsmeyer, S. J. BHl and BH2 domains of Bcl-2 are required for inhibition of apoptosis and heterodimerization with Bax. Nature (London) 369, 321-323. 211. Yonehara, S., Nishimura, Y., Kishil., S., Yonehara, M. Takazawa, K., Tamatani, T., and Ishii, A. (1994). Involvement of apoptosis antigen Fas in clonal deletion of human thymocytes. Int. Immunol. 6, 1849-1856. 212. Yuan, J., Shaham, S., Ledoux, S., EUis, H. M., and Horvitz, H. R. (1993). The C. elegans cell death gene ced-3 encodes a protein similar to mammalian interleukin1 beta-converting enzyme. Cell (Cambridge, Mass.) 75, 641-652. 213. Zhang, H., Saeed, B., and Ng, S. C. (1995). Combinatorial interaction of human Bcl-2 related proteins: Mapping of regions important for Bcl-2/Bcl-Xs interaction. Biochem. Biophys. Res. Commun. 208, 950-956.

This Page Intentionally Left Blank

CURRENT TOPICS IN CELLULAR REGULATION, VOLUME 35

Aging and Regulation of Apoptosis HuBER R. W A R N E R

Biology of Aging Program National Institute on Aging National Institutes of Health Bethesda, Maryland 20892

I. Introduction Cell death in biological systems can be separated into two distinct forms: necrotic death and programmed death or apoptosis (10,38). Necrosis results from massive cell injury and often is accompanied by inflammation. Apoptosis is a more subtle process, and until recently it v^as generally assumed to be a biological process whose major function is to destroy unwanted cells during development (72). This perception began to change in the 1980s, particularly as roles emerged for apoptosis in the negative selection of thymocytes and lymphocytes (53), in the attenuation of autoimmunity (55), and as a balancing factor in maintaining proliferative homeostasis (71,74,83). When cells become extensively damaged, apoptosis may also become either a preferred or an essential alternative to repair. This may be particularly important during the use of irradiation and drugs to damage and ultimately destroy cancer cells. Increasing evidence suggests that most mammalian cells may exist in a state of unstable equilibrium, poised to either proliferate or die, depending on the balance of factors impacting on them. Thus, most, if not all, of the cell death machinery may be present at all times, but whether it actually gets used depends on a variety of both extracellular and intracellular signals. The four most active areas of research on cell death concern the following questions: What are the components of the cell death machinery? What are the signals that trigger cell death? How are these signals transduced to the machinery? How is the overall process regulated? The beneficial physiological roles of apoptosis are counterbalanced by deleterious cell death occurring in a variety of pathological conditions, including the neurodegenerative diseases such as Alzheimer's disease, Parkinson's disease, and amyotrophic lateral sclerosis; ischemia; and acquired immunodeficiency syndrome (AIDS). Bonfoco et al. (5) have 107

108

RUBER R. WARNER

concluded that in neurons "mild excitotoxic or free radical insults result most often in apoptotic neuronal death, whereas necrosis predominates after intense, fulminate insults." Thus, it is particularly important to learn how apoptosis is regulated so that strategies can be devised to slow down or delay it in cells which cannot be replaced by cell proliferation. Franceschi et al. {21) have proposed that apoptosis is an important cellular defense mechanism in maintaining genetic stability and that centenarians have "aged successfully" because their cells are more prone to apoptosis. Evidence now suggests the possibility that changes in apoptosis may be an important factor in aging, and the purpose of this chapter is to review the evidence that is beginning to accumulate. For the purposes of this chapter, I will assume that apoptosis and programmed cell death refer to the same ultimate biological process, which is the rapid but controlled destruction of a cell following the activation of cell death proteins by any one of a variety of signals; for brevity I will refer to it only as apoptosis (10). In most cases, the exact signals which trigger this process have not been identified, but it will be assumed that the actual killing process (i.e., the machinery employed) is fairly universal, regardless of the initiation signal transduction pathway employed.

II. Genes Involved in Apoptosis The meteoric rise in research on apoptosis and the mechanisms of cell death stems directly from the pioneering work of Robert Horvitz and colleagues with the nematode Caenorhabditis elegans (15,68). This has been a useful model system because the developmental history of every cell in this organism has been determined; of the 1090 somatic cells formed, 131 are destined to die during development (15). Assuming that the death of these cells must be genetically programmed, Horvitz has identified cell death mutants (ced mutants) with defects in this developmental program. Mutations in three genes {ced-3, ced-4, and ced-9) affect the actual death of these cells, as summarized in Table I. Because apoptosis appears to be a fairly universal property of animal species, it is not surprising that mammals contain genes similar to the ced genes in C elegans. So far, at least 3 ced-3 homologs have been found in mammalian cells. Yuan et al. (86) showed that the ced-3 gene is 29% homologous to the mammalian gene for interleukin-lj8 converting enzyme (ICE), and transgenic experiments have shown that these two genes are functionally interchangeable (51). A similar protease activity has been described by Tewari et al. (70); this protease

109

AGING AND REGULATION OF APOPTOSIS TABLE I GENES INVOLVED IN APOPTOSIS

c.

elegans gene ced-3 ced-4 ced-9 ?

nuc-1

Biological function

Mammalian liomolog(s)

Required for apoptosis in C. elegans Required for apoptosis in C. elegans Blocks apoptosis in C. elegans Triggers induction of apoptosis Degradation ofDNA

ICE, nedd-2, Yama, CPP32i8 None known

Cysteine protease

70 86

Unknown

85

bcl-2

Unknown

29

Cell surface receptor

23 67

Endonuclease

15 31 36 48

fas, Apo-1 {Ipr gene) DNase I?

Other homolog(s)

Reaper? (Drosophila)

Biochemical activity

References

specifically cleaves poly(ADP-ribose) polymerase, an enzyme involved in DNA repair in eukaryotic cells. The rationale for the role of such a protease in apoptosis might be to preclude abortive attempts to repair the DNA, either in a heavily damaged cell or once DNA fragmentation has begun. The mammalian bcl-2 gene was originally identified as an oncogene because it was overexpressed as a result of translocations between human chromosomes 14 and 18 which are associated with B-cell lymphoma (8). Ultimately, it was shown that the bcl-2 gene product not only inhibits apoptosis in a variety of mammalian cells, but that it can also substitute for the ced-9 function when introduced into nematodes, and vice versa. Thus, the bcl-2 and ced-9 gene products appear to be functionally equivalent. The product of the ced-4 gene is required for apoptosis in C. elegans. A gene equivalent to ced-4 has not yet been identified in mammalian cells, nor has a specific function of the ced-4 gene product in nematodes been established. The three genes discussed above are involved in the actual killing of the cell. In mammals they are apparently activated in response to a variety of signals, including (1) withdrawal of hormones or growth

110

HUBER R. WARNER

factors, (2) flucocorticoids, (3) various protein factors (9), and (4) intracellular damage. An example is the binding of a ligand to the cell surface receptor Fas, also known as Apo-1 or CD95, which is a member of the tumor necrosis factor (TNF) family of receptors (67). Fas also shares sequence homology with the reaper {rpr) gene, which plays a role in apoptosis in Drosophila (23,82). Thus, the genes for Fas receptor protein and its homologs may also be considered cell death genes. There is no information yet available to explain how cell damage triggers apoptosis, but presumably it does not act through a cell surface receptor. A fifth gene which might be considered to be involved in cell death is the gene which encodes the nuclease responsible for the DNA fragmentation. In C. elegans this is the nuc-1 gene, which codes for a Ca^^independent endonuclease (15,31). In the absence of the nuc-1 gene product, the appropriate cells die but the DNA of the dead cells is not degraded, indicating that DNA degradation is neither responsible for nor essential for death of these cells. Wyllie (84) has postulated the activation of a Ca^^-dependent endonuclease in the death of thymocytes; DNase I is one candidate for this activity (36). This discussion of genes involved in apoptosis has focused entirely on the machinery of the process rather than on the signal transduction pathways involved. Several reviews (39,49,54,59) have discussed the latter in some detail. Current evidence indicates that the signal transduction pathways for apoptosis and the cell cycle overlap, which would explain why manipulation of intracellular levels of various participants in the cell cycle traverse may stimulate apoptosis. For example, apoptosis can be induced by c-myc expression under some conditions (16), and c-myc may be normally involved in apoptosis but may not play a unique role per se. Similarly, p21 and p53 gene expression may sensitize cells to apoptosis inducers, but these genes are not themselves physiological inducers of apoptosis (49). Ceramide appears to mediate the apoptotic response to some inducers, such as TNF-a, and can itself induce apoptosis, but it may do this indirectly as a general inducer of growth suppression (54). Thus, while there are many ways to experimentally induce apoptosis with components of signal transduction pathways, it is not clear whch of these play critical regulatory roles in vivo. Thus, the genes for these signal transduction participants have not been included in Table I.

III. Cell Senescence and Apoptosis Although senescent cells, by definition, can no longer proliferate, they are by no means dead or dying cells (42). In fact, they are remarkably

AGING AND REGULATION OF APOPTOSIS

111

resistant to stimuli which normally induce apoptosis, such as growth factor deprivation (75), suggesting t h a t senescence may be in some way linked to the potential to undergo apoptosis. This linkage has been investigated by Eugenia Wang and colleagues (28,75-77). They suggest t h a t when quiescent h u m a n fibroblasts attempt to reenter the cell cycle, the two possible outcomes are cell division and cell death. Both outcomes would require the expression of Gi genes such as Q-fos, c-myc, and cdc-2, and pRB phosphorylation. For example, Smeyne et al. (66) have shown t h a t continuous expression of c-fos precedes apoptosis. However, Seshadri and Campisi (64) have shown t h a t senescent cells fail to induce the c-fos gene on serum stimulation, and this failure provides one mechanism whereby senescent cells could be resistant to apoptosis signals. The failure of senescent cells to induce c-fos may be only one of several early cell cycle events which prevent a senescent cell from becoming apoptotic (59). Wang (76) has obtained evidence t h a t regulation of the bcl-2 gene is another factor in the resistance of h u m a n fibroblasts to apoptosis. Removal of serum from mouse 3T3 cells results in the disappearance of Bcl-2 protein and cell death within 24 hr, whereas it takes 2 weeks of serum deprivation to deplete young h u m a n fibroblasts of Bcl-2 protein. In contrast, Bcl-2 protein levels are unchanged in senescent fibroblasts after 2 weeks of serum deprivation. This inability to downregulate bcl-2 expression provides a second mechanism explaining the resistance of senescent cells to apoptosis in response to apoptosisinducing signals, but it is not known why senescent cells are less able than young cells to downregulate bcl-2 expression. Hebert et al. (28) have carried the analysis one step further, and have related the cellular commitment of h u m a n fibroblasts and mouse 3T3 cells to cell death to the metabolism of a protein they call terminin (Tp). Proliferating cells contain terminin in two major forms: 90 kDa (Tp-90) and 60 kDa (Tp-60), with Tp-90 being the predominant form. After serum deprivation, a 30-kDa (Tp-30) form appears and eventually becomes the major form of Tp observed. Up to 12 hr after serum removal, while the major terminin form is still Tp-60, 3T3 cells can be revived by serum replacement. By 48 h r after serum removal, 3T3 cells begin to die by apoptosis and can no longer be revived. The patterns are similar for h u m a n fibroblasts, except t h a t they are more resistant to apoptosis and the appearance of Tp-30 takes much longer. In senescent cells the major form of terminin is Tp-60, and these cells are particularly resistant to apoptosis, as indicated above. These investigators have concluded from these results t h a t Tp-30 is a marker for the commitment to cell death. It is not known whether Tp-30 plays a causal

112

HUBER R. WARNER

role in apoptosis, is a consequence of apoptosis, or merely correlates with induction of the cell death program. It is also not known what protease(s) catalyzes the conversion of Tp-90 to Tp-60 and Tp-30, but cycloheximide experiments suggest that the protease must be induced. It is possible that the conversion is catalyzed by ICE or one of the ICE-like proteases thought to be involved in apoptosis. The relative resistance of human fibroblasts, and presumably other human cells, to apoptosis may have mixed consequences. On the one hand, it may extend the life span of terminally differentiated cells, particularly neurons and myocytes, which cannot readily be replaced by cell proliferation. On the other hand, it may hinder elimination of damaged cells, leading to tissue dysfunction; it has been proposed that caloric restriction may extend the life span of rodents by upregulating the expression of genes required for apoptosis (79).

IV. Immunological Aging and Apoptosis It is well known that many parameters of immune function decline with increasing age (32), but how important this is in aging remains unclear. This decline could be due either to a decrease in the total number of lymphocytes produced or to an alteration in the kinds of lymphocytes produced (or both); current evidence best supports the latter alternative. For example. Miller (50) states: "The population of T cells in aging mice can be viewed as a changing mosaic of cells with different activation requirements and functional competencies." To understand what role apoptosis might play in this changing mosaic with aging, it is necessary to review briefly how T lymphocytes are produced (53). The cells that ultimately become T lymphocytes originate in bone marrow but migrate into the thymus, where they mature into T lymphocytes. Mature lymphocytes carry surface antigens, designated by CD, and are usually CD4+CD8- (helper T cells) or CD4-CD8+ (cytotoxic T cells). If these lymphocytes are able to recognize self-antigens, they undergo apoptosis (negative selection) before leaving the thymus. Usually more than 95% of the lymphocytes produced in the thymus die there, apparently by apoptosis; the survivors leave the thymus and migrate to the peripheral lymphoid organs, where additional negative selection then occurs. The Fas protein, also known as CD95, acts as a receptor on the surface of T lymphocytes and, when activated by an appropriate ligand, triggers apoptosis. Negative selection in the thymus appears to be Fas-

AGING AND REGULATION OF APOPTOSIS

113

independent (65), in contrast to peripheral negative selection, which is Fas-dependent. Fas-dependent apoptosis does not require de novo gene expression but is inhibited by bcl-2. This suggests either that the activation of Fas obviates the need for the products of the ced-3 and ced-4 functional homologs in mammals or that these gene products are already present and merely need to be activated. The latter seems to be true because Kuida et aL (40) have shown that thymocytes are sensitive to dexamethasone-induced, but not Fas-induced, apoptosis if the ICE gene has been inactivated. Fas also plays a role in regulating the number of activated T cells produced in response to a foreign antigen. Much of the early evidence indicating that Fas expression may be important in aging comes from the study of Ipr (lymphoproliferation) mice. The Ipr mice accumulate large numbers of immature CD4"CD8" lymphocytes in peripheral tissues (53), resulting in severe autoimmune disease, and are short-lived when the mutant Ipr gene is expressed in an MRL mouse (25). The Ipr gene is now known to be the structural gene for Fas in mice (80). A related autoimmune disease in mice is generalized lymopholiferative disease (gld), which is caused by mutations in the gene for the Fas ligand (69). These results confirm that both Fas and Fas ligand play important apoptotic roles in the peripheral organs in negative selection of T lymphocytes recognizing self-antigens. It is of interest that apoptosis can still be triggered by glucocorticoids in Ipr mice (73); presumably this Fas-independent apoptosis is primarily thymic. What is not so certain is whether the decreased immune competence or the increasing autoimmunity observed later in life is due to agerelated changes in expression of either the Ipr (fas) or the gld (fasL) gene. Zhou et al. (88) have found that both the number of thymocytes in the thymus and the fraction of those thymocytes which are Fas+ decrease substantially in old mice (22-26 months) compared to young mice (2 months). Both of these defects can be totally reversed by transgenic overexpression of a fas gene linked to the CD2 promoter. It is not clear whether the switch from Fas^ lymphocytes to Fas" thymocytes in the thymus with increasing age is due to actual downregulation of the fas gene itself with increasing age, but it is clear that apoptosis is reduced in old mice, and both thymus involution and this age-related decrease in Fas-dependent apoptosis can be reversed by expressing a fas transgene. Moreover, expression of the/as transgene prevents ageassociated decline in T-cell competence, suggesting that Fas-dependent apoptosis plays a critical role in maintaining immune competence, pos-

114

HUBER R. WARNER

sibly through the eUmination of senescent or actively immunosuppressive cells. Two papers on fas mutations in humans indicate that Fas is also critical in destroying self-reactive lymphoc3rtes (and thymocytes?) in humans. Rieux-Laucat et al. (62) showed that a large deletion in the fas gene leads to abnormal lymphoproliferation and some autoimmunity. Fisher et al, (19) identified human fas mutations resulting in either frameshift or splicing mutations; these individuals also presented with autoimmune lymphoproliferative syndrome. These results suggest the importance of determining whether Fas production is also downregulated in humans with increasing age, and if so, how this contributes to the increasing severity of autoimmune disease with increasing age. V. Caloric Restriction and Apoptosis Caloric restriction (CR) is an intervention which retards aging and extends the maximum life span of not only rodents but also a wide variety of other animals (81). CR has also been shown to delay the onset of age-related disease, including cancer. Although CR prevents a large number of age-related changes, it is not known whether any of these factors play a causal role in life-span extension or retardation of disease. Two papers suggest that a common mechanism may be at least partly responsible for both of these phenomena. Grasl-Kraupp et al. (26) and James and Muskhelishvili (34) have both obtained evidence that CR reduces the number of preneoplastic cells and spontaneous hepatomas and increases the number of apoptotic bodies found in mouse liver. They interpreted this to mean that CR upregulates the spontaneous level of apoptosis in mice. Warner et al. (79) subsequently hypothesized that upregulation of apoptosis by CR stimulates removal of damaged senescent cells, which are unusually resistant to apoptosis, as discussed in Section III. It is now known from the work of Dimri et al. (12) that senescent fibroblasts and keratinocytes are actually present in skin from old human donors and do accumulate with age. Therefore, if damaged senescent cells can be destroyed by apoptosis, restoration of functions characteristic of "younger" tissue would result; this would appear as a retardation of aging of the tissue and/or organism as a whole. Muskhelishvili et al. (52) have now extended their study to the entire life span of the mouse and have demonstrated that while there are modest age-related increases in spontaneous hepatic apoptosis rates in B6C3F1 mice, the apoptotic rate is significantly higher in CR mice

AGING AND REGULATION OF APOPTOSIS

115

than in mice fed ad libitum, at all ages between 12 and 30 months. The difficulty in interpreting these experiments unequivocally is that both aging and CR may alter the rate at which cell damage occurs; therefore, it is difficult to distinguish direct effects of CR on cell damage from effects on cellular sensitivity to apoptotic cell death. Also, in support of this hypothesis that CR upregulates apoptosis. Good et al. (25), Luan et al. (46), and Makinodan and James (47) have shown that CR increases T-cell apoptosis in Ipr mice, reverses lymphoproliferative disease, and extends the life span. If this hypothesis that CR upregulates apoptosis is correct, and if the mechanism by which it does this can be elucidated, then it might be possible to selectively mimic the beneficial effects of CR in humans without the actual implementation ofCR.

VI. Neurodegenerative Disease and Apoptosis Death of neurons can be induced either by removal of survival factors such as nerve growth factor, ciliary neurotrophic factor, and brainderived neurotrophic factor (61), or as a result of damage to the cell. One of the factors in the death of neurons leading to age-related neurodegenerative diseases such as Alzheimer's disease (AD), Parkinson's disease (PD), and amyotrophic lateral sclerosis (ALS), is thought to be low-level but chronic oxidative damage occurring in specific regions of the brain and peripheral nervous system (4). Cells are continuously exposed to oxidative stress in the form of superoxide anion, hydrogen peroxide, and hydroxyl radical as a necessary consequence of aerobic metabolism (78), and this stress results in oxidative damage to macromolecules such as DNA, proteins, and lipids. In postmitotic cells such as neurons, the damage to proteins and lipids may be much more critical than damage to DNA (27), although mitochondrial DNA deletions in the brain are known to increase dramatically with increasing age (11). For example, glutamine synthetase, which converts glutamate and ammonia into glutamine, is readily damaged by reactive oxygen species (56); thus, excessive oxidative stress in the brain could lead to accumulation of both toxic substrates (37). Another example is the putative nitration of proteins in the motor neurons of familial ALS patients due to the accumulation of superoxide anion and its subsequent conversion to peroxynitrite by the mutant form of superoxide dismutase (3). Oxidative stress also leads to peroxidation of the unsaturated fatty acid side chains in lipids. Peroxidation of membrane lipids leads to significant structural changes in the membrane, possibly altering the ability of the membrane to maintain cellular integrity.

116

HUBER R. WARNER

Many questions pertinent to these age-related neuronal deaths remain to be answered, such as: How does oxidative damage induce cell death? What determines the regional specificity of these deaths? Are these deaths due to necrosis, apoptosis, or both? Can these deaths be prevented, and if so, what inpact would this have on the course of the disease? Once again, the nematode C. elegans provides a possible paradigm for understanding the causes of neuronal death (33). A mutation called mec'4 results in the specific loss of mechanosensory neurons, which permit the nematode to respond to touch. All mec-4 mutations occur in a specific amino acid residue in the transmembrane region of a subunit of an ion channel (14). The normal amino acid at this site is alanine, and when amino acids with larger side chains are placed at this site, the mechanosensory cells die. The model is that when these large side chains are present, the ion channel cannot close properly, allowing sodium and calcium ions to leak in, causing swelling and eventually death. It seems possible that any membrane damage, such as by lipid peroxidation, which would affect the closing of ion channels in a comparable way would induce death in mammalian cells. A hypothalamic growth hormone-releasing cell line, GTl-7 (35), and PC 12 cells (87) have been used to study the death of neurons in vitro. These studies have shown that death can be induced by a variety of treatments such as glutathione depletion, glucose withdrawal, hydroperoxides, and calcium ionophores, but that in all cases transgenic expression of bcl-2 inhibits the death of these neuronal cells, whether by necrosis or apoptosis. These results suggest the possibility that the Bcl-2 protein may play some indirect role in decreasing damage and subsequent death due to the production of reactive oxygen species. Using sensory neurons from dorsal root ganglia of bcl-2 transgenic mice, Farlie et al. (17) also obtained evidence that Bcl-2 protein can protect neurons from apoptotic death, although Gonzalez-Garcia et al. (24) have suggested that BC1-XL, rather than Bcl-2, may be the physiological regulator of death in mature central nervous system neurons. Although much cell death occurs during development of the vertebrate nervous system (58), of particular interest in aging is the slow but steady loss of neurons due to normal brain oxidative metabolism, and this death requires RNA and protein synthesis (57). The similarity of this process to that observed in simpler model systems such as C. elegans suggests that this cell death is mainly apoptotic in nature. However, unique pathological processes may be at work in various regions of the brain, and these must be understood if we are to determine the geographical specifics of the various neurodegenerative dis-

AGING AND REGULATION OF APOPTOSIS

117

eases. For example, what is the biochemical basis for the neurotoxic effects of j8-amyloid peptide (6,30,45), and why are hippocampal neurons targeted in vivo? Smale et al. (65a) have demonstrated active apoptosis in the hippocampus of AD patients but not in normal controls. Other examples are the death of motor neurons in ALS, only a small percentage of which are known to be due to dominant deleterious mutations in Cu/Zn-superoxide dismutase (60,63) and the death of dopaminergic neurons in PD in the substantia nigra (18). The hope is not only that we can understand why these deaths occur, so that biochemical interventions to prevent neuron death can be developed (7), but also that growth factor (1,41) and cell therapies (20,22) may prove useful in prevention and replenishment strategies for treating age-related neurodegenerative diseases. Vll. Summary When Lockshin and Zakeri (44) discussed the relevance of apoptosis to aging, the common view was that apoptosis had primarily a negative impact on aging by destroying essential and often irreplaceable cells. That view has now changed to one that acknowledges that there are two general ways in which apoptosis can play a role in aging: (1) elimination of damaged and presumably dysfunctional cells (e.g., fibroblasts, hepatocytes) which can then be replaced by cell proliferation, thereby maintaining homeostasis (2,79), and (2) elimination of essential postmitotic cells (e.g., neurons) which cannot be replaced, thereby leading to pathology. Evidence exists in two systems (fibroblasts and thymocytes/ lymphocytes) that there are age-related decreases in the potential for apoptosis, although the molecular bases for these decreases appear to differ (Table II). Fibroblasts (and neurons?) lose the ability to downregulate bcl-2 in response to an apoptotic signal; thus, apoptosis is blocked even though an initiating signal has been received. In contrast, thymocytes/lymphocytes lack the ability to initiate the signal due to downregulation of the cell surface receptor Fas. There is limited information available for other tissue types, and nothing is known about why and how these age-related changes occur. An interesting observation, but not necessarily a critical one, is that the frequency of upregulation of the bcl-2 gene due to chromosome translocation increases with age (43). The role of apoptosis in regulating cell number is also a promising area of research. The studies on liver damage and neoplastic lesions suggest an extremely important role for apoptosis in controlling cancer. This may be particularly important in the prostate, where hypertrophy

118

HUBER R. WARNER TABLE II AGE-RELATED CHANGES IN APOPTOSIS IN MAMMALS

Age-related change Inability to downregulate bcl-2 in response to apoptosis signals in senescent cells Increased frequency of chromosome 14/18 translocations Switch from Fas^ to Fas" thymocytes

Tissue Human

fibroblasts

Physiological impact Resistance to cell death

Human B l5niiphocytes

Increase in bcl-2 expression

Mouse thymus

L3niiphoproliferation and increased autoimmunity

and cancer are a virtual certainty with ever-increasing age. It is not known whether the abiUty to undergo apoptosis dechnes in the prostate with increasing age, but it appears hkely t h a t it does. One problem in answering questions about the actual regulation of apoptosis is the lack of a quantitative assay. Apoptosis appears to be either "on" or "off" in cells, while the basic cell-killing machinery may often be present, but in an inactive form. Most assays for apoptosis are microscopic rather than kinetic, and the rate-limiting step may be at the level of the initiating signal. Thus, if CR, which extends the life span of rodents, does upregulate apoptosis, it is not clear how to quantify the magnitude of this effect or what should be quantified. The best one can do is to measure the frequency of occurrence of apoptotic bodies. This is essentially a pool size assay which provides little knowledge about how rapidly cells are leaving and entering the pool. Nevertheless, the results currently available do suggest t h a t apoptosis is a process which may be important in aging, at least in some tissues, and the mechanism of its regulation needs to be understood. Although a variety of tumor suppressor gene and oncogene products are known to be involved in signal transduction associated with apoptosis (39,49,50), it remains to be shown which of these, if any, are actually involved in "on-off" switches for apoptosis and which might regulate the intrinsic rate of apoptosis. As DriscoU (13) has already pointed out: "regulation and execution of cell death is an absolutely critical process t h a t interfaces with nearly every aspect of life. Future investigation of the links of cell death to cellular aging and the aging of organisms should be an exciting enterprise."

AGING AND REGULATION OF APOPTOSIS

119

REFERENCES 1. Baringa, M. (1995). Science 264, 112-llA. 2. Barr, P. J., and Tomei, L. D. (1994). Bio I Technology 12, 487-493. 5. Beckman, J. S., Carson, M., Smith, C. D., and Koppenol, W. H. (1993). Nature {London) 364, 584. 4. Benzi, G., and Moretti, A. (1995). Free Radical Biol. Med. 19, 77-101. 5. Bonfoco, E., Krainc, D., Ankarcrona, M., Nicotera, P., and Lipton, S. A. (1995). Proc. Natl Acad. Sci. U.S.A. 92, 7162-7166. 6. Butterfield, D. A., Hensley, K , Harris, M., Mattson, M., and Carney, J. (1994). Biochem. Biophys. Res. Commun. 200, 710-715. 7. Carney, J. M., and Floyd, R. A. (1991). J. Mol Neurosci. 3, 4 7 - 5 7 . 8. Chen-Levy, Z., Nourse, J., and Cleary, M. L. (1989). Mol. Cell. Biol. 9, 701-710. 9. Cleveland, J. L., and Ihle, J. N. (1995). Cell {Cambridge, Mass.) 8 1 , 479-482. 10. Columbano, A. (1995). J. Cell. Biochem. 58, 181-190. 11. Cortopassi, G. A., Shibata, D., Soong, N.-W., and Arnheim, N. (1992). Proc. Natl. Acad. Sci. U.S.A. 89, 7370-7374. 12. Dimri, G. P., Lee, X., Basile, G., Acosta, M., Scott, G., Roskelly, C , Medrano, E. E., Linskens, M., Rubelj, I., Pereira-Smith, O., Peacocke, M., and Campisi, J. (1995). Proc Natl. Acad. Sci. U.S.A. 92, 9363-9367. 13. DriscoU, M. (1995). In "Molecular Aspects of Aging" (K. Esser and G. M. Martin, eds.), pp. 4 5 - 5 8 . Wiley, Chichester, England. 14. DriscoU, M., and Chalfie, M. (1991). Nature {London) 349, 588-593. 15. EUis, R. E., Yuan, J., and Horvitz, H. R. (1991). Annu. Rev. Cell. Biol. 7, 663-698. 16. Evan, G. I., Wyllie, A. H., Gilbert, C. S., Littlewood, T. D., Land, H., Brooks, M., Waters, C. M., Penn, L. Z., and Hancock, D. C. (1992). Cell {Cambridge, Mass.) 69, 119-128. 17. Farlie, P. G., Dringen R., Rees, S. M., Kannourakis, G., and Bernard, O., (1995). Proc. Natl. Acad. Sci. U.S.A. 92, 4397-4401. 18. Fearnley, J. M., and Lees, A. J. (1991). Brain 114, 2283-2301. 19. Fisher, G. H., Rosenberg, F. J., Straus, S. E., Dale, J. K., Middelton, L. A., Lin, A. Y., Strober, W., Lenardo, M. J., and Puck, J. M. (1995). Cell {Cambridge, Mass.) 81,935-946. 20. Fisher, L. J., and Gage, F. H. (1993). Physiol Rev. 73, 583-616. 21. Franceschi, C , Monti, D., Scarfi, M. R., Zeni, O., Temperani, P., Emilia, G., Sansoni, P., Lioi, M. B., Troiano, L., Agnesini, C , Salvioli, S., and Cossarizza, A. {1992). Ann. NY. Acad. Sci. 663, 4 - 1 6 . 22. Gage, F. H., Ray, J., and Fisher, L. J., (1995). Annu. Rev. Neurosci. 18, 159-192. 23. Golstein, P., Marguet, D., and Depraetere, V. (1995). Cell {Cambridge, Mass.) 8 1 , 185-186. 24. Gonzalez-Garcia, M., Garcia, I., Ding, L., O'Shea, S., Boise, L. H., Thompson, C. B., and Nunez, G. (1995). Proc. Natl. Acad. Sci. U.S.A. 92, 4304-4308. 25. Good, R. A., Lorenz, E., Engelman, R. W., and Day, N. K. (1991). In "Biological Effects of Dietary Restriction" (L. Fishbein, ed.), pp. 147-156. Springer-Verlag, New York. 26. Grasl-Kraupp, B., Bursh, W., Ruttkay-Nedecky, B., Wagner, A., Lauer, B., and Schulte-Hermann, R. (1994). Proc. Natl. Acad. Sci. U.S.A. 91, 9995-9999. 27. Hall, N. C , Carney, J. M., Cheng, M. S., and Butterfield, D. A. (1991). Neuroscience 64, 81-89. 28. Hebert, L., Pandey, S., and Wang, E. (1994). Exp. Cell Res. 210, 10-18. 29. Hengartner, M. O., and Horvitz, H. R. (1994). Cell {Cambridge, Mass.) 76, 655-676.

120

HUBER R. WARNER

30. Hensley, K , Carney, J. M., Mattson, M. P., Aksenova, M., Harris, M., Wu, J. F., Floyd, R. A., and Butterfield, D. A. (1994). Proc. Natl. Acad. Sci. U.S.A. 9 1 , 3270-3274. 31. Hevelone, J., and Hartman, P. S. (1988). Biochem. Genet. 26, 4 4 7 - 4 6 1 . 32. Hodes, R. J. (1995). J. Exp. Med. 182, 1-3. 33. Hong, K., and DriscoU, M. (1994). Nature (London) 367, 470-473. 34. James, S. J., and Muskhelishvili, L. (1994). Cancer Res. 54, 5508-5510. 35. Kane, D. J., Ord, T., Anton, R., and Bredesen, D. E. (1995). J. Neurosci. Res. 40, 269-275. 36. Kayalar, C , Ord, T., Testa, M. P., Zhong, L.-T., and Bredesen, D. E. (1996). Proc. Natl. Acad. Sci. U.S.A. 93, 2234-2238. 37. Kelleher, J. A., Gregory, G. A., and Chan, P. H. (1995). Neurochem. Res. 19,209-215. 38. Kerr, J. F. R., Wyllie, A. H., and Currie, A. R. (1972). Br. J. Cancer 26, 239-257. 39. King, K. L., and Cidlowski, J. A. (1995). J. Cell. Biochem. 58, 175-180. 40. Kuida, K , Lippke, J. A., Ku, G., Harding, M. W., Livington, D. J., Su, M. S.-S., and Flavell, R. A. (1995). Science 267, 2000-2003. 41. Lindsay, R. M. (1995). Nature (London) 373, 289-290. 42. Linskens, M. H. K , Harley, C. B., West, M. D., Campisi, J., and Hayflick, L. (1995). Science 267, 17. 43. Liu, Y., Hernandez, A. M., Shibata, D., and Cortopassi, G. A. (1994). Proc. Natl. Acad. Sci. U.S.A. 9 1 , 8910-8914. 44. Lockshin, R. A., and Zakeri, Z. F. (1990). J. Gerontol. 45, B135-B140. 45. Loo, D. T., Copani, A., Pike, C. J., Whittemore, E. R., Walencewicz, A. J., and Cotman, C. W. (1993). Proc. Natl. Acad. Sci. U.S.A. 90, 7951-7955. 46. Luan, X., Zhao, W., Tomar, V., Chandrasekar, B., Hui, W., Mountz, J. D., and Fernandes, G. (1994). Arthritis. Rheum. 37, 5368. 47. Makinodan, T., and James, S. J. (1995). In "Dietary Restriction: Implications for the Design and Interpretation of Toxicity and Carcinogenicity Studies" (R. W. Hart, D. A. Neumann, and R. T. Robertson, eds.), pp. 311-325. ILSI Press, Washington, DC. 48. McConkey, D. J., Hartzell, P., Duddy, S. K., Hakansson, H., and Orrenius, S. (1988). Science 242, 256-259. 49. Meikrantz, W., and Schlegel, R. (1995). J. Cell. Biochem. 58, 160-174. 50. Miller, R. A. (1994). Exp. Gerontol. 29, 2 1 - 3 5 . 5L Miura, M., Zhu, H., Rotello, R., Hartwieg, E. A., and Yuan, J. (1993). Cell (Cambridge, Mass.) 75, 653-660. 52. MuskhehshviH, L., Hart, R. W., Turturro, A., and James, S. J. (1995). Am. J. Pathol. 147, 20-24. 53. Nagata, S., and Golstein, P. (1995). Science 267, 1449-1456. 54. Obeid, L. M., and Hannum, Y. A. (1995). J. Cell. Biochem. 58, 191-198. 55. Ogawa, N., Dang, H., and Talal, N. (1995). J. Autoimmun. 8, 1-19. 56. Oliver, C. N., Starke-Reed, P. E., Stadtman, E. R., Liu, G. J., Carney, J. M., and Floyd, R. A. (1990). Proc. Natl Acad. Sci. U.S.A. 87, 5144-5147. 57. Oppenheim, R. W., Prevette, D., Tytell, M., and Homma, S. (1990). Dev. Biol. 138, 104-113. 58. Oppenheim, R. W. (1991). Annu. Rev. Neurosci. 14, 4 5 3 - 5 0 1 . 59. Pandey, S., and Wang, E. (1995). J. Cell. Biochem. 58, 135-150. 60. Rabizadeh, S., Gralla, E. B., Borchelt, D. R., Gwinn, R., Valentine, J. S., Sisodia, S., Wong, P., Lee, M., Hahn, H., and Bredesen, D. E. (1995). Proc. Natl. Acad. Sci. U.S.A. 92, 3024-3028. 6L Raff, M. C , Barres, B. A., Burne, J. F. Coles, H. S., Ishizaki, Y., and Jacobson, M. D. (1993). Science 262, 695-700.

AGING AND REGULATION OF APOPTOSIS

121

62. Rieux-Laucat, F., Le Deist, F., Hivroz, C, Roberts, I. A. G., Debatin, K. M., Fischer, A., and de Villartay, J. P. (1995). Science 268, 1347-1349. 63. Rosen, D. R., Siddique, T., Patterson, D. et al. (1993). Nature (London) 362, 59-62. 64. Seshadri, T., and Campisi, J. (1990). Science 247, 205-209. 65. Singer, G. G., and Abbas, A. K. (1994). Immunity 1, 365-371. 65a. Smale, G., Nichols, N. R., Brady, D. R., Finch, C. E., and Horton, W. E. (1995). Exp. Neurol. 133, 225-230. 66. Smeyne, R. J., Vendrell, M., Hayward, M., Baker, S. J., Miao, G. G., Schilling, K., Robertson, L. M., Curran, T., and Morgan, J. I. (1993). Nature (London) 363,166-169. 67. Smith, C. A., Farrah, T., and Goodwin, R. G. (1994). Cell (Cambridge, Mass.) 76, 959-962. 68. Steller, H. (1995). Science 267, 1445-1449. 69. Takahashi, T., Tanaka, M., Brannan, C. I., Jenkins, N. A., Copeland, N. G., Suda, T., and Nagata, S. (1994). Cell (Cambridge, Mass.) 76, 969-976. 70. Tewari, M., Quan, L. T., O'Rourke, K , Desnoyers, S., Zeng, Z., Beidler, D. R., Poirier, G. G., Salvesen, G. S., and Dixit, V. M. (1995). Cell (Cambridge, Mass.) 81, 801-809. 7L Thompson, C. B. (1995). Science 267, 1456-1462. 72. Touchette, N. (1995). J. NIH Res. 7, 91-92. 73. Van Houten, N., and Budd, R. C. (1992). J. Immunol. 149, 2513-2517. 74. Vaux, D. L., Cory, S., and Adams, J. (1988). Nature (London) 335, 440-442. 75. Wang, E., Lee, M.-J., and Pandey, S. (1994). J. Cell. Biochem. 54, 432-439. 76. Wang, E. (1995). Cancer Res. 55, 22S4-2292. 77. Wang, E., and Pandey, S. (1995). J. Cell. Physiol. 163, 155-163. 78. Warner, H. R. (1994). Free Radical Biol. Med. 17, 249-258. 79. Warner, H. R., Fernandes, G., and Wang, E. (1995). J. Gerontol. 50A, B107-B109. 80. Watanabe-Fukunaga, R., Brannan, C. I., Copeland, N. G., Jenkins, N. A., and Nagata, S. (1992). Nature (London) 356, 314-317. 81. Weindruch, R., and Walford, R. (1988). "The Retardation of Aging and Disease by Dietary Restriction." Thomas, Springfield, IL. 82. White, K , Grether, M. E., Abrams, J. M., Young, L., Farrell, K., and Steller, H. (1994). Science 264, 677-683. 83. WiUiams, G. T. (1991). Cell (Cambridge, Mass.) 65, 1097-1098. 84. Wyllie, A. H. (1980). Nature (London) 284, 555-556. 85. Yuan, J., and Horvitz, H. R. (1990). Dev. Biol. 138, 3 3 - 4 1 . 86. Yuan, J., Shaham, S., Ledoux, S., EUis, H. M., and Horvitz, H. R. (1993). Cell (Cambridge, Mass.) 75, 641-652. 87. Zhong, L.-T., Sarafian, T., Kane, D. J., Charles, A. C , Mah, S. P., Edwards, R. H., and Bredesen, D. E. (1993). Proc. Natl. Acad. Sci. U.S.A. 90, 4533-4537. 88. Zhou, T., Edwards, C. K., and Mountz, J. D. (1995). J. Exp. Med. 182, 129-137.

This Page Intentionally Left Blank

CURRENT TOPICS IN CELLULAR REGULATION, VOLUME 35

Gene Regulation by Reactive Oxygen Species FiLIBERTO CiMINO FRANCA ESPOSITO

RosARio

AMMENDOLA

TOMMASO RUSSO Dipartimento di Biochimica e Biotecnologie Mediche Universitd di Napoli "Federico 11" Naples, Italy

I. Introduction Reactive oxygen species (ROS) are free radical species containing an unpaired electron in their outer orbital. Because of this unstable electronic configuration, ROS can easily capture an electron from other molecules. The main ROS are the superoxide anion (02^), the hydroperoxyl radical (H02-), and the hydroxyl radical (H0-). H2O2, although not a radical, plays an important role in causing oxidative damage because it is easily transformed into the highly reactive HO*. The intracellular concentration of ROS results from the balance between their production, which occurs spontaneously via mitochondrial electron transfer or is specifically catalyzed, and their inactivation by scavenging molecules (glutathione, ascorbate, tocopherol, j8-carotene, urate, etc.), or by antioxidant reactions catalyzed by such enzymes as superoxide dismutases, catalase and glutathione peroxidase (Halliwell and Gutteridge, 1986; Halliwell, 1990). Moreover, Bcl-2 protein, which is involved in the prevention of cell death, has been implicated in the antioxidant defenses of the cell (Hockenbery et al,, 1993; Hockenbery, 1994). However, it is not known where Bcl-2 acts to block oxidative damage to cellular constituents. Elevated levels of ROS can be responsible for or implicated in development of several h u m a n diseases: ischemia-reperfusion injury, cancer, inflammation (Cross, 1987; Ames, 1989), degenerative diseases (Halliwell and Gutteridge, 1985; Olanow, 1992), and most of the metabolic alterations observed in aging (Harman, 1978; Stadtman, 1992). Increased concentrations of ROS cause dramatic changes in the structure and biochemical properties of many biological molecules. Thus, in the case of lipids, such as linoleic acid, arachidonic acid, and other 123

Copyright © 1997 by Academic Press. All rights of reproduction in any form reserved. 0070-2137/97 $25.00

124

FILIBERTO CIMINO et al.

polyunsaturated fatty acids, their oxidation results in the formation of lipid peroxides through a series of "chain" reactions (for a review, see Yu, 1994). Nucleic acids are also affected by ROS, and DNA oxidation causes chemical modifications, including the formation of glycols, hydantoins, deamination products, and rearrangements in the pyrimidine ring which lead to DNA strand breaks (Reid and Loeb, 1993). The concentration of 8-hydroxy-2'-deoxyguanosine is increased in aged subjects and is derived from DNA oxidation. Consequently, this modified nucleoside is a biochemical marker of DNA oxidative damage (Richter et al., 1988). Exposure of a cell to elevated concentrations of ROS can induce significant damage in lipids, particularly in the membrane bilayer, and in mitochondrial and nuclear DNA. These molecular lesions can be counteracted only by eliminating the damaged components of the biological molecules and structures. The effect exerted by ROS on proteins is of two kinds: the first, consisting of HO- radical-mediated damage to proline, histidine, arginine, and other amino acid residues (Stadtman, 1993), irreversibly alters the protein structure, as observed in the case of lipids and DNA. The second type of ROS-mediated chemical modification of proteins is the oxidation of cysteine residues, which, within a defined range, can be reversed by enzyme systems that catalyze the reduction of oxidized cysteines (Stadtman, 1993). This suggests that protein oxidation-reduction may be a molecular mechanism by which the cell regulates protein functions. Among the mechanisms evolved by cells to respond to drastic oscillations in redox conditions is the activation of the expression of several genes, particularly those whose increased expression prevents oxidative damage and favors adaptive responses in living organisms. Gene expression is modulated by the activity of transcription factors, which, in turn, is regulated by the assembly or breakup of homo- and/or heterodimers, the control of the nuclear translocation of protein subunits, phosphorylation processes, and other posttranslational modifications (for a review, see Mitchell and Tjian, 1989). The observation that the DNA-binding activity of some transcription factors is modulated by oxidizing and/or reducing agents led to the notion that gene expression in eukaryotic and prokaryotic organisms is under "redox control." The modified DNA binding could also affect the gene expression controlled by the "redox-sensitive" transcprition factors. II. Sensitivity of Transcription Factors to Intracellular Redox Changes A. Prokaryotic Transcription Factors The relationship between transcription regulation and intracellular redox conditions has been well studied in prokaryotes. The hydrogen

GENE REGULATION BY REACTIVE OXYGEN SPECIES

125

peroxide-inducible genes in Salmonella typhimurium and Escherichia coli have provided a useful model with which to study how bacterial cells respond to oxidative stress. Treatment of bacterial cells with low doses of hydrogen peroxide induces the synthesis of at least 30 proteins (Morgan et al., 1986; VanBogelen et al., 1987). The expression of nine of the proteins (including catalase, alkyl hydroperoxide reductase, and glutathione reductase) induced within the first 10 min is controlled by the positive regulator oxyR (Tartaglia et al., 1989), which appears to act as an oxygen sensor (Storz et al., 1990). Strains carrying deletions of the oxyR gene are hypersensitive to hydrogen peroxide and do not show induction of the expression of the nine aforementioned proteins. The OxyR protein is an activator of a regulon of genes and a repressor of its own expression; the oxidized form of OxyR, but not its reduced form, activates transcription, seemingly as a consequence of a redoxdependent conformational change of OxyR. In fact, the oxidized and reduced forms of this transcription factor interact differently with the DNA (Storz et al, 1990). The reducing conditions do not denature the OxyR protein because, under these conditions, the protein still represses its own expression and is still bound to the promoter (Toledano etal., 1994). Mutational analysis of OxyR led to the identification of the protein domains t h a t are responsible for redox sensitivity and for the redox-dependent regulation of transcription activation (KuUik et al, 1995). The OxyR-controUed regulon is similar to other E. coli regulons that are activated as a consequence of DNA damage or various types of environmental stress. The best characterized of these groups of genes is the recA regulon, which controls the expression of 20 genes, a process known as the SOS response (Walker, 1985). The functions of all the genes activated during the SOS response are still unknown, but they are clearly essential for the repair of damaged DNA and for growth control. In E. coli during aerobic growth, the presence of superoxidegenerating compounds induces the synthesis of a distinct group of proteins (Greenberg and Demple, 1989). At least nine of these proteins are controlled by the soxR regulon (Tsaneva and Weiss, 1990); they include Mn-containing superoxide dismutase (Touati, 1988); endonuclease IV, which repairs radical-generated DNA damage (Chan and Weiss, 1987); and glucose-6-phosphate dehydrogenase, which is crucial for the generation of reduced nicotinamide-adenine dinucleotide phosphate (NAPDH) (Tsaneva and Weiss, 1990). The soxR regulon was identified in mutants showing increased resistance to menadione, constitutive expression of a subset of superoxide-inducible proteins, and elevated expression of reporter genes fused to the promoters of the

126

FILIBERTO CIMINO et al.

superoxide-inducible genes (Tsaneva and Weiss, 1990). The fusion constructs are not expressed in mutants carrying a deletion of the soxR locus, suggesting that the latter activates the transcription of superoxide-inducible genes. Two forms of soxR have been identified: Fe-SoxR containing a redox active iron-sulfur (Fe-S) center and apoSoxR (Hidalgo and Demple, 1994). Both of them bind the soxS promoter with the same affinities, but only Fe-SoxR stimulates transcription initiation of soxS, which suggests that oxidized Fe-SoxR changes the conformation of the soxS promoter. The variable redox state of the Fe-S center may thus be used to modulate the transcription activity of soxR in response to specific types of oxidative stress (Hidalgo and Demple, 1994). B. Zinc-Finger Transcription Factors

Among the various structures of DNA-binding domains, the zincfinger motif is particularly sensitive to redox changes, even though many factors bearing leucine zipper domains or homeodomains are also susceptible to redox changes. The first observations of the sensitivity of transcription factors to redox changes were made in vitro: exposure of purified proteins or of cellular extracts to such oxidants as H2O2, Nethylmaleimide (NEM), and diamide decreased the DNA-binding efficiency of almost all the transcription factors examined. In the case of the zinc-finger transcription factors, in which the cysteine residues involved in finger formation are coordinated with the Zn^^ ion only when these residues are in a reduced form, an oxidative environment can cause this structure to break up, abrogating the binding to DNA (see Fig. lA). Spl is a three-zinc finger transcription factor, with two cysteines and two histidines coordinating a Zn^^ ion in each finger (Kadonaga et al., 1987). The DNA-binding activity of this factor was markedly decreased in nuclear extracts prepared from aged (30-month-old) rat tissues as compared with their young counterpart (4-month-old tissues) (Ammendola et al., 1992). We postulated a relationship between this phenomenon and the increased levels of oxidizing radicals associated with aging. This hypothesis was supported by the observation that Spl DNAbinding decreased remarkably in young nuclear extracts exposed to H2O2 and that dithiothreitol (DTT) reversed this effect. A stronger argument in favor of an age-dependent redox impairment of Spl came from the demonstration that exposure of the nuclear extracts from aged rats to reducing agents completely restored the Spl DNA-binding activity (Ammendola et al, 1994). It has been demonstrated that the thiol groups of the cysteines involved in the zinc-finger motif confer

GENE REGULATION BY REACTIVE OXYGEN SPECIES

127

redox sensitivity to Spl; apo-Spl, contained in metal-depleted nuclear extracts, is more sensitive to oxidizing and alkylating agents than is holo-Spl. However, Spl bound to its DNA cis-element is significantly protected from oxidation (Knoepfel et al., 1994). The glucocorticoid receptor (GR) is also sensitive to the effect of oxygen free radicals. Like the other members of the steroid-thyroid hormone receptor family, this transcription factor has a multicysteine zinc-finger domain (for a review, see Beato, 1989). The ability of such oxidizing agents as H2O2 and methylmethane thiosulfonate (MMTS) to prevent the binding of GR to DNA was demonstrated by using a 15-KDa tryptic fragment containing the DNA-binding domain of the mouse GR (Hutchinson et al., 1991); DNA-binding activity was restored by the addition of DTT and Zn^+. These results demonstrated that ROS induce the formation of disulfide bonds between thiols involved in finger stabilization, thereby disrupting the structure. To investigate the in vivo relevance of the effects exerted on Spl and GR by ROS, we manipulated the intracellular redox conditions in intact cells. The DNA-binding efficiency of both transcription factors was demonstrated to be sensitive to the increased concentrations of ROS consequent to the depletion of reduced glutathione (GSH) induced in intact cells. Diethyl maleate (DEM) yields an oxidizing intracellular environment through conjugation with GSH in a reaction catalyzed by glutathione S-transferase (Boyland and Chasseaud, 1970). We treated Cos? cells, transfected with an expression vector for GR, with different concentrations of DEM for different times, and found that the efficiency of the DNA-binding activity in DEM-treated cells was much lower than that of untreated cells. We obtained the same results with two other agents, buthionine sulfoximine (BSO) and H2O2, which increase the intracellular concentration of ROS by mechanisms different from that of DEM. The Spl present in HeLa cells behaves the same way: its DNA binding is decreased in DEM-treated cells, albeit to a lesser extent than that observed with GR (Esposito et al., 1995). Pretreatment of the cell cultures with A^-acetylcysteine (NAG), a GSH precursor, counteracted the effects of DEM and almost totally restored the GR DNAbinding efficiency. Finally, the transcription of a CAT reporter gene controlled by a GR-dependent promoter was decreased under oxidative conditions, and pretreatment with NAG protected GR transactivation efficiency from the effects of DEM (Esposito et aL, 1995). In the experiments demonstrating the sensitivity of Spl and GR to redox changes, the DNA-binding activity of C-EBP, a transcriptional factor insensitive to redox manipulations (Bannister et aL, 1991), was unaltered in vitro and in vivo.

O

6upnp9j-[oiq:|[a

D)

la i

O «

GENE REGULATION BY REACTIVE OXYGEN SPECIES

129

A third redox-sensitive transcription factor is Egrl: it possesses three zinc-finger motifs and regulates the activity of a wide range of genes, including itself (Cao et al, 1990), c-fos, and other members of the jun family. The DNA-binding activity of Egrl is influenced by redox changes, as demonstrated in vitro by the abolishment of its DNA binding using the oxidized or the metal-free protein (Huang and Adamson, 1993). A conformational change of the zinc-finger region, caused by redox changes, seems to be responsible for the alteration of the DNA binding. Interestingly, Egrl DNA binding is restored by the addition of the reducing agents DTT or mercaptoethanol, or of Refl protein, a ubiquitous nuclear redox factor that controls the redox conditions of several transcription factors and, in turn, facilitates DNA-binding activity (Xanthoudakis et aL, 1992). Egrl is also sensitive to redox changes induced in intact cells. This transcription factor is precociously induced in human leukemic cell lines on exposure to differentiating treatment (Nguyen et aL, 1993), and GSH depletion, provoked in these cells by DEM, abolishes the DNA-binding activity of Egrl (Esposito et aL, 1994). The removal of the oxidant from the cells completely reverses this effect. In accordance with the observation that Egrl is responsible for most of the characteristics of the differentiated phenotype of these cells, DEM treatment completely prevented their differentiation (Esposito et aL, 1994). C. p53

Similarly to zinc-finger transcription factors, reduced cysteines are crucial to the formation of the DNA-binding domain of p53. In fact, within this structure, a Zn^+ ion is coordinated with one histidine residue and three cysteine residues, thereby contributing to the assembling of the DNA-binding domain (Cho et aL, 1994). p53 is a cell cycle regulator, being able to control the Gi checkpoint to allow time for DNA damage to be repaired (Nelson and Kastan, 1994). It is a sequencespecific DNA-binding protein (Funk et aL, 1992) that regulates the expression of several genes, including p2V^^^^^^^'^ and gadd45 (Kastan et aL, 1992; El-Deiry et aL, 1993; Donehower and Bradley, 1993). The addition to in vitro translated p53 of metal-chelating agents such as phenanthroline or oxidizing agents such as diamide alters the conformation of the protein and decreases its DNA-binding efficiency.

FIG. 1. Models of ROS-mediated regulation of transcription factors. Panels: A, a Znfinger transcription factor (see p. 126); B, N F - K B (see pp. 130 and 131); C, heat shock factor (HSF) (see p. 131); D, yeast AP-1 (YAP-1) (see p. 134).

130

FILIBERTO CIMINO et al.

Other oxidizing and/or DNA-damaging agents such as NEM, MMTS, infrared (IR), and ultra-violet (UV) radiations have the same effects (Lu and Lane, 1993; Hupp et al, 1993; Zhan et al, 1993; Hainaut and Milner, 1993). The addition of reducing agents to in vitro translated p53 facilitates the refolding of the protein and restores its DNA binding. To examine the effect of DEM-induced oxidative stress in intact cells on site-specific p53 binding in vivo, we treated three cell lines (Cos7, Hep3B, and HeLa cells), transfected with an expression vector encoding the human p53, with DEM, and compared the reaction of nuclear extracts from DEM-treated and untreated cells in gel shift experiments. The intensity of the p53-shifted band was significantly decreased in all cell lines exposed to the oxidant (Russo et al., 1995). Thus, the oxidative stress caused by DEM inhibited site-specific p53 binding in intact cells in vivo. The p53 transactivation ability was also impaired, as shown by the decreased transcriptional efficiency of a luciferase reporter gene under wild-type p53 control. Taken together, these experiments suggest that oxidative stress may block the ability of p53 to activate transcription in vivo by inactivating its ability to bind specific DNA sequences. The redox-dependent alteration of p53 DNA binding is strictly related to the three cysteine residues mentioned above. The role of individual cysteines in the regulation of p53 function has been investigated in an elegant study in which Ser-Cys substitutions were made at different positions of the p53 structure (Rainwater et al, 1995). Three cysteines (173,235, and 239) implicated in zinc-binding were found to be essential for the binding of this transcription factor to DNA, as demonstrated by a strong decrease in p53 DNA-binding and transactivation efficiencies. Four other cysteines (121,132,138, and 272) located in the loop-sheethelix of the site-specific DNA-binding domain may be involved in the modulation of DNA binding because their substitutions partially influence DNA-binding and transactivation activities. D. NF-KB and AP-1

Numerous studies have documented the sensitivity of NF-KB to oxidation and the tight relationship between its regulation and the intracellular redox conditions. NF-/<:B consists of two subunits, p50 and p65, which are restricted to the cytoplasm through association with an inhibitory subunit, I-ACB. The nuclear translocation of NF-ACB occurs after the phosphorylation of I-zcB, which is released from the complex and degraded (for a review, see Thanos and Maniatis, 1995). NF-KB DNA binding is markedly decreased by oxidation, as demonstrated by in vitro treatment of nuclear extracts with alkylating agents such as

GENE REGULATION BY REACTIVE OXYGEN SPECIES

131

NEM or oxidizing agents such as diamide, while reducing agents such as DTT or mercaptoethanol counteract the effects of oxidation (Toledano and Leonard, 1991). This redox sensitivity seems to depend on a specific cysteine residue of the p50 subunit. In fact, the mutation of this cysteine residue into a serine dramatically decreases N F - K B D N A binding (Matthews et al., 1992). Surprisingly, changes of the redox state in intact cells have an opposite effect on N F - K B , i.e., the reducing agent NAC prevents TPA- or tumor necrosis factor-« (TNFa)induced activation of NF-xB (Staal et aL, 1990), whereas H2O2 activates N F - K B (Schreck and Baeuerle, 1991). Given these conflicting results, one may speculate t h a t NF-zcB redox regulation occurs at two levels: (1) the nuclear level, where the Ref-thioredoxin system counteracts the oxidation-dependent inactivation of p50 (Matthews et al., 1992), and (2) the translocation level, with oxidation-induced N F - K B nuclear translocation (see Fig. IB). The latter process implies t h a t I-KB phosphorylation is modulated by ROS through a pathway involving the increase of tyrosine phosphorylation (Anderson et al., 1994). The speculation t h a t both phosphorylation and oxidation are required for N F - K B nuclear translocation is further supported by the observation t h a t the phosphatase inhibitor okadaic acid (OA) is able to induce N F - K B nuclear translocation only in transformed cells, and in normal cells only after pretreatment with H2O2 or BSO. NAC pretreatment prevents the OA-induced activation of N F - K B in transformed cells (Menon et aZ./1993). Furthermore, the antioxidant pyrrolidine dithiocarbamate (PDTC) blocks the phosphorylation of I-KB induced by TNF but does not prevent the proteasome-directed degradation of this inhibitor (Traencken BrittaMareen et al., 1994). Taken together, these results support the hypothesis t h a t ROS, by inhibiting phosphatases, increase the 1-KR phosphorylation t h a t leads to N F - K B nuclear translocation. Similarly to the N F - K B complex, the nuclear translocation of the heat shock factor (HSF) is inhibited by reducing agents. This transcription factor activates heat shock protein promoters, and its activation, soon after the heat shock, consists of various steps: trimerization, phosphorylation, and nuclear translocation (see Fig. IC). DTT inhibits all these phenomena. Therefore, it can be speculated t h a t oxidation plays a key role in the activation of HSF (Huang et al., 1994), which is in agreement with the observation t h a t H2O2 mimics the effects of the heat shock (Becker et al, 1990). The transcription factor AP-1 is also sensitive to redox changes. The c-fos/c-jun protooncogene heterodimer plays a central role in signal transduction pathways (Morgan and Curran, 1991). Morgan and Curran demonstrated t h a t AP-1 binds to the DNA consensus element

132

FILIBERTO CIMINO et al.

only if two cysteines (154 in Fos and 272 in Jun) are in a reduced form, and that incubation of AP-1 protein with NEM or diamide strongly inhibits its DNA-binding activity (Abate et al., 1990a). Mutation of the two cysteine residues increases DNA-binding efficiency and enhances the transforming activity of these protooncogenes (Abate et ah, 1990a). Furthermore, a Cys-Ser mutation naturally occurring in Y-jun is probably involved in the transforming activity of the oncogene (Xanthoudakis et al., 1992). All these observations indicate that the redox regulation of AP-1 is mainly related to the cysteine residues present in the DNAbinding domain of AP-1. Accordingly, AP-1 DNA binding was greatly decreased in HL60 and KGl cells treated with DEM, and the binding efficiency was fully restored when this drug was removed from the culture medium (Esposito et al., 1994). Furthermore, AP-1 DNA binding and AP-1 mediated transactivation are induced on treatment of HeLa cells with the reducing agents NAG and PDTC (Meyer et al, 1993), and PDTC-induced activation of AP-1 DNA binding resulted in the transcription activation of c-/bs through the serum responsive factor (Karin, 1991). Moreover, these agents enhanced the efficiency of DNA binding and transactivation of AP-1 after TPA stimulation in intact HeLa and Jurkat cells (Meyer et al., 1993). Surprisingly, while H2O2 is able to decrease the TPA-mediated activation of AP-1 DNA binding, it is a very weak activator of AP-1 in unstimulated cells. An AP-1-related protein of yeast, designed YAP-1, is similarly redoxsensitive. In fact, H2O2, diamide and DEM markedly increase the DNA-binding efficiency of YAP-1 in vivo. Interestingly, the three DNA-YAP-1 complexes observed in gel shift assays are differently affected by the various oxidants used, thus suggesting that posttranslational modifications of YAP-1 could play a role in the sensitivity to various oxidants (Kuge and Jones, 1994; see also Fig. ID). E. Other Redox-Sensitive Transcription Factors

Other transcription factors are sensitive, at least in vitro, to redox modifications. Such is the case with HoxB5, a sequence-specific DNAbinding protein that possesses a homeodomain structure. HoxB5 binds DNA in vitro in both oxidized and reduced forms, but it binds cooperatively only in oxidized form. Analogously to other transcription factors, the cysteine residues in the DNA-binding domain of HoxB5 are responsible and necessary for DNA binding and redox regulation (Galang and Hauser, 1993). Members of the ROI/KB family are also sensitive to changes in redox conditions. This family includes the v- and c-Rel oncoproteins that bind the KB motifs, form heterodimers with other members of the Rel family.

GENE REGULATION BY REACTIVE OXYGEN SPECIES

133

and modulate the expression of genes linked to K: B motifs. The RxxRxRxxC motif is conserved in all ROI/KB proteins, and a reduced cysteine residue is required for both protein-DNA contact and redox control (Kumar et al,, 1992). Similarly, the DNA-binding capacity of the helixloop-helix/leucine repeat transcription factor USF is dramatically decreased under nonreducing conditions. The phenomenon is due to the redox modification of the two cysteine sulfhydryl groups, both present in the helix-loop-helix domain of the protein. These modifications result in intra- and intermolecular covalent cross-links that inhibit DNA-binding potential, presumably by blocking structural changes required for interaction with DNA. The redox modulation of USF DNAbinding efficiency could be translated in vitro as a specific modulation of the ability of USF to activate transcription from a USF-responsive promoter (Pognonec et al., 1992). The reduction of the conserved cysteine-130 is also essential for the DNA-binding of Myb. Purified recombinant c-Myb is unable to bind DNA unless supplemented with DTT, and this effect is dose-dependent; on the other hand, oxidation by diamide or alkylation by NEM of the Myb protein in vitro inhibits its interaction with DNA, and the diamide effect is reversed by the addition of an excess of DTT. Moreover, sitedirected mutagenesis of the conserved cysteine to serine almost completely abolishes the DNA binding of Myb (Guehman et al, 1992). Similarly, the presence of the conserved cysteine-12 within the basic DNA-binding motif of the BZLFl transcription factor renders its in vitro DNA-binding ability sensitive to changes in the redox state. Under oxidizing conditions, an intermolecular disulfide bridge is formed between the cysteine residues of two adjacent basic motifs, and this inhibits the DNA-binding ability of BZLFl. The oxidation/reduction state of Cys-12 and its effect on DNA binding are reversible since it can be manipulated by reduced and oxidized glutathione (Bannister et al., 1991), suggesting that the sensitivity of Cys-12 to changes in the redox state may be part of a regulatory pathway controlling the DNA-binding ability of BZLFl. The DNA-binding efficiency of the OTF-1 transcription factor is decreased on redox changes induced either by H2O2 or by NEM, both in metal-depleted nuclear extracts and in nondepleted extracts (Knoepfel et al., 1994). By contrast, the DNA-binding activity of NF-1 is stimulated by millimolar concentrations of oxidized glutathione (GSSG), although this transcription factor, similarly to holo-Spl and OTF-1, is inhibited by millimolar concentrations of nonphysiological oxidants such as H2O2 or diamide (Knoepfel et al., 1994).

134

FILIBERTO CIMINO et al.

TTFl, a thyroid- and lung-specific transcription factor containing a homeodomain, is sensitive in vitro to redox changes. In fact, exposure of nuclear extracts to oxidizing agents decreases TTFl-binding activity, and reducing agents fully restore it; the specific amino acid residues involved in this oxidative-regulated process are two cysteines located outside the TTFl homeodomain, and their oxidation favors the formation of high-order oligomers (Arnone et al., 1995). The thyroid-specific transcription factor TTF2 is involved in the regulation of thyroglobulin and thyroperoxidase genes; it is a zinc-finger protein whose DNAbinding efficiency is decreased as a consequence of oxidation (Civitareale et al, 1994). III. Effects of Redox Changes on Regulation of Gene Expression Most of what we know about the regulation of gene expression by ROS comes from studies on genes involved in cellular responses to increased concentrations of ROS or to oxidant-induced injuries. However, in many cases, it is unclear whether the altered gene expression results directly from the redox changes or is a consequence of oxidative stress-induced damage to a target molecule. Furthermore, in several instances, changes in gene expression are also seen at a posttranscription level. A. Cellular Responses That Counteract Increases of ROS Concentration The activation of YAP-1 is a typical example of how the redox regulation of transcription factors leads to a change in gene expression, which in turn counteracts ROS-induced damage (or effects). In fact, YAP-1~ cells are more sensitive to H2O2 and 62^ than are wild-type cells (Schnell et al,, 1992), and the overexpression of the YAP-1 gene results in an increased resistance of yeast to oxidative stress (Kuge and Jones, 1994). In this context, it was demonstrated that H202-mediated activation of YAP-1 results in the transcription of the TRX2 gene, which encodes one of the two thioredoxins present in yeast through the binding of YAP-1 to the TRX2 promoter (Kuge and Jones, 1994). Unconjugated bilirubin is an efficient scavenger of singlet oxygen and is able to react with superoxide anion and peroxyl radicals, suggesting that bilirubin may be a physiological antioxidant in plasma and in the extravascular space. Because many of the potential cellular forms of the products of heme catabolism react efficiently with peroxyl radicals, these products may play a direct role in the cellular defense against

GENE REGULATION BY REACTIVE OXYGEN SPECIES

135

oxidant damage (Stocker and Ames, 1987). Heme oxygenase (HOx) plays an essential role in heme catabolism by cleaving heme to form biliverdin, which is subsequently converted to bilirubin by biliverdin reductase (Tenhunen et al., 1969). An increase in HOx activity can increase the cell's capacity to generate both biliverdin and bilirubin. The induction of HOx mRNA in cultured skin cells exposed to hydrogen peroxide is another example of the redox regulation of protective pathways specifically activated under conditions of oxidant stress (Keyse and Tyrrell, 1987, 1989), mRNA analysis has demonstrated that the HOx gene is also induced in cultured human skin fibroblasts by UV, sodium arsenite, cadmium chloride, iodoacetamide, and menadione (Keyse and Tyrrell, 1989), and the rate of induction suggests a posttranscriptional mechanism of regulation. Furthermore, HOx activity is increased in the liver of rats treated with heavy metal salts and sulfhydryl reactive compounds (DEM) (Maines and Kappas, 1976; Kikuchi and Yoshida, 1983), which, as described above, increase the levels of ROS by impairing the GSH-dependent scavenging mechanism. Another example of an oxidative stress-induced protein involved in the protection of the cell against the adverse effects of oxidation is ferritin. This protein, by regulating the free intracellular iron pool, controls the iron-catalyzed generation of ROS. Ferritin synthesis is significantly increased in the liver of rats subjected to oxidative stress by treatment with phorone, a glutathione-depleting drug. Northern blot analysis performed with RNA from these rats revealed an accumulation of the heavy and light ferritin mRNAs, and run-on experiments provided evidence of transcriptional activation (Cairo et al., 1995). Oxidative stress further increases ferritin synthesis posttranscriptionally through inhibition of iron regulatory factor (IRF) binding activity. RNA band-shift experiments performed with cytoplasmatic extracts from phorone-treated rats showed reduced activity of IRF, in particular IRFB, the cytoplasmic protein that posttranscriptionally controls ferritin mRNA translation (Cairo et aL, 1995). IRF is rapidly activated by exposure of cells to H2O2, but the activation is not a direct consequence of the effect of H2O2 on IRF. In fact, it is not prevented by hydroxyl radical scavengers or by the iron chelator 1,10-phenanthroline, indicating that the Fenton reaction is not involved in the process (Martins et al, 1995). B. Cellular Responses to ROS-lnduced Molecular Changes

Given the effects of intracellular redox changes on the activity of so many transcription factors, it is feasible that regulation of the transcription of many genes is similarly affected. There is evidence that the

136

FILIBERTO CIMINO et al.

expression of some genes is governed by redox changes. Only in a few cases is the redox sensitivity of the transcription factor directly linked to the changes in promoter activity. One such case involves genes that are controlled by NF-KB and regulated by intracellular redox conditions, e.g., the NF-K:B-regulated human immunodeficiency virus (HIV) promoter that is inhibited by such antioxidants as NAC (Eylar et al., 1993; Staal et al., 1993), butylated hydroxyanisole, nordihydroacid, and ce-tocopherol (Israel et al., 1992) and is activated by oxidative stress (Schreck et al., 1991; Legrand-Poels et al., 1993). Another example of an NF-zcB-regulated gene is interleukin-8 (IL-8) (Stein and Baldwin, 1993). Radical scavengers suppress IL-8 production in lipopolysaccharide (LPS)-stimulated human whole blood, whereas hydrogen peroxide stimulates IL-8 release (DeForge et al., 1992). The ROS-mediated induction of IL-8 occurs at the level of both IL-8 mRNA and protein, and can be prevented by the addition of catalase (DeForge et al., 1993). The production of IL-8 appears to be specifically related to an oxidant stress, because neither heat shock nor a chemcial stress induces the expression of IL-8 (DeForge et al, 1993). Mammalian cells respond to a wide variety of adverse conditions, both chemical and physical, by inducing the synthesis of a number of stress proteins. A survey of the reagents/conditions that induce the heat shock response reveals that many of them are capable of either generating ROS or depleting cellular sulfhydryl function. In fact, the observations that (1) heat shock induces the expression of superoxide dismutase (SOD) (Privalle and Fridovich, 1987); (2) inhibition of antioxidant defense mechanisms increases susceptibility to killing by heat shock (Mitchell and Russo, 1983); and (3) heat shock proteins confer resistance to oxidative stress (Spitz et al., 1987) indicate that heat and oxidative stresses may converge on common cellular response pathways. This suggests that heat could increase the production of ROS and/ or promote cellular oxidation events. An example of this relationship between heat shock and oxidative stress came from the differential screening of a cDNA library of human skin fibroblasts treated with H2O2. This screening resulted in the isolation of a cDNA (CLIOO) corresponding to an mRNA that is highly inducible by either oxidative stress or heat shock (Keyse and Emslie, 1992). Its nucleotide sequence is highly homologous to that of a mouse cDNA encoding a growth factorinducible immediate early gene (Charles et al., 1992) and has an amino acid sequence significantly similar to that of a Ser/Tyr phosphatase encoded by the late gene HI of vaccinia virus (Guan et al, 1991). Furthermore, the carboxyl terminal portion of the CLIOO polypeptide contains a single copy of the highly conserved active-site sequence of

GENE REGULATION BY REACTIVE OXYGEN SPECIES

137

protein-tyrosine phosphatases, and the purified protein expressed in bacteria has intrinsic phosphatase activity. CLIOO is highly specific in its activity toward mitogen-activated protein kinases (MAPK), at least in vitro, suggesting t h a t these proteins are probably the physiological substrates of the CLIOO phosphatase in vivo (Keyse, 1995). As mentioned above, one of the most dramatic effects exerted by oxidizing radicals on cell integrity is genome damage. There are numerous examples of genes whose activation follows DNA damage, and in many cases they are also activated by an oxidative stress. How these genes are activated is unclear. Does the oxidative stress per se activate them or is the activation mediated through the oxidation-induced DNA damage? Eukaryotic cells respond to DNA damage with the induction of numerous genes, and in yeast more t h a n 80 genes (about 1% of the total genome) are activated by DNA damage (Ruby and Szostak, 1985). Some of them have been identified, including the genes involved in DNA repair (EUedge and Davis, 1990), replication (Barker et aL, 1985), and growth control (Johnston et al., 1987). Several results suggest a relationship between UV radiation and oxygen radicals, and in t u r n t h a t UV-induced changes in cellular metabolism can be due at least in part to the effects of the ROS produced. Treatment of cells with UV light, ionizing radiation, or H2O2 causes a rapid increase in several kinase activities (Devary et al., 1992; Uckun et ah, 1992; Radler-Pohl et aL, 1993). These kinases activate several transcription factors, including c-Jun, E g r l , and N F - K B ; furthermore, the activation of MAPK by ionizing radiation can be prevented by pretreating cells with the reducing agent NAC (Schreck et al, 1992; Datta et al., 1993). A burst of ROS transcriptionally induces c-fos in mouse epidermal cells (Amstad et al., 1992), and the expression of both c-jun and c-fos is rapidly induced by H2O2 (Devary et al, 1991). On the other hand, phenolic antioxidants trigger the expression of fral and fra2 but are unable to activate the transcription of c-fos and FosB (Yoshioka et al, 1995). Twenty cDNA clones encoding DNA damage-inducible genes have been isolated by hybridization subtraction (Fornace et al, 1988). These clones, designated gadd (growth arrest and DNA damage-inducible) genes, encode transcripts t h a t are increased as a consequence of UV radiation-induced DNA damage (Fornace et al, 1989). Some of these genes (class II) are also induced by the alkylating agent methylmethane sulfonate (MMS) and in most cases are also under the control of an oxidative stress provoked by H2O2: the genes most sensitive to MMS also seem to be the ones most sensitive to H2O2. Among the latter, gadd45, whose transcription is at least in part under the control of

138

FILIBERTO CIMINO et al.

p53, is also strongly activated by H2O2 (Fornace et al., 1988) but not by DEM (Russo et al, 1995). The gaddl53 gene, a transcription factor of the CCAAT enhancer-binding protein family, is highly induced by H2O2 and by a variety of DNA-damaging agents, as well as by certain growth arrest conditions, but is not responsive to X-irradiation or to treatment with X-ray mimetic agents (Luethy and Holbrook, 1992), nor is it responsive to phorbol esters despite the presence of a consensus AP-1 site within its promoter region (Park et al, 1992). Cell growth and differentiation also appear to be affected by redox changes. p53 plays a key role in cell growth control in response to DNA damage, and the ^2V^^^^^^^^ gene is one of the most important targets of p53. In fact, the p53-mediated induction of p21^^^^^^^\ an inhibitor of cycline-dependent kinases regulating the Gi-S transition, is involved in the Gi arrest observed on DNA damage (Harper et al., 1993). Oxidative stress induced by DEM treatment causes an increase in p2iwafi/cipi mRNA and protein, an effect that is completely prevented by NAC (Russo et al., 1995). We also observed this phenomenon in cells lacking p53 and were thus able to demonstrate a p53-independent pathway of activation of p21^^^^^^^\ which is triggered by oxidative stress. The DEM-induced p2V^^^^^^^ increase is accompanied by growth arrest in the Gi phase and in cells lacking p53 (Russo et al, 1995). Among the other molecules that increase the levels of p2lw^^/^^Pi in cells lacking p53 are TPA and OA. Because TPA exerts its effect through protein kinase C (PKC) and because OA is a phosphatase inhibitor, we investigated whether DEM treatment could affect PKC. However, DEM treatment did not increase membrane-bound PKC activity. On the contrary, TPA, OA, and DEM strongly activated MAPK, a finding that suggests this enzyme is a direct or indirect target of ROS action and in turn is responsible for p2V^^^^^^^ induction (Russo et al., 1995; Michieli et al, 1994). These observations also suggest that some effects exerted by TPA could be a consequence of an increased ROS concentration. In fact, cytofluorimetric studies showed that TPA treatment of neutrophils stimulated the formation of oxidative products (Bass et al., 1983). However, TPA and ROS effects are not overlapping, as demonstrated by a study of the sensitivity to DEM on monoc3^ic differentiation. The TPA-induced differentiation of human leukemic cells toward monocytes results in a downregulation of the genes encoding myeloperoxidase (MPO) in HL60 cells and CD34 in KGl cells. The level of expression of both markers decreases on exposure of the cells to TPA. The addition of DEM to both cell lines prevents the downregulation of the transcription of two markers, so that the levels of MPO and CD34 mRNA are indistinguishable from those of the control cells. Accord-

GENE REGULATION BY REACTIVE OXYGEN SPECIES

139

ingly, the TPA-induced modifications of the cell phenotype are abolished by pretreatment with DEM (Esposito et al,, 1994). The dramatic phenotypic modifications that follow oxidative stress are probably related to numerous still uncharacterized changes in gene expression. We used the differential display technique (Liang and Pardee, 1992) to examine the changes in cell transcription pattern after oxidative stress induced by DEM (Ammendola et aL, 1995). Ten differentially expressed tags have been identified, four of which are identical or highly homologous to sequences contained in the human cDNA encoding c-Fos, vimentin, cytochrome oxidase IV, and ribosomal protein L4; another tag corresponds to a transcript of the mitochondrial genome whose function is unknown. The remaining five cDNAs are not recorded in any known sequence data bank. In Northern blot experiments one of these, named Rox3, lights up two mRNA species of 3400 and 3600 bp that are significantly increased after treatment with DEM or BSO; this increase appears immediately after exposure to DEM and is completely prevented by pretreatment with NAC. Rox3 is 100% homologous to a human guanine nucleotide regulatory protein named nepl (Chan et al., unpublished results; Accession No. U02081). The finding that a guanine nucleotide regulatory protein might be induced on oxidative stress is a further argument in favor of ROS as modulators of the membrane signal transduction machinery.

IV. ROS and Extracellular Signal Transduction Important cellular functions are significantly altered by changes in intracellular redox conditions. Studies in this area have long focused on the toxic effects of the oxidizing radicals, but more recently it has been suggested that the modulation of intracellular redox conditions could also be a molecular mechanism whereby the cell regulates various functions (Schreck and Baeuerle, 1991; Remade et aL, 1995). In this context, oxidizing radicals can be considered second messengers involved in the transduction of extracellular stimuli. Considering what we know about such intracellular second messengers as cAMP and Ca^^ ions, ROS to function as second messenger must fulfill the following requirements: (1) Efficient molecular systems are needed to generate a second messenger. The existence of enzyme systems as efficient as adenylate cyclase, or of membrane Ca^^ channels that function in the presence of a very high [Ca^^] gradient between the cytoplasm and the Ca^^ stores, allows the rapid increase of second messenger levels, which results in a manyfold amplification of extracellular signals. NADPH oxidase

140

FILIBERTO CIMINO et al.

catalyzes the formation of ©2"; this is a membrane-bound enzyme t h a t was originally found in phagocytic cells in which it generates ROS to kill the pathogens (Babior, 1978). Subsequently, NADPH oxidase was found in many other cells, e.g., endothelial cells (Matsubara and Ziff, 1986; Rosen and Freeman, 1984) and fibroblasts (Meier et al, 1989, 1990, 1991), which do not possess phagocytic activity. Furthermore, there is evidence t h a t ROS, including those generated by membranebound NADPH oxidase, are also released inside the cell (Meier et aL, 1989; Rosen and Freeman, 1984; Tiku et al, 1990; Radeke et al, 1990). (2) The molecular systems t h a t generate second messengers must be regulated by signal receptors and transducers. The systems t h a t generate second messengers are regulated by complex mechanisms t h a t represent the interface between the messengers and the extracellular signals. These mechanisms are often based on the activation of guanine nucleotide binding proteins, consisting of guanosine diphosphate/guanosine triphosphate (GDP/GTP) exchange, as in the cases of the regulation of adenylate cyclase or of phospholipases by heterotrimeric G proteins, or the regulation of kinase cascades triggered by p21'^^' on the activation of membrane-bound tyrosine kinases. These mechanisms resemble the regulatory mechanism proposed for NADPH oxidase, which requires two cytosolic proteins, p47P^°'' and p67P^°'', and a small GTP-binding protein, p21'^'^ In fact, NADPH oxidase is present in the membrane in inactive form and is activated after its interaction with phosphorylated p47P^°'' and p67P^°''. p21'^^^\ which is stimulated on its binding to GTP, is involved in this interaction, and after the hydrolysis of this nucleotide to GDP, catalyzed by the p21'^^''^ itself, the multiprotein complex dissociates, and p21'^^'^^ binds rhoGDI (a GDP-dissociation inhibitor) (Segall and Abo, 1993). The integration of this complex regulatory mechanism with membrane receptors could be the molecular link between extracellular signals and the generation of ROS. Evidence accumulated so far indicates t h a t NAPDH oxidase could be a membrane-bound element of the signal transduction system able to generate a second messenger (see Fig. 2). As yet no results implicate other enzymes in this function. (3) Second messengers must have a short half-life. To ensure t h a t the presence of a second messenger within a cell and its metabolic effects are restricted to the time during which the cell is exposed to a stimulus, efficient machineries exist t h a t inactivate (e.g., cAMP phosphodiesterase) or remove (e.g., the Ca^^ATPase pump) the second messenger. The half-life of ROS is very short due to their elevated reactivity and the numerous scavenging systems devoted to their elimination from the cell. O2" is, in fact, efficiently removed by SOD, and H 0 - ,

141

GENE REGULATION BY REACTIVE OXYGEN SPECIES Extracellular

4.

y

>

stimuli

S^

r.

r

SOD

H2O

o.

CT

H2O2 (diffusible)

. 1

HO*

PTPh

QO

iiiiiiiiiiiiiiii

v4y

iiiiiiiiiiiiiiiiii^^^

MAPK

nO

HllllllillBlllli;^^'

PTK

MO

niiiiniiiiiiiiii^^^

NFKB, Y A P - 1 , H S F

QO

'•niiiiiiiiiiiiiij^^

Spl, Egrl,GR, p53

^^^

FIG. 2. Possible targets of ROS generated on extracellular stimuli. STM, Signal transduction machinery; SOD, superoxide dismutases; CT, catalase; PTPh, protein-tyrosine phosphatases; MAPK, mitogen-activated protein kinases; PTK, protein-tjnrosine kinases.

generated from H2O2, due to its remarkable reactivity, also has a very short half-life (Halliwell and Gutteridge, 1986). However, the effect of the second messenger on the target proteins is a consequence of a noncovalent interaction, so t h a t a decrease in the concentration of the second messenger results in the dissociation of the target-second messenger complex and thus in the disappearance of the second messenger-induced activation. By contrast, the simple inactivation of ROS leaves the target proteins stably modified (i.e., oxidized). Therefore, to abrogate a ROS-induced effect, an accessory machinery must catalyze the restoration (i.e., reduction) of the oxidized target. Dithiol reducing enzyme systems, such as t h a t based on thioredoxin, could play an important role in this respect. In fact, NADPH thioredoxinthioredoxin reductase is a disulfide reductase, found both in prokaryotes and in all eukaryotic cells (Holmgren, 1989), and thioredoxin is an abundant protein, manyfold more active t h a n DTT in the reduction of various proteins (Holmgren, 1979). NADPH thioredoxin-thioredoxin reductase affects transcription factors by way of an unusual mechanism in which DNA binding of AP-1 and other transcription factors (e.g., N F - K B , Myb, and members of the ATF/CREB family) is modulated by a redox mechanism (Abate et al., 1990a, b). The essence of this mechanism is a nuclear protein, named Refl, which keeps these transcription

142

FILIBERTO CIMINO et al.

factors in a reduced form in the absence of other reducing agents. Refl is a 37-kDa protein that facihtates the DNA binding of Fos/Jun heterodimers, and does not itself directly bind to DNA (Xanthoudakis and Curran, 1992). Interestingly, Refl is itself under redox control; in fact, its activity is increased in the presence of thioredoxin, thioredoxin reductase, and NADPH (Abate et al., 1990a). The glucocorticoid receptor copurifies with the thioredoxin-thioredoxin reductase complex and is thus probably governed by the reducing activity of the thioredoxin system (Grippo et al., 1985). Similarly, the activity of AP-1 transcription factor is highly induced in the presence of thioredoxin (Schenk et al., 1994). (4) The targets of second messengers exert pleiotropic effects. All known second messengers act directly or indirectly on enzyme targets, e.g., protein kinases (PKA and PKC), which have a wide pattern of substrates, so that they coordinately regulate a large number of cellular functions. The possible pathways through which ROS could exert their effects as second messengers are illustrated in Fig. 2. Extracellular stimuli could induce an increase in 0<^, possibly by activating membrane-bound enzymes, such as NADPH oxidase. The subsequent generation of HO* would lead to the oxidation of sulfhydryl moieties, causing conformational changes in several proteins. The catal3^ic activity of phosphotyrosine protein phosphatases (PTPh) requires the existence of a reduced cysteine in the active site (Fisher et al., 1991). Accordingly, the treatment of Jurkat T cells with the PTPh inhibitor phenylarsine oxide (PAO) or with the thiol-oxidizing agent diamide results in a remarkable increase in tyrosine phosphorylation (Staal et al., 1994). This increase of tyrosine phosphorylation has been related to the growth factor signaling machinery. In fact, it was demonstrated that (1) PDGF treatment of smooth muscle cells induces an increase in H2O2 intracellular concentration; (2) exposure of these cells to exogenous H2O2 causes an increase in overall protein tyrosine phosphorylation similar to that provoked by exposure to PDGF; (3) high levels of intracellular catalase activity counteract the PDGF-induced increase in both H2O2 concentration and tyrosine phosphorylation (Sundaresan et al., 1995). Therefore, it is not inconceivable that ROS significantly affect the balance between protein tyrosine phosphorylation and dephosphorylation and probably act by inhibiting PTPh in response to extracellular stimuli. Similarly, MAPK activity is enhanced by oxidative stress; this was demonstrated in HeLa and T98G cells (Russo et al, 1995), in NIH 3T3 cells (Stevenson et al, 1994), and in HL60 cells (Fialkow et al, 1994). Again, this effect could be a consequence of PTPh inhibition or of a direct effect on MAPKK or ras, which are upstream regulators of the MAPK cascade (Fialkow et al, 1994; Lander et al.

GENE REGULATION BY REACTIVE OXYGEN SPECIES

143

1995). Lastly, because the transcription of many genes is regulated by transcription factors whose activity is sensitive to redox changes, a transient increase in ROS, in response to an extracellular stimulus, would probably significantly modify the gene transcription pattern of a cell. V. Concluding Remarks Studies conducted in the last few decades have produced a wealth of information on the chemistry and high reactivity of ROS, on the intracellular sites of their generation, and on the mechanisms developed by cells to protect against ROS-induced injury. We now know much about the toxic effects exerted by ROS on DNA, proteins, lipids, and such important cellular processes as apoptosis, in which lipid peroxidation and DNA fragmentation precede cell death. In other words, the emphasis of these studies has been on the toxicity of ROS and on mechanisms developed by the cell to counteract them. Numerous antioxidant compounds have been identified, many of which have therapeutic implications for the treatment or prevention of diseases in which oxidative stress seems to play a role. The discoveries of the properties and functions of ROS have also contributed to our understanding of the molecular basis of aging. In this context, attempts have been made to increase the antioxidant capacity of organisms so as to try to control the progression of aging or, at least, to reach a condition in which a senescent organism would be less exposed to ROS-induced damage, thereby containing the development of degenerative diseases. Mainly as a consequence of the tremendous progress made in the field of gene expression and signal transduction systems, the emphasis of ROS studies has shifted and a new scenario is opening. In fact, rather than investigating their toxicity, efforts are now aimed at understanding the molecular basis of a new function of ROS, i.e., the regulation of important cellular processes. The attractive hypothesis that ROS could function as second messengers and could thus be implicated in the amplification and transduction of several extracellular signals is supported by a number of experimental results. Among the most telling are those demonstrating that ROS regulate the DNA-binding activity of transcription factors, stimulate gene expression, and activate elements of signal transduction pathways. All these findings strongly suggest that the modulation of ROS levels is a new and easy mechanism by which the cell regulates several important biological processes. ACKNOWLEDGMENTS The experiments performed in the laboratory of the authors were supported by grants from C.N.R., Special Projects "Invecchiamento" and "Ingegneria Genetica," from AIRC and from MURST (40% and 60%).

144

FILIBERTO CIMINO et

al.

REFERENCES Abate, C , Patel, L., Rauscher, F. J., Ill, and Curran, T. (1990a). Science 2 4 9 , 1 1 5 7 - 1 1 6 1 . Abate, C , Luk, D., and Curran, T. (1990b). Cell Growth Differ. 1, 455-462. Ames, B. N. (1989). Free Radical Res. Commun. 7, 121-128. Ammendola, R., Mesuraca, M., Russo, T., and Cimino, F. (1992). J. Biol. Chem. 267, 17944-17948. Ammendola, R., Mesuraca, M., Russo, T., and Cimino, F. (1994). Eur. J. Biochem. 225, 483-489. Ammendola, R., Fiore, F., Esposito, F., Caserta, G., Mesuraca, M., Russo, T., and Cimino, F. (1995). FEBS Lett. 371, 209-213. Amstad, P. A., Krupitza, G., and Cerutti, P. A. (1992). Cancer Res. 52, 3952-3960. Anderson, M. T., Staal, F. J. T., Gitler, C , Herzenberger, L. A., and Herzenberger, L. A. (1994). Proc. Natl. Acad. Sci. U.S.A. 91, 11527-11531. Arnone, M. I., Zannini, M., and Di Lauro, R. (1995). J. Biol. Chem. 270, 12048-12055. Babior, B. M. (1978). N. Engl. J. Med. 298, 659-668. Bannister, A. J., Cook, A., and Kouzarides, T. (1991). Oncogene 6, 1243-1250. Barker, D. G., White, J. M., and Johnston, L. H. (1985). Nucleic Acids Res. 13, 8 3 2 3 8337. Bass, D. A., Parce, J. W., Dechatelet, L. R., Szejda, P., Seeds, M. C , and Thomas, M. (1983). J. Immunol. 130, 1910-1917. Beato, M. (1989). Cell (Cambridge, Mass.) 56, 335-344. Becker, J., Mezger, V., Courgeon, A., and Best-Belpomme, M. (1990). Eur. J. Biochem. 189, 553-558. Boyland, E., and Chasseaud, L. F. (1970). Biochem. Pharmacol. 19, 1526-1528. Cairo, G., Tacchini, L., Pogliaghi, G., Anzon, E., Tomasi, A., and Bernelli-Zazzera, A. (1995). J. Biol. Chem. 270, 700-703. Cao, X., Koshi, R. A., Gashler, A., McKienan, M., Morris, C. F., Gaffney, R., Hay, R. V., and Sukhatme, V. P. (1990). Mol. Cell. Biol. 10, 1931-1939. Chan, E., and Weiss, B. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 3189-3193. Charles, C. H., Abler, A. S., and Lau, L. F. (1992). Oncogene 7, 187-190. Cho, Y., Gorina, S., Jeffrey, P. D., and Sauletich, N. P. (1994). Science 265, 346-355. Civitareale, D., Saiardi, A., and Falasca, P. (1994). Biochem. J. 304, 981-985. Cross, C. E. (1987). Ann. Intern. Med. 107, 526-545. Datta, R., Taneja, N., Sukhatme, V. P., Quresshi, S. A., Weichselbaum, R., and Kufe, D. W. (1993). Proc. Natl. Acad. Sci. U.S.A. 90, 2419-2422. DeForge, L. E., Fantone, J. C , Kenney, J. S., and Remick, D. G. (1992). J. Clin. Invest. 90,2123-2129. DeForge, L. E., Preston, A. M., Takeuchi, E., Kenney, J., Boxer, L. A., and Remick, D. G. (1993). J. Biol. Chem. 26S, 25568-25576. Devary, Y., Gottlieb, R. A., Lau, L. F., and Karin, M. (1991). Mol. Cell. Biol. 11, 2 8 0 4 2811. Devary, Y., Gottheb, R. A., Smeal, T., and Karin, M. (1992). Cell (Cambridge, Mass.) 71, 1081-1091. Donehower, L. A., and Bradley, A. (1993). Biochim. Biophys. Acta 1155, 181-205. El-Deiry, W. S., Tokino, T., Velculescu, V. E., Levy, D. B., Parson, R., Trent, J. M., Lin, D., Mercer, W. E., Kinzler, K. W., and Vogelstein, B. (1993). Cell {Cambridge, Mass.) 75,817-825. EUedge, S. J., and Davis, R. W. (1990). Genes Dev. 4, 740-751. Esposito, F., Agosti, V., Morrone, G., Morra, F., Cuomo, C , Russo, T., Venuta, S., and Cimino, F. (1994). Biochem. J. 301, 649-653.

GENE REGULATION BY REACTIVE OXYGEN SPECIES

145

Esposito, F., Cuccovillo, F., Morra, F., Russo, T., and Cimino, F. (1995). Biochim. Biophys. Acta 1260, 308-314. Eylar, E., Rivera-Quinones, C , Molina, C , Baez, I., Molina, F., and Mercado, C. M. (1993). Int. Immunol. 5, 9 7 - 1 0 1 . Fialkow, L., Chan, C. K , Rotin, D., Grinstein, S., and Downey, G. P. (1994). J. Biol. Chem. 269, 31234-31242. Fisher, E. H., Charbonneau, H., and Tonks, N. K. (1991). Science 253, 401-406. Fornace, A. J., Jr., Alamo, I., Jr., and Hollander, M. C. (1988). Proc. Natl. Acad. Sci. U.S.A. 85, 8800-8804. Fornace, A. J., Jr., Nebert, D., Hollander, M. C , Papathanasiou, M., Fargnoli, J., and Holbrook, N. (1989). Mol. Cell. Biol. 9, 4196-4203. Funk, W. D., Pak, D. T., Karas, R. H., Wright, W. E., and Shay, J. W. (1992). Mol. Cell. Biol. 12, 2866-2871. Galang, C. K., and Hauser, C. A. (1993). Mol. Cell. Biol. 13, 4609-4617. Greenberg, J. T., and Demple, B. (1989). J. Bacteriol. 171, 3933-3939. Grippo, J. F., Holmgren, A., and Pratt, W. B. (1985). J. Biol. Chem. 260, 93-97. Guan, K., Broyles, S. S., and Dixon, J. E. (1991). Nature (London) 350, 359-362. Guehman, S., Vorbrueggen, G., Kalkbrenner, F., and Moelling, K. (1992). Nucleic Acids Res. 20, 2279-2286. Hainaut, P., and Milner, J. (1993). Cancer Res. 53, 4469-4473. Halliwell, B. (1990). Free Radical Res. Commun. 9, 1-32. Halliwell, B., and Gutteridge, J. M. C. (1985). Trends Neurosci. 8, 22-26. Halliwell, B., and Gutteridge, J. M. C. (1986). Arch. Biochem. Biophys. 246, 501-514. Harman, D. (1978). Age 1, 143-150. Harper, J. W., Adams, G. R., Wei, N., Keyomarsi, K., and EUedge, S. J. (1993). Cell (Cambridge, Mass.) 75, 805-816. Hidalgo, E., and Demple, B. (1994). EMBO J. 13, 138-146. Hockenbery, D. M. (1994). J. Cell. Sci. 18, 5 1 - 5 5 . Hockenbery, D. M., Oltvai, Z. N., Yin, X. M., Milliman, C. L., and Korsmeyer, S. J. (1993). Cell (Cambridge, Mass.) 75, 241-251. Holmgren, A. (1979). J. Biol. Chem. 254, 9627-9632. Holmgren, A. (1989). J. Biol. Chem. 264, 13963-13966. Huang, E. L., Zhang, H., Bae, S. W., and Liu, A. Y. C. (1994). J. Biol. Chem. 269, 3071830725. Huang, R. P., and Adamson, E. D. (1993). DNA Cell Biol. 12, 265-273. Hupp, T. R., Meek, D. W., Midgley, C. A., and Lane, D. P. (1993). Nucleic Acids Res. 21, 3167-3174. Hutchinson, K. A., Matic, G., Meshinchi, S., Bresnick, E. H., and Pratt, W. B. (1991). J. Biol. Chem. 266, 10505-10509. Israel, N., Gougerot-Pocidalo, M. A., Aillet, F., and Virelizier, J. L. (1992). J. Immunol. 149, 3386-3393. Johnston, L. H., White, J. H. M., Johnson, A. L., Lucchini, G., and Pleviani, P. (1987). Nucleic Acids Res. 15, 5017-5030. Kadonaga, J. T., Garner, K. R., Masiarz, F. R., and Tjian, R. (1987). Cell (Cambridge, Mass.) 51, 1079-1090. Karin, M. (1991). Curr. Opin. Cell Biol. 3, 467-473. Kastan, M. B., Zhan, Q., El-Deiry, W. S., Carrier, F., Jacks, T., Walsh, W. V., Plunkett, B. S., Vogelstein, B., and Fornace, A. J., Jr. (1992). Cell (Cambridge, Mass.) 71, 587-597. Keyse, S. M. (1995). Biochim. Biophys. Acta 1265, 152-160.

146

FILIBERTO CIMINO et

al.

Keyse, S. M., and Emslie, E. A. (1992). Nature (London) 359, 644-647. Keyse, S. M., and Tyrrell, R. M. (1987). J. Biol. Chem. 262, 14821-14825. Keyse, S. M., and Tyrrell, R. M. (1989). Proc. Natl. Acad. Sci. U.S.A. 86, 99-103. Kikuchi, G., and Yoshida, T. (1983). Mol. Cell. Biochem. 53/54, 163-183. Knoepfel, L., Steinkuhler, C , Carri, M. T., and Rotillo, G. (1994). Biochem. Biophys. Res. Commun. 201, 871-877. Kuge, S., and Jones, N. (1994). EMBO J. 13, 655-664. KuUik, I., Toledano, M. B., Tartaglia, L. A., and Storz, G. (1995). J. BacterioL 111, 1 2 7 5 1284. Kumar, S., Rabson, A. B., and Gelinas, C. (1992). Mol. Cell. Biol. 12, 3094-3106. Lander, H. M., Ogiste, J. S., Teng, K. K , and Novogrodsky, A. (1995). J. Biol Chem. 270, 21195-21198. Legrand-Poels, S., Hoebeke, M., Vaira, D., Rentier, B., and Piette, J. (1993). J. Photochem. Photohiol. BiBiol. 17, 229-237. Liang, P., and Pardee, A. B. (1992). Science 257, 967-971. Lu, X., and Lane, D. P. (1993). Cell {Cambridge, Mass.) 75, 765-778. Luethy, J. D., and Holbrook, N. J. (1992). Cancer Res. 52, 5-10. Maines, M. D., and Kappas, A. (1976). Biochem. J. 154, 125-131. Martins, E. A., Robalinho, R. L., and Meneghini, R. (1995). Arch. Biochem. Biophys. 316, 128-134. Matsubara, T., and Ziff, M. (1986). J. Cell Physiol. 127, 207-210. Matthews, J. R., Wakasugi, N., Virelizier, J.-L., Yodoi, J., and Hay, R. T. (1992). Nucleic Acids Res. 20, 3821-3830. Meier, B., Radeke, H. H., Selle, S., Younes, M., Sies, H., Resche, K , and Habermehl, G. G. (1989). Biochem. J. 263, 539-545. Meier, B., Radeke, H. H., Selle, S., Habermehl, G. G., Resch, K , and Sies, H. (1990). Biol. Chem. Hoppe-Seyler 371, 1021-1025. Meier, B., Cross, A. R., Hancock, J. T., Kaup, F. J., and Jones, O. T. (1991). Biochem. J. 275, 241-245. Menon, S. D., Quin, S., Guy, G. R., and Tan, Y. H. (1993). J. Biol. Chem. 268, 2 6 8 0 5 26812. Meyer, M., Schreck, R., and Baeuerle, P. A. (1993). EMBO J. 12, 2005-2015. Michieh, P., Chedid, M., Lin, D., Pierce, J. H., Mercer, W. E., and Givol, D. (1994). Cancer Res. 54, 3391-3396. Mitchell, J. B., and Russo, A. (1983). Radiat. Res. 95, 471-485. Mitchell, P. J., and Tjian, R. (1989). Science 245, 371-378. Morgan, J., and Curran, T. (1991). Annu. Rev. Neurosci. 14, 421-451. Morgan, R. W., Christman, M. F., Jacobson, F. S., Storz, G., and Ames, B. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 8059-8063. Nelson, W. J., and Kastan, M. B. (1994). Mol. Cell. Biol. 14, 1815-1823. Nguyen, H. Q., Hohffmann-Liebermann, B., and Liebermann, D. (1993). Cell (Cambridge, Mass.) 72, 197-209. Olanow, C. W. (1992). Ann. Neurol. 32, S 2 - S 9 . Park, J. S., Luethy, J. D., Wang, M. G., Fargnoh, J., Fornace, A. J., Jr., McBride, O. W., and Holbrook, N. J. (1992). Gene 116, 259-267. Pognonec, P., Kato, H., and Roeder, R. G. (1992). J. Biol. Chem. 267, 24563-24567. Privalle, C. T., and Fridovich, I. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 2723-2726. Radeke, H., Meier, B., Topley, N., Floge, J., Habermehl, G., and Resch, K. (1990). Kidney Int. 37, 767-775.

GENE REGULATION BY REACTIVE OXYGEN SPECIES

147

Radler-Pohl, A., Sachsenmaier, C , Gebel, S., Auer, H. P., Bruder, J. T., Rapp, U., Angel, P., Rahmsdorf, H. J., and Herrlich, P. (1993). EMBO J. 12, 1005-1012. Rainwater, R., Parks, D., Anderson, M. E., Tegtmeyer, P., and Mann, K. (1995). Mol. Cell. Biol 15, 3892-3903. Reid, T. M., and Loeb, L. A. (1993). Proc. Natl. Acad. Sci. U.S.A. 90, 3904-3907. Remade, J., Raes, M., Toussaint, O., Renard, P., and Rao, G. (1995). Mutat. Res. 316, 103-122. Richter, C , Park, J. W., and Ames, B. N. (1988). Proc. Natl. Acad. Sci. U.S.A. 85, 6 4 6 5 6467. Rosen, G. M., and Freeman, B. A. (1984). Proc. Natl. Acad. Sci. U.S.A. 81, 7269-7273. Ruby, S. W., and Szostak, J. W. (1985). Mol. Cell. Biol. 5, 75-84. Russo, T., Zambrano, N., Esposito, F., Ammendola, R., Cimino, F., Fiscella, M., Jackman, J., O'Connor, M., Anderson, C. W., and Appella, E. (1995). J. Biol. Chem. 270, 2938629391. Schenk, H., Klein, M., Erdbrugger, W., Droge, W., and Schultze-Osthoff, K. (1994). Proc. Natl. Acad. Sci. U.S.A. 9 1 , 1672-1676. Schnell, N., Krems, B., and Entian, K. D. (1992). Curr. Genet. 21, 269-273. Schreck, R., and Baeuerle, P. A. (1991). Trends Cell Biol. 1, 39-42. Schreck, R., Rieber, P., and Baeuerle, P. A. (1991). EMBO J. 10, 2247-2258. Schreck, R., Albermann, K , and Baeuerle, P. A. (1992). Free Radical Res. Commun. 17, 221-237. Segall, A. W., and Abo, A. (1993). Trends. Biochem. Sci. 18, 43-47. Spitz, D. R., Dewey, W. C , and Li, G. C. (1987). J. Cell. Physiol. 131, 364-373. Staal, F. J. T., Roederer, M., Herzenberg, L. A., and Herzenberg, L. A. (1990). Proc. Natl. Acad. Sci. U.S.A. 87, 9943-9947. Staal, F. J. T., Roederer, M., Raju, P. A., Anderson, M. T., Ela, S. W., and Herzenberg, L. A. (1993). AIDS Res. Hum. Retroviruses 9, 299-306. Staal, F. J. T., Anderson, M. T., Staal, G. E. J., Herzenberg, L. A., Gitler, C , and Herzenberg, L. A. (1994). Proc. Natl. Acad. Sci. U.S.A. 9 1 , 3619-3622. Stadtman, E. R. (1992). Science 257, 1220-1224. Stadtman, E. R. (1993). Annu. Rev. Biochem. 62, 797-821. Stein, B., and Baldwin, A. S., Jr. (1993). Mol. Cell. Biol. 13, 7191-7198. Stevenson, M. A., Pollock, S. S., Coleman, C. N., and Calderwood, S. K. (1994). Cancer Res. 54, 12-15. Stocker, R., and Ames, B. N. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 8130-8134. Storz, G., Tartaglia, L. A., and Ames, B. N. (1990). Science 248, 189-194. Sundaresan, M., Yu, Z., Ferrans, V. J., Irani, K , and Finkel, T. (1995). Science 270, 296-299. Tartaglia, L. A., Storz, G., and Ames, B. N. (1989). J. Mol. Biol. 210, 709-719. Tenhunen, R., Marver, H. S., and Schmid, R. (1969). J. Biol. Chem. 244, 6388-6394. Thanos, D., and Maniatis, T. (1995). Cell (Cambridge, Mass.) 80, 529-532. Tiku, M. L., Liesch, J. B., and Robertson, F. M. (1990). J. Immunol. 145, 690-696. Toledano, M. B., and Leonard, W. J. (1991). Proc. Natl. Acad. Sci. U.S.A. 88,4328-4332. Toledano, M. B., KuUik, L, Trinh, F., Baird, P. T., Schneider, T. D., and Storz, G. (1994). Cell (Cambridge, Mass.) 78, 897-909. Touati, D. (1988). J. Bacteriol. 170, 2511-2520. Traencken Britta-Mareen, E., Wilk, S., and Baeuerle, P. A. (1994). EMBO J. 22,5433-5441. Tsaneva, I. R., and Weiss, B. (1990). J. Bacteriol. 172, 4197-4205. Uckun, F. M., Tuel-Ahlgren, L., Song, C. W., Waddick, K , Myers, D. E., Kirihara, J., Ledbetter, J. A., and Schieven, G. L. (1992). Proc. Natl. Acad. Sci. U.S.A. 89,9005-9009.

148

FILIBERTO CIMINO et al.

VanBogelen, R. A., Kelley, P. M., and Neidhardt, F. C. (1987). J. Bacteriol. 169, 26-32. Walker, G. C. (1985). Annu. Rev. Biochem. 54, 425-457. Xanthoudakis, S., and Curran. T. (1992). EMBO J. 11, 653-665. Xanthoudakis, S., Miao, G., Wang, F., Pan, Y. E., and Curran T. (1992). EMBO J. 11, 3323-3325. Yoshioka, K., Deng, T., Cavigelli, M., and Karin, M. (1995). Proc. Natl Acad. Sci. U.S.A. 92, 4972-4976. Yu, B. P. (1994). Physiol. Rev. 74, 139-162. Zhan, Q., Carrier, F., and Fornace, A. J. (1993). Mol. Cell. Biol. 13, 4242-4250.

CURRENT TOPICS IN CELLULAR REGULATION, VOLUME 35

Regulation of N F - K B and Disease Control: Identification of a Novel Serine Kinase and Thioredoxin as Effectors for Signal Transduction Pathway for N F - K B Activation TAKASHI OKAMOTO SHINSAKU SAKURADA JiAN-PiNG Y A N G JOCELYN P . M E R I N

Department of Molecular Genetics Nagoya City University Medical School Nagoya, Aichi 467, Japan

I. Introduction One of the principal chemical reactions for life on this planet is oxidoreduction. Considering the strong chemical reactivity of the thiols on protein molecules, it is believed t h a t the primordial life form was dependent on the energy generated by oxidoreductive reactions (de Duve, 1991). For modern living organisms, however, oxygen plays unbivalent roles: while oxygen is indispensable for the cell to obtain the essential chemical energy as a form of adenosine triphosphate (ATP), it is often transformed into highly reactive forms, radical oxygen intermediates (ROI), which are often toxic to the cell (Stadtman and Oliver, 1991). In order to defend themselves from the toxic actions of ROI, living organisms have acquired multiplicated endogenous antioxidant systems during their long evolutionary periods. These defense mechanisms include glutaredoxin, thioredoxin, and superoxide dismutases. Interestingly, studies of cell biology and biochemistry have revealed t h a t these molecules are involved in cell signaling (Holmgren, 1985, 1989; Ziegler, 1985; Allen, 1993). The term redox regulation has been proposed, indicating the active role of oxidoreductive modifications of proteins in regulating their functions. In this chapter we will describe the nature of redox regulation of one transcription factor, nuclear factor kappa B ( N F - K B ) , as such an example and discuss its applicability in biology and medicine. 149

Copyright © 1997 by Academic Press. All rights of reproduction in any form reserved. 0070-2137/97 $25.00

150

TAKASHI OKAMOTO et al.

II. Transcription Factor N F - K B and its Activation Pathways A. NF-KB and Diseases

Recognition of regulatory mechanisms of gene expression by transcription factors has indicated a new strategy to control disease processes in which transcriptional control of pathogenetic genes plays a major role. Transcriptional control involves signal transduction pathways to transcription factors and recognition of the specific DNA sequence motif within the promoter regions of the relevant genes. Although there are various kinds of signaling pathways for individual transcription factors we will summarize our recent observations regarding one such example for NF-ACB. N F - K B regulates expression of a wide variety of cellular and viral genes (for reviews, see Gilmore, 1990; Baeuerle, 1991; Baeuerle and Henkel, 1994; Thanos and Maniatis, 1995). These genes include cytokines such as interleukin-2 (IL-2), IL-6, IL-8, GM-CSF, and tumor necrosis factor (TNF), cell adhesion molecules such as ICAM-1 and Eselectin, inducible nitric oxide synthase (iNOS) and viruses such as h u m a n immunodeficiency virus (HIV) and cytomegalovirus (CMV) (Schindler and Baichwal, 1994; Okamoto et al., 1989; Maekawa et al., 1989; Stade et al, 1990; Mukaida et al, 1990; Whelan et al, 1991; Roebuck et al, 1995; Donnelly et al., 1993; Schreck and Baeuerle, 1990; Staynov et al, 1995; Xie et al, 1994). Because these gene products are also known to play major roles in the pathogenetic processes of various diseases, N F - K B is regarded as a good therapeutic target. For example. Fig. 1 indicates the relationship between transcription factors and diseases. As shown in Fig. 1, the relationships between N F - K B and rheumatoid arthritis, hematogenic cancer metastasis, and acquired immunodeficiency syndrome (AIDS) are interconnected by the sets of genes t h a t are under the control of N F - K B . Through the causal relationship with these genes, N F - K B is considered to be involved in these currently intractable diseases. Although the genes induced by N F - K B vary with the context of cell lineage and are under the control of other transcription factors, N F - K B plays a major role in regulation of these genes and thus contributes a great deal to pathogenesis. Therefore, biochemical intervention of N F - K B should interfere with the pathogenic process and would be effective for the treatment. B. Regulation of N F - K B through Signal Transduction Pathways 1. ACTIVATION OF N F - K B

N F - K B consists of two subunit molecules, p65 and p50, and usually exists as a molecular complex with an inhibitory molecule, I-KB, in the

151

REGULATION OF NF-KB AND DISEASE CONTROL

Transcript!on Factors (

^

Rheumatoid l l l l C U I I I v l li\^l VI

r IL-2 . IL-6 IL-8 NF-IL6^-' GM-CSF ' V\wS/N ^ TNF NF"AT1< ^ \ ^ ^ ^ ICAM-1 * s V \ \ ^ VCAM-1 > ^ E-selectm< '^\ •^ HIV ^^ . J NF-KB

v

Diseases

Genes

Arthritis (RA)

^

Hematogenic l\/letastasis of Cancer Acquired Immunodeficiency Syndrome (AIDS) v

>

FIG. 1. Relationships between transcription factors and diseases: a pathogenetic paradigm regarding NF-/
cytosol (Sen and Baltimore, 1986; Nable and Baltimore, 1987; Baeuerle and Baltimore, 1988a,b; Ghosh ei^aZ., 1990; Ghosh and Baltimore, 1990; Read et aL, 1994). On stimulation of the cells such as by proinflammatory cytokines, IL-1, and TNF, I-KB is dissociated and NF-fcB is translocated to the nucleus and activates expression of target genes (Fig. 1). Thus, activity of NF-ACB itself is regulated by the upstream regulatory mechanism. Not much is known about the upstream signaling cascade. However, there are at least two independent steps in the N F - K B activation cascade: kinase pathways and redox-signaling pathways. These two distinct pathways are involved in the NF-/<:B activation cascade in a coordinate fashion, which may contribute to a fine-tuned, as well as fail-safe, regulation of NF-/<:B activity. In the following sections, we will summarize what is known with regard to the regulatory cascades leading to the activation of N F - K B . 2. SIGNALING EFFECTORS FOR N F - K B ACTIVATION

a. Kinase Pathways.

At least two distinct types of kinase pathways

are known to be involved in N F - K B activation: N F - K B kinase and IKB kinase. We found a 43-kDa serine kinase, N F - K B kinase, t h a t is associated with N F - K B (Hayashi et al, 1993b). This kinase phosphorylates both subunits of N F - K B and dissociates it from I - K B . Molecular

152

TAKASHI OKAMOTO et al.

cloning of the gene for this kinase is now underway. There is another kinase (or kinases) t h a t is known to phosphorylate I-/cB (Shirakawa and Mizel, 1989; Meichle et al, 1990; Feuillard et al, 1991; Ostrowski et al, 1991; Schutze etal, 1992), although the nature of I-ACB kinase(s) is not clear. Consistent with these findings, N F - K B was shown to be phosphorylated in some cell lines, and I-KB was phosphorylated in others in response to stimulation with TNF or IL-1 (Naumann and Scheidereit, 1994; Li et al, 1994). In most cases, N F - K B dissociation by the kinase cascade is a primary step in N F - K B activation. b. Thioredoxin and Redox Regulation of NF-KB. After dissociation from I - K B , however, NF-ACB must go through redox regulation by the cellular reducing catalyst thioredoxin (Okamoto et al, 1992; Hayashi et al, 1993a). Interestingly, h u m a n thioredoxin was initially identified as a factor responsible for induction of the a subunit of the IL-2 receptor, which is now known to be under the control of NF-fcB (Tagaya et al, 1989; Wakasugi et al, 1990). It is known t h a t N F - K B cannot bind to the KB D N A sequence of the target genes until it is reduced (Okamoto et al, 1992; Schreck et al, 1991; Molitor et al, 1991; Toledano and Leonard, 1991; Matthews et al., 1992). In a previous paper, we identified the cysteine residue at amino acid position 62 of the p50 subunit as a target of redox regulation based on the high p / values near the conserved cysteine residues (Hayashi et al, 1993a). Structural biological approaches have shed new light on the redox regulation of NF-zcB by thioredoxin. First, in early 1995, two groups independently demonstrated the three-dimensional structure of the N F - K B subunit p50 homodimer cocrystallized with the target DNA (Ghosh et al, 1995; Mtiller et al, 1995). N F - K B appears to have a novel DNA-binding structure called /3 barrel, a group of jS sheets stretching toward the target DNA. There is a loop in the tip of the j8 barrel structure t h a t intercalates with the nucleotide bases and is believed to make direct contact with the DNA. This DNA-binding loop contains the cysteine-62 that we predicted to be the target of redox regulation as a proton donor from thioredoxin. Although in both studies this cysteine was replaced with alanine, presumably for technical reasons for crystallization, these observations confirm our earlier speculations (Hayashi et al, 1993a). Furthermore, another group at the National Institutes of Health has determined the three-dimensional structure of the thioredoxin molecule associated with the DNA-binding loop of p50 by using nuclear magnetic resonance (NMR) (Qin et al, 1995). A boot-shaped hollow on the surface of thioredoxin containing the redox-active cysteines can

REGULATION OF NF-KB AND DISEASE CONTROL

153

recognize the DNA-binding loop of p50 and is likely to reduce the oxidized cysteine by donating protons in a structure-dependent way. Therefore, the reduction of N F - K B by thioredoxin is considered to be specific. 3. INVOLVEMENT OF RADICAL OXYGEN INTERMEDIATES

What triggers these NF-ACB signaling cascades? Not much is known about what happens immediately downstream of the cell surface receptor. However, involvement of ROI is suggested upstream of the N F - K B activation pathway since signaling is efficiently blocked by pretreatment of the cells with antioxidants such as A^-acetylcysteine (NAG) or a-lipoic acid (Roederer et ah, 1990; Schreck et aL, 1991; Suzuki et al., 1992; Meyer et al, 1993; Biswas et al, 1993; Suzuki and Packer, 1994; Packer et al., 1995). Therefore, antioxidants are now considered to be effective N F - K B inhibitors. Moreover, we found t h a t NAG blocks the induction of thioredoxin (Sachi et al., 1995; Okamoto et al., submitted). Therefore, anti-NF-fcB actions of antioxidants are considered to be twofold: (1) blocking the signaling immediately downstream of the signal elicitation and (2) suppression of induction of the redox effector thioredoxin (Fig. 2). We found t h a t even the fully activated N F - K B could be blocked by gold ion by a redox mechanism (Yang et al., 1995). We found t h a t the zinc ion is a necessary component of the active N F - K B and t h a t addition of monovalent gold ion could efficiently block its activity by oxidizing the redox-active cysteines on NF-/<:B. Because gold did not appear to replace zinc (Yang et al., 1995), it is likely t h a t gold ion oxidizes these thiolate anions on N F - K B into disfulfides and thus abrogates the DNAbinding activity because of its higher oxidation potential over zinc ion. It is notable t h a t gold compounds have been successfully used for the treatment of rheumatoid arthritis (Skosey, 1993; Insel, 1990). Our finding could explain why gold is effective in rheumatoid arthritis and suggests t h a t N F - K B might have a crucial role in the disease process (Sakurada et al., 1996). It may be t h a t gold compound is potentially effective in other diseases where N F - K B plays a pathological role.

III. Involvement of N F - K B in Disease Processes A. Acquired Immunodeficiency Syndrome The pivotal role of N F - K B in the h u m a n immunodeficiency virus (HIV) life cycle, especially in the virus reactivation process within latently infected cells, has been widely accepted. After activation through intracellular signaling pathways such as those elicited by the

154

TAKASHI OKAMOTO et al.

1) Dissociation of IKB by NF-KB kinase or

IKB

kinase

IKB Kinasj

NF-KI

Kinase

NF-KB

2) Redox regulation of NF-KB by thioredoxin (Trx)

DNA binding (+)

FIG. 2. Signal transduction pathways for NF-/cB activation. The first step involves kinase pathways such as by NF-KB and I-KB kinases. The second step involves redox regulation by thioredoxin (Trx). See text for details.

T-cell receptor-antigen complex or by receptors for IL-1 or TNF, N F KB initiates HIV gene expression by binding to the target DNA element within the promoter region of the HIV long terminal repeat (LTR) (Nabel and Baltimore, 1987; Bohnlein et al, 1988; Okamoto et al, 1989, 1990) (Fig. 3b). Then the virus-encoded i^razis-activator Tat is produced and triggers explosive viral replication (Arya et al, 1985; Sodroski et al, 1985; Okamoto and Wong-Staal, 1986; Peterlin et al, 1986). Because the activation pathway of HIV gene expression by cellular transcription factor NF-KB conceptually precedes activation by viral transactivators, it is conceptual to ascribe NF-KB as a determinant of the maintenance and breakdown of viral latency. Attempts to control the

155

REGULATION OF NF-ZCB AND DISEASE CONTROL (a)

N F - K B kinase, p105 kinase

®

IKB kinase Receptors . Cytokines ^^ ^ ^ ^ ^ ~ ] '^^^ Signals (antigens, ^^ ' superantigens) Others

®

HIV virions

RNA poi 11

FIG. 3. Transcriptional regulation of HIV replication involving NF-KB and Tat. (a) Activation of NF-KB by various extracellular stimuli. Extracellular stimuli include cytokines such as TNF and IL-1, signals from the T-cell receptor (in the case of T cells), phorbor esters, and ultraviolet irradiation. On transduction of signals to kinases (NFKB kinase and other putative kinases), either directly or indirectly, 1-KB is dissociated from the NF-KB/I-KB complex. The liberated NF-/cB translocates to the nucleus, where it activates the target genes, including HIV provirus. The site of action of Trx on NFKB has not been clarified. However, as shown in Fig. 2, NF-KB must go through redox regulation by Trx to be fully activated, (b) Major regulatory events on the HIV-1 LTR in the transcriptional activation program by NF-KB and Tat. Induction of the initiation of viral transcription by NF-KB is believed to be followed by Tat production from the viral genome. The Tat protein then further augments the rate and efficiency of viral mRNA transcription by RNA pol3mierase II. Spl appears to be involved in the action of Tat (Okamoto et al., 1990). Other transcription factors interacting with the cis-regulatory elements within the HIV-1 LTR, such as NFAT-1 and USF, are believed to have minor roles compared with NF-ACB and Tat.

156

TAKASHI OKAMOTO et

al.

signaling pathways to NF-ACB activation will give us a clue to biochemical intervention in the clinical development of AIDS. B. Cancer Metastasis

Another situation in which NF-KB plays a role is hematogenic cancer cell metastasis (Tozawa et al, 1995). NF-KB induces E-selectin (also known as ELAM-1) on the surface of vascular endothelial cells (Montgomery et al, 1991; Whelan et aL, 1991). Because some cancer cells constitutively express a ligand for E-selectin, called sialyl-Lewis^ antigen, on their surface, induction of E-selectin is considered to be a ratedetermining step of cancer cell-endothelial cell interaction (Dejana et al, 1988; Takada et al, 1993). We examined this phenomenon in detail with regard to the role of NF-KB in induction of E-selectin (Tozawa et al, 1995). When primary human umbilical venous endothelial cells (HUVEC) were treated with IL-1 or TNF, nuclear translocation of NF-KB was observed, followed by the augmented expression of E-selectin. We examined the cell-tocell interaction between HUVEC and QG90 cell, a tumor cell line derived from human small cell carcinoma of the lung expressing sialylLewis^ antigen and found that IL-1 was able to induce the attachment of cancer cells to HUVEC (Fig. 4). However, pretreatment of HUVEC Endothelial Cell

E-selectin

Sialyl Lewis^

Proinflammatory Cytokines

Cancer Cell

FIG. 4. Involvement of NF-/cB in hematogenic cancer cell metastasis. NF-KB is activated by proinflammatory cytokines produced from macrophages or T cells (which may be attacking the cancer cells), and E-selectin is induced by the activated NF-KB in endothelial cells. Cancer cells may attach to the endothelial cells through the interaction of E-selectin on the surface of endothehal cells and the ligand for E-selectin, sialyl-Lewis^ antigen, on the surface of cancer cells (called the rolling and tethering processes), thus escaping from the bloodstream and invading the metastatic tissues.

REGULATION OF NF-fCB AND DISEASE CONTROL

157

with iV-acetylcysteine, aspirin, or pentoxyphillin efficiently blocked the cell-to-cell attachment in a dose-dependent manner. IV. Screening Strategy for Anti-NF-KB Compounds As an application of basic studies on transcriptional regulation of HIV, we have constructed a plasmid to genetically transduce the regulatory cascade of virus replication (Kira et al., 1995). The pKO plasmid has been created containing the Tat gene under the control of cytomegalovirus (CMV) immediate early promoter which is positively controlled by N F - K B . This plasmid also contains the hygromycin B resistance gene (hygromycin B phosphotransferase gene) under the control of HIV LTR, which is induced by Tat and NF-KB. The cells stably transfected with the pKO plasmid can survive in the presence of the antibiotic hygromycin B as long as the hygromycin B resistance gene is expressed by the combinatorial actions of NF-KB and Tat (Kira et al., 1995). This system could be used not only for anti-Tat but also for anti-NF-^B compounds (Kira et al., 1996; Merin et al., 1996). In using pKO for antiNF-KB screening, we assume that Tat is acting simply as an amplifier of NF-KB action, as in the case of HIV replication (Fig. 5). Because the pKO plasmid can endow any cell with drug resistance to hygromycin B, irrespective of the NF-KB activation cascade that is now known to be variable in different cell lineages, this strategy would be feasible for the screening of anti-NF-AcB drugs in various disease conditions by using the relevant cells. We confirmed the therapeutic efficacies of known anti-Tat (Hsu et al, 1991, 1993) and anti-NF-/<:B (Suzuki et al, 1992) compounds in blocking HIV-1 replication using these cells. V. Summary We have identified novel signal transduction cascades in activating NF-KB, as well as its pathogenetic roles in various disease processes. By applying the basic knowledge obtained through these studies, we hope to find new therapeutic measures against currently incurable diseases such as hematogenic cancer cell metastasis, rheumatoid arthritis, and AIDS. We also propose a novel strategy in screening effective inhibitors against transcription factors. Elucidation of the cisregulatory element for expression of pathogenetic genes and identification of the responsible transcription factor will not only facilitate the study of pathogenesis but will also promote the development of effective therapy. Recognition of control mechanisms of the NF-f<:B activation pathway has explained the therapeutic efficacy of various compounds

158

TAKASHI OKAMOTO et al.

Hygromycin (non-toxic)

Hygromycin (toxic)

Hygromycin phosptiotransferase

FIG. 5. Strategy of an anti-HIV drug screening system using the viral regulatory network of gene expression. The pKO plasmid was created to render the cells with resistance to hygromycin B by expressing the hygromycin B phosphotransferase gene (hygromycin') under the control of Tat and NF-KB (Kira et al, 1995). The anti-Tat and anti-NF-zcB compounds were shown to reverse the phenotype of the cells, making them sensitive to hygromycin B. Cell viability is easily monitored by non-RI method such as using MTT dye.

with different pharmacologic actions. A similar strategy may be applicable for other inducible transcription factors. From the medical point of view, one of the purposes of these approaches is to find small molecular weight compounds that can be administered orally and that are effective in controlling gene expression of pathogenetic genes. REFERENCES Allen, J. F. (1993). FEES Lett 332, 203-207. Arya, S. K, Guo, C, Josephs, S. F., and Wong-Staal, F. (1985). Science 229, 69-73. Baeuerle, P. A. (1991). Biochim. Biophys. Acta 1072, 63-80. Baeuerle, P. A., and Baltimore, D. (1988a). Cell (Cambridge, Mass.) 53, 211-217. Baeuerle, P. A., and Baltimore, D. (1988b). Science 242, 540-546. Baeuerle, P. A., and Henkel, T. (1994). Annu. Rev. Immunol. 12, 141-179.

REGULATION OF NF-KB AND DISEASE CONTROL

159

Biswas, D. K., Dezube, B. J., Ahlers, C. M., and Pardee, A. B. (1993). J. AIDS 6, 778-786. Bohnlein, E., Lowenthal, J. W., Siekevitz, M., Ballard, D. W., Franza, B. R., and Greene, W. C. (1988). Cell (Cambridge, Mass.) 53, 827-836. de Duve, C. (1991). "Blue Print for a Cell: The Nature and Origin of Life." Neil Patterson Publ., North Carolina. Dejana, E., Bertocci, F., Bortolami, M. C , Regonesi, A., Tonta, A., Breviario, F., and Giavazzi, R. (1988). J. Clin. Invest. 82, 1466-1470. Donnelly, R. P., Crofford, L. J., Freeman, S. L., Buras, J., Remmers, E., Wilder, R. L., and Fenton, M. J. (1993). J. Immunol. 151, 5603-5612. Feuillard, J., Gouy, H., Bismuth, G., Lee, L. M., Debre, P., and Korner, M. (1991). Cytokine 3, 257-265. Ghosh, G., Van Duyne, G., Ghosh, S., and Sigler, P. B. (1995). Nature (London) 373, 303-310. Ghosh, S., and Baltimore, D. (1990). Nature (London) 344, 678-682. Ghosh, S., Gifford, A. M., Riviere, L. R., Tempst, P., Nolan, G. P., and Baltimore, D. (1990). Cell (Cambridge, Mass.) 62, 1019-1029. Gilmore, T. D. (1990). Cell (Cambridge, Mass.) 62, 841-843. Hayashi, T., Ueno, Y., and Okamoto, T. (1993a). J. Biol. Chem. 268, 11380-11388. Hayashi, T., Sekine, T., and Okamoto, T. (1993b). J. Biol. Chem. 268, 26790-26795. Holmgren, A. (1985). Annu. Rev. Biochem. 54, 237-271. Holmgren, A. (1989). J. Biol. Chem. 264, 13963-13966. Hsu, M., Schutt, A. D., Holly, M., Slice, L. W., Sherman, M. I., Richman, D. D., Potash, M. J., and Volsky, D. J. (1991). Science 254, 1799-1802. Hsu, M., Dhingra, U., Earley, J. V., Holly, M., Keith, D., Nalin, C. M., Richou, A. R., Schutt, A. D., Tam, S. Y., Potash, M. J., Volsky, D. J., and Richman, D. D. (1993). Proc. Natl. Acad. Sci. U.S.A. 90, 6395-6399. Insel, P. A. (1990). In "Autacoids: Drug Therapy of Inflammation" (G. Oilman et al., eds.), pp. 670-681. Macmillan, New York. Kira, T., Merin, J. P., Baba, M., Shigeta, S., and Okamoto, T. (1995). AIDS Res. Hum. Retroviruses 11, 1359-1366. Kira, T., Hashimoto, K , Baba, M., Okamoto, T., and Shigeta, S. (1996). Antiviral Res. 32,55-62. Li, C.-C. H., Dai, R.-M., Chen, E., and Longo, D. L. (1994). J. Biol. Chem. 269, 3008930092. Maekawa, T., Itoh, F., Okamoto, T., Kurimoto, M., Imamoto, F., and Ishii, S. (1989). J. Biol. Chem. 264, 2826-2831. Matthews, J. R., Wakasugi, N., Virelizier, J.-L., Yodoi, J., and Hay, R. T. (1992). Nucleic Acids Res. 20, 3821-3830. Meichle, A., Schutze, S., Hensel, G., Brunsing, D., and Kronke, M. (1990). J. Biol. Chem. 265, 8339-8343. Merin, J. P., Matsuyama, M., Kira, T., Baba, M., and Okamoto, T. (1996). FEBS Lett.

394,9-13. Meyer, M., Schreck, R., and Baeuerle, P. A. (1993). EMBO J. 12, 2005-2015. Molitor, J. A., Ballard, D. W., and Greene, W. C. (1991). New Biol. 3, 987-996. Montgomery, K. F., Osborn, L., Hession, C , Tizard, R., Goff, D., Vassallo, C , Tarr, P. I., Bomsztyk, K , Lobb, R., Harlan, J. M., and Pohlman, T. H. (1991). Proc. Natl. Acad. Sci. U.S.A. 88, 6523-6527. Mukaida, N., Mahe, Y., and Matsushima, K. (1990). J. Biol. Chem. 265, 21128-21133. Mtiller, C. W., Rey, F. A., Sodeoka, M., Verdine, G. L., and Harrison, S. C. (1995). Nature

(London) SIS,

Sll-Sn.

160

TAKASHI OKAMOTO et

al.

Nabel, G., and Baltimore, D. (1987). Nature {London) 326, 711-713. Naumann, M., and Scheidereit, C. (1994). EMBO J. 13, 4597-4607. Okamoto, T., and Wong-Staal, F. (1986). Cell (Cambridge, Mass.) 47, 2 9 - 3 5 . Okamoto, T., Matsuyama, T., Mori, S., Hamamoto, Y., Kobayashi, N., Yamamoto, N., Josephs, S. F., Wong-Staal, F., and Shimotohno, K. (1989). AIDS Res. Hum. Retrovir. 5, 131-138. Okamoto, T., Benter, T., Josephs, S. F., Sadaie, M. R., and Wong-Staal, F. (1990). Virology 177, 606-614. Okamoto, T., Ogiwara, H., Hayashi, T., Mitsui, A., Kawabe, T., and Yodoi, (1992). Int. Immunol. 4, 811-819. Okamoto, T. et al. (1997). In preparation. Ostrowski, J., Sims, J. E., Sibley, C. H., Valentine, M. A., Dower, S. K , Meier, K. E., and Bomsztyk, K. (1991). J. Biol. Chem. 266, 12722-12733. Packer, L., Witt, E. H., and Tritschler, H. J. (1995). Free Radical Biol. Med. 19,227-250. Peterlin, B. M., Luciw, P. A., Barr, P. J., and Walker, M. D. (1986). Proa. Natl. Acad. Sci. U.S.A. 83, 9734-9738. Qin, J., Clore, G. M., Kennedy, W. M. P., Huth, J. R., and Gronenborn, A. M. (1995). Structure 3, 289-297. Read, M. A., Whitley, M. Z., WiUiams, A. J., and Collins, T. (1994). J. Exp. Med. 179, 503-512. Roebuck, K. A., Rahman, A., Lakshminarayanan, V., Janakidevi, K., and Malik, A. B. (1995). J. Biol. Chem. 270, 18966-18974. Roederer, M., Staal, F. J. T., Raju, P. A., Ela, S. W., Herzenberg, L. A., and Herzenberg, L. A. (1990). Proc. Natl. Acad. Sci. U.S.A. 87, 4884-4888. Sachi, Y., Hirota, K., Masutani, H., Toda, K , Okamoto, T., Takigawa, M., and Yodoi, J. (1995). Immunol. Lett. 44, 189-193. Sakarada, S., Kato, T., and Okamoto, T. (1996). Int. Immunol. 8, 1483-1493. Schindler, U., and Baichwal, V. R. (1994). Mol. Cell. Biol. 14, 5820-5831. Schreck, R., and Baeuerle, P. A. (1990). Mol. Cell. Biol. 10, 1281-1286. Schreck, Rieber, R. P., and Baeuerle, P. A. (1991). EMBO J. 10, 2247-2258. Schutze, S., Potthoff, K , Machleidt, T., Berkovic, D., Wiegmann, K., and Kronke, M. (1992). Cell (Cambridge, Mass.) 7 1 , 765-776. Sen, R., and Baltimore, D. (1986). Cell (Cambridge, Mass.) 46, 705-716. Shirakawa, F., and Mizel, S. B. (1989). Mol. Cell. Biol. 9, 2424-2430. Skosey, J. L. (1993). In "Arthritis and Alhed Conditions" (D. J. McCarty and W. J. Koopman, eds.), pp. 603-614. Lea & Febiger, Philadelphia. Sodroski, J., Patarca, R., and Rosen, C. (1985). Science 229, 74-77. Stade, B. G., Messer, G., Riethmuller, G., and Johnson, J. P. (1990). Immunobiology 182, 79-87. Stadtman, E. R., and Oliver, C. N. (1991). J. Biol. Chem. 266, 2005-2008. Staynov, D. Z., Cousins, D. J., and Lee, T. H. (1995). Proc. Natl. Acad. Sci. U.S.A. 92, 3606-3610. Suzuki, Y. J., and Packer, L. (1994). J. Immunol. 153, 5008-5015. Suzuki, Y. J., Aggarwal, B. B., and Packer, L. (1992). Biochem. Biophys. Res. Commun. 189, 1709-1715. Tagaya, Y., Maeda, Y., Mitsui, A., Kondo, N., Matsui, H., Hamuro, J., Brown, J., Arai, K. L, Yokota, T., Wakasugi, H., and Yodoi, J. (1989). EMBO J. 8, 757-764. Takada, A., Ohmori, K , Yoneda, T., Tsuyuoka, K., Hasegawa, A., Kiso, M., and Kannagi, R. (1993). Cancer Res. 5S, 354-361. Thanos, D., and Maniatis, T. (1995). Cell (Cambridge, Mass.) 80, 529-532.

REGULATION OF N F - K B AND DISEASE CONTROL

161

Toledano, M. B., and Leonard, W. J. (1991). Proc. Natl Acad. Sci. U.S.A. 88,4328-4332. Tozawa, K , Sakurada, S., Kohri, K , and Okamoto, T. (1995). Cancer Res. 55,4162-4167. Wakasugi, N., Tagaya, Y., Wakasugi, H., Mitsui, A., Maeda, M., Yodoi, J., and Tursz, T. (1990). Proc. Natl Acad. Sci. U.S.A. 87, 8282-8286. Whelan, J., Ghersa, P., Huijsduijnen, R. H., Gray, J., Chandra, G., Talabot, F., and DeLamarter, J. F. (1991). Nucleic Acids Res. 19, 2645-2653. Xie, Q.-W., Kashiwabara, Y., and Nathan C. (1994). J. Biol. Chem. 269, 4705-4708. Yang, J. P., Merin, J. P., Nakano, T., Kato, T., Kitade, Y., and Okamoto, T. (1995). FEES Lett. 361, 89-96. Ziegler, D. M. (1985). Annu. Rev. Biochem. 54, 305-329.

This Page Intentionally Left Blank

CURRENT TOPICS IN CELLULAR REGULATION, VOLUME 35

Regulation of Bacterial Responses to Oxidative Stress JUDAH L. ROSNER* GiSELASTORzt * Laboratory of Molecular Biology National Institute of Diabetes and Digestive and Kidney Diseases National Institutes of Health Bethesda, Maryland 20892 f Cell Biology and Metabolism Branch National Institute of Child Health and Human Development National Institutes of Health Bethesda, Maryland, 20892

I. Introduction Oxidative stress occurs when organisms encounter elevated levels of reactive oxygen species such as superoxide anion (O2T), hydrogen peroxide (H2O2), and hydroxyl radical (0H-). The reactive oxygen species are produced at low rates during normal aerobic respiration in both prokaryotes and eukaryotes. For example, the intracellular superoxide anion concentrations for aerobically growing Escherichia coli cells have been measured to be 10"^^ M, while the concentrations of hydrogen peroxide are maintained at around 10"^ M (27,35). A variety of environmental conditions, however, can lead to increased levels of these reactive oxygen species. Shifts between aerobic and anaerobic environments and exposure to radiation, metals, and xenobiotic drugs capable of reacting with oxygen species can all result in elevated superoxide, hydrogen peroxide, and hydroxyl radical concentrations. Another source of oxidants is the reactive species generated by phagocytes in a defense against microorganisms. Reactive oxygen species can cause DNA mutations, enzyme inactivation, and membrane damage, but all bacteria, even strict anaerobes, appear to have mechanisms to detoxify these deleterious oxidants. Superoxide dismutases which convert superoxide anion to hydrogen peroxide and catalase/peroxidases which eliminate hydrogen peroxide are central to the defense against oxidative stress and are very con163

Copyright © 1997 by Academic Press. All rights of reproduction in any form reserved. 0070-2137/97 $25.00

164

JUDAH L. ROSNER AND GISELA STORZ

served. Studies have shown that an alkyl hydroperoxide reductase which can convert peroxides such as hneohc hydroperoxide to their corresponding alcohols is also a ubiquitous defense activity (11). In addition, enzymes and DNA binding proteins that repair or protect against oxidative DNA damage are critical and appear to be conserved. The expression of many of the defense activities is induced by changes in the levels of the hydrogen peroxide or superoxide, suggesting that many cells have mechanisms to sense reactive oxygen species. In this chapter, we review the properties of transcriptional regulators that are important for the induction of antioxidant defense genes in bacteria. The regulators have been best characterized in E. coli, but the studies of the oxidative stress responses in other bacterial species are pointing to some interesting similarities and differences between bacteria. Here we compare the oxidative stress responses of E. coli, Salmonella, Haemophilus, Mycobacterium,, and Bacillus; discuss interesting connections between oxidative stress and pathogenesis and drug resistance in these organisms; and propose directions for future studies.

II. Regulators of Escherichia coli Responses to Oxidative Stress Like many genetic responses, the defenses against oxidative stress have been best studied in E. coli (see refs. 17 and 22 for comprehensive reviews). Escherichia coli cells have two catalase/peroxidase activities, denoted hydroperoxidase I (HPI, encoded by katG) and hydroperoxidase II (encoded by katE), as well as three superoxide dismutase activities, manganese superoxide dismutase (encoded by sodA), iron superoxide dismutase (encoded by sodB), and copper-zinc superoxide dismutase (7). TheE. coli alkyl hydroperoxide reductase activity is composed of two subunits: a 22-kDa subunit (encoded by ahpC) and a 52-kDa subunit (encoded by ahpF). DNA binding or repair activities that appear to be critical for protection against oxidative damage include a nonspecific DNA binding protein (encoded by dps), exonuclease III {xthA), endonuclease IV (encoded by nfo), and the RecA recombinase. A. OxyR

Escherichia coli cells show an adaptive response to hydrogen peroxide and approximately 30 proteins are induced when treated with low concentrations of hydrogen peroxide (17,22). The expression of nine of the hydrogen peroxide-inducible proteins is controlled by OxyR. Several of the proteins whose expression is activated by OxyR have been identified and include HPI catalase, the alkyl hydroperoxide reductase, the

REGULATION OF BACTERIAL RESPONSES TO OXIDATIVE STRESS

165

Dps protein, and glutathione reductase (encoded by gorA). All of these activities have understandable roles in protecting the cell against oxidative damage. OxyR also activates the expression of a small untranslated regulatory RNA denoted OxyS, but the role of this RNA in the defense against oxidative stress is not yet understood (S. Altuvia, D. Weinstein, A. Zhang, L. Postow, and G. Storz, unpublished). OxyR has also been shown to be a repressor and negatively autoregulates its own expression, so a constant level is maintained in the cell. In addition, two studies have shown that OxyR represses the expression of the Mu phage mom gene and the E, coli flu gene (30), neither of which has an understandable role in the oxidative stress response. An interesting direction for further research will be to elucidate all of the roles of OxyR within the cell, as well as to characterize the other proteins induced by hydrogen peroxide. The tetrameric OxyR protein is a member of the LysR family of transcriptional activators and has been characterized extensively (17,22,41,42). The protein appears to exist in two forms, reduced and oxidized, but only the oxidized form is able to activate transcription. Direct oxidation of OxyR is therefore the likely mechanism whereby the cells sense hydrogen peroxide and induce the OxyR regulon. The redox-active center in OxyR has been proposed to be a single cysteine, and the next challenge will be to determine the nature of the oxidative reaction that activates OxyR. OxyR has been shown to bind to promoters by contacting four major grooves on one face of the DNA, and a putative consensus of four repeats of ATAGnt has been defined for OxyR (63). Interestingly, the two forms of OxyR appear to have different binding specificities. The reduced form is able to bind the oxyR and mom promoters, but not the^a^G and ahpC promoters, and contacts ATAG nucleotide repeats in two pairs of adjacent major grooves separated by one helical turn. In contrast, oxidized OxyR has been found to bind all OxyR-regulated promoters that have been tested and binds in four adjacent major grooves. The differences in binding may allow OxyR to carry out different functions under different conditions. Therefore OxyR can repress the oxyR and mom promoters during normal growth and activate katG and ahpC in response to oxidative stress. OxyR activates transcription by increasing the binding of RNA polymerase to the promoters and has been shown to require specific surfaces on the C-terminal domain of the a subunit of RNA polymerase {a CTD) to activate transcription (61,62). B. SoxRS Escherichia coli cells also adapt to increased levels of superoxide, and key regulators of this response are the SoxR and SoxS proteins (17,22). The genes activated by SoxS include sodA, nfo, micF, zwf, acnA,

166

JUDAH L. ROSNER AND GISELA STORZ

fumC, fpr, and acrAB, in addition to a few other genes identified by lacZ fusions or by two-dimensional polyacrylamide gel electrophoresis (PAGE) (Table I). The corresponding gene products help to eliminate superoxide {sodA), repair damaged DNA {nfo), reduce outer membrane permeability (micF mRNA interferes with the translation of ompF, which encodes an outer membrane porin), increase the reducing power of the cell (zwf), and encode superoxide-resistant isozymes of fumarase (fumC) and aconitase (acnA). Unexpectedly, the soxRS regulon also confers multiple antibiotic resistance as well as resistance to certain organic solvents and heavy metals, but the genes responsible for the latter defenses are not known. Regulation of the soxRS regulon occurs by a two-stage process. SoxR is first converted to an active form which enhances soxS transcription (17,22). The increased level of SoxS, in turn, activates expression of the regulon. Curiously, the genes encoding the two regulators overlap each other, with the soxR promoter embedded in the soxS structural gene and transcribed in the opposite direction. The constitutively expressed SoxR protein resembles MerR, a regulator of mercury resistance. Like MerR, SoxR is a dimer and has four C-terminal cysteines that are critical for activity. SoxR can be isolated as Fe-free or Fecontaining forms, and both forms can bind the soxS promoter, but only the Fe form with two [2Fe:2S] centers per dimer is able to activate transcription in vitro (31,32,64). The mechanism of SoxR activation and the nature of the signaling molecule are still under debate. Possibly SoxR exists as an apoprotein, and the full [2Fe:2S] clusters in SoxR are assembled with the iron released from superoxide-sensitive enzymes in the cell (31). Alternatively, the SoxR protein is normally in a reduced [2Fe:2S]+ state and is activated by oxidation to a [2Fe:2S]2+ state (64). This oxidation may occur through direct exposure to superoxide anion, although Liochev et al. argue that SoxR responds to changes in reduced nictotinamide-adenine dinucleotide phosphate (NADPH) or to reduced flavodoxin or ferredoxin levels (46). They observe that when zwf~ mutants, which have less ability to generate NADPH, are treated with low concentrations of paraquat, the expression of fumC and sodA is elevated, yet overexpression of manganese superoxide dismutase does not diminish these levels. However SoxR is activated, the regulator appears to distort the soxS promoter and thereby increases soxS transcription (31). The SoxS protein activates the promoters of the soxRS regulon by mechanisms which involve binding near or at the - 3 5 hexamer. SoxS and a MalE-SoxS fusion protein have been purified and shown to bind to several SoxS-regulated promoters (23,44). The core sequence of a

o

m o o

B 3 2

T3

1=3

"3 .-^

fj

o ft ft o

«-^

^

.1 CD

O

+ + +

+ + + + + +

+ + + ^ + + + + + + + + + + +

+ + + + + +

^

+ + + +

+ ^ + + + + + + + + + ^ + + + ^ + + + + + +

^ O B3 ! 2 ; pLH

PLH

o a

;2^

1

^ ^ + + + + + + + ^ + + + + + + + + + +

+ + + + + 0 + + +

PH

^ + + + + + + + + + + +

O V

a, 03 CO

•^^^

c3 o

5

^-

00

CD ^ Cd

g

A

CO

.

b e PLH S —<

o

I

-I I (D 0)

^ CD

02

^

bo

•S §

Cd

^ i=l

1

i ml

c^ S

CQ

Sal

^ CD ^

^.|

si 0

CO h (N c^ fN (N

+

1—I -

«3

5H

0

S ^ 0 r^ M 0 ;H o; r/i 0 ?^ f^ 0 -r^ iH

a

^ fl

W

P PI

CO (D (N"

0

<.«5

^

+3 +e

168

JUDAH L. ROSNER AND GISELA STORZ

proposed SoxS box is AnnGCAY. For some promoters such as zwf and fpr, the sequences bound by SoxS do not overlap the - 3 5 promoter sequence, and in vitro transcription experiments have shown that activation of these promoters requires the a CTD of RNA polymerase, indicating direct contact between SoxS and RNA polymerase (36). In contrast, at other promoters such as micF and fumC, SoxS binds at sites overlapping the - 3 5 hexamer, and the a CTD of RNA polymerase is not required for activation. Thus, SoxS, like the E, coli cyclic adenosine monophosphate (cAMP) receptor protein (CRP), is an "ambidextrous" transcriptional activator and activates RNA polymerase differently at different promoters. SoxS also binds to its own promoter, where it appears to downregulate its own transcription (54). C. MarA, Rob A surprise has been the finding that the soxRS regulon overlaps the regulon controlled by MarA, a regulator identified as part of an operon conferring multiple antibiotic resistance, and the regulon controlled by Rob, a protein detected originally by its ability to bind DNA near the bacterial origin of replication (3,17) (see Table I). Transcription of the marRAB operon is repressed by MarR, which binds to two sites between the - 3 5 transcriptional signal and the initial MarR methionine (49). The operon is induced by a variety of phenolic compounds, including salicylate and 2,4-dinitrophenol, and apparently by certain antibiotics, such as tetracycline. Salicylate, but not chloramphenicol or tetracycline, has been shown to bind MarR and thereby reduce the binding of MarR to the promoter (49). Derepression of the operon then results in synthesis of MarA, a transcriptional activator with about 40% sequence identity to SoxS and 50% identity to Rob. What regulates Rob and why it has a high constitutive level in the cell (5000 molecules) is not known (60). SoxS, MarA, and Rob resemble each other is a number of ways. They share the DNA binding motif characteristic of the AraC subfamily of transcriptional activators; are ambidextrous activators of many, if not all, of the same promoters in vitro and in vivo; bind to these promoters at very similar, in some cases identical, sequences; bend the DNA; and bind as monomers (36-38). Each of these transcriptional activators is functional in the absence of the others. Nevertheless, the degree to which the mar, soxRS, and rob regulons differ from each other has not been systematically studied. It may be that SoxS activates functions involved in superoxide defense more than MarA or Rob does (Table I), but the basis for the differences in expression is not known. Furthermore, it has not been clearly established which of the functions are

REGULATION OF BACTERIAL RESPONSES TO OXIDATIVE STRESS

169

required for antibiotic or superoxide resistance, and none of the genes have been evaluated with regard to the heavy metal- and organic solvent-resistance phenotypes seen for strains that overexpress SoxS or Rob (52). D. RpoS

One additional regulator that cannot be excluded from discussions of the E. coli response to oxidative stress is the rpoS-encoded a% subunit of RNA polymerase (see ref. 47 for a comprehensive review). This sigma (a) factor is important for the expression of a large group of genes that are induced when cells encounter a number of different stresses, including starvation, osmotic stress, and acid stress, as well as on entry into stationary phase. Starved and stationary phase cells are intrinsically resistant to a variety of stress conditions, including high levels of hydrogen peroxide, and RpoS has been shown to regulate the expression of the antioxidant defense activities encoded by katG, katE, dps, xthA, and gorA (1,6,47). The katG, dps, and gorA genes are activated by OxyR, suggesting that E. coli cells have two regulons that can protect against exosure to hydrogen peroxide: the OxyR regulon during exponential growth and the a^ regulon in stationary phase. It seems likely that some of the SoxS/MarA/Rob-regulated genes are also regulated by cr', and one SoxS target, acrAB, has been shown to be induced in stationary phase (48). The regulation of a^ levels occurs at multiple levels, and much remains to be learned (47). The transcription, translation, and stability of cr^ are all modulated in response to different signals, including the starvation signal ppGpp, a cell density signal homoserine lactone, cAMP, and uridine diphosphate (UDP)-glucose (9,34,43,47). The regulators that transmit the different stress signals are now being identified. They include CRP, which regulates transcription, and the general regulator H-NS, which affects both the translation and stability of cr^ (5,65). Studies have also shown that the stability of a^ is controlled by the ClpXP proteases and a response regulator protein alternatively denoted Hns/RssB/SprE (50,55,58). Among the interesting questions to be addressed in the future are how the different responses are integrated and whether oxidative stress has a direct impact on cr' levels.

III. Oxidative Stress Responses in Salmonella, Haemophilus, Mycobacterium, and Bacillus Antioxidant defense activities have been characterized in a number of different bacterial species, and it will be interesting to learn more

170

JUDAH L. ROSNER AND GISELA STORZ

about how the expression of the corresponding genes is regulated in the different organisms. In this section, we review the oxidative responses in prokaryotes in which regulators have been identified (see Table II). A. Salmonella typhlmurlum

All studies of the responses to oxidative stress in S. typhimurium suggest that they are very similar to the responses iriE. coli. Salmonella typhimurium has homologs of the peroxidases and superoxide dismutases encoded by katG, katE, ahpCF, sodA, and sodB, and mutational studies and sequence analysis have shown that the OxyR, SoxRS, MarA, and RpoS regulators are also present in S. typhimurium (17,22,47, 60a; E. A. Martins and B. Demple, personal communication). The S. typhimurium responses to oxidative stress are interesting in that several connections to virulence have been reported. Strains carrying either an oxyi? constitutive mutation or an oxyR deletion mutation are less virulent than the corresponding wild-type parent in vivo, suggesting that the OxyR-regulated response plays a role in virulence (24). The role of RpoS in Salmonella virulence is even more clearly established. Like JB. coli rpoS~ mutants, the S. typhimurium mutant strains show increased sensitivity to nutrient limitation, acid stress, and DNA-damaging agents, as well as oxidative stress (21). The rpoS~ TABLE II REGULATORS OF RESPONSE TO OXIDATIVE STRESS IN PROKARYOTES"'

Regulator

Map position of E. coli

Homologs found in

OxyR

89.6'

SoxR/SoxS

92.2'

MarA

34'

RpoS

59'

Escherichia coli Salmonella typhimurium Haemophilus influenzae Mycobacterium Escherichia coli Salmonella typhimurium Escherichia coli Salmonella typhimurium Escherichia coli Salmonella typhimurium Haemophilus influenzae Klebsiella pneumoniae Shigella flexneri Pseudomonas aeruginosa

'^From refs. 17, 18, 22, 25, 47, 59, and 60a; E. A. Martins and B. Demple, personal communication.

REGULATION OF BACTERIAL RESPONSES TO OXIDATIVE STRESS

171

mutants also show significantly reduced virulence in mice. The oral lethal dose for the mutant strain is 1000-fold greater than for the wildtype parent. The role of RpoS in virulence is probably complex. Most likely, many of the chromosomally encoded RpoS-regulated genes which help the cells survive against general stress help the bacteria survive in the host environment. However, RpoS also modulates the expression of the spvRABCD genes carried on virulence plasmids (21,53). Studies of several different lacZ fusion constructs in combination with mutant backgrounds suggest that RpoS controls the level of SpvR, a LysR family-type transcriptional regulator, which, in turn, activates expression of the spvABCD genes (40). B. Haemophilus influenzae A scan of the completed sequence of the entire genome oiH. influenzae suggests that H. influenzae encodes homologs oikatE (denoted hktE) and sodA. (25). Surprisingly, the M^£^-encoded KatE homolog has been shown to be regulated like E, coli katG. The HktE catalase levels are higher in exponential cells than in stationary phase cells, and the hktE message and protein are induced by treatment with hydrogen peroxide or ascorbic acid, which generates hydrogen peroxide in the presence of oxygen (8). The regulator of hktE induction has not been reported, and SoxS, SoxR, MarA, Rob, and RpoS homologs are not apparent from the genomic sequence. However, homology searches do indicate the presence of a gene encoding a protein that is more than 70% identical to E, coli OxyR. Interestingly, this gene was identified as tbpR in a screen for multicopy clones that lead to expression of a transferrin binding activity in j5. coli (48a). The reason for the effect on transferrin binding activity is not known, but H, influenzae strains carrying mutations in tbpRI oxyR do show increased sensitivity to hydrogen peroxide. Therefore, TbpR/OxyR appears to play a role in regulating the response of H. influenzae to oxidative stress, and it is interesting that, as noted by Bishai et al., the hktE promoter has some similarity to the OxyR consensus sequence (8). C. Mycobacterium It has been long been known that Mycobacterium strains have hydroperoxidase I that is similar to the E. coli katG-encoded peroxide. This activity has been the focus of several studies since strains carrying mutations that inactivate katG are resistant to isonicotinic acid hydrazide (isoniazid, INH), one of the main antimycobacterial drugs (for a review, see ref. 66). For wild-type M. smegmatis, the minimum inhibi-

172

JUDAH L. ROSNER AND GISELA STORZ

tory concentration of INH is 32 ^cg/ml compared to 512 /xg/ml for katG~ mutant strains. Escherichia coli cells are constitutively more resistant to INH, and katG mutants do not show increased resistance. Therefore, an observation that S. coli oxyR~ mutants or katG~ahpCF~ double mutants are more sensitive to INH seemed almost contradictory. Further studies, however, showed that the oxyR~ mutants as well as the katG~ ahpCF' double mutants had increased endogenous hydrogen peroxide levels, and hydrogen peroxide alone could potentiate the effects of INH (57). These and other results are consistent with the view that INH is a prodrug that can be activated by peroxides or by mycobacterial hydroperoxidase I and subsequently inactivate other targets within the cell. In Mycobacterium, the primary target of the activated INH appears to be mycolic acid biosynthesis (4), while in JB. coli, DNAis likely to be a target since INH treatment of susceptible cells results in filamentation and mutagenesis (J. L. Rosner, R. G. Martin, and P. M. R. Achary, unpublished data). The different effects of the E. coli and Mycobacterium katG~ mutations can be explained by the finding that M. tuberculosis hydroperoxidase I is more effective at INH-dependent generation of radicals than is E. coli hydroperoxidase I (33). The regulation of the &a^G-encoded hydroperoxidase and other antioxidant enzymes in Mycobacterium is not well understood. However, two studies have provided some interesting insights (18,59). Sequence homology searches showed that M. avium andM. leprae encode proteins with strong similarity to the AhpC protein of S. typhimurium and E. coli (11). A gene encoded directly upstream of both the M. avium and M. leprae ahpC-Uke genes showed homology to the LysR family of transcriptional regulators (18,59). The close proximity to ahpC led to the suggestion that this LysR family member encodes the Mycobacterium OxyR, but this remains to be proven rigorously. The putative homologs show only 33% identity to E. coli OxyR, in contrast to the 70% identity seen fori?, influenzae; however, the amino acids surrounding the critical cysteine in E. coli OxyR are conserved. Intriguingly, whereas the putative oxyR genes from M. avium and M. leprae appear to be intact, the genes from M. tuberculosis and other members of the M. tuberculosis complex, such as M. bovis BCG, M. africanum, and M. microti, contain numerous deletions and frameshifts and are probably nonfunctional. Studies of the oxidative stress responses of various myocobacteria show that the different strains have significantly different responses to hydrogen peroxide (59). Only the saphrophytic strain M. smegmatis showed adaptation analogous to the OxyR-regulated response. Mycobacterium smegmatis bacilli pretreated with 50 iiM hydrogen peroxide

REGULATION OF BACTERIAL RESPONSES TO OXIDATIVE STRESS

173

became resistant to 5 mM hydrogen peroxide and showed the induction of several proteins, as seen on two-dimensional gels. Mycobacterium avium and M. bovis BCG strains were constitutively more resistant to 10 mM hydrogen peroxide but did not show any adaptation. In M. avium, only three proteins were induced by hydrogen peroxide, and only the expression of the Aai^G-encoded hydroperoxidase I expression was induced in M. bovis. These findings have led to the hypothesis that pathogenic mycobacteria may continuously encounter reactive oxygen species in their host environments and therefore constitutively express antioxidant defense activities, eliminating the need for a functional OxyR protein. D. Bacillus subtllis

Several proteins that are counterparts to E. coli antioxidant defense activities have been identified in JB. subtilis. These include a vegetative catalase (encoded by katA), a catalase present in spores (encoded by katE), and proteins with similarity to the two alkyl hydroperoxide reductase subunits, as well as a 16-kDa protein (encoded by mrgA) that is similar to Dps (12,20). The presence of an adaptive response to hydrogen peroxide has also been known in B. subtilis for many years. Pretreatment of exponentially growing B. subtilis cells with 50 /xM hydrogen peroxide results in protection against killing by 10 mM hydrogen peroxide and leads to the induction of eight proteins, as detected on one-dimensional gels (51). The regulators of this response are now being identified by mutational studies. One mutant was isolated by screening for resistance to hydrogen peroxide (29). While the wild-type parent lysed in 100 mM hydrogen peroxide, the mutant strain grew with a doubling time of 85 min in minimal medium containing 150 mM hydrogen peroxide. The mutant was also more resistant to organic peroxides and synthesized a number of proteins at a much higher rate than the wild type. The constitutively expressed proteins included the KatA catalase, the two subunits of alkyl hydroperoxide reductase, MrgA/Dps, flagellin, and all the proteins which were induced by hydrogen peroxide in the wildtype strain (29). In a complementary screen, Chen et al. isolated mutants with increased mrgA expression in the presence of Mn(II) (13). Trans-acting mutants identified in this screen were also resistant to hydrogen peroxide and expressed increased levels of KatA, AhpCF, and MrgA/Dps, indicating that the mutations isolated in both screens affected the same locus. The outcome of these screens also suggested that the Mn(II)-dependent repression and hydrogen peroxide-dependent induction of these genes are mediated through the same pathway. Chen et

174

JUDAH L. ROSNER AND GISELA STORZ

al. noticed that the katA and mrgA promoters both have inverted repeats of the sequence TTAtAAt. Because point mutations in this region of the mrgA promoter lead to derepression of mrgA, the redox-sensitive regulator of katA, ahpCF, and mrgA is likely to be a repressor. Therefore, in contrast to OxyR inE. coli, the presence of hydrogen peroxide inB. subtilis may be sensed by a peroxide-sensitive repressor. Like E. coli, stationary phase and starved B. subtilis cells are much more resistant to hydrogen peroxide than exponentially growing cells, and some of the proteins induced by protective concentrations of hydrogen peroxide are, also induced on entry into the late log phase (19). Because the starvation and sporulation responses are well characterized in JB. subtilis, the response to hydrogen peroxide was examined in sporulation (spoO) mutants. Five spoO mutants (spoOB, E, F, H, J) were indistinguishable from wild-type cells; however, strains with spoOA mutations showed altered resistance to hydrogen peroxide. The stationary phase induction of a katA-lacZ fusion was also shown to be dependent on spoOA (10). The SpoOA DNA binding protein controls the expression of the negative regulator encoded by abrB, and the SpoOA mutant phenotypes can be suppressed by mutations affecting this downstream regulator. These studies suggest that the expression of the B. subtilis katA gene may be similar to the expression of the E. coli katG gene, with induction in both exponential phase and stationary phase, but the corresponding regulators are likely to be different. Studies have also shown that B. subtilis cells have a second catalase with homology to the E. coli katE-encoded hydroperoxidase. The expression of this katE gene is regulated by the sJ^J5-encoded a^, which shows similarity to the rpoS-encoded a^ (20).

IV. Concluding Remarks Although much remains to be learned about the bacterial responses to oxidative stress, some general themes have emerged. From the studies ofE. coli, it appears that bacterial cells perceive superoxide differently from hydrogen peroxide. However, while the responses to hydrogen peroxide and superoxide are distinct, the two regulons may both overlap with the general stress response that is induced by starvation or entry into stationary phase. It is also noteworthy that mutations affecting antioxidant defense genes or their corresponding regulators alter antibiotic resistance in both E. coli and Mycobacterium. In addition, a comparison ofE. coli and B. subtilis suggests that the oxidative stress regulators may not necessarily be conserved, and bacteria may use different mechanisms for sensing the same oxidant.

REGULATION OF BACTERIAL RESPONSES TO OXIDATIVE STRESS

175

Several interesting questions for future studies are raised by our current understanding of the bacterial responses to oxidative stress: What are the chemical reactions that lead to SoxR and OxyR activation? Why do diverse treatments such as salicylate and superoxide trigger the same set of diverse responses, superoxide resistance and multiple antibiotic resistance? What are the roles of the still unidentified proteins that are induced by the different oxidative stress conditions, and how do they protect the cell? The answers to these questions should help elucidate the mechanisms that are used to sense and defend against oxidative stress in all organisms. ACKNOWLEDGMENTS We t h a n k B. Demple, E. Hansen, and P. Miller for discussing results prior to publication.

REFERENCES 1. Altuvia, S., Almiron, M., Huisman, G., Kolter, R., and Storz G. (1994). Mol. Microbiol. 13, 265-272. 2. Ariza, R. R., Cohen, S. P., Bachhawat, N., Levy, S. B., and Demple, B. (1994). J. Bacteriol. 176, 143-148. 3. Ariza, R. R., Li, Z., Ringstad, N., and Demple, B. (1995). J. Bacteriol 177,1655-1661. 4. Banerjee, A., Dubnau, E., Quemard, A., Balasubramanian, V., Um, K. S., Wilson, T., Collins, D., de Lisle, F., and Jacobs, W. R. Jr. Science 263, 227-230. 5. Barth, M., Marschall, C , Muffler, A., Fischer, D., and Hengge-Aronis, R. (1995). J. Bacteriol. 177, 3455-3464. 6. Becker-Hapak, M., and Eisenstark, A. (1995). FEMS Microbiol. Lett. 134, 39-44. 7. Benov, L. T., and Fridovich, I. (1994). J. Biol. Chem. 269, 25310-25314. 8. Bishai, W. R., Smith, H. 0., and Barcak, G. J. (1994). J. Bacteriol. 176, 2914-2921. 9. Bohringer, J., Fischer, D., Mosler, G., and Hengge-Aronis, R. (1995). J. Bacteriol. 177, 413-422. 10. Bol, D. K., and Yashin, R. E. (1994). J. Bacteriol. 176, 6744-6748. 11. Chae, H. Z., Robison, K , Poole, L. B., Church, G., Storz, G., and Rhee, S. G. (1994). Proc. Natl. Acad. Sci. U.S.A. 9 1 , 7017-7021. 12. Chen, L., and Helmann, J. D. (1995). Mol. Microbiol. 18, 295-300. 13. Chen, L., Keramati, L., and Helmann, J. D. (1995). Proc. Natl. Acad. Sci. U.S.A. 92, 8190-8194. 14. Chou, J. H., Greenberg, J. T., and Demple, B. (1993). J. Bacteriol. 175, 1026-1031. 15. Cohen, S. P., Levy, S. B., Foulds, J., and Rosner, J. L. (1993). J. Bacteriol. 175, 78567862. 16. Compan, I., and Touati, D. (1993). J. Bacteriol. 175, 1687-1696. 17. Demple, B. (1991). Annu. Rev. Genet. 25, 315-337. 18. Deretic, V., Philipp, W., Dhandayuthapani, S., Mudd, M. H., Curcic, R., Garbe, T., Heym, B., Via, L. E., and Cole, S. T. (1995). Mol. Microbiol. 17, 889-900. 19. Dowds, B. C. A., Murphy, P., McConnell, D. J., and Devine, K. M. (1987). J. Bacteriol. 169, 5771-5775. 20. Engelmann, S., Lindner, C , and Hecker, M. (1995). J. Bacteriol. Ill, 5598-5605. 21. Fang, F. C , Libby, S. J., Buchmeier, N. A., Loewen, P. C , Switala, J., Harwood, J., and Guiney, D. G. (1992). Proc. Natl. Acad. Sci. U.S.A. 89, 11978-11982.

176

JUDAH L. ROSNER AND GISELA STORZ

22. Farr, S. B., and Kogoma, T. (1991). Microbiol Rev. 55, 561-585. 23. Fawcett, W. P., and Wolf, R. E. Jr. (1994). Mol Microbiol. 14, 669-679. 24. Fields, P. L, Swanson, R. V., Haidaris, C. G., and Heffron, F. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 5189-5193. 25. Fleischmann, R. D., Adams, M. D., White, O., Clayton, R. A., Kirkness, E. F., Kerlavage, A. R., Bult, C. J., Tomb, J.-F., Dougherty, B. A., Merrick, J. M., McKenney, K , Sutton, G., FitzHugh, W., Fields, C., Gocayne, J. D., Scott, J., Shirley, R., Liu, L.-L, Glodek, A., Kelley, J. M., Weidman, J. F., Phillips, C. A., Spriggs, T., Hedblom, E., Cotton, M. D., Utterback, T. R., Hanna, M. C , Nguyen, D. T., Saudek, D. M., Brandon, R. C , Fine, L. D., Fritchman, J. L., Fuhrmann, J. L., Geoghagen, N. S. M., Gnehm, C. L., McDonald, L. A., Small, K. V., Eraser, C. M., Smith, H. O., and Venter, J. C. (1995). Science 269, 496-512. 26. Gambino, L., Gracheck, S. J., and Miller, P. F. (1993). J. Bacteriol. 175, 2888-2894. 27. Gonzalez-Flecha, B., and Demple, B. (1995). J. Biol. Chem. 270, 13681-13687. 28. Gruer, M. J., and Guest, J. R. (1994). Microbiology 140, 2531-2541. 29. Hartford, O. M., and Dowds, B. C. A. (1994). Microbiology 140, 297-304. 30. Henderson, I. R., Meehan, M., and Owen, P. (1997). In "Mechanisms in the Pathogenesis of Enteric Diseases" (P. S. Paul, D. H. Francis, and D. A. Benfield, eds.) Plenum, New York. 31. Hidalgo, E., Bollinger, J. M., Jr., Bradley, T. M., Walsh, C. T., and Demple, B. (1995). J. Biol. Chem. 270, 20908-20914. 32. Hidalgo, E., and Demple, B. (1994). EMBO J. 13, 138-146. 33. Hillar, A., and Loewen, P. C. (1995). Arch. Biochem. Biophys. 323, 438-446. 34. Huisman, G. W., and Kolter, R. (1994). Science 265, 537-539. 35. Imlay, J. A., and Fridovich, I. (1991). J. Biol. Chem. 266, 6957-6965. 36. Jair, K.-W., Fawcett, W. P., Fujita, N., Ishihama, A., and Wolf, R. E. Jr. (1996). Mol. Microbiol. 19, 307-317. 37. Jair, K.-W., Martin, R. G., Rosner, J. L., Fujita, N., Ishihama, A., and Wolf, R. E. Jr. (1995). J. Bacteriol. Ill, 7100-7104. 38. Jair, K.-W., Yu, X., Skarstad, K., Thony, B., Fujita, N., Ishihama, A., and Wolf, R. E. Jr. (1996). J. Bacteriol. 178, 2507-2513. 39. Koh, Y.-S., and Roe, J.-H. (1995). J. Bacteriol. Ill, 2673-2678. 40. Kowarz, L., Coynault, C , Robbe-Saule, V., and Norel, F. (1994). J. Bacteriol. 176, 6852-6860. 41. KuUik, I., Stevens, J., Toledano, M. B., and Storz, G. (1995). J. Bacteriol. Ill, 1 2 8 5 1291. 42. KuUik, I., Toledano, M. B., Tartaglia, L. A., and Storz, G. (1995). J. Bacteriol. Ill, 1275-1284. 43. Lange, R., and Hengge-Aronis, R. (1994). Genes Dev. 8, 1600-1612. 44. Li, Z., and Demple, B. (1994). J. Biol. Chem. 269, 18371-18377. 45. Liochev, S. I., and Fridovich, I. (1992). Proc. Natl. Acad. Sci. U.S.A. 89, 5892-5896. 46. Liochev, S. I., Hausladen, A., Beyer, W. F„ Jr., and Fridovich, I. (1994). Proc. Natl. Acad. Sci. U.S.A. 9 1 , 1328-1331. 47. Loewen, P. C , and Hengge-Aronis, R. (1994). Annu. Rev. Microbiol. 48, 53-80. 48. Ma, D., Alberti, M., Lynch, C , Nikaido, H., and Hearst, J. E. (1996). Mol. Microbiol. 19,101-112. 48a. Maciver, I., and Hansen, E. J. (1996). Infect. Immun. 64, 4618-4629. 49. Martin, R. G., and Rosner, J. L. (1995). Proc. Natl. Acad. Sci. U.S.A. 92, 5456-5460. 50. Muffler, A., Fischer, D., Altuvia, S., Storz, G., and Hengge-Aronis, R. (1996). EMBO J. 15, 1333-1339.

REGULATION OF BACTERIAL RESPONSES TO OXIDATIVE STRESS

177

51. Murphy, P., Dowds, B. C. A., McConnell, D. J., and Devine, K. M. (1987). J. Bacteriol. 169, 5766-5770. 52. Nakajima, H., Kobayashi, K., Kobayashi, M., Asako, H., and Aono, R. (1995). Appl. Environ. Microbiol. 61, 2302-2307. 53. Norel, F., Robbe-Saule, V., Popoff, M. Y., and Coynault, C. (1992). FEMS Microbiol. Lett. 99, 271-276. 54. Nunoshiba, T., Hidalgo, E., Li, Z., and Demple, B. (1993). J. Bacteriol. 175, 74927494. 55. Pratt, L. A., and Silhavy, T. J. (1996). Proc. Natl. Acad. Sci. U.S.A. 93, 2488-2492. 56. Rosner, J. L., and Slonczewski, J. L. (1994). J. Bacteriol. 176, 6262-6269. 57. Rosner, J. L., and Storz, G. (1994). Antimicrob. Agents Chemother. 38, 1829-1833. 58. Schweder, T., Lee, K.-H., Lomovskaya, O., and Matin, A. (1996). J. Bacteriol. 178, 470-476. 59. Sherman, D. R., Sabo, P. J., Hickey, M. J., Arain, T. M., Mahairas, G. G., Yuan, Y., Barry, C. E., Ill, and Stover, C. K. (1995). Proc. Natl. Acad. Sci. U.S.A. 92,6625-6629. 60. Skarstad, K, Thony, B., Hwang, D. S., and Romberg, A. (1993). J. Biol. Chem. 268,5365-5370. 60a. Sulavik, M. C., Dazer, M., and Miller, P. F. (1997). J. Bacteriol. 179 (in press). 61. Tao, K, Fujita, N., and Ishihama, A. (1993). Mol. Microbiol. 7, 859-864. 62. Tao, K., Zou, C., Fujita, N., and Ishihama, A. (1995). J. Bacteriol. 177, 6740-6744. 63. Toledano, M. B., KuUik, I., Trinh, F., Baird, P. T., Schneider, T. D., and Storz, G. (1994). Cell (Cambridge, Mass.) 78, 897-909. 64. Wu, J., Dunham, W. R., and Weiss, B. (1995). J. Biol. Chem. 270, 10323-10327. 65. Yamashino, T., Ueguchi, C., and Mizuno, T. (1995). EMBO J. 14, 594-602. 66. Zhang, Y., and Young, D. B. (1993). Trends Microbiol. 1, 109-113.

This Page Intentionally Left Blank

CURRENT TOPICS IN CELLULAR REGULATION, VOLUME 35

Mechanism and Regulation of Bone Resorption by Osteoclasts NOBUHIKO K A T U N U M A

Institute for Health Sciences Tokushima Bunri University Tokushima 770, Japan

I. Historical Background and Future Prospects With the aging of society, osteoporosis is rapidly becoming a serious problem. The osteoclasts are known to be mainly responsible for degradation of bone type-1 collagen during the process of bone resorption. Lysosomal proteinases of osteoclasts, possibly certain cysteine proteinases, may play the main role in osseous coUagenolysis. Compared with cathepsin B, H, and N, the coUagenolytic activity of cathepsin L and L family is particularly strong in vitro. Delaissi Vaes et al. (1,2) showed that there is no correlation between the resorption of cultured mouse bone cells and their secretion of coUagenase. In addition, in experimental bone resorption, pit formation was not inhibited by Cl-1, a specific inhibitor for coUagenase, while a good correlation was obtained between resorption and excretion of lysosomal cysteine proteinase induced by parathyroid hormone (PTH). Furthermore, the authors demonstrated that the group-specific inhibitors of cysteine proteinases, such as leupeptin, antipain, E-64 [^ra/zs-epoxysuccinylL-leucylamido(4-guanidino)butane], and tosyllysyl chloromethane, markedly inhibited the bone resorption induced by PTH, as shown in Fig. 1. The structures and properties of several intralysosomal proteinases have been identified, including those of the cysteine proteinases, such as cathepsins B, H, L, J, S, and N. Moreover, their physiological roles in intracellular protein degradation processes have received attention. The proteinases responsible for particular protein degradation processes appear to differ with the physiological function. However, the particular lysosomal cysteine proteinase responsible for bone collagen degradation has not yet been determined. The development of specific inhibitors of these cysteine proteinases would facilitate identification of their physiological functions. Katunuma et al. (3,4) developed specific 179

Copyright © 1997 by Academic Press. All rights of reproduction in any form reserved. 0070-2137/97 $25.00

180

NOBUHIKO KATUNUMA pit number formed/cell 0 1

0.2

0.4

1

1

0.6

\

control Z-Phe-Ala-CHNH2

1 ",'

1 '^' '^' '',

E-64

1 control \

''7 ; /

/. ' ' '

a-1

FIG. 1. Effect of coUagenase inhibitor and cysteine proteinase inhibitors on pit formation by bone resorption. From Dellaisse et al. (2).

inhibitors for cathepsin B, CA-074 and CA-030, and also for cathepsin L, CLIK-37 and CLIK~ 52, which are epoxysuccinate derivatives. The suppression of pit formation was tested using a specific inhibitor for cathepsin L, such as pig leukocyte cysteine proteinase inhibitor (PLCPI) (5), chymostatin and the group-specific inhibitors of cysteine proteinases, E-64 and leupeptin (6,7). The molecular mechanisms explaining the specific inhibition of CA-030 were clarified by analysis of the tertiary structure of cocrystals of cathepsin B and CA-030, as determined by X-ray crystallography (8). It is also important to know the mechanisms and regulation of cathepsin L secretion from osteoclasts to explain the bone resorption mechanisms. We investigated the relationship between pit formation and the increase in cathepsin secretion into the bone cell culture medium induced by PTH, tunlor necrosing factor a (TNF-ce), and la,25-(OH)2D3 [la,25-(OH)2 vitamin Da], and the suppression of the release by calcitonin and monensin (9). Cathepsin L was secreted in the precursor form into lacunae and then processed to a mature active form in the lacunae (10). The mechanisms arid regulation of bone resorption by secreted cathepsins from osteoclasts kt the molecular level are reviewed here. The schematic illustration of the total process of bone resorption is shown in Fig. 2.

II. Suppression of Bone Resorption by Cathepsin L Family Inhibitors Lysosomal cysteine proteinases have been shown to degrade bone collagen type 1 efficiently in acidified lacunae. Cathepsin L family

181

BONE RESORPTION BY OSTEOCLASTS

Osteoclast

PTH

Osteoblast Howshlp's Lacuna

procathepsin L

H demineralizatlon

\

pH 45-4.8

I

cathepsin L

collagenolysi

FIG. 2. Schematic illustration of the mechanisms of bone resorption.

proteases has particularly strong coUagenolytic activity among the other types of lysosomal cathepsins. The pit formation test is used as the quantitative experiment for bone resorption and is performed as follows: Ivory is sliced and punched out to make thin, round pellets. The pellets are incubated with an osteoblast and osteoclast mixture, and the pits formed on the bone surface are stained and counted by microscopy. Vaes and Delaisse (1,2) have reported that there is no correlation between the resorption of cultured bone and the secretion of coUagenase. Furthermore, pit formation was not suppressed by the coUagenase inhibitor Cl-1. However, a good correlation existed between bone resorption and the secretion of cathepsin B induced by PTH. Furthermore, these authors demonstrated that several inhibitors of cysteine proteinases, leupeptin, antipain, and E-64, markedly inhibited the resorption induced by PTH in cultured bone. The effects of cysteine proteinase inhibitors on pit formation induced by PTH were systematically investigated by Katunuma's group (6,7,9), and the suppression was compared with the inhibition profiles of cathepsin L and B activities, as Fig. 3 shows. E-64, leupeptin, and cystatin A, which strongly inhibit both cathepsins B and L, markedly suppressed

I

I

I

I

I

I

I

I

oo

c

<

t^

tp

to

CO

O Q.

E5:

t^

1©

I I I I I I I o o o o o o o X X X X X X X X X X X X X X X

6

o o o o o o o o

?

•^.s

g^-2

BONE RESORPTION BY OSTEOCLASTS

183

pit formation. Because cystatin A does not inhibit calpain, calpain does not participate in bone resorption. CA-074, a specific inhibitor of cathepsin B, did not inhibit pit formation, suggesting that cathepsin B does not participate in osteoclastic bone resorption. Also, aprotinin and soybean trypsin inhibitor, specific inhibitors of serine proteinases, did not suppress pit formation. It follows that trypsin-type serine proteinases are not involved in bone resorption. Chymostatin, which inhibits cathepsin L but not cathepsin B, strongly inhibited pit formation; moreover, PLCPI isolated from pig leukocytes, a specific inhibitor of cathepsin L, suppressed pit formation. These inhibitor studies revealed that cathepsin L or cathepsin L-like proteinase is the main proteinase responsible for bone resorption. Endogenous cathepsin L inhibitors, such as cystatins and PLCPI, may participate in the natural regulation of bone resorption. By means of polyacrylamide gel electrophoresis and Western blotting techniques by using anticathepsin L antibody, it was established that osteoclast secretions contain two proteins with molecular weights corresponding to procathepsin L and cathepsin L; both proteins reacted with anticathepsin L antibody.

III. Functional Differentiation of Osteoclasts and Macrophages with Respect to Cathepsins L and B E. Kominami et al. (unpublished data) demonstrated immunohistochemically the presence of cathepsins B and L in rat osteoclasts of the bone surface. The concentrations of cathepsins B, L, and H in osteoclasts and macrophages were determined by our differential assay method of these cathepsins (11). Osteoclasts and macrophages are differentiated from the same origin, the monocyte, but the concentration of cathepsin L in osteoclasts is much higher than that in the other cells. The functional differentiation in the contents of cathepsins B and L in osteoclasts and macrophages (9) is compared in Table I. This also supports the TABLE I DIFFERENCES IN ACTIVITIES OF CATHEPSINS B AND L IN OSTEOCLASTS AND MACROPHAGES

Activity (mU/nig) Component

Osteoclasts

Macrophages

Z-Phe-Arg-MCA (total hydrolytic activities) Cathepsin B Cathepsin L

1.48 0.62 0.30

1.45 1,39 0.05

184

NOBUHIKO KATUNUMA

important role in coUagenolytic activity of cathepsin L family protease(s) on bone resorption.

IV. Mechanism and Regulation of Procathepsin L Secretion from Osteoclasts The degradation of bone matrix mainly consisting of type 1 collagen during the process of bone resorption occurs in the extracellular acidified lacunae under the ruffled border of the osteoclasts, as Fig. 4 shows. Addition of calcitonin or E-64 caused the release of osteoclast from the bone. Therefore, it is important to clarify the functions of the various cells responsible for the secretion of cysteine proteinases related to bone resorption, such as osteoclasts, osteoblasts, and osteocytes. The lysosomal cysteine proteinases of osteoclasts are synthesized in the endothelial reticulum, transported toward the ruffled border membrane via Golgi^ and secreted into the bone-resorbing lacunae mediated by various hormones and biofactors. We investigated the relationship between pit formation and secretion of cathepsins L and B induced by PTH into the culture medium of rat bone marrow cells and the suppression by calcitonin, (see Table II). The Western blotting of procathepsin L secretion regulated by PTH and calcitonin into culture medium is demonstrated in Fig. 5. Addition of PTH enhanced the excretion of the 39-kDa preform of cathepsin L, as indicated in lane C (Fig. 5), and further addition of calcitonin suppressed the secretion enhance by PTH, as shown in lane D (Fig. 5). To clarify the mechanism and regulation of secretion of 39-kDa procathepsin L and its processing into lacunae, we examined the stimulation of procathepsin L secretion induced by l«,25-(OH)2D3, TNF-ce, and PTH, as well as the effects of the cysteine proteinase inhibitors on pit formation induced by these effectors, as shown in Fig. 6 (10). It is known that la,25-(OH)2D3 stimulates the differentiation of osteoclasts and also that TNF-Qj causes malignant hypercalcemia. The stimulation effects of lce,25-(OH)2D3 and TNF-o; on the secretion of procathepsin L were also examined by Western blotting. These secretion stimulators of procathepsin L enhanced bone pit formation, which was inhibited by E-64 but not by CA-074. However, the mechanism of stimulation of procathepsin L by PTH, 1Q:,25-(OH)2D3, and TNF-ce has not been sufficiently clarified. The specific receptor of PTH is located on osteoblasts but not on osteoclasts. Therefore, some unknown second messenger should be transferred from osteoblasts to osteoclasts; interleukin-6 (IL6) has been reported to play such a role.

BONE RESORPTION BY OSTEOCLASTS

185

Control

Calcitonin

E-64

FIG. 4. Release of osteoclast (DC) from bone surface due to the addition of calcitonin or E-64. (From Professor Osawa, with kind permission.)

186

NOBUHIKO KATUNUMA TABLE II

CORRELATIONSHIP BETWEEN Z - P H E - A R G - M C A HYDROLYTIC ACTIVITIES SECRETED FROM RAT BONE CELLS

Cysteine proteinase activities in culture medium (mU/ml) Inhibitor

Minus PTH

Z-Phe-Arg-MCA (total hydrolytic activities) Cathepsin B Cathepsin L

6.85 ± 1.33**

Plus PTH

Plus PTH and calcitonin

14.00 ± 2.16

9.15 ± 1.01**

4.29 ± 0.94** 9.68 ± 1.69 6.12 ± 1.03** 1.43 ± 0.16 2.50 ± 0.44 1.55 ± 0.17** Bone resorption (number of pits^one) 668.4 ± 94.3** 133.7 ± 41.1** 211.8 ± 73.5**

"^ Each value of Z-Phe-Arg-MCA hydrolytic activity indicates the mean ± SD of 5 observations. Each value of bone resorption indicates the mean ± SD of 20 observations. **p < .01; significant difference from PTH-plus group (Student's ^test). Note: Concentration of PTH and calcitonin were 1 X 10~^ and 1 X 10""^ M, respectively.

kOa 49.5

39.0 32.5

27.5

•25.0

A B C D FIG. 5. Effect of PTH and calcitonin on procathepsin L secretions in bone culture medium using Western blotting detecting systems. Lane A, purified mature cathepsin L; lane B, minus PTH; lane C, plus PTH; lane D, plus PTH and calcitonin.

187

BONE RESORPTION BY OSTEOCLASTS

-r

loo-

XM r n 1

X

II II

a >

se-

An

1

1 B ** 1 1 ^3 fS

LU

X

1 11

-7 -6 -7 -6 E-64a CA-074

VD3

T 1

II

^

1

B ^11 1

III i 1 ^^ 1 11 " i f--

-7 -6 -7 -6 E-64a CA-074

ITNF-a

fl

1

^H igi^ I j

O S

Jl

d1 [ 1

-7 -6 -7 -6 E-64a

CA-074

iPTHi

FIG. 6. Induction of pit formation by addition of PTH, la,25-(OH2D3, and TNF-o; and the suppression by E-64a or CA-074. All values are given as the percent of control with PTH alone. The numbers - 7 and - 6 represent an inhibitor concentration of lO""^ and 10"^. * p < .05, ** p < .01: significant differences from control group.

V. Cathepsin L Secreted from Osteoclasts as Precursor Form and Processed by Cysteine Proteinase(s) in Bone Lacunae The 39-kDa cysteine proteinase, which seems to be a precursor form, was secreted into culture medium of a bone cell mixture during the process of pit formation. Procathepsin L purified from rat long bones under cold alkaline conditions was rapidly converted to the mature form under acidic conditions at room temperature. However, this conversion was inhibited by the addition of E-64 during the process, suggesting that the procathepsin L secreted into lacunae (10) is autocatalytically converted to the mature form, as shown in the Western blot in Fig. 7. All procathepsins can be activated by the same processes.

188

NOBUHIKO KATUNUMA

kDa

49.5

•

-

;,-•

,

."

;xi

-

•

" I

I

;52. 5

27.5 pK

A

B

FIG. 7. Inhibition of processing of procathepsin L by addition of E-64 during purification and detection using Western blotting analysis by anti-cathepsin L antibody. Procathepsin L was processed in 0.1 M acetate buffer, pH 4.5, for 30 min in the presence of 10"^ M E-64 (lane A) and the absence (lane B) of E-64.

VI. Inhibitory Mechanisms of H+-ATPase, Carbonic Anhydrase II, and Monensin on Pit-Forming Assay Bone resorption depends on the secretion of proton and procathepsin L from osteoclasts into the extracellular lacunae. The proton is generated by carbonic anhydrase II in osteoclasts and activity in the lacunae by a proton pump driven by vacuolar-type H+-ATPase at the ruffled border, and the secreted proton participates in the bone resorption. However, the interrelated regulation between the proton and procathepsin L secretions from osteoclasts is still unclear. Addition of Bafilomycin Ai, an H+-ATPase inhibitor, completely suppressed pit formation stimulated by PTH but did not suppress procathepsin L secretion.

189

BONE RESORPTION BY OSTEOCLASTS Bone resorption "Sj

-^ CO •-« a.

Enzyme activity

1000 1800 h

o k. 0)

E

600 h

3 C

• ^ - ^

c o

400 [

Mtf

a. o e n Q ]•>

200 H

h.

a> c o

m

0 '•

PTH

monensin FIG. 8. Influences of monensin on bone resorption number induced by PTH and Z-PheArg-MCA hydrolytic activities of rat osteoclasts. Concentrations of PTH and monesin was 5 X 10"^ M. Each value of pit formation indicates the mean ± SD offiveobservations. **p < .01; significant difference from monensin minus group (Student's ^test).

Administration of acetazolamide, a carbonic anhydrase II inhibitor, suppressed both bone resorption and procathepsin secretion. These findings suggest that the secretion of procathepsin L and the activation (processing) of the secreted procathepsin L may be regulated by production and secretion of proton in the osteoclasts (12). The effects of monensin on bone resorption induced by PTH and ZPhe-Arg-MCA hydrolj^ic activity in isolated osteoclasts were assayed. Figure 8 shows the inhibition of pit formation and proteolytic activity in osteoclasts (10). Monensin itself does not inhibit cysteine proteinase activity, but it does inhibit the targeting of cysteine proteinases into lysosomes to prevent the excretion of cathepsins. These findings confirm that lysosomal cysteine proteinase in osteoclasts secreted into the lacunae plays an important role in the process of bone resorption.

VII. Suppression of Bone Resorption by Cathepsin L Family Inhibitors in Vivo The results of pit formation in vitro reflex directly affect the serum calcium level derived from bone resorption in vivo. We examined the in vivo effects of E-64, cystatin A, and CA-074 on the changes in serum

190

NOBUHIKO KATUNUMA

CA-074

E-64-a 11.0

10.0

9.0

8.0 4

7.0

T 10

Time (hr) FIG. 9. Effect of E-64 or CA-074 on serum calcium levels of rats on low-calcium diets. (•) No inhibitor; (A) 3 mg/100 g body weight of inhibitor; (•) 6 mg/100 g body weight of inhibitor. Levels measured 0-10 hr after inhibitor administration. Each value indicates the mean ± SEM of five observations, p < .05; significant difference from control (Student's ^test).

calcium of rats on a low-calcium diet and also with experimental malignant hypercalcemia as shown in Figs. 9 and 10, respectively. At 3 and 6 hr after the injection of E-64 (doses of 3 and 6 rag/ 100 g body weight) into rats on a low-calcium diet, serum calcium was significantly decreased in a dose-dependent manner. Administration of cystatin A (dose of 8 mg/100 g body weight) also significantly decreased the calcium level in the serum. At 1 hr after injection of E-64 or cystatin A, all cysteine proteinases in femur bone were inhibited effectively. By contrast, CA-074 did not affect the serum calcium level, although cathepsin B activity in the bone was specifically inhibited. These findings reconfirmed in vivo that cathepsin L rather than cathepsin B plays a central role in bone collagen degradation.

Vlli. Possible Strategies for New Drug Design to Protect Bone Resorption I propose the following two kinds of strategy to protect bone resorption. For suppression of cathepsin L secretion from osteoclasts: (1) inhibit the PTH effect mediated by the receptor located on the membrane of osteoblasts, (2) inhibit the second messenger information

191

BONE RESORPTION BY OSTEOCLASTS

>5-__

14,0h

J. ^ ^ - ^

ConUoi

13.0 10 mg/kg/day

12.0 11.0

25 mg/kg/day

CO 10.0!

o

O)

E

9.0| 8.0 80 mg/kg/day 7.0 I

I

Days FIG. 10. Inhibition of serum calcium levels by E-64 in rats with experimental malignant hypercalcemia. Oral administration of E-64 (dose in mg/kg). Levels measured 0 - 5 days after E-64a administration.

from osteoblast to osteoclast, (3) suppress procathepsin L secretion by administering a monensin-like drug, and (4) suppress proton formation and secretion by inhibition of the ATP-proton pump and of carbonic anhydrase. For inhibition of cathepsin L activity: (1) inhibit cathepsin L activity by a specific inhibitor and (2) inhibit conversion from precursor to the mature form. ACKNOWLEDGMENT I sincerely t h a n k Miss Yoshiko Nagai for preparation of this manuscript.

REFERENCES 1. Delaisse, J. M., Boyde, A., Maconnachie, E., Ali, N. N., Sear, C. H. J., Eedkhout, Y., Vaes, G., and Jones, S. J. (1987). Bone 8, 305.

192

NOBUHIKO KATUNUMA

2. Delaisse, J. M., and Vaes, G. (1992). In "Biology and Physiology of the Osteoclast" (B. R. Rifkin and C. V. Gay, eds.), p. 290. CRC Press, Boca Raton, FL. 5. Towatri, T., Hanada, K., Katunuma, N., et al. (1991). FEES Lett. 280, 311. 4. Katunuma, N., and Kominami, E. (1995). In "Methods in Enzymology" Vol. 251, Chaper 37, pp. 382-397. Academic Press, San Diego, CA. 5. Lenarcic, B., Ritonja, A., Dotenc, J., Stoka, V., Berbic, S., Pungercar, J., Stukelj, B., and Turk, V. (1993). FEES Lett. 336, 289. 6. Kakegawa H., Nikawa, T., Tagami, K, Kamioka, T., Drobnic-Kosorok, M., Lenarcic, B., Turk, v., and Katunuma, N. (1993). FEES Lett. 321, 247. 7. Katunuma, N., Kakegawa, H., Matsunaga, Y., Nikawa, T., and Kominami, E. (1993). Agents Actions, Suppl. 42, 195. 8. Turk, D., Podobnik, M., Popovic, T., Katunuma, N., Bode, W., Huber, R., and Turk, V. (1995). Biochemistry 34(14), 4791-4797. 9. Tagami, K, Kakegawa, H., Kamioka, H., Sumitani, K., Kawata, T., Lenarcic, B., Turk, v., and Katunuma, N. (1994). FEES Lett. 342, 308. 10. Kakegawa, H., Tagami, K, Ohba, Y., Sumitani, K., Kawata, T., and Katunuma, N. (1995). FEES Lett. 370, 78-82. 11. Inubushi, T., Kakegawa, H., Kishino, Y., and Katunuma, N. (1994). J. Eiochem. (Tokyo) 166, 282-284. 12. Ohba, Y., Yamamoto, T., and Katunuma, N. (1996). FEES Lett. 387, 175-178.

Indx

Abdominal muscles, apoptosis in, 81 Acetazolamide, effect on bone resorption, 189 Acid phosphatases, catalj^ic mechanism of, 31 Aconitase, iron-sulfur cluster of, 14 Acquired immunodeficiency syndrome. See AIDS Actinomycin D, apoptosis induction by, 70 Adenosine triphosphate (ATP), biochemical importance of, 149 Aging apoptosis role in, 70, 77, 107-121 benefits of caloric restriction in, 114-115 elevated levels of reactive oxygen species in, 123, 143 immune function decline in, 112-114 osteoporosis as problem of, 179 Agrobacterium rhizogenes, rolB oncogene from, 24 AIDS apoptosis in, 107 Fas-mediated killing in, 89 N F - K B role in, 150, 153-156, 157 Alkaline phosphatases, catalytic mechanism of, 31 Alzheimer's disease (AD), apoptosis and, 107, 115, 117 Amyotrophic lateral sclerosis (ALS), apoptosis and, 107, 115, 117 Antibiotics, bacterial resistance to, 174, 175 Antioxidants, in organisms, 149 Antipain, as cysteine proteinase inhibitor, 179 AP-1, redox sensitivity of, 131-132, 141 APO-1. See Fas and Fas ligand Apopain. See Yama Apoptosis accelerants for, 72, 76 aging and, 70, 77, 107-121 antagonists of, 72, 75, 76, 80 caloric restriction and, 114-115 cell response in, 71-72

cell senescence and, 110-112 description of, 69 Fas-mediated, 24, 69-105, 110, 112-114 function of, 69-70, 107-108 genes involved in, 108-110 immunological aging and, 112-114 induction of, 7 0 - 7 1 neurodegenerative disease and, 115-116 nucleosomal ladder of, 69, 70, 71, 80, 81, 91 proteolysis in, 83 regulation of, 7 2 - 9 1 Apoptotic bodies, formation of, 69, 71 Aprotinin, as serine proteinase inhibitor, 183 Arachidonic acid, oxidative damage to, 123 Arginine oxidative damage to, 124 in PTPase signature motif, 22, 27, 28, 30,32-34,38,46 Articular chondrocytes, apoptosis in, 117 Ascorbate, as free radical scavenger, 123 Aspartic acid in ICE-cleaved substrates, 84 in PTPases, 37, 38, 39, 40, 46-47, 61 ATF/CREB family, DNA binding of, 141, 142 Autoimmune diseases apoptosis and, 107 defective signaling network in, 21 increase in aging, 113 PTPase role in, 21, 57 B Bacillus suhtilis, oxidative stress responses of, 164, 173-174 Bacteria antibiotic resistance of, 174, 175 oxidative stress responses of, 163-177 Bad, as accelerant for apoptosis, 76 BAG-1, as antagonist of apoptosis, 76 Bak, as accelerant for apoptosis, 76 Bax, as accelerant of apoptosis, 73-75 193

194 Bcl-2 as antagonist of apoptosis, 7 2 - 7 3 , 91, 123 Mcl-l interaction with, 76 Bcl-2 family control of expression of, 77-78 list of members of, 72 as regulators of apoptosis, 72-78, 92, 111 bcl-2 gene, role in apoptosis, 109, 111, 113, 116, 117 Bcl-x, as antagonist of apoptosis, 75 Benzylphosphonic acid, as PTPase inhibitor, 59 Bilirubin, as oxygen scavenger, 134-135 B lymphocytes, apoptosis in, 118 Bone resorption drug protection of, 190-191 mechanism of, 181 pit formation test for, 180, 181, 182, 183, 187 procathepsin L regulation of, 179-191 suppression by cathepsin L inhibitors, 180-183 Brain, apoptosis in, 116-117 Breast carcinoma, apoptosis in, 87 Burst kinetics, of phosphatase activity, 35, 36 BZLFl transcription factor, redox sensitivity of, 133

CA-074, effect on bone resorption, 189-190 Caenorhadbitis elegans, apoptosis studies on, 70, 76, 108, 109, 116 Calcitonin, suppression of cathepsin release by, 180 Caloric restriction (CR), in upregulation of apoptosis, 114-115, 118 Cancer apoptosis role in control of, 107, 117-118 defective signaling network in, 21 deregulated protein-tyrosine kinase role in, 23 elevated levels of reactive oxygen species in, 123 metastatic, NF-zcB role in, 150, 1516-157 PTPase role in, 21, 57

INDEX Carbonic anhydrase II, effect on bone resorption, 188 ^-Carotene, as free radical scavenger, 123 Catalase, antioxidant reactions catalyzed by, 123 Cataracts, IRE mutation role in, 13 Cathepsins, inhibitors of, 179, 180 Cathepsin L regulation of secretion of, 184-187 suppression of bone resorption by inhibitors of, 180-183, 189-190 c-cdc gene, role in apoptosis. 111 CD45 biological function of, 24 as receptor-like PTPase, 22, 32 Cdc25, as dual-specificity phosphatase, 24 CD95 protein. See Fas and Fas ligand Ced-3, role in apoptosis, 78-79 ced genes, role in apoptosis, 76, 108, 113 Cell death. See also Apoptosis necrosis as, 107 Cell growth and differentiation PTPase role in, 21, 62 redox sensitivity of, 138 Cell senescence, apoptosis and, 110-112 Cell signaling, antioxidant role in, 149 Ceramide, role in apoptosis, 84-86, 91, 92, 110 c-fos/c-jun gene, role in signal transduction, 131-132 c-fos gene expression of, 137 role in apoptosis. 111 Chemotherapeutic agents, apoptosis induction by, 70, 107 Chromosome 11, H chain gene on, 2 Chromosome 19, L chain gene on, 2 Chromosomes 14 and 18, bcl-2 gene and, 109 Cinnamic acid, structure of, 60 c-Jun transcription factor, activation of, 137 c-myc gene, role in apoptosis, 110, 111 Colonic epithelium, apoptosis in, 74 corkscrew gene, 24 Corticosteroids, apoptosis induction by, 70 Covalent phosphoenzyme intermediate, in PTPase kinetic pathway, 63

195

INDEX Cowpox virus, CrmA of, 82 CrmA, as ICE inhibitor, 82 Cubanes, iron-sulfur clusters as, 4, 14 Cu/Zn-superoxide dismutase, deleterious mutation affecting, 117 Cyclins, role in cell division, 92 Cycloheximide, apoptosis induction by, 70,73 Cystatin, as cathepsin inhibitor, 181, 189 Cysteine in degradation domain of iron regulatory protein 2, 19 in PTPase signature motif, 22, 2 7 - 2 8 , 30, 31, 38, 46 Cysteine proteinase inhibitors, bone-resorption inhibition by, 179 Cysteinylphosphate enz5mrie, in phosphatase pathway, 32 Cytokines apoptosis induction by, 70 N F - K B in regulation of, 150, 151, 157

Cjrtomegaloviruses, N F - K B regulation of, 150 C5^oplasm, apoptosis effects on, 69 D Death domain, of Fas, 87 Death proteases definition of, 82 role in apoptosis regulation, 78, 79, 8 0 - 8 1 , 88, 92 substrates for, 8 3 - 8 4 Death signaling cascade, propagation of, 81, 82, 86 Degenerative diseases, elevated levels of reactive oxygen species in, 123 Diabetes defective signaling network in, 21 PTPase role in, 21, 57 Diacylglycerol, as ceramide inhibitor, 8 4 - 8 5 , 86 Disease apoptosis in, 73, 87, 89 defective signaling network in, 21 elevated levels of reactive oxygen species in, 123 IRE mutation role in, 13 DNA apoptosis effects on, 69, 70, 110

oxidative damage to, 1, 115, 124, 125, 143, 163, 164 D128 mutant, acid-base catalysis by, 40 D129 mutant, acid-base catalysis by, 39 Double-displacement pathway, of phosphatase action, 30 Drosophila apoptosis in, 110 succinate hydrogenase of, 15, 16 Dual-specificity phosphatases, catalytic mechanism of, 2 8 - 2 9 E eALAS protein, IRE-IRP regulation of, 14, 16 E g r l transcription factor activation of, 137 redox sensitivity of, 129 E - P formation and E - P breakdown steps, in PTPase-catalyzed reactions, 42-46 Epstein-Barr virus protein, as apoptosis inhibitor, 77 Erythrocyte amino levulinic acid synthase, role in heme biosynthesis, 4 Escherichia coli aconitase of, 8, 9, 14 oxidative stress response in, 125, 163, 164-169, 172, 174 E-selectin, N F - K B regulation of, 150 E-64, as cysteine proteinase inhibitor, 179, 180, 181, 189 Eukaryotes, iron-metabolism regulation in, 1-19 Evolution, iron-sulfur cluster role in, 13-14 F FADD, reaction with Fas death domain, 88 Fas and Fas ligand apoptosis mediated by, 24, 69-105, 110, 112-114 expression of, 88-89 structure and activity of, 87-89 Fas-associated phosphatase (FAP-1), biological function of, 24 fas gene, overexpression of, 113 Fenton reaction, 1, 135

196 Ferritin expression of, 5 role in oxidant stress reversal, 135 structure of, 3 translational regulation of, 1-3, 4, 16 Ferritin repressor protein, 3. See Iron regulatory protein 1 (IRPl); Iron regulatory protein 2 (IRP2) Ferrous iron, in Fenton reaction, 1 Fibroblasts apoptosis in, 111-112, 117, 118 senescent, in aged persons, 114 FLICE, role in apoptosis, 82 Flucocorticoids, role in ced activation, 109-110 4-( Fluoromethyl)phenylphosphate (FMPP), as PTPase inhibitor, 6 0 - 6 1 fra genes, expression of, 137 Free radicals. See also Reactive oxygen species (ROS) neuronal death from, 108 G gadd45 gene, transcription of, 137 Gene transcription, PTPase role in, 21 Genome, oxidant radical effects on, 137 gld gene, role in apoptosis, 113 Gliomas, apoptosis in, 91 Glucocorticoid receptor (GR), redox sensitivity of, 127 Glutamate-ammonia ligase, degradation of, 10 Glutamic acid, in PTPases, 37, 38, 40, 51 Glutamine synthetase of E. coli, degradation of, 10 as target of reactive oxygen, 115 Glutaredoxin, in cell signaling, 149 Glutathione, as free radical scavenger, 123 Glutathione peroxidase, antioxidant reactions catalyzed by, 123 Glycerophosphate, as PTPase substrate, 56 Glycine, in degradation domain of iron regulatory protein 2, 19 Gold ion, NF-/cB blockage by, 154 G proteins, in signal transduction, 92 Granzymes, apoptosis induction by, 70, 81 Growth factors effect on apoptosis, 7 0 - 7 1 , 111, 117 role in ced activation, 109-110

INDEX Guanidium group, in arginine of PTPase signature motif, 3 3 - 3 4 H Haemophilus influenzae, oxidative stress responses of, 164, 171 H+-ATP, effect on bone resorption, 188 H chain, of ferritin, 2 Heat shock, oxidant stress and, 136 Heat shock factor (HSF), oxidation role in activation of, 131 Hematopoietic tissue, Bcl-x expression in, 75 Heme biosynthesis, eALAS protein role in, 14 Hinge-linker, in mitochondrial aconitase binding site, 7, 14 Hippocampus, apoptosis in, 116-117 Histidine, oxidative damage to, 124 HIV. See AIDS; Human immunodeficiency virus (HIV) HIV-1 Tat, upregulation of Fas by, 89 Homeoboxes, embryogenesis and, 92 HoxB5, redox sensitivity of, 132 H u m a n immunodeficiency virus (HIV) anti-HIV drug screening for, 158 NF-/cB role in, 136, 150, 154-156, 157 Hydrogen peroxide cell damage from, 115, 135, 137 ferrous iron reaction with, 1 Hydroperoxidases, of bacteria, 164, 171 Hydroperoxyl radical, as reactive oxygen species, 123, 163 8-Hydroxy-2'-deoxyguanosine, as marker for DNA oxidation, 124 Hydroxyl radical cell damage from, 115 formation in Fenton reaction, 1 as reactive oxygen species, 123, 163 Hyperferritinemia-cataract syndrome, IRE mutation role in, 13

ICAM-1, NF-/
197

INDEX ICH-IL, as apoptosis inhibitor, 80 Immune response apoptosis role in, 70, 112-114 decline in aging, 112-114 PTPase role in, 21 Inducible nitric oxide synthase (iNOS), N F - K B regulation of, 150

Inflammation elevated levels of reactive oxygen species in, 123 necrotic cell death from, 107 Integrins, apoptosis effects on, 69 Interleukins, N F - K B in regulation, 150, 151 Interleukin-1/3 converting enzyme. See ICE Interleukin-8 (IL-8), oxidant stress induction of, 136 Iron in Fenton reaction, 1 metabolic regulation in eukaryotes, 1-19 toxic by-products from, 1, 16 Iron regulatory factor (IRF). See Iron regulatory protein 1 (IRPl); Iron regulatory protein 2 (IRP2) Iron regulatory protein 1 (IRPl) aconitase comparison to, 14 binding site of, 12-13 cellular expression of, 11-12, 135 cloning of, 4, 9 evolution of, 13-14, 16 function of, 8.16 iron regulatory protein 2 comparison to, 9-10, 11, 12-13 iron-responsive elements as binding sites for, 3 iron-sulfur cluster of, 4, 5-6, 8-9, 16 regulation of ferritin and transferrin receptor by, 2 steric hindrance model for, 4 - 5 Iron regulatory protein 2 (IRP2) binding sites of, 3, 12-13 cellular expression of, 11-12, 135 cloning of, 9 degradation domain of, 10-11 iron regulatory protein 1 comparison to, 9-10, 11, 12-13 regulation and function of, 9-10

regulation of ferritin and transferrin receptor by, 2 Iron-responsive elements (IREs) as binding sites for iron regulatory proteins, 3-4, 8 in eALAS protein, 15 in ferritin, 2 - 3 in Krebs cycle enzymes, 15 mutations in, disease and, 13 Iron-responsive elements (IRE) binding protein (IRE-BP). See Iron regulatory protein (IRPl); Iron regulatory protein 1 (IRP2) Iron sequestration protein. See Ferritin Iron-sulfur clusters as early protein prosthetic groups, 13 function of, 16 in iron regulatory protein 1 and aconitase, 4, 5-6, 7^9 redox activity of, 126 Ischemia, apoptosis in, 70, 107 Ischemia-reperfusion injury, elevated levels of reactive oxygen species in, 123 K Kinase pathway, role in N F - K B regulation, 151-152 "Kiss of death," in apoptosis, 71 Krebs cycle enzymes iron regulation of, 15 iron-sulfur clusters in, 14

Lactacystin, as inhibitor of iron regulatory protein 2 degradation, 10 Lamin, apoptosis effects on, 69 L chain, of ferritin, 2 Ick gene, as oncogene, 23 Leukocyte antigen-related PTPase (LAR) catalysis by, 31, 32 inhibition of, 61 Leupeptin, as cysteine proteinase inhibitor, 179, 180, 181 Linoleic acid, oxidative damage to, 123 Lipids, oxidative damage to, 115, 123-124, 143 Low Mr protein-tyrosine phosphatases active site of, 39, 57 biological function of, 23, 2 4 - 2 5

198 enzymatic activity of, 35, 36, 47 PTPase signature motif in, 22, 28, 29, 32, 34, 46 Lung carcinoma, tumor suppressor gene for, 24 Lymphomas, apoptosis in, 73, 109 Lymphoproliferation, fas role in, 114 M Macrophages, functional differentiation of, 183-184 Malonylt3rrosine, structure of, 60 MarA, oxidative activation of, 167, 168-169, 170 Master genes, embryogenesis and, 92 Mcl-l, as antagonist of apoptosis, 76 mec-A gene, role in neuronal death, 116 Menstrual cycle, role in apoptosis regulation, 77 Metaphosphate, formation in phosphatase action, 4 0 - 4 1 MG132, as inhibitor of iron regulatory protein 2 degradation, 10 Mitochondrial aconitase active site of, 8 function of, 11 iron regulatory protein 1 similarity to, 4, 7, 9 model of, 7 Mitogenesis, PTPase role in, 21 MOLSCRIPT program, use for enz3mie structure studies, 52 Monensin, suppression of cathepsin release by, 180, 188, 191 MORT-1, reaction with Fas death domain, 88 motheaten phenotype, in mice, 24 Myb protein, redox sensitivity of, 133, 141 Mycobacterium spp., oxidative stress responses of, 164, 171-.173, 174 N Nerve growth factor, 86 neu gene, as oncogene, 23, 24 Neurodegenerative disease, apoptosis in, 107, 115-116 Neurons apoptosis of, 108

INDEX Bcl-x expression in, 75 PTPase role in development of, 21 NF-KB

activation pathways of, 137, 150-154 drugs based on blockage of, 157 redox sensitivity of, 130-131, 141 regulation of, 149-161 role in disease, 150, 153-157 nuc gene, role in apoptosis, 109, 110 Nuclear factor kappa B. See N F - K B Nucleic acids, oxidative damage to, 124 Nucleophilic displacement reactions, on phosphate monoesters, 41 Nucleosomal ladder, of apoptosis, 69, 70, 71, 80, 81, 91 O 1Q:,25-(OH)2D3, effect on cathepsin secretion, 180 Oncogenes deregulated protein-tyrosine kinases as, 2 3 - 2 4 products of, role in apoptosis, 118 Osteoclasts, procathepsin L from, 179-191 OTF-1 transcription factor, redox sensitivity of, 133 Oxidative stress bacterial responses to, 163-177 cell damage from, 115-116 Oxidizing agents, apoptosis induction by, 70 OxyR, oxidative activation of, 164-165, 170, 171, 175 oxyR gene, role in oxidative stress, 125

Parathyroid hormone (PTH), bone resorption induction by, 179, 180, 181, 190 Parkinson's disease, apoptosis and, 107, 115 Perforin, apoptosis induction by, 70 p53 protein gene expression for, 110, 138 redox sensitivity of, 129 Phagocytes, role in apoptosis, 69 Phenylarsine oxide, as PTPase inhibitor, 59

199

INDEX Phosphate monoester hydrolysis, mechanism of, 21, 30, 4 6 - 4 8 Phosphatidylserine, apoptosis effects on, 69 Phosphocysteine, as phosphatasecatalysis intermediate, 31, 32 Phosphohistidine, as phosphatase intermediate, 31 PhosphoHpases, in signal transduction, 92 Phosphonic acid derivatives, as PTPase inhibitors, 59-60 Phosphonofluoromethylphenylalanine, structure of, 60 Phosphonomethylphenylalanine, structure of, 60 Phosphoserine as phosphatase intermediate, 31 as phosphatase substrate, 22, 50 Phosphothreonine, as phosphatase substrate, 22, 50 Phosphotyrosine as phosphatase substrate, 22, 50, 51 structure of, 60 Pit formation test, for bone resorption, 180, 181, 182, 183, 187 P-loop, of PTPases, 25, 26, 28, 48 Porcine mitochondrial aconitase, iron-responsive elements in, 15 PrICE, effect on apoptosis, 80 Procathpsin L, regulation of bone resorption by, 179-191 Proline in degradation domain of iron regulatory protein 2, 19 oxidative damage to, 124 Prostate, apoptosis in old cells of, 70, 118 Proteases, role in apoptosis, 78-84 Proteasome iron regulatory protein 2 degradation by, 1 0 - 1 1 role in apoptosis, 81 Proteins, oxidative damage to, 115, 124, 143 Protein kinases role in apoptosis, 8 9 - 9 0 in signal transducing network, 21 Protein phosphatases, in signal transducing network, 21

Protein-tyrosine kinases (PTKs) biological function of, 21, 62 deregulated, as oncogenes, 2 3 - 2 4 dual-specificity type of, 2 4 - 2 5 signature motif of, 22, 23, 28, 29, 32-37 substrate specificity of, 48 Protein-tyrosine phosphatases (PTPases), 21-68 acid-base catalysis by, 37-40 active sites of, 22, 25, 32, 53-57 amino acid sequence specificity of, 4 9 - 5 3 , 63 biological functions of, 21, 2 3 - 2 5 catalytic mechanism of, 30-48 classification of, 2 2 - 2 3 crystal structures of, 5 0 - 5 1 disease and, 57 domains of, 22, 29, 63 inhibitors of, 5 7 - 6 1 intracellular type, 22, 23 low Mr type. See Low My. proteintyrosine phosphatases phosphate monoester hydrolysis by, 21, 30, 4 6 - 4 8 P-loop of, 25, 26, 28 receptor-like, 22, 23 signature motif of, 22, 23 structures of, 2 5 - 2 8 substrate specificity of, 48-62 transition state for hydrolysis by, 40-46 Protoantioncogenes, effect on Fas-mediated apoptosis, 90 Protooncogenes, effect on Fas-mediated apoptosis, 90 p35 protein, as ICE inhibitor, 82-83 PTPQ!

ectopic expression of, 24 inhibition of, 61 PTPy, tumor suppressor gene for, 24 PTPl acid-base catalysis by, 37, 42 inhibition of, 61 substrate specificity of, 51, 52, 54-56, 57,58 as tyrosine-specific phosphatase, 62 PTPase. See Protein-tyrosine phosphatases (PTPases)

200 PTPIB acid-base catalysis by, 37 as intracellular PTPase, 22 structure of, 22, 25, 28, 29, 34, 46 substrate specificity of, 48, 51, 52, 53 tungstate inhibition of, 26 p21 protein, gene expression for, 110, 138 Pyridoxal phosphate, as PTPase substrate, 56 R Radical oxygen intermediates. <See Reactive oxygen species (ROS) Radiotherapy apoptosis following, 107 oxidative stress from, 137 Reactive oxygen species (ROS) cellular responses to, 134-139 gene regulation by, 123-148 role in h u m a n disease, 123 role in N F - K B activation, 153-154 signal transduction and, 139-143, 153 targets of, 141 Redox changes, effects on gene expression, 134-139 Redox regulation, of NF-kB, 149-161 Redox-sensitive transcription factors, 124-134 models of, 129 Rel/zcB family, redox sensitivity of, 132-133 Renal carcinoma, tumor suppressor gene for, 24 R409 mutants, of PTPases, 33, 34 Rheumatoid arthritis, N F - K B role in, 150 RIP, reaction with Fas death domain, 88 Rob, oxidative activation of, 167, 168-169 rolB oncogene, 24 RpoS, oxidative activation of, 169, 170 rpr gene, role in apoptosis, 110 s Salmonella typhimurium, oxidative stress responses of, 125, 164, 170-171 Second messengers

INDEX inhibition in bone resorption, 190-191 requirements for, 139-140 targets of, 142-143 SELEX procedures, application to binding by iron regulatory proteins, 12, 13 Serine protease, role in apoptosis, 81 Serine/threonine, in PTPase signature motif, 34-37, 46 Serpin, as death protease inhibitor, 79-80, 91 SH2 domains, in intracellular PTPase, 22, 4 8 - 4 9 , 62 SH-PTPl, as negative regulator of cytokine signaling, 24 SH-PTP2, substrate specificity of, 4 8 - 4 9 Signaling mechanisms antioxidant role in, 149 in apoptosis, 110 N F - K B regulation by, 150-157 PTPase role in, 21, 23-24, 62 Single-displacement pathway, of phosphatase action, 30 Small nuclear ribonucleoprotein (SNURP), enzymatic cleavage of, 81 soxR gene, role in oxidative stress, 125, 126 SoxRS, oxidative activation of, 165-168, 170, 175 Soybean trypsin inhibitor, as serine proteinase inhibitor, 183 S p l transcription factor, redox sensitivity of, 126-127 src gene, as oncogene, 23, 24 Steroid hormones, apoptosis induction by, 70, 72 Substantia nigra, apoptosis in, 116-117 Succinate dehydrogenase iron-responsive elements in, 15 translational regulation of, 16 Sulfotyrosine, structure of, 60 Superoxide anion cell damage from, 115 effect on aconitase iron-sulfur cluster, 9 formation in Fenton reaction, 1 as reactive oxygen species, 123, 163, 175 scavengers for, 134

201

INDEX Superoxide dismutases (SOD) antioxidant reactions catalyzed by, 123, 136, 163-164 in cell signaling, 149

Tat gene, drugs based on blockage of, 157 Terminin (Tp), role in apoptosis, 111-112 Thiol phosphate covalent intermediate, in phosphatase catalysis, 31-32 Thioredoxin in cell signaling, 149 role in N F - K B regulation, 152-153 T lyinphoc5rtes, apoptosis of, 112 Tocopherol, as free radical scavenger, 123 Tosyllysyl chloromethane, as cysteine proteinase inhibitor, 179 Transcription factors, redox-sensitive. See Redox-sensitive transcription factors Transferrin, binding activity of, 171 Transferrin receptor (TfR), expression and regulation of, 2, 3, 5, 16 Transglutaminase, apoptosis activation of, 69 trx2 gene, YAP-1 activation effects on, 134 T T F l transcription factor, redox sensitivity of, 134 Tumor necrosis factor (TNF), 82 apoptosis induction by, 70, 83, 110 effect on cathepsin secretion, 180 N F - K B regulation of, 150

receptor for, 8 6 - 8 7 Tumor suppressor genes PTPases as possible products of, 24 role in apoptosis, 118 Tungstate, as PTPase inhibitor, 26-27, 33 TX, effect on apoptosis, 81 Tyrosine, phosphorylation of, enzymes for, 21 Tyrosine kinases, role in apoptosis, 89-90

U UBC-FAP, role in apoptosis, 81 Ultraviolet radiation apoptosis induction by, 70 role in oxidative stress, 135, 137 Urate, as free radical scavenger, 123 5'UTR sequence of ferritin transcripts, 2, 4 iron-responsive elements in, 15

Vaccinia V H l phosphatase biological function of, 24 dual-specificity phosphatases related to. See VHR phosphatases PTPase signature motif in, 22 Vaccinia virus, H I gene of, 136 Vanadate, as PTPase inhibitor, 57, 59 VHR phosphatases active site substrate specificity of, 54, 57,58 dual specificity of, 62 enz3maatic activity of, 35, 36 structure of, 28-30, 46 Viral hepatitis. Fas and FasL in host defense against, 89 Viral proteins, effect on apoptosis, 76-77

Yama, role in apoptosis, 81-82 YAP-1, redox sensitivity of, 132, 134 Yersinia protein-tyrosine phosphatase acid-base catalysis by, 37, 38, 42 active site substrate specificity of, 54 biological function of, 24 enzymatic activity of, 35, 36 signature motif of, 32, 34, 46 structure of, 22, 25, 26, 29 substrate specificity of, 4 9 - 5 0 transition-state structure of hydrolysis by, 42

Zinc-finger transcription factors, redox sensitivity of, 126-129

ISBN D-lE-lSEflaS-T 90018

9 n 780121 n 528355