Advances in Botanical Research, Vol. 15

Advances in

BOTANICAL RESEARCH VOLUME 15

Advances in

BOTANICAL RESEARCH Editor-in-Chief J. A. CALLOW

Department of Plant Biology, University of Birmingham, Birmingham, England

Editorial Board H. W. WOOLHOUSE W. D. P. STEWART E. G. CUTTER W. G. CHALONER

E. A. C. MAcROBBIE

John Innes Institute, Norwich, England Department of Biological Sciences, The University, Dundee, Scotland Department of Botany, University of Manchester, Manchester, England Department of Botany, Royal Holloway & Bedford New College, University of London, Egham Hill, Egham, Surrey, England Department of Botany, University of Cambridge, Cambridge, England

Advances in

BOTANICAL RESEARCH Edited by

J. A. CALLOW Department of Plant Biology University of Birmingham Birmingham, England

VOLUME 15

1988

ACADEMIC PRESS Harcourt Brace Jovanovich, Publishers

London San Diego New York Berkeley Boston Sydney Tokyo Toronto

ACADEMIC PRESS LIMITED 24/28 Oval Road London NW17DX

United States Edition published by ACADEMIC PRESS INC. San Diego, CA 92101

Copyright 0 1988by ACADEMIC PRESS LIMITED

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.

British Library Cataloguing in Publication Data Advances in botanical research.-Vol. 15 1. Botany-Serials 581'.05 ISBN 0-12-005915-0

Typeset by Paston Press, Loddon, Norfolk and printed in Great Britain by T.J. Press (Padstow), Cornwall

CONTRIBUTORS TO VOLUME 15

THOMAS BJORKMAN, Department of Botany, Universityof Washington, Seattle, Washington, U S A H. GRIFFITHS, Department of Biology, University of Newcastle, Newcastle upon Tyne, N E l 7RU, UK LEON V. KOCHIAN, US Plant, Soil and Nutrition Laboratory, USDAA R S , Cornell University, Ithaca, New York, U S A WILLIAM J. LUCAS, Department of Botany, University of California, Davis, California, U S A ROGER I. PENNELL, John Znnes Institute and A F R C Institute of Plant Science Research, Colney Lane, Norwich, NR4 7UH, U K

This Page Intentionally Left Blank

PREFACE

The major part of this volume of the Advances is devoted to three of the most enduring themes in plant physiology, the mechanisms by which plants perceive and respond to gravity (Bjorkman), Crassulacean Acid Metabolism (CAM) and its regulation (Griffiths), and the mechanisms of K+ uptake and transport in roots (Kochian and Lucas). All three have been reviewed many times, what justification is there for further treatments of these perennial favourites? In his evaluation of the various hypotheses that seek to explain gravity perception and response, T. Bjorkman seeks a more critical application of the laws of physics than hitherto. It appears, however, that a basic lack of information still limits our ability to advance sound models at the subcellular and molecular level. Ways of rectifying the situation are presented with particular emphasis on incorporating recent advances in the understanding of cellular signalling mechanisms. Much of the pioneering biochemical work on CAM and its regulation was carried out by the Newcastle Group, most notably involving M. Thomas and S . L. Ranson. In continuing the Newcastle tradition, H. Griffiths’ review is concerned with the ecophysiological aspects of CAM regulation as found in plants of very diverse terrestrial and aquatic habitats. The emphasis here is on the integrated study of the various primary aspects of CAM and its diverse secondary consequences. The author also seeks to rectify the conventional view that CAM is all about malate, by raising the, as yet, not entirely explained role of citrate. Despite a vast literature on potassium uptake and accumulation by plant roots, it seems that our understanding of the mechanisms involved and their regulation is still far from complete. In their highly authoritative review, L. V. Kochian and W. J. Lucas give appropriate consideration to the historical dimension in attempting a synthesis of the current models, but they are also concerned to look to the future to identify the most profitable lines of research through the integration of the physiological approach with advanced biophysical techniques such as “patch-clamping” , and with improved immunological and gene cloning methods for characterizing membranes and their protein components. vii

...

Vlll

PREFACE

Volume 15 is not entirely devoted to physiology. Although the cellular processes involved in the formation of microspores and megaspores are fairly well documented for flowering plants other groups of seed-bearing plants are more neglected and, taking Tuxus as his main example, R. Pennell outlines recent work on this subject. I thank all the authors for their endeavours and efforts to minimize the editor’s task. J. A. CALLOW

CONTENTS

CONTRIBUTORS TO VOLUME 15 . . . . . . . . . . PREFACE . . . . . . . . . . . . . . . . . . . .

V

vii

Perception of Gravity by Plants THOMAS BJORKMAN I. I1.

Objectives

. . . . . . . . . . . . . . . . . . .

Susception . . . . A . How Gravity Acts B . Thermal Motion

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I11.

Transmission . . . . . A . Electrical Transmission B . Chemical Transmission

IV .

Perception . . . . . . . . A . Signal Transduction Overview B . Multiple Systems . . . . C . Statolith Sensors . . . . D . Nonstatolith Perception . .

V.

Integration and Conclusion

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . . . . . . . . . .

1 3 3 4

7 8 10 12 12 16 18 32 36

Crassulacean Acid Metabolism: a Re-appraisal of Physiological Plasticity in Form and Function H . GRIFFITHS I. I1.

Introduction

. . . . . . . . . . . . . . . . .

Modification and Regulation of Constitutive CAM . . . . . . A . Occurrence and Distribution of Constitutive CAM . . . .

ix

44 45 45

CONTENTS

X

B. C. D. E.

Characterization of the Die1 Cycle . . . . . . . . . Biochemical Regulation . . . . . . . . . . . . . Modification of CAM Phases by the Environment . . . . Plant Water Status: Is there Regulation both of and by Solute Accumulation? . . . . . . . . . . . . . . . . Stable Isotope Ratio Analysis . . . . . . . . . . . Regulation of Respiratory C 0 2 Recycling During CAM . . .

46 50 51

111.

Plasticity of Metabolic Response: Shades of CAM . . . . . . A . CharacteristicsofC,- CAMIntermediates . . . . . . . B . Occurrence, Distribution and Evolution . . . . . . . . C . Respiratory COz Recycling by C,- CAM Intermediates . . . D . Physiological Characteristics of the C,- CAM Transition . .

67 68 70 74 76

IV .

Significance of Respiratory COz Utilization During CAM . . . . A . RecyclingintheTerrestrialEnvironment . . . . . . . B . Recycling in the Aquatic Environment . . . . . . . . C . Organic Acid Speciation: The Newcastle Hypothesis Revisited

78 79 82 83

. . . . . . . .

85

F.

G.

V.

CAM: Development of Integrated Research

53 55 64

Potassium Transport in Roots LEON V . KOCHIAN and WILLIAM J . LUCAS

. . . . . . . . . . . . . . . . . .

94

I1 .

Plasma Membrane Transport of K+in Roots . . . . . . . . A . Early Work: thecarrier-Kinetic Approach . . . . . . . B . Are Root K+Fluxes Coupled to H+? . . . . . . . . . C . Uptake at High K+ Concentrations: the Linear Component . D . Summary . . . . . . . . . . . . . . . . .

94 94 99 120 128

111.

Redox-coupledPlasmalemmaTransportofK' . . . . . . . A . Influence of Exogenous NADH on K+ Influx . . . . . . B . Membrane Transport and the Wound Response . . . . . C . DevelopmentofanIntegratedNADHModel . . . . . .

129 130 131 132

IV .

Radial K+Transport to the Xylem . . . . . . . . . . . A . Site of K+ Entry into the Symplasm . . . . . . . . . B . Radial Pathway . . . . . . . . . . . . . . . C . Lag Phase in Xylem Loading . . . . . . . . . . . D . K+Transport into the Xylem . . . . . . . . . . . Regulation of K+ Fluxes within the Plant . . . . . . . . . A . Allosteric Regulation of K+Transport . . . . . . . . B . K+ Cycling within the Plant: an Integration of Regulatory Mechanisms . . . . . . . . . . . . . . . .

136 137 140 143 145 151 152

I.

V.

VI .

Introduction

Future Research and Prospects

. . . . . . . . .

163 169

CONTENTS

xi

Sporogenesis in Conifers ROGER I . PENNELL I.

. . . . . . . . . . . . . . . . . .

179

I1.

Sporogenesis in the Pollen-bearing Cone . . . . . . . . . A . The Archaesporium and Differentiation within the Sporangium B . Sporogenous Cells and Tapetum . . . . . . . . . . C . Meiosis . . . . . . . . . . . . . . . . . . D . ExinePatterningandtheFreeSporePeriod . . . . . .

181 181 182 183 185

I11.

Megasporogenesis . . . . . . . . . . . . . . . . A . The Origin of the Reproductive Cell Lineage within the Ovule B . Mitochondria, Plastids and Planes of Division within the Megaspore Mother Cell . . . . . . . . . . . . . . . C . Megaspore Viability . . . . . . . . . . . . . .

190 190 191 193

AUTHOR INDEX . . . . . . . . . . . . . . .

197

SUBJECT INDEX . . . . . . . . . . . . . . .

207

Introduction

This Page Intentionally Left Blank

Perception of Gravity by Plants

THOMAS BJORKMAN

Department of Botany KB-15 University of Washington, Seattle, Washington 98195, USA

I.

11.

111.

IV.

. . . . . . . . . . . . . . . . . . Susception . . . . . . . . . . . . . . . . . . A. How Gravity Acts . . . . . . . . . . . . . . B. Thermal Motion . . . . . . . . . . . . . . Transmission . . . . . . . . . . . . . . . . . A. Electrical Transmission . . . . . . . . . . . . Objectives

1

. . . . , .

3 3 4

. . 7 . . 8 B. Chemical Transmission . . . . . . . . . . . . . . 10 Perception . . . . . . . . Signal Transduction Overview B. Multiple Systems . . . . C. Statolith Sensors . . . . D. Nonstatolith Perception . . A.

V.

. .

Integration and Conclusion

. . . . . . . . . . . . . . .

. . .

. . . . . . . . . . . . . . .

. . . . .

. . . . . 12 . . . . . 12 . . . . . 16 . . . . . 18 . . . . . 32

. . . . . . . . . . .

36

I. OBJECTIVES An environmental cue widely used by plants to guide development is gravity. Although many things are known about plant responses to gravity, one fundamental aspect remains obscure, i.e. the means by which the physical stimulus of gravity is transduced into a physiological response which the plant can use to guide development. Copyright 01988 Academic Press Limited All rights of reproduction in any form reserved.

Advances in Botanical Research Vol. 15 ISBN 0-12-005915-0

1

2

T. BJORKMAN

In order to analyse that transduction, both the physics and physiology must be considered. Although the physical aspects have been considered before, notably by Audus (1979), some reiteration and expansion of earlier discussions is useful to interpret recent data and to indicate valuable concepts to pursue. In this chapter the intention is to give these physical laws deeper attention than they have received in the past. This discussion of physics and gravity perception should point out paradigms, some provocative, which will be helpful in developing new hypotheses of gravity perception. The physical behaviour of objects on a human scale is very different from that on the subcellular scale. It is easy, when imagining how gravity perception might work, to make erroneous assumptions about how various components will act. It is hoped to give the reader an intuitive grasp of how gravity acts on plant cells and how that may be turned into physiological information. The perception of other physical stimuli, light and sound, has recently been well characterized in animals. Metabolic syndromes in these transduction mechanisms may have parallels in gravity perception by plants. The experimental approaches which led to the elucidation of the animal transducers also provide a useful guide to promising future approaches. Gravitropism is the result of a series of events, and the terms used for each step are as defined by Hensel(1986a). Susception is the initial physical reaction by amass in the gravitational field. Perception is the conversion of the physical signal to a physiological one. Transmission is how the signal moves from the cells where gravity is sensed to those where growth occurs. The response is the differential growth which results in curvature. The term transduction is sometimes used to describe the step called perception, but here a more specific meaning will be used: the carrying over of energy or information from one form (or place) to another. By that definition, transduction occurs in each of the four steps of gravitropism. Gravity sensing (or gravisensing) is also a common term used to describe part of gravitropism, usually corresponding to susception and perception. There is not yet a consensus on the terminology describing the steps of gravitropism; as these steps become physiologically better defined, so will the words used to describe them. In order to discuss potential mechanisms of perception, gravity will first be considered from a physical perspective and related to a plant’s susception of gravity. Second, the types of signals which may be involved in transmission will be reviewed. In the main section, the means by which the information provided by susception can be transduced to the kinds of signals which may be transmitted will be considered. Some potential perception mechanisms will be evaluated in terms of their likelihood of performing as rapidly and as sensitively as does the true gravity-sensing system.

PERCEPTION OF GRAVITY BY PLANTS

3

11. SUSCEPTION The first step in gravitropism is a physical action of gravity on some element in the plant, which is called susception. Gravity is an attractive force on a mass; it is that force which a graviresponding plant must use to orient itself relative to the gravitational field. Although gravity acts on every atom in the plant, the number of relevant interactions is limited. This section will cover the action of the gravitational force and limitations to its detection. A. HOW GRAVITY ACTS

To analyse how a plant senses gravity one must know how gravity interacts with physical objects. Newton’s law of gravitation holds that two objects attract each other with a force proportional to the product of their masses and inversely proportional to the square of the distance between them:

Fg = Gmm’/r2 where m and m’ are the two masses and r the distance between them. G is the experimentally determined gravitational constant. The attractive gravitational force will tend to move two particles towards each other, which we can observe, for example, as the attraction between an apple and the earth. In the case of gravity sensing by plants on the earth’s surface, one particle is the earth, and the other is within the plant. The distance between the particles is the radius of the earth because the gravitational force of a sphere acts as if all the mass is at its centre. Thus the values for G, m and r are constant and the gravitational equation reduces to: Fg = (9.8 m s-’) m’. The gravitational force then depends on the mass (m‘)of whatever particle we are considering. A plant must sense gravity by detecting the attraction between the earth and the mass of some object associated with the plant, which will be referred to as the sensing particle. For that attraction to be detectable, the sensing particle must do work (in the thermodynamic sense) on something to cause a change in the physiological activity of the plant. Displacement of the particle in the gravitational field is required to convert gravitational potential energy to work. If the particle does not move, there is no energy to alter the physiology of the plant. A simple example of work, as defined in physics, is a mass moving against gravity. The work done is the force applied ( g x m ’ ) times the distance the mass is moved. This can be illustrated by a seesaw (Fig. 1). With mass A on the low side, if a larger mass, B, is placed on the high side it will exert more force and drop. As it drops, it loses gravitational potential energy and does work on mass A. The work done on mass A causes the

4

T. BJORKMAN

Fig. 1. Displacement of a mass in a gravitational field does work. (a) Mass A is at rest on a seesaw; a larger mass (B) is placed on the high side of the seesaw. (b) Mass B exerts more force and therefore moves the seesaw. As it descends, it loses potential energy (B x g X d ) and does work on mass A by raising it in the gravitational field. (c) The mechanical work done on mass A increases its gravitational potential energy by A X g X d. The work could also be converted to: (d) electrical energy by running a generator; or (e) chemical energy by running a reverse osmosis unit which separates solutes from water.

mass to rise, increasing its gravitational potential energy. The rising arm of the lever could also cause a generator to turn, creating electrical energy; or it could push the piston in a reverse osmosis unit, creating chemical potential energy by purifying water. Work can convert the gravitational potential energy to many other forms of energy, depending on the transduction mechanism. Therefore, a simple model of susception can lead to many possibilities for perception. Susception can occur in a number of possible ways: the sensor may do work either by being denser than the surrounding medium and sinking, or by being lighter and rising. The sensor may be inside or outside the cell. There are many objects within the cell which may move relative to each other due to their differing densities. The cell could perceive motion, displacement or position of the sensor. We will see what evidence there is for each of these things happening. B. THERMAL MOTION

The mass acted on by gravity is also being constantly agitated by collision with water and other molecules which have kinetic energy due to heat.

PERCEPTION OF GRAVITY BY PLANTS

5

This random thermal motion is commonly seen as Brownian motion. For the gravity sensor to be effective, it must be relatively insensitive to the random thermal energy but be very sensitive to changes in the direction of gravity. An important limiting factor for a sensing mechanism, then, is thermal noise. The magnitude of thermal energy on a particle is +kT in each dimension, where k is Boltzman’s constant (1.38 X lovz3J K-’)and T is the absolute J. The thermal temperature. At room temperature, f k T = 2 x agitation of a particle is independent of the particle’s mass, so the effect of gravity relative to the thermal noise is greater the more massive the particle. Thermal noise sets one lower limit on the minimum work which must be done by a sensor during susception. The rate of a chemical reaction is limited by the activation energy of the reaction. Thermal motion of the reactant provides the energy needed to overcome the activation energy of a chemical reaction. For a reaction caused by a mechanical stimulus, in contrast, the sensor should be selectively activated by the stimulus rather than thermal motion. This can be accomplished if the activating reaction has an activation energy high enough that it is rarely stimulated spontaneously. If the activation energy is high, the minimum stimulus must be correspondingly large. There is a trade-off between a sensor’s sensitivity and its selectivity. The effect of activation energy on the spontaneous reaction rate can be calculated (Fig. 2). The frequency of activation by thermal energy decreases very rapidly as the activation energy increases. The figure is based on the Arrhenius equation: Rate = Ae-E’RT.The light sensor in vision is rhodopsin, which is physically activated by the energy in photons. Rhodopsin is stimulated only very rarely by heat. If the enzyme reduced the activation energy for the reaction only two-fold, thermal activation could occur 10” times faster. Rhodopsin functions well in light perception because its activation energy is high enough to make spontaneous triggering very rare (about once every 1000 years per molecule), but is low enough that the light stimulus contains ample energy. By a similar approach, this figure can be used to estimate the activation energy of gravity perception. The rate of the first step in gravity perception by the thermal motion of the suscepting body must be much less than the rate during gravistimulation. The activation energy of that step must thereto fore be high enough that the thermal energy (hkT) produces activations for each one produced by a small gravistimulation. From Fig. 2 it can be determined that the activation energy of the first step of perception which fits this criterion is 3 4 x lO-”J. Thus the activation energy, and hence the amount of work required during susception, can be fairly precisely estimated. Presumably there would be many activating events per cell per second, each using about 4 X lo-’’ J. This estimate applies regardless of the specific mechanisms of susception and perception.

6

T.BJORKMAN

-5

-

-10

-

1/2 kT

Enzymatic reactions

W c

e

c .-0

t -150

p!

m

-0

-20

-25

-

Rhcdopsinactivation

Activation energy ( J 1 Fig. 2. The effect of activation energy on the relative rate of reaction. Enzymic reactions have activation energies clustered over a small range, yet one where the rate of a reaction can vary by many orders of magnitude. Spontaneous thermal activation of the light-sensing molecule rhodopsin occurs at a very low rate. A gravity sensor must be insensitive to thermal energy (IkT), so its activation energy would be expected to be 3 4 X lo-*' J. Intermediate marks on the abscissa are 2 and 5 times the order of magnitude. The formula for this curve is: log R = -E,/kT.

Although small effects of gravity can be amplified through various biochemical cascades, these amplifiers will not discriminate between the desired signal and noise. This lack of discrimination is why amplification alone does not provide the necessary sensitivity to detect weak stimuli. For an amplifier to be useful, the total signal must be filtered or averaged using some criterion to reduce the noise. The possibilities for such processing by the perceiving system will be discussed in a later section. The only measurable physiological response of a plant which yields information about susception is the presentation time, which is the threshold gravistimulation required to elicit a growth response. There is a reciprocity between the presentation time and acceleration when the acceleration is changed from 1 x g by centrifuging or clinostatting (Johnsson, 1965). A metabolic process would not exhibit this sensitivity to the gravitational force, so the presentation time must reflect the physical process of susception. The presentation time may also include some time

PERCEPTION OF GRAVITY BY PLANTS

7

required for the physiological steps of perception (Johnsson, 1965). The inverse relationship between the force and the presentation time is consistent with a threshold displacement of a sensor required to stimulate perception. This threshold may be thermal noise. Hair cells in the cochlea (the auditory receptor in animals) can detect extremely small stimuli through tuning and time-averaging the repetitive stimulus of a sound wave. The limit of perception corresponds to a motion of atomic dimensions (Harris, 1967), even though sensitivity is limited by thermal motion. A measure of the ability of the sensor in susception to overcome thermal noise is the minimum amount of stimulation which will produce a gravitropic response. Avena roots respond to 3 X 10-4g when stimulated as long as 68 h on a clinostat (Shen-Miller et al., 1968). In lettuce seedlings grown in centrifuges aboard the Salyut 7 space station, the shoots had a threshold response at 3 x 10-3g, and the roots at much lower gravity (Merkis et al., 1985). At 1 X g the presentation time can be as short as 7 s for Lepidium roots (Larsen, 1969). Such high sensitivity can be achieved only by signal averaging and with a substantial responding mass. The potential sensors in a plant are limited to those that are large enough to move a perceptible amount relative to thermal noise within the time it takes a plant to detect a gravistimulus. That much motion must be produced even by the very small stimuli to which plants are capable of responding.

111. TRANSMISSION Gravity sensing occurs only in certain regions of the plant, but gravitropic curvature rarely occurs in those cells which sense gravity. A signal which indicates how the responding cells must alter their growth passes from one group of cells to the other. Transmission has been excellently reviewed by Audus (1979). In roots, for example, gravity is sensed in the root cap, but growth occurs in the elongating zone of the root apex, several millimetres away (Darwin, 1899). In etiolated beans, gravity is sensed in the cotyledonary hook, but curvature occurs 2-3 cm lower in the stem (Hart and Macdonald, 1984; Verbelen et al., 1985). The sensing and responding cells may also be separated radially. In coleoptiles, gravity is sensed in the inner mesophyll, but growth is controlled by cells in and near the epidermis (Thimann and Schneider, 1938; Kutschera et al., 1987). Curvature can only be expressed in cells which are growing or which may be induced to grow; sensing cells need not be near a growing zone. A signal must therefore travel from one group of cells where gravity is sensed to another where growth is controlled. The signal may move symplasmically or apoplasmically;in either case the signal must leave the cell where it originates by crossing the plasma membrane.

8

T. BJORKMAN

Membrane transport is tightly regulated, and therefore is a good place to focus attention when trying to narrow down the type of signal which is elicited from the sensing cell. In roots, transmission of the gravistimulus appears to be through modulation of a growth-retarding factor in the elongating zone (Shaw and Wilkins, 1973), potentially an inorganic ion rather than a plant growth regulator (Mertens and Weiler, 1983). In the shoot it is likely to be modulation of a growth-stimulating factor (Dolk, 1933). In grass nodes, a growth initator moves to the epidermis (Kaufman and Dayanadan, 1984). Perception therefore occurs by a mechanism which causes this kind of modulation. Below, the biological mechanisms which elicit such signals are discussed and those which promise to be relevant to gravity perception are identified. Intercellular communication can be either electrical or chemical. The most dramatic example of bioelectrical communication is the nervous system of animals. Chemical communication is typified by hormones, second messengers and plant growth regulators. The nature of the transmitted signal indicates the kind of reaction which is the final step of perception. A. ELECTRICAL TRANSMISSION

Signals are often transmitted electrically, both in biological and engineered systems. A biological electrical signal can have many manifestations: an action potential, an electrical gradient or electrophoresis in an electrical field. Each of these manifestations could transmit a physiological signal. 1. Action Potentials Action potentials allow rapid communication. The signal for leaf folding in Mimosa pudica (sensitive plant) and Dionea muscipula (Venus flytrap) are carried by action potentials (Pickard, 1973a). In Mimosa, vibration causes an action potential which is transmitted to pulvini at the base of petioles. There, the action potential triggers ion fluxes which cause turgor changes, folding the leaf. In Dionea, stimulation of a trigger hair sends an action potential to the leaf base (Burdon-Sanderson, 1873; Benolken and Jacobson, 1970), initiating rapid growth (Williams and Bennett, 1982). Action potentials are rapid and transient changes in the membrane potential in response to a stimulus. The change in potential is due to increased permeability of an ion which is far from its equilibrium and which normally has a low permeability. The membrane potential then approaches the equilibrium potential for that ion. The potential difference between a stimulated cell and its neighbour is detected, presumably across plasmadesmata, and triggers an action potential in the second cell. This process continues down the line of excitable cells. Cells which can transmit an action potential are termed excitable because they respond actively to a stimulus.

PERCEPTION OF GRAVITY BY PLANTS

9

Action potentials have been measured in many plants, especially in association with rapid movements including growth (Pickard, 1973a), but there are no examples analogous to the processes in gravitropism. Apparent action potentials have been noted in gravitropic Allium roots (Berry and Hoyt, 1943) but not in Lepidium roots (Behrens and Gradmann, 1985) Action potentials remain a possibility for transmission of the gravitropic stimulus, but evidence to support this possibility is lacking.

2. Electrical Gradient A potential applied apoplasmically across a tissue can affect growth (Schrank, 1948; Evers and Lund, 1947; Moore et al., 1987). Gravistimulation does produce an electrical gradient (Schrank, 1947; Grahm and Hertz, 1964; Tanada and Vinten-Johansen, 1980; Behrens et al., 1982; Bjorkman and Leopold, 1987a), raising the possibility that curvature is the result of an electrical field. Although ion currents flowing in such an electrical field gradient have been proposed as guides to development in many tissues (Jaffe and Nuccitelli, 1974), there is no evidence that the field is sensed directly. A direct sensor would not be unheard of: magnetosensors, which sense magnetic fields, are known in biology (Lohmann and Willows, 1987). 3. Electrophoresis An electrical field applies an attractive force to charged particles; therefore, an apoplasmic electrical gradient across a tissue causes mobile ions to move across the tissue through the cell walls. The ions, redistributed by electrophoresis, may then influence growth or act allosterically to modulate growth regulators (Hasenstein and Evans, 1986). There is evidence against electrophoretic translocation of Ca2+ in the establishment of tissue asymmetry, and it is applicable to other ions as well. Curvature is towards the positive pole of an applied potential gradient in both positively gravitropic roots (Bjorkman, 1987) and negatively gravitropic shoots and coleoptiles (Schrank, 1948; Woodcock and Wilkins, 1969a). Transverse movement of Ca2+ is towards the slower-growing, positively charged side both in roots (Lee et al., 1983) and in coleoptiles (Slocum and Roux, 1983). Calcium ions are therefore moving against the electrical gradient; they are not being electrophoresed. Auxin anions would electrophorese towards the slower-growing side in both cases. That is compatible with its role as a growth inhibitor in roots, but there is apparently no auxin redistribution in roots (Mertens and Weiler, 1983). In shoots, auxin stimulates growth and is redistributed against its electrical gradient (Mertens and Weiler, 1983; Bandurski et al., 1984). Auxin would perhaps not be susceptible to apoplasmic electrophoresis in any case because at the low wall pH it would be in the uncharged, protonated form. Thus, existing data disprove such simple models of growth modulation through electrophoresis

10

T. BJORKMAN

of biologically active ions. Nevertheless, electrical polarization appears to be a common feature of the gravitropic response. An electrical gradient is established by differences in electrogenic ion transport across the plasma membrane. Electrogenic transport occurs when a charge is moved across a membrane without a balancing charge, resulting in a change in the electrical potential across the membrane. The electrogenic transport could be varied by changing the activity of an electrogenically ion-transporting ATPase, such as the proton pump. The membrane potential could either hyperpolarize or depolarize, depending on whether the activity of the pump were increased or decreased. Depolarization could also be effected by increasing the permeability to some ion which normally has a large electrochemical gradient across the plasmalemma, as in action potentials. Such a permeability increase would most likely result from opening of a specificion channel and ion movement through it down the electrochemical gradient without balancing counterions. Regulation of ion transport is an area of active research, and there are experimental means of determining how electrical effects of gravistimulation are generated. B. CHEMICAL TRANSMISSION

The unequal signal to the growth zone may be generated by direct chemical transport, with the electrical polarization which results from ion movement being incidental. Signal substances transmitted by this type of mechanism would be those which are actively transported by the cells using polarly distributed carriers for directional facilitated diffusion, or even specific ATP-driven transport proteins.

1. Ion Pumping Directional active transport of ions is widespread in plants to serve their mineral nutrition needs. The systems which are in place to distribute mineral ions in the plant may be adapted to transmit information, or similar independent transport systems could be dedicated to this role. Applied Ca2+ can cause curvature in corn roots, even if they are decapped (Lee et al., 1983), suggesting that an apoplasmic Ca2+gradient may be the transmitted signal. The increased apoplasmic Ca2+ concentration may sensitize the tissue to auxin which is present in growth-inhibiting concentrations (Hasenstein and Evans, 1986; Salisbury et al., 1985). It may also act directly on the wall to inhibit growth (Cleland and Rayle, 1977). Apoplasmic Ca2+ ions therefore fulfil the requirements of a signal substance. Transverse Ca2+fluxes have been measured (Lee et al., 1983) and wall calcium has been shown to redistribute (Slocum and ROUX,1983) following gravistimulation. The redistribution of only one other inorganic ion, K+,has been measured. Although K+ redistribution is important in pulvinus move-

PERCEPTION OF GRAVITY BY PLANTS

11

ment in seismonastic Mimosa, it does not generate the response of the pulvini to gravistimulation (Roblin and Fleurat-Lessard, 1987). If Ca2+ions are electrogenically pumped towards the slower-growingside of a tissue, that side will become electropositive relative to the faster-growing side. Such polarization has been notoriously difficult to measure (Woodcock and Wilkins, 1969b). Reliable measurements in coleoptiles show that polarization is a consequence, not a cause, of auxin redistribution (Grahm, 1964). Apparent upward currents in root caps following gravistimulation (Behrens et al., 1982; Bjorkman and Leopold, 1987a) imply that the upper side is positive relative to the lower. This would be consistent with electrogenic cation transport towards the upper side. However, Ca2+move towards the lower side of root tips (Lee et al., 1983). A calcium-transporting ATPase has been discovered on the plasma membrane (Rasi-Caldogno etal., 1987), but the effect on the charge distribution is unclear because the Ca2+ are exchanged for protons in an unknown stoichiometry. Unfortunately, there is not quite enough information available to determine whether direct pumping of an ion may be the transmitter of the gravity signal.

2. Growth Regulator Pumping Polar (or directional) transport of growth-regulating substances would more directly effect differential growth than ionic messengers. Evidence is weak for growth regulator redistribution in roots (Mertens and Weiler, 1983; Jackson and Barlow, 1981), but good for above-ground parts (Pickard, 1985) with some exceptions (Trewavas, 1981). Polar transport of auxin occurs through an auxin-translocating carrier in the plasma membrane, down an electrochemical gradient maintained through protonation of indoleacetate with protons supplied by the plasma membrane proton pump (Rubery and Sheldrake, 1974). The net result is ATP-driven auxin transport. The transport is polar because the IAA carrier is located only on one flank of the cell. This transport requires Ca" transport from the auxin sink (Niedergang-Kamien and Leopold, 1957), though the biochemical basis for this requirement is unknown. A gradient across the tissue may be established through two kinds of cellular response. Either the top and bottom of each sensing cell respond differently, with each sensing cell behaving similarly, or the cells in the upper part of the tissue may act differently from those in the lower part. The membrane potential of sensing cells in Lepidium root caps respond to gravistimulation; those on the lower side hyperpolarize slowly whereas those on the upper side depolarize rapidly yet transiently (Behrens et al., 1985). This is a clear example of different responses by cells in different parts of the sensitive region. This result is expected if the cell is organized so that the flank towards the epidermis differs from the side towards the centre of the tissue (Volkmann and Sievers, 1979). In grass nodes, such an organiza-

12

T. BJORKMAN

tion is believed to induce auxin transport only from cells on the lower part of a horizontal node (Wright and Osborne, 1977). Whether an electrical gradient is established through change in electrogenic ion transport or a chemical gradient is established by polar transport of some substance, modulation of transport across the plasma membrane is necessary. The tissue gradient can be established either by asymmetric transport across each sensing cell or by cells on the bottom of the tissue responding differently from those on top. Because there is intercellular signalling, in all cases a change in plasma membrane transport would be involved. Based on presently available data we cannot definitely eliminate any of the transmission methods: ion channels, plasmadesmata, proton pump modulation, action potentials, polar transport of ions or growth regulators, and activation of a specific porter.

IV. PERCEPTION The section on susception described ways in which perception could be triggered. The section on transmission described the type of signal produced by perception. Taking the inferences about perception from those sections into account, this section will establish how perception is likely to work and will consider some mechanisms which may safely be rejected. The mass with which gravity interacts in susception must not only detect gravity, but must cause some physiological response. The effect of perception is to transmit a polar signal across the pIasma membrane of some or all of the cells in the sensing tissue. The transduction from an intracellular stimulus to an extracellular signal, perception, could occur through a direct interaction between the sensing mass and a cell component which elicits the transmitted signal. More likely, it is mediated through a cascade of intracellular messengers to trigger transmission of a signal. An intermediate reaction cascade amplifies the signal and provides an opportunity for modulation of the message by other cellular processes. To consider the ways susception could produce a transmitted signal, biological signal transduction in general will be discussed, and then compared to what we know about gravity perception. Some hypotheses which have been proposed for gravity perception in plants will also be discussed. A. SIGNAL TRANSDUCTION OVERVIEW

As a basis for discussing the mechanism of gravity perception in plants, it is worthwhile to first review signal transduction in some sensory systems which are better understood. There may be analogous patterns among the sensory systems which can be useful in predicting perception of gravity by plants.

PERCEPTION OF GRAVITY BY PLANTS

13

1, Other Sensory Systems

The photoreceptors in eyes of invertebrate animals are probably the sensory system which is best characterized at the physiological level (Steive, 1986). Perception of light begins with the absorption of photons by rhodopsin in the plasma membrane of retinal rod cells. Activation of a rhodopsin molecule J (Yau et al., 1979), 100 times the thermal noise in each requires 2 x dimension, so spontaneous triggering is rare. A photon contains 4 X J, easily activating the rhodopsin (Yau et a l . , 1979). The activated rhodopsin causes hydrolysis of polyphosphatidylinositol in the plasma membrane to yield inositol triphosphate in the cytoplasm. The inositol triphosphate stimulates cGMP release, amplifying the signal. The cGMP binds to ion channels, causing their opening. In a dark-adapted cell, a single photon causes the opening of 1000 channels. The resulting depolarization stimulates neurotransmitter secretion, The neurotransmitter binds to the adjacent neuron and causes it to trigger. The initial stimulus produces a second messenger which starts an amplifying cascade to produce a larger stimulus to the transmission apparatus. Light also causes the cytoplasmic Ca” concentration in the rod cells to rise. This increase was previously thought to be part of the perception sequence. However, the release of CaZt from endoplasmic reticulum by inositol triphosphate (Payne and Fein, 1987) serves to desensitize the cell and is in fact the means of light adaptation; the gain of the cascade is attenuated at least 1000-foldfrom the dark-adapted state (Yau et al., 1986). The cytoplasmic Ca2+ concentration serves as a short-term memory of previous light levels, not as an amplificaton step of signal transduction. The transduction of sound in the ear is less well established, but is relevant because it occurs through a mechanotransducer, as must gravity perception. A modified form of the sound-detecting cells is used by animals to sense acceleration and maintain balance. In this vestibular system, a calcified sphere called a statolith or an otolith provides the mechanical stimulus in place of sound waves. In hearing, sound waves reaching the ear enter the cochlea, a remarkable fluid-filled resonance chamber. The cochlea acts as a spectrum analyser by establishing standing waves at different positions for each wavelength. This initial filtering greatly simplifies the amount of information which must be elicited from each sensing cell, since the location of the cell identifies the wavelength, and the intensity of the signal corresponds to the intensity of the sound at that wavelength. The hair cells of the cochlear epithelium have rod-like microtubule bundles which are displaced by the sound waves in the cochlear fluid. The displacement of the bundles is believed to open stretch-sensitive K+ channels in the epithelial membrane (Hudspeth, 1985). The influx of K+ causes cell depolarization, which in turn triggers neurotransmitter release into the synapse with the adjacent auditory neuron.

14

T. BJORKMAN

The cytoplasmic Ca2+concentration increases during stimulation of the hair cells, as it does in retinal rod cells. The sequence of ion movements has been described by Hudspeth (1985). Calcium ions enter from the extracellular pool through voltage-gated Ca2+ channels. These channels open during the depolarization caused by K+ influx. The Ca2+in turn open Ca-gated K+ channels to the inside of the epithelium where the K+ concentration is low, allowing the excess K+ to leave the cell and restore the membrane potential. In effect, the rise in Ca2+ serves to attenuate the effect of the stimulus. What parallels may be drawn between these sensory systems and gravity perception in plants? A plant gravity sensor perceives a mechanical stimulus as do the hair cells in both hearing and balance. The vestibular system uses the inertia of a large mass (a statolith) to detect acceleration. The stimulus is amplified to produce a large signal out of the cell. There is also a means for resetting the conditions in the cell when the stimulus ceases. The perceiving cells transmit a signal to other cells where the final response is elicited. Signalling of this type predominantly uses ion channels to create a rapid response (Methfessel and Sakmann, 1986). A difference between these sensors and plant gravity perception is that nerve cells do not have a plant analogue, so the nature of the transmitted signal will be different. Although the mechanotransducer in Dionea trigger-hair bases produces an action potential which may be considered analogous to nervous transmission, the evidence for an action potential being the elicited signal in gravity perception is weak. Also, evidence that cyclic nucleotides have a physiological role in higher plants is not strong, so amplification of the stimulus must occur by some other means. There is one clear commonality between the signals leaving gravity-perceiving cells and those discussed above, and that is polar transmembrane transport. 2. Calcium Ions in Transduction In the sensory transducers just described, Ca2+ play an ancillary role. There are also cases where the ion is central in the transduction chain. The intracellular action of Ca2+ in regulation is often mediated by Ca-binding proteins, whereby it acts as a second messenger. The best known of these proteins is calmodulin (CaM). Calmodulin, when activated by Ca2+,often regulates enzymes by stimulating their phosphorylation. In this manner, an amplifying cascade is established. Calcium+almodulin can also directly regulate ion transport proteins (Dieter, 1984). In the sensory transducers discussed above, Ca2+ attenuated the signal, but this action was not through CaM. The possible role of Ca-CaM as an amplifier in gravity perception also deserves discussion. Many enzyme reactions in plants are known to be under Ca-CaM control (Dieter, 1984). Typically, the cytosolic Ca2+concentration is below 1 PM. If

PERCEPTION OF GRAVITY BY PLANTS

15

the concentration is elevated beyond this by some stimulus, Ca2+ binds CaM. This complex can then bind other enzymes in a regulatory manner. Perception of light by phytochrome involves Ca-CaM (Roux et al., 1986). One example is the rotation of the chloroplast of Mougeotia (Wagner et al., 1984). Red light stimulates phytochrome in or near the plasma membrane, inducing a Ca2+influx from the extracellular medium. This Ca2+binds CaM, in turn stimulating a microfilament-associated protein, perhaps myosin light-chain kinase. The actomyosin then contracts differentially, according to the relative irradiance on phytochrome in the cell, to reorient the chloroplast. There is good biochemical evidence that Ca-CaM also regulates NAD kinase, NAD-quinate oxidoreductase and ion transport proteins (Dieter, 1984). The first two enzymes catalyse reactions unlikely to be part of gravity perception, though ion transport can be. The broad range of Ca-CaM-regulated reactions in animals suggests that there are many more reactions to be discovered in plants, and that these will have many essential functions. There is some evidence that CaM may be involved in gravitropism. The evidence is of two kinds: localization of CaM and effects of CaM inhibitors. Immunocytochemical labelling of CaM has identified particularly high concentrations of CaM in gravity-sensingcells of root caps and shoot apices (Roux and Dauwalder, 1985;Lin et al., 1986). In maize roots which require light induction to become positively gravitropic, the CaM activity and graviresponsivenessrise in parallel (Stinemetz and Evans, 1988). Calmodulin inhibitors block gravitropism (Bjorkman and Leopold, 1987b;Stinemetz and Evans, 1987). Furthermore, CaM inhibitors block the bioelectrical response of maize roots associated with gravity perception (Bjorkman and Leopold, 1987b). Signal amplification in perception could occur through Ca-CaM interaction. There is also evidence that gravity perception requires CaM activity. More rigorous experiments are necessary to reveal whether CaM has a direct role in signal amplification during gravity perception. 3. Phosphoinositides in Transduction Signal transduction can also involve phosphoinositides, which were discussed in relation to vision above. There is evidence for the phosphoinositide pathway being present in plants [for review see Poovaiah et al. (1987)l. The phosphoinositide pathway is generally initiated by a receptor in the plasma membrane which causes phosphatidylinositol bisphosphate to be cleaved into inositol triphosphate (IP,) and diacylglycerol by phospholipase C. The IP3 causes Ca2+ release from the endoplasmic reticulum with the consequences described in the discussion of CaM. Diacylglycerol stimulates protein kinase C, which can regulate enzymes by phosphorylation. Protein kinase C has been shown to turn on a class of ion channels which are

16

T. BJORKMAN

completely silent in the unstimulated cell (Strong et al., 1987). Through this pathway there are numerous ways to establish polar transport by a sensing cell, but each requires activation of some receptor in the plasma membrane. There are no experiments to date which have tested the involvement of phosphoinositides in gravity perception. Signal transduction is accomplished in biological systems in a number of ways. Calcium ions, which have been associated with many aspects of gravitropism, have different roles in different transduction apparatus. In the following sections, I will discuss ways gravity perception may work, using these common transduction pathways as guides.

B. MULTIPLE SYSTEMS

From an engineering standpoint the most efficient way to detect the direction of gravity is to use the displacement by an object which is attracted by gravity-something heavy. Sensors which work in this way are called statoliths. Efficiency and simplicity are two reasons why the statolith theory (Haberlandt, 1900) has been widely accepted for gravity sensing. There are, however, data which are difficult to reconcile with a statolith theory. In fungi, though gravisensing is poorly studied, it is known that the graviresponse is rapid and that there are no detectable sedimenting bodies (Burnett, 1976). The moss Physcornitrella responds to gravity, but no statoliths are apparent (Jenkins et al., 1986). In these, statoliths may have gone unobserved if they are in unexpected places. Bean seedlings provide an example where a simple observation could have failed to find statoliths because the gravisensitive region has been difficult to identify, Etiolated bean shoots perceive gravity using statoliths in the cotyledonary hook (Verbelen et al., 1985), but when de-etiolated they perceive gravity using statoliths along much of the hypocotyl (Heathcote, 1981); in both instances they curve 2-3 cm below the cotyledons. Nevertheless, gravity can apparently be perceived in the absence of statoliths. Nonstatolith gravity sensing has been proposed in higher plants also. Making amyloplasts too light to sediment by depleting the starch usually eliminates gravisensitivity (Haberlandt, 1902; Iversen, 1969), but there are exceptions. In starch-depleted, excised wheat coleoptiles a graviresponse is observable aft& 5 h (Pickard and Thimann, 1966). The lighter plastids may have taken longer to have an effect than the normal amyloplasts, which produced a response in excised coleoptiles within 2 h, or a slower alternative mechanism could have been responsible for the graviresponse. A starch-free Arabidopsis mutant (Caspar et a l . , 1985) appears to be a more interesting exception because it responds more quickly. However, changing the intensity of the gravistimulus produces a response fully consistent with the statolith theory (Sack and Kiss, 1988).

PERCEPTION OF GRAVITY BY PLANTS

17

In discussions of graviresponsive organs which apparently lack statoliths, the assumption is often made that the same gravisensing system is operating as in organs which do have statoliths (Pickard and Thimann, 1966; Moore and McClelen, 1985). This may not be a safe assumption. The presence of a second mechanism for gravitropism in addition to a statolith mechanism has been proposed (Shen-Miller and Hinchman, 1974); it would not be a novel case of a plant having more than one mechanism for getting an important task done. Gravitropism is essential for many organisms, and it would not be surprising if plants and other organisms are “overbuilt” for responding to gravity, as suggested by George Malacinski (personal communication), because failure to respond to gravity would often have lethal consequences. There may be a sensing mechanism which evolved before plastids which is potentially present in all cells. In organisms with specialized organs and diverse cellular constituents, an efficient statolith-based gravisensing system may operate. While amyloplasts act as statoliths in angiosperms, the green alga Cham uses membrane-bound barium sulphate crystals, and both invertebrates and vertebrates use crystalline organic calcium complexes as statoliths in their vestibular (balance) systems. Jellyfish contain calcium sulphate statoliths to detect their orientation relative to the gravitational field. The wide occurrence of statoliths is perhaps an example of convergent evolution. If there are parallel gravity-sensing systems present in an organism and the statolith sensor is disabled, the “primitive” system may take over. Characterizations of systems with disabled or absent statoliths have concentrated on the rate and extent of curvature, an indicator of how growth is controlled (Moore and McClelen, 1985; Caspar et al., 1985). However, there are no data which indicate whether the gravity-perceiving step has characteristics similar to those in the normal, statolith-containing plants. To learn something about the susception of gravity, the most interesting difference in apparent nonstatolith systems, susception (and perception) must be studied rather than growth. The parameters to measure would be the site of perception, sensitivity to small stimuli, the presentation time and whether the reciprocity rule (Johnson, 1965) holds. There are very few data on gravity sensing in organisms which appear never to use statoliths, so there is very little basis for guessing how they might perceive gravity. A critical survey of such mechanisms would be very welcome. Whatever the mechanism is, it must obey the physical laws discussed in this chapter. If an organism truly has no sedimenting intracellular component, the inclination would be to look for relatively large (>20 pm) structures outside the cell, or interactions between cells. If there are indeed parallel gravity-sensing mechanisms, it is important to determine to which mechanism a measured parameter applies. Data gathered from two gravity-sensing systems probably cannot be reconciled as consistent with a single mechanism.

18

T. BJORKMAN C. STATOLITH SENSORS

1. Identifying the Stutolith The most obvious candidates for sensing bodies are those which sediment. Sedimentation of amyloplasts is well documented, and has been the focus of studies on statolith sensors in higher plants. Most cellular components are unable to move freely, being secured by the cytoskeleton. Components which move by cytoplasmic streaming do so by attachment to the cytoskeleton (Wagner, 1979). In concentrating on amyloplasts, have we overlooked &her components of the cell which could act as statoliths in addition to, or instead of, amyloplasts?

Zntrurnernbrune statolith. If intercellular communication is accomplished by effecting a change in the membrane properties, the plasma membrane may be a good place to look for a sensing body. An intramembrane sensor would be a protein, most probably a transport protein capable of migrating to the bottom of the cell. If a membrane protein could act as a statolith, sedimentation could occur without microscopically detectable changes. Since proteins are more dense than lipids, they would move towards the bottom and not be buoyed to the upper side (Fig. 3a). The sedimentation rate for a protein in the membrane can be calculated from Stokes' Law, using a measured diffusion coefficient for proteins in the membrane (5 x m2 s-') (Schlessinger et ul., 1977) and a typical value for the density of a protein (1.33) and of the plasma membrane (1.03). The time for such a protein to settle half a cell diameter (5 pm) would be 88 years! Because equilibrium occurs in much less than 88 years, the sedimentation equilibrium is more appropriate to consider than sedimentation kinetics. By this method, we find the difference in concentration of a freely diffusing protein at the bottom of the cell (c) divided by the mean concentration (co):

Cleo = M ( l - @)gh/2RT = 6 x lo-'

where M = molecular weight (25 kg mol-'); i j = specific volume of prom3 kg-'); Q = density of plasma membrane (1.10 X lo3 kg tein (7 x m); m-3); g = gravity (9.8 m s - ~ ) ; h = half the height of the cell (5 X R = gas constant (8 J K-'mol-'); and T = temperature (300 K). Thus, if there were lo7 molecules of the relevant protein in the membrane there would only be one more molecule in the bottom half of the cell than in the top half at equilibrium. In comparison, a hair cell in the ear contains about 300 transducer channels (Hudspeth, 1985). Clearly gravity will not cause transport proteins to sediment in the plane of the membrane, so sedimentation of proteins will not influence gravitropism.

PERCEPTION OF GRAVITY BY PLANTS

19

Fig. 3. Hypothetical gravity susception and perception by proteins in the plasma membrane. (a) Dense transport proteins sediment to the lower part of the cell; unequal transport of the substrate creates a gradient. (b) Ion channels are pulled out of the plane of the membrane by gravity, and are active only in this position. This channel would be active only on the lower flank of the cell. Neither mechanism is energetically possible.

In addition to moving in the plane of the plasma membrane, a membrane protein can move across the membrane bilayer. A protein in a horizontal membrane would tend to be pulled out of the plane of the membrane by gravity. If that displacement were great enough to expose a catalytic site, on an ion channel for example, it could act as a gravity sensor (Fig. 3b). As described above, the force of gravity on a single protein is very small, whereas intrinsic membrane proteins are held in position by hydrophobic forces greater than gravity, the thermal motion greatly exceeds the effect of gravity, and the membrane potential also has a stronger influence on the position of the protein than gravity. Gravity therefore cannot be perceived by displacement of proteins across the plane of the membrane. Membrane components are clearly too small to be effective sedimenting bodies. To find a statolith, it will be necessary to look at larger components, inside the cell. Zntrucellulur stutoliths. To consider which intracellular components could act as statoliths, we will further consider the Stokes equation. The settling rate depends on the difference between the density of the particle and the density of the medium, the particle’s volume, and the effective viscosity of the medium. The density of amyloplasts is approximately 1.5 x lo3kg m-3; that of proteins, 1.3 X lo3kg m-3; of mitochondria, 1.2 x lo3kg m-3; and of Golgi bodies, 1.1X lo3kg m-3 (Audus, 1962).The cytoplasm is non-Newtonian, so the viscous drag is difficult to calculate. The effective viscosity and density of the cytoplasm would, however, act similarly on all organelles.

20

T. BJBRKMAN

The total work potentially done by a sedimenting particle is the force of gravity times half the cell diameter (for a 90” rotation). Because the cytoplasm must be displaced, the effective sedimenting mass is not density times volume, but density difference times volume. For any amyloplast, force = density difference x volume x gravity, = [(1.50 x lo3 kg m-’) - (1.03 x lo3 kg m-’)I x (1.4 x lo-’’ m’) x (9.8 m s-’) = 4.92 x 10-14 N Potential energy = force x distance, so where distance =

N) x energy = (4.9 x = 4.9 x 10-19 J

m,

m)

The gravitational potential energy in one sedimentable amyloplast is, therefore, 250 times the thermal noise (4kT = 2 x J), about 15 times the stimulus estimated from Fig. 2, and similar to the energy in a photon. A change in angle of 90”was used in this example but plants detect stimuli far smaller. To account for the observed sensitivity, a 90” rotation should generate a response well above the detection limit. The amount of energy involved in amyloplast sedimentation seems reasonable for its proposed role as a statolith. Mitochondria are smaller (0.5 pm) and less dense (1.2 g cm-’) than amyloplasts. They are unlikely to be sensors (Audus, 1962), losing only 1.1 x J (0.3kT) of potential energy in a 10 pm fall. Thermal agitation has a larger effect on mitochondria1motion than does gravity. If a perceiving structure were triggered by such a low energy, it would be triggered by thermal motion far more often than by gravistimulation, even with large movements. In fact, a mitochondrion would not sediment 10 pm within the presentation time: they have not been observed to sediment at all. Stokes’ law predicts that other organelles are too small or light to sediment detectably within the presentation time, and microscopy confirms this (Griffiths and Audus, 1964; Edwards and Pickard, 1987). Sedimenting amyloplasts have been the focus of most studies of the gravity sensor. In addition to the obvious sedimentation of amyloplasts, the great attention given to them is due to the presence of sedimenting amyloplasts at the site of gravisensitivity. Amyloplasts are plastids in which starch storage is the overriding function. Starch storage serves to maintain an energy source without the osmotic consequences of accumulating small molecules such as hexoses and sucrose. There are many instances of amyloplasts serving only for storage, such as in potato tuber cells and in bean parenchyma, but such amyloplasts maintain a fixed position in the cell (Verbelen et al., 1985). However, in the gravisensing cells of angiosperms,

PERCEPTION OF GRAVITY BY PLANTS

21

such as the node of grass stems, the root cap, and the gravisensing part of seedling shoots and coleoptiles, there are amyloplasts which sediment in response to gravity. If a maize root is decapped, the quiescent centre will rapidly produce amyloplasts, but gravisensitivity returns only after a change in the cytoplasm which allows them to sediment (Hillman and Wilkins, 1982). The amyloplasts in the gravisensing region are unlike amyloplasts elsewhere in that they are larger, multigranular, will not produce chlorophyll, and have more RNA (Gaynor and Galston, 1983). All gravisensing regions in higher plants normally have sedimenting amyloplasts, and amyloplasts sediment only in gravisensing cells. Rhizoids of the green alga Chara contain membrane-bound crystals of barium sulphate whose sedimentation produces curved growth (Sievers and Schroter, 1971; Schroder, 1904). These statoliths are able to move in a restricted zone in the growing tip of each rhizoid. When the rhizoid is turned from a vertical position, the statoliths sediment rapidly, and collect at the side wall. The requirements of a statolith are that it be massive enough for displacement to be detected against thermal noise and that it sediments rapidly. The barium sulphate-filled vesicles of the green alga Chara meet these requirements, and there may be other kinds in plants. However, in higher plants, the amyloplast is by far the best and apparently only reasonable candidate for a statolith.

2. Statolith Action When susception occurs by sedimentation of a large organelle, what part of sedimentation is actually perceived? There are several aspects of sedimentation which could be detected. These are the position ( x ) , displacement (dx), velocity (dxldt) and acceleration (dxldt2). The statolith balance systems in animals are optimized to detect acceleration. Plants, being stationary, do not need a bioaccelerometer to correct their travel as animals do. The other three aspects of sedimentation can be considered, however. In this section, how each of these aspects of statolith sedimentation might be the one which is perceived will be discussed. The concept of a pressure sensor is included with displacement, because a pressure transducer works by measuring the displacement of something with specified elastic behaviour. 3. Statolith Motion If statolith velocity is the parameter which is detected, perception would be due to an effect which depends on conditions varying with time. The statolith could affect the structural or the electrical conditions in the cell. Cytoskeletal shear. The change in gravitational potential energy when an amyloplast sediments was calculated above. How much of the potential

22

T. BJORKMAN

.. '.

Statolith

Mechanotransducer

Intracellular signal

Transmitter

Fig. 4. If a statolith is doing work on a transducer, it will sediment more slowly than predicted by the viscosity of the medium. The transducer could then modulate the cell's communication to other cells.

energy is available to stimulate the transducer? Some portion of the potential energy is dissipated as heat by the viscous drag of the cytoplasm, leaving the rest available to do work to trigger perception. An analogy is illustrated in Fig. 4.A ball falls through a viscous medium, and a string attached to it turns a shaft on a generator-the analogue of the transducer. The gravitational potential energy is converted to work which is converted to electrical energy in this example. The electrical energy then modulates the signal from the transmitter in proportion to the sedimentation rate. Knowing the mass and volume of the ball and the viscosity of the medium, one can calculate how fast the ball would be expected to fall if it were not attached to a transducer. If it is attached to the transducer, it will descend more slowly. Can this analogy be used to determine whether sedimenting amyloplasts may be doing work as they sediment? The observed rate of sedimentation of an amyloplast, 5-20 p m min-' (Sack et al., 1985a), reflects an apparent cytoplasmic viscosity of 350-1400 cp (water is 1 cp at 20°C) if the cell is considered analogous to a falling-ball viscometer. This viscosity is similar to that of glycerol. Is it appropriate for cytoplasm, or would an unrestrained particle fall faster? The rheological properties of the cytoplasm are believed to be determined largely by the cytoskeleton (MacLean-Fletcher and Pollard, 1980). Studies of the rheology of actin and microtubule solutions reveal a very peculiar behaviour. The shear force is essentially constant at all shear velocities; i .e. the apparent viscosity decreases in inverse proportion to the amount of stress (Buxbaum et al., 1987). An object moving rapidly through such a medium will tend to cause its viscosity to decrease, a phenomenon called shear thinning. For a fluid with dynamic viscosity characteristics like

PERCEPTION OF GRAVITY BY PLANTS

23

those of actin solutions, the behaviour of a falling ball viscometer is difficult to predict, but changes in viscosity would tend to be exaggerated (Rockwell et al., 1984). Therefore, even if the rheological nature of the cytoplasm were known, the expected sedimentation kinetics of amyloplasts could not be calculated accurately. Nevertheless, qualitative assessments can be made. Experiments on isolated cytosol with a falling-ball viscometer reveal an apparent viscosity of about 1 cp for the sol state, and >lo00 cp for the gel state (MacLean-Fletcher and Pollard, 1980). Since the viscometer used a metal ball of different size and density from an amyloplast, the shear forces are different and therefore the viscosity values are not quantitatively comparable. Nevertheless, these measurements indicate that amyloplasts in vivo sediment more slowly than isolated amyloplasts would in a sol state cytosol in vitro. The possibility remains that amyloplasts are restrained, and may thereby modulate the perceiving mechanism. An interesting observation is that the amyloplasts slow down as they fall. In corn root statocytes, the initial rate is 19 pm min-', but after moving about 2 pm, they have slowed to 4 p m min-' (Sack et al., 1985a). Shear thinning of the cytoplasm by Brownian motion of the amyloplast may cause the cytoplasm to have a lower effective viscosity in the region just around the amyloplast. The amyloplast would sediment faster in this thinned region, and slow down when it leaves this region (Fig. 5a). This may in part be the explanation for the observed decrease in the sedimentation rate during the early part of an amyloplast's fall. A region of thinned cytoplasm around the amyloplasts could indicate the gravity vector if it perturbed cell metabolism, which is dependent on the cytoskeletal structure. If this is the case, statolith motion in response to gravistimulation is not detected. Rather, Brownian motion would be used to create a signal which indicates the statolith position. Alternatively, the obscured slowing could be explained if amyloplasts, like chloroplasts (Witztum and Parthasarathy, 1985), are directly attached to the cytoskeletal matrix. During the initial part of the fall, the amyloplast would not be restrained, but after a short distance, the cytoskeletal elements would become taut and slow the descent of the amyloplast (Fig. 5b). Through this restraint, the motion of the amyloplasts could modulate a signal through localized perturbations as discussed above. Still another way to account for the slowing of the amyloplasts is compression of the cytoskeletal matrix as the amyloplasts settle on it (Fig. 5c). In Chara statoliths are held in position away from the tip of the rhizoid by a cytoskeletal matrix. If the rhizoid is treated with colcemid, the statoliths descend to the extreme apex (Friedrich and Hertel, 1973). If such a compression is involved in perception, displacement rather than motion is detected; this mechanism is discussed later.

24

T. BJORKMAN t=O

t =5s

Fig. 5. Models which could explain the slowing of sedimenting amyloplasts. The left column is immediately after gravistimulation, the right column is about five seconds later. (a) Shear caused by Brownian motion of the amyloplasts will reduce the effective viscosity of microfilaments in the cytoplasm. They will therefore sediment more rapidly through this layer than through the remaining matrix. (b) Hypothesized cytoskeletal connections to amyloplasts will become taut after some displacement. (c) Cytoskeletal elements may be compressed below the sedimenting amyloplast, and increase the effective viscosity.

It is difficult to make more than general suggestions about the significance of amyloplast sedimentation kinetics based on available data. These models could be tested by measuring the sedimentation kinetics when the statocytes are reinverted. A comparison of these kinetics with sedimentation kinetics of isolated amyloplasts through appropriate actin suspensions in vitro would make it easier to assess the likelihood of connections between amyloplasts and the cytoskeleton in detecting amyloplast sedimentation.

25

PERCEPTION OF GRAVITY BY PLANTS

Electricalfield. An amyloplast carries a charge (Sack and Leopold, 1982), so one can envisage its motion being detected from the changing electrical field. An electrical generator works through electrical and magnetic fields moving past a coil. Bandurski et al. (1985) proposed that the amyloplasts deform the electrical field around plasmadesmatal openings and thereby open these intercellular connections, if they are voltage-sensitive. The distortion of the field around the amyloplast would only extend as far as the Debye layer-about 2 nm (Starzak, 1984); even with electron microscopy, this distance is indistinguishable from actual contact, which does not occur (Perbal, 1978; Sack and Leopold, 1985). Also, the potential energy in the charge separation caused by the negatively charged amyloplast moving in the cytoplasm is much less than the potential energy due to gravity. This would be inefficient conversion of the potential energy to work in an instance where almost all the energy is required for sensing to occur within the presentation time. If none of the potential energy of the amyloplast is converted to work as it sediments, then the sensing must occur when the amyloplast contacts the bottom flank of the cell. The energy available to do work on a sensor by this scheme is the kinetic energy in the moving amyloplast. That energy is also easily calculated using some of the values cited above: kinetic energy depends on the mass and the velocity. If mass = 2 x

kg

and velocity = 20 pm min-' = 3.3 x 1 0 - ~ m

s-l

then kinetic energy = 1/2 x mass x (velocity)' = (0.5) x (2 x kg) x (3.3 x = 1.1 x 10-27 J

m s-')

The kinetic energy in a moving amyloplast is four million times less than the thermal noise. This comparison shows that amyloplasts sediment so slowly that there is simply not enough kinetic energy for the motion of an amyloplast to be detected. Although physiological models to detect statolith motion can be described, energetic evaluations rule them out. Perception based on statolith motion would also fail to explain the continued differential growth after sedimentation is complete until the tissue is again in its preferred orientation. Perception must come about through detection of some parameter of sedimentation other than motion of statoliths.

26

T. BJORKMAN

4. Statolith displacement An obvious event in gravity-sensing cells is the movement of statoliths from one position in the cell towards another when the tissue is displaced. The indicator of orientation in the gravitational field could be the change in statolith position. Specifically, the distance statoliths move from their normal (e.g. tissue vertical) position, or the amount some structure is displaced by statoliths, could be the graded stimulus which elicits a response. A statolith exerts a force on any structure with which it comes into contact. The pressure is detected by displacement of an elastic structure which responds in proportion to the amount of displacement. A pressure sensor is therefore a special case of displacement perception. The effect of thermal noise on a receptor can be reduced if the receptor averages the input over a period of time. If statolith sedimentation does work on a receptor by displacing it, the displacement required for triggering should be high enough that the triggering by gravity occurs much more frequently than triggering by thermal agitation of the statolith. This can be accomplished by averaging the signal over time. Although the probability of a large thermal displacement increases as the observation period increases, the effect of gravity increases more during statolith sedimentation. This can

Fig. 6. The effect of integration time on the ability to distinguish the effect of gravity on a statolith from the effect of thermal noise. The left axis is statolith motion due to each force. The right axis is the amount of work done by the corresponding displacement of an amyloplast. Indicated along the right axis is the work required for activation of sensors on vision [rhodopsin (Yau ef al., 1979)] and hearing [hair cell (Corey and Hudspeth, 1983)] as well as thermal noise ( M )Values . used in this graph were estimated from data in Sack etal. (1985a): r] = 300 cp = 0.3 Pa, r = 1 pm, sedimentation rate = 4 p m min-'. The uncertainty of each line is about three-fold due to imprecision in the estimation of one of the parameters and assumptions about the behaviour of the cytoplasm.

PERCEPTION OF GRAVITY BY PLANTS

27

be calculated from Einstein’s equation of Brownian motion (Einstein, 1907):

a

where is the net distance moved, z is the integration time, 7 is the viscosity of the medium and r the radius of the particle. The relative effects of sedimentation due to gravity and random thermal motion can be seen in Fig. 6. An integration of several seconds is needed for sedimentation to be the dominant source of the signal, and for work done by sedimentation to be equivalent to that which triggers other sensitive sensors. Evidence of such integration is that gravistimulation need not be imposed continuously. If short intermittent stimulations are frequent enough, they have the same effect as the same amount of stimulation given continuously (Pickard, 1973b). One-second stimulations every five seconds are summed; halfsecond stimulations must be repeated every second to be summed. Significantly, the smaller stimuli are “remembered” for a shorter time. The perception mechanism probably averages stimuli over a time period of one or a few seconds, and the minimum stimulus involves a displacement of at least 100 nm.

Cytoskeleton stretching. The lack of contact between amyloplasts and the plasma membrane (Witztum and Parthasarathy, 1985; Heathcote, 1981) suggests that an indirect interaction causes a signal to be passed out of the cell. This indirect interaction could be through cytoskeletal members attached to amyloplasts transmitting, by tension, energy to receptors in a membrane. In the extreme case, where the amyloplasts are completely restrained, they would do no work at all. Amyloplasts are often restrained to this extent in cells which do not function as gravity sensors. The cytoskeleton is certainly a promising agent for transmitting the stimulus, but not by immobilizing the amyloplasts, although they would be partially restrained. This concept has been raised by Larsen (1969) who proposed that the amyloplasts function as pendula attached to the distal part of the statocyte by the cytoskeleton. Stimulus transmission by the cytoskeleton has also been proposed by Shen-Miller and Hinchman (1974) and by Friedrich and Hertel (1973). If the cytoskeleton is attached to a stretch-sensitive Ca2+channel in the amyloplast envelope, displacement of amyloplasts could cause tension in the stationary cytoskeleton. There is a precedent for microfilament-plastid interaction (Witztum and Pathasarathy, 1985). The tension could cause a conformational change in the channel to which the cytoskeleton is attached, analogous to the way the membrane potential causes a conformational change in voltage-gated channels, by exerting a force on dipoles in the protein. This conformational change decreases the amount of energy needed

28

T. BJORKMAN

to open the channel (Honig et al., 1986). Amyloplasts contain large amounts of Ca2+(Chandra et al., 1982). More frequent opening of a Ca2+channel in the amyloplast membrane would cause a locally elevated Ca2+concentration. This region of higher Ca2+would be a directional signal when amyloplasts are only in the lower part of the cell. A Ca2+channel in the amyloplast membrane, attached to microfilaments anchored at the end of the cell where the amyloplasts are in vertical tissues, would cause a rise in Ca2+in the lower part of the cell. A localized release of Ca2+would result in an elevated Ca2+ concentration only in a restricted region of the cytoplasm (Keith et al., 1985; Brownlee and Wood, 1986; Weir et al., 1987). Should such channels be at the other end of the cytoskeletal connection, in the plasma membrane, sedimenting amyloplasts would not create a directional signal. In that case the local rise in intracellular Ca2+ would occur in the same location in the cell regardless of the direction of stimulation. A locally higher Ca" concentration around displaced amyloplasts could stimulate ion transport across the plasma membrane in that region of the cell. Elevated cytoplasmic Ca2+ may stimulate ion transport proteins directly or through CaM (Dieter, 1984). The directional ion transport is the type of signal which the perceiving cell would be expected to elicit. If sedimenting amyloplasts do work on an ion channel via microfilaments or microtubules, they can elicit a physiological asymmetry by stimulating channels in only one part of the cell. The operation of this type of microfilament-membrane channel interaction has been found (Horwitz et al., 1986) and is likely to be common (Geiger, 1985). Lawton et al. (1986) noted disruption of microtubules around amyloplasts at the onset of sedimentation. The interaction of the cytoskeleton with channels is a promising candidate for perceiving mechanical stimuli.

Displacement of endoplasmic reticulum. Volkmann and Sievers (1979) have proposed that perception is by the interaction of amyloplasts with the endoplasmic reticulum. In that case, amyloplasts would begin to act only after reaching the endoplasmic reticulum at the lower surface of the cell. The presentation time could then be interpreted as including the time required for the amyloplasts to reach a position where they had an effect. Amyloplasts sediment rapidly, as fast as 40 p m min-' in Taraxacum stalks (Clifford and Barclay, 1980). Sack et al. (1985a,b) measured both the presentation time and the sedimentation of amyloplasts in two different kinds of statocytes-the root cap and the coleoptile of Zea mays. In both cases, the amyloplasts sedimented rapidly enough that the first ones had reached the new lower flank of the cell within the presentation time. Sedimentation of amyloplasts thus occurs in an appropriate time period for perception to occur through interaction with the endoplasmic reticulum. The motion of amyloplasts at the new lower flank of the cell also bears

PERCEPTION OF GRAVITY BY PLANTS

29

on this possibility. Heathcote (1981) observed that amyloplasts in Phaseolus hypocotyls slow down when they are about 1 p m from the lower wall of the cell. The slowing may be caused by amyloplasts deforming the endoplasmic reticulum which underlies the plasma membrane. Heathcote makes the comment that during sedimentation, the amyloplasts “appeared to be slowed by invisible cytoplasmic structures”. Observations of live tissue, like these and also those of Sack (Sack et al., 1985a,b; Sack and Leopold, 1985), are very helpful when considering the interactions of statoliths with the perception mechanism. Volkmann’s observation that graviresponse is proportional to pressure prompted him to propose that the endoplasmic reticulum senses pressure directly (Volkmann, 1974). If the endoplasmic reticulum is deformed elastically, Hooke’s law holds that displacement will be proportional to pressure. Hence Volkmann’s data support displacement detection equally well. Sievers et al. (1984) propose that the signal which is elicited when amyloplasts settle on the endoplasmic reticulum is intracellular Ca2+ released from the endoplasmic reticulum, a storage site for CaZf in the cell. In their model, statocytes depolarize the cell as a result of Ca2+ release when amyloplasts deform the distal beds of endoplasmic reticulum. The proposed involvement of a specific endoplasmic reticulum structure is supported by the observation that a pea mutant in which the endoplasmic reticulum is uniformly distributed is not graviresponsive (Olsen and Iversen, 1980). The endoplasmic reticulum is arranged so that maximum amyloplast contact is with beds on the outer side of the cell (Juniper and French, 1970), explaining the depolarization only of cells on the lower side of the tissue (Behrens et al., 1982). The product of this Caz+-induced depolarization of the lower cells is an electrical asymmetry across the root cap. The initial perceiving step remains difficult to explain, namely how amyloplasts cause Ca2+ release from the endoplasmic reticulum. Amyloplasts deform the membrane extensively, bringing the endoplasmic reticulum cisternae into contact (Volkmann and Sievers, 1979). Perhaps this intermembrane interaction can induce Ca2+ release. The question of how amyloplast action elicits a molecular response is unanswered for this specific model, but the model provides a good tool to answer this question.

5. Statolith Position Finally, the position of a statolith as the detected aspect of sedimentation will be considered. The argument that statolith position can be detected seems to go against the preceding discussion of thermal noise and of work done during sedimentation, but those restrictions still apply though in a somewhat different way. To detect the position of a statolith there would be a sensitive area on the lower surface of the cell which recognizes the statoliths on some basis other than mass. The statolith would still need to be

30

T. BJORKMAN

massive enough to sediment to the bottom surface without being extensively agitated by Brownian motion. Also, a high-affinity recognition site could use electrical or chemical potential energy rather than gravitational for the recognition (Volkmann, 1974), but the gravitational force would have to be large enough to break the attraction when the tissue is reoriented. The specific recognition site would be near the outer flank of the cell. Although there is no contact between the amyloplast and the plasma membrane (Perbal, 1978; Heathcote, 1981), there are several structures just inside the plasma membrane with which the amyloplast may interact: one or more layers of endoplasmic reticulum (Juniper and French, 1970; Volkmann, 1974; Sievers and Hensel, 1982) held in place by microfilaments (Hensel, 1984,1985, 1986b); highly stable cortical microtubules (Kakimoto and Shibaoka, 1986); a meshwork of microfilaments (Parthasarathy, 1985) not involved with cytoplasmic streaming (Derksen et al., 1986); desmotubules which extend into the cytoplasm. There are many structures of importance with which statoliths may interact, but it is unclear which are involved in gravity perception. Electrostatic attraction. If electrical potential energy is used to trigger perception, it could be through electrostatic attraction between the charged amyloplast membrane (Sack and Leopold, 1982) and charged sites on the sensitive surface. The electrical field created by the surface charge of a membrane decays rapidly away from the membrane surface, being negligible more than 1 or 2 nm from the surface (Starzak, 1984). An electrostatic interaction is not a long-distance one on a cellular scale, requiring an approach microscopically indistinguishable from contact. Without postulating characteristics of the binding site, it is impossible to calculate whether the energetics of electrostatic binding are reasonable, and there is not enough information for productive speculation about the properties of such a binding site. One generalization which can be applied is that an electrostatic interaction must be strong enough to elicit a reaction, but weak enough to allow the gravitational force on the statolith to move it away if the tissue is reoriented. There is evidence for electrostatic effects of gravistimulation. In maize roots, the surface charge of the plasma membrane changes (Pilet, 1985). This change may affect the transport properties of the charged membrane through screening of substrates and through electrostatic effects on transport proteins within the membrane (Mdler and Lundborg, 1985). On the other hand, the plasma membrane surface charge could be altered by changes in the extracellular Ca2+ activity (Moller and Lundborg, 1985), which also occurs on gravistimulation (Lee et al., 1983). Ligand binding. The location of statoliths in a sensing cell may also be detected by releasing chemical potential energy if exothermic binding occurs. The energy to trigger a physiological change would be released by a

PERCEPTION OF GRAVITY BY PLANTS

31

reaction such as ligand binding. As a simplistic example, if a ligand on the amyloplast envelope binds to a receptor on the endoplasmic reticulum, an associated ion channel could be caused to open. Cytological evidence suggests that amyloplasts approach the endoplasmic reticulum more closely than they do the plasma membrane (Perbal, 1978). Ligand-activated channels are a common type (Hille, 1984); though ligands are usually small molecules which diffuse quickly in the cytoplasm, none have been found attached to a large organelle. One difficulty in using ligand binding as a perception mechanism in gravity sensing, as with electrostatic binding, is that the binding energy would make it hard for the amyloplasts to come loose again. If the binding is exothermic and the contribution of gravity is energetically negligible, the force of gravity on the amyloplast will be insufficient to release it from the binding site. A direct role of statolith position in altering growth to produce curvature has been proposed in Chum rhizoids (Sievers and Schroter, 1971). An important distinction must be made between this alga and angiosperms, however. Sensing and response occur in the same cell, with the statoliths sedimenting to the exact position in the cell where active growth is occurring. Thus no transmission step is necessary, and perception is presumably simpler. Sievers and Schroter (1971) suggested that the role of the statoliths is to prevent Golgi vesicles from fusing with the plasma membrane and thereby preventing wall growth and membrane expansion in that region of the cell. This contention is strengthened by the observation that growth at the cell apex of vertical roots stops if the statoliths are caused to settle there due to reduced turgor (Sievers and Schroter, 1971), or by treatment with the anti-microtubule agent colcemid (Friedrich and Hertel, 1973). This very straightforward action of statolith position works well in a tip-growing cell, but is difficult to apply to multicellular responses. A statolith acting in this manner must be large in order to remain at the lower flank of the cell despite thermal agitation. Specifically, as described in Section 11, it must take about 3x J to move the statolith away in order for thermal displacement to be insignificant, yet for a change in the gravity vector to move the statolith. Specific chemical or electrostatic interactions are intriguing possibilities. A combination of these attractions with displacement due to gravity would form the basis of such an interaction. There are very few data which are useful in evaluating amyloplast binding, although those which suggest that amyloplast position is important support the possibility indirectly. Based on microscopic evidence, any such direct interaction would most likely be with endoplasmic reticulum or cytoskeletal elements, rather than the plasma membrane or its components. The identity of the relevant statolith action can thus be limited somewhat. The kinetic energy of statolith motion is too small to be perceived. Statolith position appears to act directly in the case of Cham rhizoids. In multicellular tissues, displacement is much more likely. The total displacement of a

32

T. BJORKMAN

statolith on gravistimulation may not be relevant; rather, it may be the displacement of some transducer, occurring only during a part of sedimentation. The relevant physical displacement may appear as if statolith position were detected, as in the model in Lepidium root caps involving endoplasmic reticulum beds. The larger response to a sliding action across a cell flank than to just sedimentation to it (Iversen and Larsen, 1971) implies that a signal is elicited by deformation of a structure on the lower cell flank. Just as there are several types of biological statoliths, there may be more than one way to perceive their action. For multicellular plants, gravity perception by statoliths is likely to be through the work done by statolith displacement.

D. NONSTATOLITH PERCEPTION

The preceding discussion of statolith action involves fairly straightforward principles and can refer to a large body of published work because the statolith theory has dominated research in gravitropism. Gravity sensing without statoliths would necessarily be more subtle. There is nevertheless strong evidence that it does occur and the question of how must be addressed. The ability of organisms to detect subtle signals easily exceeds our ability to explain it. A remarkable example is the marine mollusc Tritionia which, without any ferromagnetic particles, detects not only the earth’s magnetic field but also the phase of the moon, from the bottom of Puget Sound (Lohmann and Willows, 1987)! 1 . Pressure Differential As an alternative to a sedimenting body, Pickard and Thimann (1966) have proposed that the weight of a cell’s protoplasm stimulates the sensor by exerting pressure on the side of the cell towards gravity. The rationale for this hypothesis is that the protoplasm is more massive than any substituent of the cell and therefore can exert more force on a sensor. A discussion of this mechanism may be found in Audus (1979). The resultant change in pressure across the cell plasma membrane and wall, higher on the lower side, and lower on the upper side, would be instantaneous. This pressure difference can be calculated; it is the density of the cytoplasm times gravity times the diameter of the cell. Thus, for a cell 10 p m in diameter, the pressure will be 0.1 Pa higher at the bottom than at the top. For a large stimulation (90’ rotation), the pressure change at the new lower side would be 5 X Pa. Plants respond to stimuli that are at least 100 times smaller. The pressure change would have to be detected against the background turgor pressure, which is typically 5-15 X 10’ Pa. To complicate matters further, the turgor pressure is not static, but is constantly changing with the evaporative demand on the plant and, to a smaller extent, with fluctuations in solute exchange in and out of the cell. In pea stems held in a uniform, humid

PERCEPTION OF GRAVITY BY PLANTS

33

environment, the turgor pressure varied over a range of 5 X lo4 Pa in the period of 1 min (Cosgrove and Steudle, 1981; Cosgrove and Cleland, 1983). To detect a change in pressure due to gravity against a static background pressure at least ten million times larger, and a rapid fluctuation in that pressure one million times larger, would require an amplifier which would selectively amplify the signal to overcome the noise with only a few seconds of sampling time. Such an amplifier would be unprecedented in both biology and engineering. 2. Membrane Tension The differential volume change resulting from changes in the osmotic potential inside or outside the cell result in changes in the plasma membrane tension (tangential vector). Because the elastic modulus of the membrane is high, the tension is much more sensitive than the turgor to changes in volume. For reasons very similar to those responsible for the difference in protoplasmic pressure, gravity would cause a difference in membrane tension between the upper and lower surface of the cell. Is there a way this difference in tension could have metabolic consequences leading to cell polarization? Guharay and Sachs (1984) have discovered a tension-sensitive ion channel which would respond to the tension changes resulting from a change in volume. It represents an attractive candidate for the elusive turgor-sensing mechanism. A direct effect of pressure has been proposed as a turgor pressure sensor (Coster and Zimmermann, 1976), which might be relevant to gravity sensing, but the direct effect of pressure on membrane permeability is relevant only at pressures of lo8 to lo9 Pa (Aldridge and Bruner, 1985). A change in turgor always involves a change in cell surface area because cell walls are elastic. This surface area change causes large changes in the membrane tension (Wolfe and Steponkus, 1981) which, through a tension-sensitive channel, may be the basis for turgor sensing. Although the cell surface area would not change with reorientation in a gravitational field, there would be a differential membrane tension between the top and bottom of the cell. This tension differential has been proposed as a gravity sensor (Edwards and Pickard, 1987). Could the energy from the difference in tension be sufficient to overcome the activation energy for changing the state of the channel? The tension-sensitive ion channel in patch-clamped tobacco protoplast membranes (Falke et al., 1986) opens at a pressure difference of 2500 Pa. If the radius of the patch pipette is 0.5 pm and there is one channel per patch, then the tension for opening and activation energy can be calculated: Tension = 1/2(pressure difference) X (radius of curvature) = 1/2(2.5 x lo3Pa)(5 x lo-’ m) = 625 p N m-l

34

T. BJORKMAN

Energy = (tension) = (6.25 x =

X

(area of patch) N m-') x 1/2(4 n)(5x

m)'

J

The activation energy for channel opening is about lo-'' J per channel, occurring at a tension of 625 pN m-'. For comparison, the resting tension of a plant cell membrane is about 100 p N m-l and the critical tension for lysis is 4000 p N m-' (Wolfe and Steponkus, 1981). The difference in tension due to gravity, based on a turgor difference of 0.05 Pa and a cell radius of 5 p m , is about 0.25 pN m-'. Even if the energy in the differential tension caused by gravity across a whole cell's membrane (<3 x J) could be focused on a single channel, it would not overcome the activation energy for opening (10-15 J). This particular channel would not be useful as a gravity sensor, but could a more sensitive channel work, e.g. a channel with an activation energy around 3 x J? A channel with this higher sensitivity, at a density of 100 per cell, would be activated at 0.002 pN m-'. Such a channel would be constantly activated by the resting tension of the membrane (>lo0 pN m-'). A tension-sensing channel is only effective if it is sensitive to changes relative to the resting tension, so the minimum tension for opening must be greater than the resting tension. A very specialized channel could conceivably open over a narrow range, e.g. between 100 and 100.25 pN m-'. However, normal fluctuations in the resting tension are much larger than this (Wolfe and Steponkus, 1981), so the channel would not really be capable of detecting changes due to gravity. A channel which opens at a tension meaningfully higher than the resting tension would require more energy for activation than that which gravity provides to the cell. There is no design by which a change in the membrane tension due to gravity can be detected by a tension-sensitive channel. There are other difficulties with perceiving gravity using membrane tension. The cell tension is constantly being adjusted through addition and removal of membrane material, so that small differences such as that due to gravity would be difficult to distinguish. Further, while it is clear how the absolute tension of a structural part of the cell can transmit energy to a channel, there is no obvious way to interpret a differential tension in a similar way. In walled cells, the radius of curvature of the membrane is much smaller than in protoplasts because it is appressed to the microfibril meshwork of the wall. Therefore the tension in the membranes of turgid walled cells is much higher than described above. If tension sensors were connected to the wall (Edwards and Pickard, 1987), the tension would be higher than that calculated for the membrane, but the differential tension would be the same. Since wall tension is directly related to turgor, the arguments regarding turgor in the previous section apply to this model as well. Detection of

PERCEPTION OF GRAVITY BY PLANTS

35

changes in wall tension with a stretch-sensitive channel is even more difficult than detection of changes in membrane tension. The mechanotransducer of Dionea appears to detect pressure on sensing cells from the trigger hair, but the pressure change is much larger, being produced by deflection of a multicellular trigger hair which acts as a lever. These pressure changes would be of the same order as detectable turgor changes, so a turgor sensor based on membrane or wall tension is a physically reasonable mechanotransducer in Dionea. Moore et al. (1984) imply that there is a nonstatolith sensor in decapped corn roots which keeps them from regenerating caps in space. There are two explanations which have not been tested. First, without a 1 X g control in space, the effects of radiation cannot be determined. That these effects are likely to be important is illustrated by the effect of gravity on space-grown lettuce hypocotyls. Relative to ground controls, the growth rate of seedlings in space was inhibited by 30% whether at microgravity or at 1 x g (Merkis et al., 1985). Thus some condition in space other than gravity depressed the growth rate. Differences between earth- and spacegrown material are not necessarily due to gravity. Cell division-which is initiated in quiescent zone cells during cap regeneration-appears to be particularly sensitive to radiation in space vehicles (Kostina et al., 1984). Second, the mitotic apparatus may be large enough to act as a statolith. This does not appear to have been tested. In light of the many reasons why root cap regeneration might fail in spaceflight, it appears unnecessary in this case to invoke a nonstatolith gravity sensor. The basic assumption in the above two schemes involving the mass of the protoplasm is that the mass of the whole cell is greater than any individual part, and can therefore be the source of more energy than any individual part. These schemes then depend on focusing that energy to do work on one or a few sites. Both schemes are unreasonable because, based on the above calculations, energy depends both on mass and displacement. While the whole cell is a larger mass, it is not displaced relative to itself; no work is done, so there is no energy to focus on a receptor.

3. Multicellular Sensors A nonstatolith gravity sensor would not have to be intracellular. The entire organ could be the sensor of gravity, with the weight of the organ compressing the lower tissue and stretching the upper tissue. This effect would be apparent immediately upon gravistimulation, and would provide a large force. Sliwinski and Salisbury (1985) investigated this possibility by restraining the shoot so that it was curved upward, thus producing compression on the upper surface whereas the stem would normally be compressed on the lower side. Despite this reversal of tension and compression, the shoot curved normally. Thus this mechanism has been rejected experimentally for a higher plant shoot.

36

T. BJORKMAN

A gravisensing root is normally supported by the soil matrix, and does not experience stresses similar to those of a horizontal plant part in air. Pressure of the tissue against the matrix below it is not the signal to roots, since roots curve similarly in soil and in humid air. Also, by bending the terminal 1-2 mm of root tips (the cap but not the elongating zone) in glass tubes, the two parts could be gravistimulated separately. Subsequent curved growth in these roots occurred if the root cap was horizontal and the elongating zone vertical, but not if the cap was vertical and the elongating zone horizontal (Pfeffer , 1894). This experiment elegantly excludes stresses on the whole tissue in roots. When discussing a multicellular gravity sensor it is important to distinguish the gravitropic reaction from reaction wood formation which is indirectly a response to gravity. Reaction wood formation appears to be associated with responses like thigmotropism and seismonasty, where mechanical stress causes altered growth mediated by ethylene formation. It appears that an extracellular gravity sensor based on tissue stresses does not exist in higher plants. However, there appear to be no experiments in fungi or mosses, where there are no obvious statoliths, which test such a hypothesis. In large plants, the force of gravity produces a thigmic stress, such as that produced by wind or contact. The response to thigmic stresses occurs by a different mechanism. The physical constraints on how a multicellular gravity sensor would operate are great, and make it difficult to design a hypothetical model of such a system.

V. INTEGRATION AND CONCLUSION There has been much speculation about what the gravity sensor may be; speculation will continue until a definitive causal role for sedimenting statoliths (amyloplasts in higher plants) has been proven or another scheme is found which requires no statoliths. Much of this speculation has ignored fundamental rules of the physical world: (1) gravity can only interact with a mass; (2) a mass must move to do work on a sensor; (3) the energy of the motion must be detectable in the presence of thermal noise. In a system as small as a cell the energy of thermal noise dominates; in a system as large as a human being, thermal noise is practically unnoticeable. Therefore, our conception of what can be detected inside a single cell could easily be wrong. A mass which senses gravity must be large on a cellular scale, and it must move observable distances under the influence of gravity, if it is to affect the physiology of a gravisensing cell. There are two lower limits to activation of perception-thermal noise and energy of activation. Susception must produce a signal large enough to trigger perception. To prevent frequent spontaneous triggering, the necessary stimulus must be several times greater than thermal energy. Specifically,

PERCEPTION OF GRAVITY BY PLANTS

37

to meet these requirements gravity perception will have an activation energy of about 4 x J. There is compelling evidence for the existence of both statolith and nonstatolith gravity perception. It is not known whether these can be simultaneously operating in one graviresponsive issue. Models of perception without statoliths have yet to satisfy the basic physical requirements of signal transduction. The scant information available to date is not sufficient to create relevant models. The interaction of a statolith with the mechanism which perceives sedimentation must use the potential energy which is released during sedimentation. The kinetic energy of a moving amyloplast is too small to be detected; the work done by statolith displacement is the most reasonable interaction. This work is most likely done by deforming a cellular structure during some part of sedimentation, resulting in a physiological response. The integration of the gravitational stimulus through discontinuous stimulation experiments (Pickard, 1973b) can be correlated with intracellular events to test hypotheses about the initial physiological steps of gravity perception. One such hypothesis is that membrane depolarization is the integrator as in retinal rod cells. The cytoplasmic Ca2+ concentration can also act as an integrator if statolith sedimentation produces a graded Ca2+ release. The mechanism of stimulus integration appears to be an accessible step of perception to study. For progress in the study of gravity perception, it is essential to be able to study individual steps in gravitropism. Although it is unlikely that steps of gravity sensing can be isolated and studied in vitro, it must be possible to detect the function of individual steps. Of particular interest would be to correlate a signal intensity with details of statolith behaviour. It would then be possible to test hypotheses which predict that specific interactions between statoliths and other cellular components give rise t o perception. To test models of gravity perception invoking elevated cytoplasmic Ca2+ concentrations as an intracellular signal, it is imperative to determine whether the Ca2+concentration really changes within the cell during gravistimulation, and, if so, where in the cell. The role of Ca2+,phosphoinositides and other second messengers in other examples of signal transduction has been assessed by microinjecting the substance into the sensing cell, and measuring the elicited signal. This technique could easily be applied to gravity perception if the elicited signal could be measured. Curvature is a poor parameter to use because it is a highly variable, delayed response which requires optimum function of many unknown intermediate steps. The various models which include release of Ca2+ into the cytoplasm invoke different sources for that Ca2+: extracellular, endoplasmic reticulum and plastid. If cytoplasmic Ca2+does increase with gravistimulation, it would be very useful to find the immediate source of that Ca2+. Finally, a robust study of sensitivity and noise discrimination has yielded

38

T. BJORKMAN

a groundwork for evaluating susception in hearing. An equivalent study of graviperception could restrict the possible mechanisms to a few, and would put to rest much of the controversy about the statolith hypothesis. In this paper, the author has been able to impose only very wide limits using available data.

REFERENCES Aldridge, B. B. and Bruner, L. J. (1985). Biochim. Biophys. Acta 817,343-354. Audus, L. J. (1962). Symp. SOC.Exp. Biol. 16, 197-226. Audus, L. J. (1979). J. EXP.Bot. 30,1051-1073. Bandurski, R. S., Schulze, A. and Momonoki, Y. (1984). Physiologist 27, S123S126. Bandurski, R. S . , Schulze, A. and Reinecke, D. (1985). Physiologist28, Slll-S112. Behrens, H. M. and Gradrnann, D. (1985). Plants 163,453-462. Behrens, H. M., Weisenseel M. H. and Sievers, A. (1982). Plant Physiol. 70, 1079-1083. Behrens, H. M., Gradrnann, D. and Sievers, A. (1985). Planta 163,463-472. Benolken, R. M. and Jacobson, S. L. (1970). J. Gen. Physiol. 56,6489. Berry, L. J. and Hoyt, R. C. (1943). Plant Physiol. 18,372-396. Bjorkrnan, T. (1987). Ph.D. Thesis, pp. 56-57. Cornell University, Ithaca, NY. Bjorkrnan, T. and Leopold, A. C. (1987a). Plant Physiol. 84,841-846. Bjorkrnan, T. and Leopold, A. C. (1987b). Plant Physiol. 84,847-850. Brownlee, C. and Wood, J. W. (1986). Nature 320,624-626. Burdon-Sanderson, J. (1873). Proc. R . SOC.Lond. 21,495496. Burnett, J. H. (1976). “Fundamentals of Mycology”, p. 398. Edward Arnold, London. Buxbaum, R. E . , Dennerll, T., Weiss, S . and Heidemann, S. R. (1987). Science235, 1511-1514. Caspar, T., Sornerville, C. and Pickard, B. G. (1985). Plant Physiol. 77, S105. Chandra, S . , Chabot, J. F., Morrison, G. F. and Leopold, A. C. (1982). Science 216, 1221-1223. Cleland, R. E. and Rayle, D. L. (1977). Plant Physiol. 60,709-712. Clifford, P. E. and Barclay, G. F. (1980). Plant Cell Environ. 3,381-386. Corey, D. P. and Hudspeth, A. J. (1983). J. Neurosci. 3,962-976. Cosgrove, D. J. and Cleland, R. E. (1983). Plant Physiol. 72,332-338. Cosgrove, D. and Steudle, E. (1981). Planta 153,343-350. Coster, H. G. L. and Zirnmerrnann, U. (1976). 2. Naturforsch. 31C, 461-463. Darwin, F. (1899). Ann. Bot. 13,567-574. Derksen, J., Trass, J. A. and Oostendorp, T. (1986). Plant Sci. 43,77-81. Dieter, P. (1984). Plant Cell Environ. 7,371-380. Dolk, H. E. (1933). Rec. Trav. Bot. Neerl. 33,509-585. Edwards, K. L. and Pickard, B. G. (1987). In “NATO Advanced Research Workshop Proceedings: The Cell Surface in Signal Transduction” (H. Greppin, B. Millet and E. Wagner, eds). Plenum/Springer, Heidelberg. Einstein, A. (1907). 2. Elektrochem. 13,4142. Evers, J. and Lund, E. J. (1947). In “Bioelectric Fields and Growth” (E. J. Lund, ed.), pp. 197-217. University of Texas Press, Austin. Falke, L., Edwards, K . L., Misler, S. and Pickard, B. G. (1986). Plant Physiol. 80, s9.

PERCEPTION OF GRAVITY BY PLANTS

39

Friedrich, U. and Hertel, R. (1973). 2.Pflanzenphysiol. 70,173-184. Gaynor, J. J. and Galston, A. W. (1983). Plant Cell Physiol. 24,411-421. Geiger, B. (1985). Trends Biochem. Sci. 10,456461. Grahm, L. (1964). Physiol. Plant. 17,231-261. Grahm, L. and Hertz, C. H. (1964). Physiol. Plant. 17,186-201. Griffiths, H. J. and Audus, L. J. (1964). New Phytol. 63,319-333. Guharay, F. and Sachs, F. (1984). J. Physiol. 352, 685-701. Haberlandt, G. (1900). Ber. Dtsch Bot. Ges. 18,261-272. Haberlandt, G. (1902). Ber. Dtsch Bot. Ges. 20,189-195. Harris, G. G. (1967). J. Acoust. SOC.Amer. 44,176-186. Hart, J. W. and Macdonald, I. R. (1984). Plant Physiol. 74,272-277. Hasenstein, K.-H. and Evans, M. L. (1986). Plant Physiol. 81,439-443. Heathcote, D. G. (1981). Plant Cell Environ. 4,131-140. Hensel, W. (1984). Protoplusmu 119,121-134. Hensel, W. (1985). Protoplasma 129,178-187. Hensel, W. (1986a). Prog. Bot. 48,205-214. Hensel, W. (1986b). Naturwissenschaften 73,510-511. Hille, B. (1984). “Ionic Channels of Excitable Membranes.” Sinauer, Sunderland, MA. Hillman, S. K. and Wilkins, M. B. (1982). Planta 155,267-271. Honig, B. H., Hubbell, W. L. and Flewelling, R. F. (1986). Annu. Rev. Biophys. Chem. 15, 163-193. Horwitz, A., Duggan, K., Buck, C., Beckerle, M. C. and Burridge, K. (1986). Nature 320,531-533. Hudspeth, A. J. (1985). Science 230, 745-752. Iversen, T.-H. (1969). Physiol. Plant. 22, 1251-1262. Iversen, T.-H. and Larsen, P. (1971). Physiol. Plant. 25,23-27. Jackson, M. B. and Barlow, P. W. (1981). Plant Cell Environ. 4, 107-123. Jaffe, L. F. and Nuccitelli, R. (1974). Annu. Rev. Biophys. Bioeng. 6, 445-476. Jenkins, G. I., Courtice, G . R. M. and Cove, D. J. (1986). Plant Cell Environ. 9, 637-644. Johnson, A. (1965). Physiol. Plant. 18,945-966. Juniper, B. E. and French, A. (1970). Plants 95, 314-329. Kakimoto, T. and Shibaoka, H. (1986). Plant Cell Physiol. 27,91-101. Kaufman, P. B. and Dayanadan (1984). In “Hormonal Regulation of Plant Growth and Development” (S. S. Purohit, ed.), vol. 1, pp. 369-385. Agro Botanical Publishers, Bikaner, India. Keith, C. H., Ratan, R., Maxfield, F. R., Bajer, A . and Shelanski, M. L. (1985). Nature 316, 848-850. Kostina, L., Anikeeva, I. and Vaulina, E. (1984). Adv. Space Res. 4,65-70. Kutschera, U., Bergfeld, R. and Schopfer, P. (1987). Planta 170, 168-180. Larsen, P. (1969). Physiol. Plant. 22, 469-488. Lawton, J. R., Juniper, B. E. and Hawes, C. R. (1986). J . Exp. Bot. 37, 693-704. Lee, J. S., Mulkey, T. J. and Evans, M. L. (1983). Plant Physiol. 73, 874-876. Lin, C.-T., Sun, D., Song, G.-X. and Wu, J.-Y. (1986). J . Histochem. Cytochem. 34, 561-567. Lohmann, K. J. and Willows, A. 0. D. (1987). Science 235,331-334. MacLean-Fletcher, S. D. and Pollard, T. D. (1980). J. Cell Biol. 85,414428. Merkis, A. I., Laurinavichyus, R. S. and Shvyagzhdene, D. V. (1985). Izvest. Akad. Nauk SSSR: Ser. Fizich. 49,715-723. Mertens, R. and Weiler, E . W. (1983). Planta 158, 339-348. Methfessel, C. and Sakmann, B. (1986). Fortschr. 2001.33, 141-153. Mdler, I . M. and Lundborg, T. (1985). Planta 164, 354-361.

40

T. BJORKMAN

Moore, R. and McClelen, C. E. (1985). Plant Physiol. 77,5-105. Moore, R., McClelen, C. E., Fondren, W. M. and Wang, C.-L. (1984). Am. J. Bot. 74, 216-221. Moore, R., Fondren, W. M. and Marcum, H. (1987). Am. J. Bot. 14, 329-336. Niedergang-Kamien, E. and Leopold, A. C. (1957). Physiol. Plant. 10,29-37. Olsen, G. M. and Iversen, T.-H. (1980). Physiol. Plant. 50,275-284. Parthasarathy, M. V. (1985). Eur. J. Cell Biol. 39, 1-12. Payne, R. and Fein, A. (1987). J. Cell Biol. 104, 933-937. Perbal, G . (1978). J. Exp. Bot. 29,631-638. Pfeffer, W. (1894). Ann. Bot. 8,317-320. Pickard, B. G. (1973a). Bot. Rev. 39, 172-201. Pickard, B. G. (1973b). Can. J. Bot. 51, 1003-1021. Pickard, B. G. (1985). In “Encyclopedia of Plant Physiology”, vol. 11, “Hormonal regulation of development. 111. Role of environmental factors” (R. P. Pharis and D. M. Reid, eds), pp. 193-281. Springer Verlag, Berlin. Pickard, B. G. and Thimann, K. V. (1966). J. Gen. Physiol. 49, 1065-1086. Pilet, P. E. (1985). “The Physiological Properties of Plant Protoplasts” (P. E. Pilet, ed.), pp. 226-235. Springer Verlag, Berlin. Poovaiah, B. W., Reddy, A. S. M. and McFadden, J. J. (1987). Physiol. Plant. 69, 569-573. Rasi-Caldogno, F., Pugliarello, M. C. and DeMichelis, M. I. (1987). Plant Physiol. 83,994-1000. Roblin, G. and Fleurat-Lessard, P. (1987). Planta 170, 242-248. Rockwell, M. A., Fechheimer, M. and Taylor, D . L. (1984). Cell Motil. 4, 197-2 13. Roux, S. J. and Dauvalder, M. (1985). Physiologist 28, S115-S116. Roux, S. J., Wayne, R. 0. andDatta, N. (1986). Physiol. Plant. 66, 344-348. Rubery, P. H. and Sheldrake, A. R. (1974). Planta 118, 101-121. Sack, F. D. and Leopold, A. C. (1982). Physiologist 25, S97-S100. Sack, F. D. and Leopold, A. C. (1985). Planta 164,56-62. Sack, F. D., Suyemoto, M. M. and Leopold, A. C. (1985a). Planta 165,295-300. Sack, F. D., Suyemoto, M. M. and Leopold, A. C. (1985b). Planta 161,459-464. Sack, F. D. and Kiss, J. (1988). Plant Physiol. 86,529. Salisbury, F. B., Rorabaugh, P. A. and White, R. (1985). Physiologist 28, S95S96. Schlessinger, J . , Axelrod, D., Koppel, D. E., Webb, W. W. and Elson, E . L. (1977). Science 195, 307-309. Schrank, A. R. (1947). In “Bioelectric Fields and Growth” (E. J. Lund, ed.), pp. 75-123. University of Texas Press, Austin. Schrank, A. R. (1948). Plant Physiol. 23, 188-200. Schroder, H. (1904). Beih. Bot. Zentralbl. 16, 269-287. Shaw, S. and Wilkins, M. B. (1973). Planta 109,ll-26. Shen-Miller, J. and Hinchman, R. R. (1974). Bioscience 24,643-651. Shen-Miller, J., Hinchman, R. R. and Gordon, S. A . (1968). Plant Physiol. 43, 338-344. Sievers, A. and Hensel, W. (1982). “Plant Growth Substances 1982” (R. F. Wareing, ed.), pp. 497-506. Academic Press, London. Sievers, A. and Schroter, K. (1971). Planta 96, 339-353. Sievers, A., Behrens, H. M., Buckhout, T. J. and Gradmann, D. (1984). Z . Pflanzenphysiol. 114, 195-200. Sliwinski, J. E. and Salisbury, F. B. (1985). Plant Physiol. 76, 1000-1008. Slocum, R. D. and Roux, S. J. (1983). Planta 157,481492.

PERCEPTION OF GRAVITY BY PLANTS

41

Starzak, M. E. (1984). “The Physical Chemistry of Membranes.” Academic Press, Orlando. Steive, H. (1986). Fortschr. 2001.33,377-380. Stinemetz, C. L., Kuzmanoff, K. M., Evans, M. L. and Jarrett, H. W. (1988). Plant Physiol. 84, 1337-1342. Stinemetz, C. L. and Evans, M. L. (1987). In “Molecular Biology of Plant Growth Control” (J. E. Fox and M. Jacobs, eds), UCLA Symposia on molecular and cellular biology #44, pp. 361-370. Alan R. Liss, New York. Strong, J. A., Fox, A. P., Tsien, R. W. and Kacmarek, L. K. (1987). Nature 325, 714-717. Tanada, T. and Vinten-Johansen, C. (1980). Plant Cell Environ. 3,127-130. Thimann, K. V. and Schneider, C. L. (1938). A m . J. Bot. 25,627-641. Trewavas, A. (1981). Plant Cell Environ. 4,203-228. Verbelen, J.-P., Spruyt, E. and DeGreef, J. A. (1985). A m . J. Bot. 72,1054-1060. Volkmann, D. (1974). Protoplasma 79, 159-183. Volkmann, D. and Sievers, A. (1979). In “Encyclopedia of Plant Physiology”, vol. 7, “Physiology of Movements” (W. Haupt and M. E. Feinleib, eds), pp. 573-600. Springer, Berlin. Wagner, G. (1979). In “Encyclopedia of Plant Physiology”, vol. 7, “Physiology of Movements” (W. Haupt and M. E . Feinleib, eds), pp. 114-126. Springer, Berlin. Wagner, G., Valentin, P., Dieter, P. and Marme, D. (1984). Plantu 162,6247, Weir, W. G., Cannell, M. B., Berlin, J. R.. Marban, E. and Lederer, W. J. (1987). Science 235,325-328. Williams, S. E. and Bennett, A. B. (1982). Science 218, 1120-1122. Witztum, A. and Parthasarathy, M. V. (1985). Eur. J. Cell Biol. 39,21-26. Wolfe, J. and Steponkus, P. L. (1981). Biochim. Biophys. Acta 643,663-668. Woodcock, A. E. R. and Wilkins, M. B. (1969a). J. Exp. Bot. 20,156-169. Woodcock, A. E. R. and Wilkins, M. B. (1969b). J. Exp. Bot. 20,687-697. Wright, M. and Osborne, D. J. (1977). Biochem. Physiol. Pflanz. 171,479-492. Yau, K.-W., Matthews, G. and Baylor, D. A. (1979). Nature 279,806-807. Yau, K.-W., Haynes, L. W. andNakatani, K. (1986). Fortschr. Zool.33,343-366.

This Page Intentionally Left Blank

Crassulacean Acid Metabolism: a Re-appraisal of Physiological Plasticity in Form and Function

H . GRIFFITHS

Department of Biology. The University. Newcastle upon Tyne N E l 7 R U . UK

I. I1.

Introduction

. . . . . . . . . . . . . . . . . . . 44

ModificationandRegulationofConstitutiveCAM . . . . . . . 45 A . Occurrence and Distribution of Constitutive CAM . . . . . 45 B . Characterization of the Die1 Cycle . . . . . . . . . . 46 C . Biochemical Regulation . . . . . . . . . . . . . . 50 D . ModificationofCAMPhasesbytheEnvironment . . . . . 51

E . Plant Water Status: Is there Regulation both of and by Solute Accumulation? . . . . . . . . . . . . . . . . . 53 F. Stable Isotope Ratio Analysis . . . . . . . . . . . . 55 G . Regulation of Respiratory C 0 2 Recycling During CAM . . . . 64

I11.

IV .

V.

PlasticityofMetabolicResponse: Shadesof CAM . . . . . . . A . Characteristics of C,- CAM Intermediates . . . . . . . . B . Occurrence, Distribution and Evolution . . . . . . . . . C . Respiratory C 0 2 Recycling by C,- CAM Intermediates . . . . D . Physiological Characteristics of the C3-CAM Transition . . .

70

Significance of Respiratory C 0 2Utilization During CAM . . . . . A . Recycling in the Terrestrial Environment . . . . . . . . B . Recycling in the Aquatic Environment . . . . . . . . . C . Organic Acid Speciation: the Newcastle Hypothesis Revisited

78 79 82 83

67 68 74 76

CAM: Development of Integrated Research . . . . . . . . . 85

Advances in Botanical Research Vol . 15

Copyright 01988 Academic Press Limited All rights of reproduction in any form reserved .

ISBN 0-12-005915-0

43

44

H. GRIFFITHS

I. INTRODUCTION Crassulacean acid metabolism (CAM) is traditionally illustrated by succulent plants from arid, semi-desert regions whereby nocturnal stomata1 opening, C 0 2 uptake and malic acid storage are associated with reduced transpiration and improved water economy (Kluge and Ting, 1978). In terms of the contribution that is made to global carbon balance or crop productivity, it could be argued that CAM has received disproportionate attention when compared to conventional C3 photosynthesis or even C4 metabolism [the latter being found in many economically important crops or weeds (Edwards and Walker, 1983)l. However, the reasons for this interest in CAM may be briefly summarized by a number of recent developments. CAM is phylogenetically and geographically more widely distributed than previously thought, with 20,00& 30,000 species ranging from alpine succulents to aquatic “isoetids” (Cockburn, 1985; Keeley, 1987; Osmond, 1984; Winter, 1985). In particular, many CAM plants are found (and are yet to be found) growing in the humid tropics, often as epiphytes (Griffiths and Smith, 1983; Smith et al., 1986a; Ting, 1985; Winter, 1985). Furthermore, biomass productivity of CAM plants may be much greater than previously accepted and may equal or exceed that of C3 crops (Bartholomew, 1982; Sale and Neales, 1980; Nobel and Hartsock, 1983). It has also been suggested recently that the water economy of CAM plants may be regulated indirectly, as a result of reduced transpiration, or directly, by generation of osmotica (i.e. control of and by nocturnal malic acid accumulation) (Luttge, 1986,1987,1988; Smith and Luttge, 1985; von Willert and Brinckmann, 1985; Smith et al., 1987). Consideration has also been given to the flexibility of the expression of CAM with C3-CAM intermediate plants able to induce CAM (or variations thereof) in response to water or salt stress (Lee and Griffiths, 1987; Ting, 1985; Winter, 1985), with implications for drought resistance breeding programmes. Above all, any metabolic “curiosity” such as CAM (Osmond, 1978) will continue to engender interest not only because of the stamina and fortitude required for nocturnal studies, but also because of the potential for integration of biophysical and biochemical metabolism in terms of ecophysiology, morphology and taxonomy (Luttge, 1987). It should be noted that while these areas may have been the subject of a number of recent reviews, some elegantly describing CAM in the context of a C0,-concentrating mechanism and its relation to carbon, water and nitrogen use efficiencies (Osmond, 1984, 1987; Osmond et al., 1982; Winter, 1985), others have tended to identify a number of complex terminologies (Cockburn, 1985; Neales, 1975; Ting, 1985). The expression and regulation of CAM in response to environmental stress can more simply be related to the extent that respiratory COz appears to be recycled as a substrate for malic acid formation (Griffiths etal., 1986; Lee and Griffiths, 1987; Luttge, 1987,1988;

CRASSULACEAN ACID METABOLISM

45

Szarek et al., 1973; Winter, 1985; Winter et al., 1986b). It has also been asserted by Luttge (1987) that the description of CAM could be simplified to the four phases, as defined by Osmond (1978), but including the contribution from recycling of respiratory C 0 2 , as particularly demonstrated by the epiphytic bromeliads (Griffiths et al., 1986, 1988). While the author is in complete agreement with this viewpoint (Lee and Griffiths, 1987), for the moment there is a need to maintain the distinction between those plants which can repeatedly induce or repress CAM in response to environmental variation [C,-CAM intermediates (Lee and Griffiths, 1987)] and those “irreversibly committed” to CAM (i.e. constitutive CAM) so that available data and theories can be reviewed before hopefully adopting the reductionist stance implied above. Luttge (1988) has also suggested that many recent investigations may have underestimated the role of organic acids such as citric acid, which in some species may also fluctuate diurnally as part of CAM (Ranson and Thomas, 1960). This has important implications for the source of carbon substrate utilized for carboxylation and for CAM bioenergetics (Luttge, 1988). It is thus the aim of this chapter to link the developments in ecophysiology of CAM (Cockburn, 1985; Luttge, 1987; Ting, 1985; Ting and Gibbs, 1982; Winter, 1985) to as yet neglected areas of biochemical regulation of CAM (Luttge, 1987,1988) in an attempt to provide a coherent re-appraisal of the enigma that is CAM.

11. MODIFICATION AND REGULATION OF CONSTITUTIVE CAM In order to provide a suitable introduction to the range of processes encompassed by this chapter, this section presents a brief summary of the essential features of the basic CAM pathway, as previously described as “constitutive” or “obligate” CAM. Under these conditions, once CAM has been established by mature leaves, it is never completely repressed during the plant growth period, although the extent of gas exchange and nocturnal acidification may be regulated by environmental conditions. Particular emphasis has been paid to those publications dealing with integrated studies of constitutive CAM variation (Table 11). Attention will be drawn to other relevant reviews, many of which provide more details of the background to the regulation of CAM, before going on to consider details of the plasticity of CAM expression (Sections I11 and IV). A. OCCURRENCE AND DISTRIBUTION OF CONSTITUTIVE CAM

The suggestion that CAM has multiphyletic origins, developing independently even within single families, is based on observations that 25,000 or so

46

H. GRIFFITHS

CAM species are distributed among at least 25 families, representing -10% of the angiosperms (Griffiths, 1988; Kluge and Ting, 1978; Ting, 1985; Winter, 1985). Although these estimates are sometimes based on the distribution of succulent species (see Section II.B), it seems likely that there may be -15,000 species with CAM in the Orchidaceae, together with the 2000 CAM species of the Cactaceae and the majority of Aizoaceae and Crassulaceae (Winter, 1985) [for specific reference to epiphytes, see Griffiths (1988) and Winter et al. (1983, 1986a)l. Although details are uncertain, it is probable that the majority of these plants are constitutive CAM. In the Bromeliaceae, approximately one-third (-1000 species) are constitutive CAM , with only one definite C,-CAM intermediate as yet characterized (Medina and Troughton, 1974; Medina et al., 1977; Griffiths and Smith, 1983; Griffiths et al., 1986). Despite the (probable) greater biomass produced from stem succulents, such as the cacti, in arid environments, Winter suggests that there may be a preponderance of CAM species distributed within the humid tropics (Winter, 1985; Winter et al., 1983; see also Griffiths and Smith, 1983). More extensive investigations are required into the regulation of CAM by “leaf succulent” as compared to “stem succulent” plants (see Section 1II.B). Although nearly all cacti are constitutive CAM, in terms of the evolution of CAM it is intriguing to note that the genus with most primitive characteristics (Pereskia) is a C3-CAM intermediate (Diaz and Medina, 1984; Nobel and Hartsock, 1986; Rayder andTing, 1981; Ting, 1985). This seems to represent a clear evolutionary sequence, with gradation of CAM from the Pereskioideae (net C 0 2 uptake by leaves in daytime only) through the Opuntioideae to the Cactaceae with exclusively dark C 0 2 uptake by the stem (Nobel and Hartsock, 1986). CAM is also found in epiphytic ferns (Griffiths, 1988; Winter et al., 1983, 1986a) and fern allies from submerged aquatic habitats (Boston and Adams, 1986; Griffiths, 1988; Keeley, 1987; Richardson et al., 1984), and perhaps even in a gymnosperm, Welwitschia (Ting, 1985; von Willert, 1985; Winter, 1985). However, many of these plants could be described as C3-CAM intermediates, and will be considered in more detail in Sections I11 and IV, where the significance of variation in CAM expression is discussed.

B. CHARACTERIZATION OF THE DIEL CYCLE

Since the initial characterization of the four phases of CAM (Osmond, 1978) this description of the sequence of physiological processes throughout the day-night cycle has been universally adopted. Winter (1985) lucidly reviews both the biochemical and physiological differentiation during each phase, and it is necessary only to reiterate the bare details in this chapter; these are also shown in Fig. 1.

Temperature OC

u

R H%

Cell sap osmotic pressure (0) Titratable H+ Xylem tension ( 0 ) (MPa)

fmol m.3)

CO, uptake

Conductance

Transpiration

(pin01 m-2s-11

(mmol m-2s-1)

iprnol m a 2 i ' )

Quantum flux L

Vapour pressure

(prnol rn-2s-1)

deficit (mol m-'),AW

Fig. 1. Leaf gas exchange and water relations of the CAM bromeliad Aechmeu nudicuufis, showing the timing of the CAM phases. (Phase I1 was not determined by the gas exchange measurements.) Plants were growing epiphyticallyin a forest clearing at Simla, Northern Mountain Range, Trinidad, with measurements made during the dry season (25-26 February, 1983). Redrawn from Smith et ul. (1986b).

48

H. GRIFFITHS

Phase I. This refers to the dark period when there is net uptake of C 0 2 with most incorporated into malic acid, via oxaloacetate. The initial “C4” carboxylation is catalysed by phosphoenolpyruvate carboxylase (PEPc), in a single carboxylation reaction (Ranson and Thomas, 1960; Winter, 1985). Organic acids (predominantly malic acid) accumulate in the vacuole, so that by the end of the dark period titratable acidity in the cell sap may in exceptional circumstances be greater than 500 mol H+ m-3 [e.g. epiphytic bromeliad, Aechmea nudicaulis; see Fig. 1and Smith etal. (1986b)l. During phase I, stomatal conductance is usually high (20-30 mmol m-2 s-’) when leaf-air vapour pressure difference (VPD) under natural conditions is usually lower than during the light period (0.4 kPa versus 3 kPa). It should be noted that the rate of C 0 2 assimilation at night is limited by mesophyll processes (carboxylation capacity, vacuolar storage capacity) rather than by stomatal conductance (Winter, 1985). Phase I encompasses the essential adaptations of CAM: improved water-use efficiency resulting from the temporal separation of C3and C4carboxylation pathways, with subsequent operation of a C02-concentratingmechanism in the light period. Phase ZZ. At the end of the dark period, there may be a transient phase when stomatal conductance increases and COz is incorporated with a carboxylation sequence progressing from C4 to C3, as photosynthetically active radiation (PAR) increases (Fischer and Kluge, 1984; Kluge et al., 1982; Littlejohn and Ku, 1984). This represents the transition from dark carboxylation (C,) to the light period C3carboxylation, catalysed by ribulose bisphosphate carboxylase (RUBISCO). It utilizes a combination of C 0 2 taken up and directly incorporated into photosynthetic products and C 0 2 released as a result of organic acid decarboxylation. The expression of phase I1 seems to be dependent on environmental conditions, as it may be repressed under natural conditions (Griffiths et a l . , 1986,1988; Liittge et al., 1986;Lee etal., 1988) and is perhaps the first of the four phases to be reduced in response to water limitation. Phase ZZZ. Representing the major period of decarboxylation of organic acids, following efflux from the vacuole, phase I11 is the period of lowest stomatal conductance. Following decarboxylation by either malic enzyme or PEP carboxykinase [depending on species (Winter, 1985)], the C 0 2 is refixed via RUBISCO and the conventional C3pathway. Photorespiration is suppressed by this effective “C02-concentrating mechanism”, with C 0 2 internal partial pressure increasing to over 100 times atmospheric levels (Cockburn et al., 1979; Friemert et al., 1986; Spalding et al., 1979). and C 0 2 release occurring from leaves at this time (Friemert et al., 1986). The extent of phase I11 is also regulated by environmental conditions (see Section II.F), although plants under optimal water and PAR regimes may complete decarboxylation within 3-4 h.

CRASSULACEAN ACID METABOLISM

49

Phase N.Following the decarboxylation of the nocturnally accumulated organic acids, the internal partial pressure of C 0 2 also declines, and stomata1conductance may then increase. Carbon dioxide can then be taken up and assimilated directly via RUBISCO, with a concomitant increase in transpiration rate when leaf-air VPD is high (Griffiths et al., 1986; Osmond, 1978; Winter, 1985). Towards the end of phase IV, double carboxylation may again occur (cf. phase II), although it is the C4 component which increases with time as the dark period approaches (Kluge et al., 1982; Ritz et al., 1986). As a result of this unified description of day to day CAM regulation, several criteria for CAM were defined (Kluge and Ting, 1978; Osmond, 1978; Ting and Gibbs, 1982). It is perhaps the adherence to these criteria, together with concentration on experiments on the leaf succulent Crassulaceae, which has led to certain variations in CAM expression being neglected (Ramon and Thomas, 1960; Luttge, 1987). These criteria include a definition of succulence, but it is now important that a more detailed survey be made in terms of the regulation of CAM by “stem” succulents as opposed to “leaf” succulents. Most experimental work on the biochemical and biophysical regulation of CAM has been carried out on “ideal” experimental plants (e.g. Kalanchoe spp.), where leaves consist of uniformly large, vacuolate, chlorenchymatous cells (Luttge and Smith, 1984a,b; Smith, 1984; Luttge, 1987). However, many CAM plants are stem succulents, whereby the outer layers of cells are chlorenchymatous and surround inner water storage parenchyma (WSP) as found in most Cactaceae and Agavaceae. Here, the two cell types, often replete with the gums and mucilages, may have confounded modern cellular and enzymic extraction techniques. In addition, both CAM and C3 members of the Bromeliaceae tend to have chlorenchyma distinct from WSP, with the chlorenchyma surrounding longitudinal air channels which link directly to the stomata in the lower leaf surface (Smith et al., 1985; Tomlinson, 1969). The WSP may be distributed above and between the chlorenchyma in these plants, and the regulation of CAM may vary in relation to the occurrence of WSP (Griffiths et al., 1986,1988;Luttge and Ball, 1987) (see Section IV). It is important that a study of the expression of CAM be made in terms of the type of succulence, particularly in view of the finding that WSP may undergo small changes in titratable acidity (Earnshaw et al., 1987; c.f. Luttge and Ball, 1987). The second criteria often defined is that of the nature of the day-night fluctuation in organic acids. It has been shown that the Kalanchoe (Crassulaceae) leaf succulents and some Cactaceae and Aizoaceae show the “perfect” stoichiometry for CAM, i.e. 2H’: 1 malic acid: l C 0 2 (Luttge, 1987, 1988). However, in view of the variable contribution from respiratory C 0 2 recycling (Griffiths et al., 1986, 1988: Luttge, 1987; Winter et al.,

50

H. GRIFFITHS

1986b), with large variations in the H+: C 0 2uptake ratio, this could also be affected by the organic acid composition, particularly if citric acid was involved (Luttge, 1988; Ting et al., 1985b). Many measurements of CAM have been made without analysis of all three components of the die1 cycle (titratable acidity, organic acid speciation and net COz uptake) and it is now critical for the contribution from these components to be more fully assessed. While the “criteria” have been important in furthering our definition and identification of constitutive CAM, they may have drawn attention away from more variable CAM characteristics. C. BIOCHEMICAL REGULATION

The basic mechanisms of CAM are often simply summarized as PEPC activity and malic acid accumulation in vacuoles. However, it has been succinctly pointed out that these attributes are to be found in all plant cells (Luttge, 1987). Although in CAM it is critical that malic acid, rather than the malate’- anion, is accumulated, these common features are found as responses to long-distance pH regulation (Raven, 1986; Smith and Raven, 1979), ion uptake and osmotic adjustment (Luttge and Clarkson, 1985; Luttge and Smith, 1984a) and movements of stomatal guard cells or pulvinar cells (Luttge, 1987;Zeiger, 1983). It should be noted that such arrangements have been used to both justify (Cockburn, 1983,1985) and exclude (Winter, 1985) the possibility that CAM evolved from stomatal metabolism. However, as well as malic acid, citrate is also of ubiquitous origin and importance in plant cell metabolism, again pointing to the neglect of citratexitric acid balance as a feature of CAM (Luttge, 1988). As explained in Section II.B, much work with CAM has been devoted to the bioenergetics and biophysics of malic acid accumulation (Luttge, 1987, 1988). The stoichiometry of titratable acidity has been found to account for malic acid accumulation in Kalanchoe spp. (Luttge and Ball, 1979, 1980; Luttge et al., 1982), Mesembryanthemum crystallinum (Winter et al., 198l), and Opuntia spp. (Osmond et al., 1979; Nobel and Hartsock, 1983). These aspects have been extensively reviewed recently (Luttge, 1987, 1988; Winter, 1985) and so only relevant processes will be considered here, including the regulation of substrate supply and energy required for the biosynthetic processes of CAM. The evidence to date that PEPC is the critical enzyme in CAM is unequivocal: rates of conversion of COz to H C O j (the latter being the substrate for PEPC) and specific activity of PEPC are more than 40 times greater than observed rates of carboxylation (Winter, 1985). “Futile cycling”, as PEPC activity during phase I11decarboxylation, is prevented by a modification of PEPC kinetic properties in response to malic acid (at that time diffusing out of the vacuole) and light (Winter, 1982, 1985). Indeed,

CRASSULACEAN ACID METABOLISM

51

such regulation can account for the decreasing (phase 11) or increasing (phase 111) activity of PEPC described in Section II.B, although protection from malate inhibition at night may also be conferred by glucose 6-phosphate (Winter, 1985). Phosphoenolpyruvate (PEP) as substrate for PEPC is derived from starch andor other glucans in the Crassulaceae (Black et al., 1982; Luttge, 1988a) or from hexoses stored in the vacuole of the Bromeliaceae (Black et al., 1982; Lee et al., 1988; Luttge, 1987; Medina et al., 1986). It is of interest to note that should the role of citrate be found to be an important facet of some CAM plants, then the regulation of other carbon reserves will also require investigation (Luttge, 1988; Section 1V.C). The energetic costs of malic acid accumulation have been budgeted in detail by Luttge (1987,1988), and when hexose is utilized via glycolysis there is a cost of 1 mol ATP per mol malic acid. The energy is utilized in active pumping of malic acid into the vacuole, and since the demonstration of the 2H': malic acid stoichiometry (Luttge, 1987, 1988; Luttge and Ball, 1979, 1980; Liittge et al., 1982), it has also been shown that thermodynamically, the tonoplast pump has a stoichiometry of 2H': ATP and is operating close to equilibrium (Luttge, 1987; Luttge and Ball, 1979). The final critical evidence comes from the characterization of a malate-stimulated vacuolar ATPase, which possibly, together with the activity of a vacuolar pyrophosphatase, has sufficient activity to account for observed rates of malic acid accumulation (Smith et al., 1984a,b; Luttge, 1987). More information is required on respiration during CAM in view of the occurrence of extensive recycling of respiratory COz (Griffiths et al., 1986, 1988; Table 11), and since in Kalanchoe the dark respiration rate was shown to be providing only half of the energy required (Luttge and Ball, 1987). It is intriguing that the Bromeliaceae have a greater energetic requirement in terms of respiration, since hexoses are utilized for PEP synthesis. The relationship between net dark respiration rates and recycling of respiratory COz will be considered in more detail in Section IV. D. MODIFICATION OF CAM PHASES BY THE ENVIRONMENT

Each of the phases described in Section 1I.B may respond independently to environmental variables. Figure 1shows the typical gas exchange and water relations parameters for the epiphytic CAM bromeliad A . nudicaulis. During phase I, gas exchange and malic acid accumulation may be regulated by leaf-air VPD, -temperature, water supply and the previous day's PAR. Stomata of some CAM plants respond directly to humidity, and thus gas exchange may be reduced when leaf-air VPD is reduced (Griffiths et al., 1986; Lange and Medina, 1979; Medina, 1982; Osmond et al., 1979). There is also a temperature optimum for malic acid accumulation, in the range of

52

H. GRIFFITHS

10-20°C, but this is complicated by temperature acclimation processes, alteration of leaf-air VPD and the effect on plant dark respiration rate (Winter, 1985). The direct effect of temperature on respiratory COz recycling is considered in more detail in Section IV. Gas exchange rates at night are usually lower for CAM plants (2-10 pmol COz m-’ s-l) than the photosynthetic rates quoted for C3plants (20-30 pmol COz m-’ s-’), and maximum stomata1 conductance is usually only 0.1 mmol m-’ s-l for CAM, compared to over 0.5 mmol m-’ s-l for C3 plants. Although efficiency of water use is much increased during phase I of CAM as compared to C3photosynthesis [transpiration ratios (TR) may vary from 20-600 g HzO:g COz for CAM, compared to 600-1400 for C3plants (Kluge and Ting, 1978)], it is interesting to note that the TR values for C3and CAM bromeliads during the dry season in Trinidad were within the normally quoted CAM range (Griffiths et af., 1986). In any event, the expression of phase I gas exchange can be severely modified by drought, as shown in an excellent study on water relations in K. daigremontiana (Smith and Luttge, 1985) and has also been shown in the field for Opuntia inermis (Osmond et al., 1979). Net COz uptake is reduced and occurs later in the dark period in water-stressed CAM plants (Griffiths et al., 1986; Smith and Luttge, 1985), although a concomitant reduction in acidification may be modified by the occurrence of respiratory COz recycling (Griffiths et af., 1986; Luttge, 1987). Taking the reduction in COz uptake to an extreme, it has often been noted that individual cladodes or detopped cacti can remain viable for many months, putting on new growth each year, without any apparent external water supply. Under these conditions, uptake of COz at night may cease, and yet fluctuations in titratable acidity continue, i.e. 100% of the COz fixed as malate is recycled internally from respiratory COz (Szarek et af., 1973; Ting, 1985; Winter, 1985; Brulfert et al., 1987). This phenomenon has been termed “idling”, and may be important in preventing photoinhibition (Osmond, 1982; Section IV). The work of Nobel has clearly demonstrated that acidification in stem succulents is proportional to the incident PAR during the preceding light period (Acevedo etal., 1983; Nobel and Hartsock, 1978,1983; for review see Nobel, 1982; Winter, 1985). Furthermore, cladodes in the Cactaceae are orientated so as to optimize PAR interception, although spines may also shade and limit growth (Nobel, 1982, 1983). This is in contrast to T. usneoides, where acidification saturates earlier in response to increasing PAR (Martin et af., 1986; Table 11). However, as pointed out by Winter (1985), it is also important to consider responses of CAM plants to low PAR, because of self-shading in rosette plants or shading by canopies in rain forests. Data presented in Section 1I.G show that under these conditions, the rate of respiratory COz recycling may modify the carbon balance of phase I (Table 2). The expression of phase I1 may be reduced by high temperatures and is

CRASSULACEAN ACID METABOLISM

53

also regulated by other environmental factors (Luttge et a f . , 1986; Winter and Tenhunen, 1982). Although at one time thought to be a laboratory artefact, phase I1 has been well characterized by field studies in Agave (Nobel, 1976) and Opuntia (Acevedo et al., 1983; see also Littlejohn and Ku, 1984). Phase I1 was not demonstrated by epiphytic bromeliads during the dry season in Trinidad (Griffiths et al., 1986; Luttge etal., 1986) although it was found with Brumefiahumilis in Venezuela (Lee et a f . ,1988). It appears likely that there is a direct interaction between temperature and water status in the regulation of phase 11. The onset of phase I11is therefore also affected by increased temperature, and the decarboxylation rate may be higher (Winter and Tenhunen, 1982). When decarboxylation proceeds rapidly under field conditions, there may be a loss of C 0 2 from leaves (Griffiths et al., 1986), and this is enhanced at high temperatures (Lange and Zuber, 1980). The release of C 0 2 during phase I11 has been correlated with variations in stomata1 response to Pco,, in different species, with the constitutive CAM K . tubiJlora maintaining a low leaf conductance, in contrast to Sempervivum tecturum (Friemert et a f . , 1986). These gas exchange data were not quantified in terms of rates of malic acid mobilization, and PAR was low. However, during phase I11 under high PAR, RUBISCO operates at a rate close to saturation (Spalding et al., 1979), although initial decarboxylation proceeds at a slightly faster rate, resulting in the increase in PcoZiat that time (Winter, 1985). Phase IV can be used to characterize the quantum requirement of C3 photosynthesis by CAM plants directly. In response to environmental stress, this phase is rapidly lost, since transpiration ratios tend to be higher, approaching the range for C3 plants (Griffiths et a f ., 1986; Kluge and Ting, 1978). E. PLANT WATER STATUS: IS THERE REGULATION BOTH OF AND BY SOLUTE ACCUMULATION?

In view of the large die1 changes in solute accumulation during CAM, there are clearly resultant effects on the parameters of plant water relations (see Fig. 1). It is only comparatively recently that the plant water relations of constitutive CAM plants have been quantified. With the use of the pressure probe and pressure chamber it has been possible to assess changes in turgor both directly (Steudle et al., 1980; Luttge and Nobel, 1984; Rygol et a f . , 1986) and indirectly (Smith and Luttge, 1985; Smith et al., 1985, 1986b). Leaf sap osmotic pressure (n) may increase from 200 to 500 m osmol kg-' in the Bromeliaceae (Smith et al., 1986b) with corresponding day-night changes in n of up to 0.5 MPa in A . nudicaulis (Fig. 1). There is a 1 : 1 relationship between the day-night change in osmotic pressure ( n )and malic acid content (Amal, assayed directly or as AH') in K.

54

H. GRIFFITHS

daigremontiana (Smith and Luttge, 1985) and two bromeliads (Smith et al., 1986b), indicating that in these species malic acid is the only organic acid participating in CAM and behaving in effect as a perfect osmoticum (Smith and Luttge, 1985; Smith et al., 1986b; Luttge, 1987). Xylem sap tension, as measured with the Scholander-type pressure chamber, increases during the dark period (Smith and Luttge, 1985; Smith et al., 1985,1986b) as water loss due to transpiration occurs (Fig. 1). Under some circumstances, however, dew formation can enhance plant water status [i.e. xylem tension falls (Smith et al., 1986b)], indicating that plant water supply is modulated directly as a result of organic acid accumulation. Values of xylem sap tension have been used as a measure of plant water potential, v , although strictly speaking, a correction should be applied to account for xylem sap osmotic pressure (Smith and Luttge, 1985; Luttge, 1987). Using the relationship between bulk leaf water potential (v),bulk leaf turgor ( P ) and leaf sap osmotic pressure (n),

v=P-n (1) it is then possible to calculate P. Changes in bulk leaf turgor can then be calculated throughout the day-night cycle for CAM (Smith and Luttge, 1985) and are found to correlate with those measurements made directly with the pressure probe (Steudle et al., 1980; Luttge and Nobel, 1984; Rygol et al., 1986). In general, leaf turgor is found to be higher during the early light period, and lower early in the dark period, although transient increases may be associated with stornatal closure between phase IV and phase I, as well as increases in malic acid (Luttge, 1987). Thus malic acid accumulation may enhance plant water status in two ways: firstly as a result of the increased leaf conductance at low leaf-air VPD, and secondly by the direct generation of osmotica. The significance of this latter process has been identified in ecological terms, since dewfall usually occurs pre-dawn, at the time of maximum solute accumulation (Smith and Liittge, 1985; Smith et al., 1986b). This phenomenon has been investigated in terms of leaf volume changes, measured directly by water uptake (Ruess and Eller, 1985; Eller and Ruess, 1986), or by measurements of gas exchange and turgor (Liittge, 1986). In general, it seems that cellular water storage capacity ranges from 3 to lo%, as determined by the combined methods outlined above (Luttge, 1986), and could be of particular importance to epiphytic CAM plants, and plants of coastal deserts (Luttge, 1987). The relationship between these observations and those of Chen and Black (1983) are obscure. However, changes in specific leaf weight and a slight increase in water content did occur at the time of maximum titratable acidity in Sanseveria fasciata and K . daigremontiana (Chen and Black, 1983; Luttge, 1986, 1987). In contrast, Smith et al. (1987) have demonstrated that the capacitance of Agave is such that water status would not be significantly enhanced by malic acid accumulation. This again stresses the need for a

CRASSULACEAN ACID METABOLISM

55

more detailed comparison of stem and leaf succulents in view of the much greater water storage capacity of the former. Furthermore, it may also be necessary to consider in more detail the balance between water loss during phase IV when compared to uptake of water at night (J. A. C. Smith, personal communication). It has been suggested that the role of organic acid accumulation during the induction of CAM in the C3-CAM intermediate Sedurn telephium may be more important in terms of water uptake than carbon gain (Lee and Griffiths, 1987; Teeri et al., 1986), and these adaptations will be considered in more detail in Sections I11 and IV. In any event, further studies on water storage and leaf capacitance, using the rigorous approach of Smith and Nobel (1986) and Smith et al. (1987), are required in order to elucidate this effect. F. STABLE ISOTOPE RATIO ANALYSIS

Carbon isotope ratio (dl3C) analysis has so far contributed rather superficially to our understanding of CAM, mainly through identification of CAM plants in diverse populations. However, in the future it is likely that 613C analysis and associated mass spectrometric techniques will make a major contribution to the exploration of regulatory mechanisms. 613C analysis ratios are calculated from the following formula: 613C(%O) =

[

13C/12Csample

13c/'2cPDB

-

Std

1

1 x 1000

(2)

This natural abundance ratio is calibrated against a carbonate standard, derived from a Cretaceous belemnite from the Pee Dee formation, arbitrarily set at O%o. The application of carbon isotope ratio analysis to studies of plant metabolism has been extensively reviewed by O'Leary (1981) and Raven (1987): c3plants commonly have 613Cvalues of about -27%0 (with a range of -23 to -35%,); C4have less negative 6 values, about -14%0 (with a range of - 10 to - 18%0,and thus are enriched in 13Ccompared to C3plants); CAM plants have variable 613C ratios encompassing C3 and C4 ranges dependent mainly on the relative contributions of phase I (C,) and phase I V (C,) carboxylation (see Table I). Atmospheric C 0 2 has a 6 value of -8%o [becoming more negative with the combustion of fossil fuels (Pearman, 1980)], although source 613Cmay vary in response to secondary fractionation processes (e.g. respiratory C 0 2 is depleted in 13C) and should always be measured: Apart from the basic identification of the photosynthetic pathway, 613C analysis can be used more critically to evaluate photosynthetic limitations, as a result of the elegant analyses of Farquhar and co-workers (Farquhar,

56

H. GRIFFITHS

1980; Farquhar et al. ,1982; Farquhar and Richards, 1984). The relationship between 6 13Cvalue, C 0 2concentration gradient and carboxylation discrimination can be expressed as a positive fractionation (A) which increases with decreasing water-use efficiency in C3 plants (Farquhar and Richards, 1984; Evans et al., 1986):

A =a -d

Ci + ( b - a) -

(4)

CO

where A = fractionation po0) a = fractionation due to diffusion (+4%0) b = fractionation due to RUBISCO activity (+27 to +30%0) d = fractionation due to PEPC, C 0 2 dissolution and respiration ci = intracellular C 0 2 concentration c,, = external C 0 2 concentration The magnitude and direction of these fractionation processes are summarized in Table I. A plant with a 613C value of -27%0 reflects the photosynthetic limitation imposed mainly by diffusion (k3 = 0.6: Table I).

TABLE I Fractionation during photosynthesis: C3plants Fractionation during C3photosynthesis Potential fractionation (%o) :

C020 Range of 6'3C (YW)

Carboxylation +30

Diffusion +4.4 ki

CO,,

ki

RUBISCO

+

Diffusion not limiting -8 Diffusion limiting -8

+

PGA + C3products

theoretical -38 -12.4

-8 -12.4

observed -35 -23

Fractionation during phase I of CAM Potential fractionation (%.):

Diffusion

Equilibration and Carboxylation __L_

+4.4 ki

CO,,

-8

+ 2 = -6net k3

CO,, + HC03 --$ PEPC + OAA theoretical

Range of

613c (%o)

-+

malate

k2

Diffusion not limiting -8 Diffusion limiting -8

-8 -12.4

-2 -12.4

observed CAM C3-CAM -11 -16 -18 -30

57

CRASSULACEAN ACID METABOLISM

Environmental factors which increase k3 (i.e. increase stomata1 conductance) will result in more negative Bplant, as the inherent fractionation of RUBISCO is expressed. Diffusion limitations (e.g. reduced water supply) result in less negative dplant(i.e. k3 decreases: Table I) as C 0 2 tends to be fixed irrespective of label. In the aquatic habitat, these analyses can be used to calculate boundary layer diffusion limitations and/or contributions from any C02-concentratingmechanism (Raven et a f . ,1982; Beardall e t a f . ,1982; for reviews see Kerby and Raven, 1985; Raven et al., 1987). When the initial carboxylation is catalysed by PEPC (i.e. C4 or CAM), HCO; is substrate, and fractionation results from the dissolution of C 0 2 [ -8Yi at 25°C (Mook et al., 1974)l. The latter process accounts in part for the less negative (heavier) 613C values for carbonates, including the PDB standard (Table I). PEPC discriminates slightly against 13C,and thus there is a net fractionation of -6%0 during C4 carboxylation (Table I). Should photosynthesis be markedly limited by diffusion, plant 613C should reflect and diffusion fractionation [i.e. (-8) (-4.4) = the combination of BSoUrCe - 12.4%,]; under non-limiting conditions, Bplantshould tend towards -2%0 (Table I; Griffiths, 1984; O’Leary, 1981). Thus diffusion limitation during C4 carboxylation should bring about a shift of dplantin the opposite direction to that in C3 plants, although K . daigremontiana appears to be equally limited by diffusion and carboxylation (Holtum et a f . ,1983). In practice, CAM plants rarely have Bplant in this theoretical range. Although Nalborczyk et al. (1975) report a value of - 11%0 when COz (6,,,,, = -7%0) was only supplied at night, this does not so much reflect PEPC diffusion limitation as RUBISCO fractionation during phase 111; the 613C of COz leaking during decarboxylation was correspondingly enriched in 13C02 (Nalborczyk et a f . ,1975). Thus the 613Cof plants which are C4or constitutive CAM (with little direct C3 carboxylation) are more negative than those calculated theoretically, because the inherent fractionation of RUBISCO is expressed during C 0 2leakage. Direct C3carboxylation during phase IV also tends to make dplantmore negative, as does refixation of respired C 0 2 at night (Griffiths, 1984; O’Leary and Osmond, 1980). Other approaches have included analysis of the 613Cof the 4-C of malate (Deleens et al., 1985; Holtum et a f . ,1983; O’Leary and Osmond, 1980). In this respect, CAM presents an ideal analytical system, since a large pool of malate accumulates overnight, and the 4-C can be experimentally decarboxylated and analysed, providing corrections are made for residual malate and fumarate randomization. It seems that diffusion and carboxylation equally limit CAM (Holtum et al., 1983; O’Leary and Osmond, 1980). Further analyses have indicated that C 0 2 fixation becomes more diffusionlimited and less carboxylation-limited as temperature increases [i .e. dmal becomes depleted in 13C(Deleens et a f . , 1985)l. This trend is in contrast to overall effects of higher temperatures on Bplantwhereby expression of CAM (as AH’ or Amal)increases, and Bplantbecomes less negative (Deleens et al., 1985; Medina, 1984; Table 11).

+

TABLE I1 Variation in constitutive CAM: summary of integrated studies showing response of AH', C 0 2exchange in phases I , II, III and IV,recycling of respiratory CO, and 613Cto environmental perturbation CO, exchange succuFamily and species" lence" Polypodiaceae Pyrrosia longifolia (r, 1 ) P. longifolia (r, 1&5) Drymoglossum pilosilloides (r, 1J3,5)

Asclepiadaceae Hoya carnosa (r, 2 )

H. nicholsaniae (r, 2 )

Asteraceae Senecio medleywoodii (r, I )

L

AH'"

Recycling' ACO, Phase Phases PhaseI" IIIb IIandIVb % 6H'

+

238 66 440 1960 1418 1970 2002 396

74 31 19 18 28 22 23 23

+ + +

307 102

133 15

-

-

L

30

10

-

L

121

55

110 60

34 29

L L

L

6 13c Ymd

Conditions

38 90 6 4 (11) (81) (77) (77) (77) (0)

-14.4 -14.9

l3 71

41 72

-15.4'

+H,O -H20

30

33

10

-14.3 - 18.3'

Shaded Shaded

-

38

0

-

43 51

38 3

-

+ + +

-

+

+ + +

+

+

Sunlexposed Shaded

Winter et al. (1986a) Ong et al. (1986)

Ong et al. (1986)

Old 42 1

Reference

-13.9b

Intermed Young

Rayder and Ting (1983b) Sternberg et al. (1984b) Winter et al. (1986a) 'Winteretal. (1983)

'

Eller and Ruess (1986) Luttge (1987) Sternberg et al. (1984b)

'

Cactaceae Opuntia inermis (r, 2 ) 0.ficus-indica (r, 2 )

0.ficus-indica ( r , 2 ) 0 .vulgaris (r, 3 , 4 )

S S

S S

Mammillaria woodsii S (r, 3 , 4 ) Cereus validus (r, I ) S

Crassulaceae Echeveria columbiana (r, I ) Kalanchoe pinnata (I; 2 )

K. daigremontiana (rJ2 )

L L

L

1450 1120 205 415 805 200 662 824 560 1380(150)(6) 1180(450)(6) 264(350)(6) 372 (36)(6) 140(168)(6) 460 226 96 86

684 338 0 80 306 50 265 305 130 615 365 0 168 0 181 59

+ +

+ +

-

-

-

-

-

-

-

38 25

+ +

-

+

-

6 40 100 61 24 50 20 26 54 11 38 100 5 100 21 48

410 255 193 100 132 214 300 150 450 264 36 140 98 108

+ +

21 42

20 36

-

82 444

235 200 (27)(6) 145 127

97 79 58 40

18 21 14 35

41 42 29 47

381 (7x6) 173 (12)(6)

178 77

6 11

25 19

153 212

67 101

16 5

25 10

-10.7 -10.0

+H,O -H20

Osmond et al. (1979) Nobel and Hartsock (1986)

48 Nobel and Hartsock (1984) Winter etal. (1986b) 35

30 15}night t"C +H,O

Winter et al. (1986b) Nobel et al. (1984)

+400 rnM NaC1

-12.06

18-19.5 }nightroc l3

Medina and Delgado (1976) Rundel et al. (1979) Medina (1982)

22 17)night t"C 25) night growth t"C Medina and Osrnond (1981) Medina (1984) continued

TABLE 11-Continued C 0 2exchange succuFamilyandspecies" lence" Crassulaceae (continued) K . daigremontiana L (I; 3 , 4 ) Cucurbitaceae Xerosicyos danguyi

L

(r, 4

Bromeliaceae Aechmea aquilega (1 1 A . fendleri ( 1 ) A . nudicaulis ( 1 )

L L L

AH'"

Recycling" AC02 Phase Phases PhaseI" IIIb IIandIVb % 6H+

6 '3c %d

5 130 58 570

3250(250)(6) 1560 720(570)(6) 390 190

56 13

-

+ +

113 237 230 332 301 233 322 239 360 240

6 47 61 73 43 44 60 16 98 22

-

-

-

+

+ -

+

+ + +

+ + + + + + + +

71 278 86 164

89 91 47 56 71 65 63 87 46 82

101 143 108 186 215 145 202 207 164 196

Conditions

35 nightt"C

-18.5 -17.5 -14.6'

+H,O - H 2 0 lmonth

-15.0' - 13.8'

Pt Gourde-dry Simla-wet

-14.1' - 14.1' -13.7' -14.4' -14.0'

Textel Tucker-dry Textel} wet Simla Lalaja-shaded

Reference Winter et al. (1986b)

Rayder and Ting (1983a) Sternberg et al. (1984b)

'

Griffiths et al. (1986)

'Smith et al. (1986a)

Griffiths et al. (1986) Smith et al. (1986a) Griffiths et al. (1986) Smith et al. (1986a)

'

Sale and Neales (1980) Griffiths and Smith (1983)

'

Bromelia humilis (2 1

L

Tillandsiaflexuosa (1 ) T. usneoides (1, 3 )

L

L

133 190 144 160 243 170 243 450

25 36 21 6 29 23 18 62

360 338

180 25 248 141 100 55 61 41 9 16 110

200

T. usneoides (r, 3 )

L

T. schzedana ( r , 3 )

L

T. juncea (r, 3 ) Dyckia brevifolia (r, 3 )

L L

510 612 223 258 377 260 160 220 320

S

650

Agavaceae Agave deserti (r, 1 )

+ + + + ++ +

+

-

62 62 71 93 76 73 85 72

-

0 0 85 288

-

3 54 10 57 67 68 89 85 31

+

83 118 102 148 185 124 207 326

14 330 23 148 255 178 142 188 100

38 250

-16.5 -17.7 -13.1

Green exposed + H 2 0 Lee et al. (1988) Green exposed -H20 Green shade - H 2 0 Yellow exDosed -H,O Griffiths et al. (1988) +H,O (Any) -H2O (dry) Martin et al. (1986) Medina and Troughton (1974) 3 weeks 'Medina etal. (1977) 500 Griffiths and Smith (1983) Martin et al. (1981) May August October Martin et al. (1982) Martin and Adams (1988)

'

-19.8'

-15.0'

'

'

McWilliams (1970) Griffiths and Smith (1983) McWilliams (1970) 'Medina et al. (1977)

'

Field

Nobel (1976) and Woodhouse et al. (1980) continued

TABLE 11-Continued C 0 2exchange

succuFamily and species" lence" Liliaceae Yucca baccata (r, 1 H 5 )

Orchidaceae Arachnis hookeriana (r,5 ) Dendrobium speciosum (r, 2 ) Schomburgkia humboldtiam ( I )

S

L

AH'"

Recycling' AC02 Phase Phases PhaseI" IIIb IIandIVb % 6H'

41 40 152 15

81 32 165 32

97 35

+ + +

+

181 14 119 14

5

-

+

1

-

-

L

250

6

L

281 130

50 20

4

+ +-

+

+

*

(90) (94)

6 '3c %od

-16.1 -13.5'

+H,O 32/10

-15.1'

-H20 +H,O

49 122

-15.4

65 18'

-13.4

70

90

Conditions

-H20

Reference Kemp and Gardetto (1982) Szarek and Troughton (1976) Sternberg et al. (1984b)

' '

Fu and Hew (1982)

'Neales and Hew

(1975) Winter et al. (1986b)

Rainy season Dry season

Griffiths et al. (1988)

"AH+ derived from titratable acidity, assuming stoichiometry of 2H:: 1 malate for calculations; ACO, derived by integration of gas exchange curves. (I) mmol kg-'/mol m-3; (2) mmol m-'; (3) mol (kg dry wt)-'; ( 4 ) data from AH' f CO, in air; (5) data not directly comparable; (6) AH' in CO,-free air; leaf (L) or stem (S) succulent; (r) recalculated or derived from original data. *Presence or absence of C 0 2 exchange in phases 11,111 and IV: = uptake (ACO, quoted if greater than dark period AC02), - = C 0 2 release. 'Recycling calculated from AH' as measure of absolute fixation, with stoichiometry of 2H': 1 malate: 1C02 assumed, i.e. (0.5 X AH+) - ACO, loo; 0.5 x AH+ absolute recycling, 6H+, calculated as AH+ - (2 x ACO,). Values in parentheses based on an estimate of leaf succulence. 6I3Cnot corrected for source variation; 'relates isotope ratio and source reference when other than primary reference.

+

CRASSULACEAN ACID METABOLISM

63

Values of dplantfor constitutive CAM species are listed in Table 11, which shows the variation in expression of constitutive CAM, with data compiled mainly from integrated studies which have been carried out with plants under a range of environmental conditions. Comparable values of dplant from other studies have been included where relevant. Table I1 also includes details of acidification and C 0 2 exchange, integrated throughout the daynight cycle. Under controlled experimental conditions, or defined natural environmental conditions, dplantcan be shown to reflect the proportion of C3 and C4carboxylafion, in terms of water supply (Osmond et al., 1976), growth temperature (Medina, 1984), photoperiod, leaf age, flowering and seed set (O’Leary, 1981; Osmond, 1978; Winter, 1985; Table 11). However, under natural conditions, despite apparently large variation in altitude (Szarek and Troughton, 1976); Yucca baccata, Table II), climatic zonation (Smith el al., 1986a; Bromeliaceae, Table 11) or external C 0 2 concentration (Holtum et al., 1983), there may be remarkably little intraspecificvariation of dplant from varying habitats. However, dplantmay be genetically determined (Teeri, 1982a,b; Section 1II.B). Despite apparent limitations of this technique, d13C analysis has made important contributions to integrated ecophysiological studies (Griffiths and Smith, 1983; Smith et al., 1985,1986a) and analysis of the evolutionary origins of CAM, in for example the Bromeliaceae (Griffiths, 1988; Griffiths and Smith, 1983; Medina et al., 1977; Medina and Troughton, 1974; Smith et al., 1986a), the Crassulaceae (Teeri et al., 1981; Teeri, 1982a,b) and Australian vascular epiphytes (Winter et al., 1983, 1986a). More recently, 613C analysis has been applied to residual gas analysis with following photosynthetic C 0 2uptake (O’Leary et al., measurement of dsource 1986; Evans et al., 1986; Nalborczyk et al., 1975). This method gives a much more immediate measure of isotope fractionation during photosynthesis, but has not as yet been applied to the analysis of CAM. It will be of use to evaluate diffusion limitation during phases I and IV, but also to analyse the transition from C3 to C4 carboxylation in phases I1 and 111. Additionally, such analyses will be important for identification of carboxylation pathways and diffusion limitation during the transition between C3 and CAM, as exemplified by S . telephiurn (Lee and Griffiths, 1987). d13C analysis has also been used to investigate secondary fractionation following photosynthetic processes in CAM plants (Deleens and GarnierDadart, 1977), to infer the operation of glycolytic regulation of CAM as well as to show the mobilization of carbon reserves during development (Deleens and Quieroz, 1984). A further intriguing possibility may involve the use of d13C analysis to identify lipid and carbohydrate substrates in plant metabolism. Lipids are isotopically “lighter” (i.e. depleted in 13C, d13C of -29%0) compared to carbohydrates (d13Cof -24%0), and such variation has been used to investigate the source of respiratory substrate in aroid spadices during thermogenesis (Walker et al., 1983). The finding that metabolism of

64

H . GRIFFITHS

lipids predominated was difficult to interpret initially, but can now be explained as a result of the pioneering work of Thomas and co-workers, in showing that P-oxidation of fatty acids can occur in plant mitochondria (Thomas and Wood, 1986; Wood et al., 1986). This technique may also be applicable to investigations of CAM, if indeed substrates other than carbohydrates may be utilized during citrate metabolism (Luttge, 1988; Section 1V.D). In general terms, labelled stable isotopes (as opposed to natural abundance studies) are also applicable to CAM studies. 13C02 labelling has demonstrated the existence of double carboxylation during phase IV (Ritz et al., 1986; Section II.C), and 1 8 0 2 has been used to investigate respiratory processes in CAM (AndrC et al., 1979), although these results are difficult to interpret. By using C 0 2 highly enriched in 13C and l80it is possible to analyse in vivo activities of carbonic anhydrase and PEPC (Holtum et al., 1984), and also to differentiate between malic acid derived from carboxylation of external C 0 2 and that derived internally from respiratory C 0 2 (Holtum et al., 1984; AndrC et al., 1979). This method could be of great importance in separating the respiratory C 0 2 recycling components of constitutive CAM and C,-CAM intermediates (Table 11; see Sections VI and IV). Analysis of the naturally occurring stable isotope of hydrogen, 6D, has also been used for research into CAM (Ziegler et al., 1976; for review see Ting, 1985). Although the fractionation processes are not fully understood, they appear to be specifically related to carbohydrate metabolism, with 6 D more positive (i.e. enriched) in CAM plants. This can be used to identify C,-CAM intermediates, which often have C3-like dplantbut 6 D characteristics of CAM (Sternberg et al., 1984a,b). It would now seem that surveys of plant populations and/or herbarium specimens, previously used as a means of identifying phylogenetic distribution of C4and CAM, have now perhaps been superseded by the promise of integrating ecophysiological and biochemical regulation of CAM using stable isotope analyses.

G. REGULATION OF RESPIRATORY CO2 RECYCLING DURING CAM

The previous sections have discussed various aspects of the regulation of constitutive CAM, and have demonstrated that environmental conditions may alter the magnitude of the CAM phases in terms of gas exchange, water relations and biochemical control. In order to investigate in more detail the contribution of these variables to overall CAM carbon balance, a comparative study has been made of constitutive CAM, compiled from a number of integrated studies in both field and laboratory (Table 11). Each study selected fulfils several criteria: gas exchange (as COz uptake) hes or can be

CRASSULACEAN ACID METABOLISM

65

integrated so as to give a measure of phase I net C 0 2uptake (AC02)directly comparable with dusk-dawn titratable acidity (AH'). Ideally, the effect of a number of environmental parameters was examined from each study, for plants which represent a wide phylogenetic range. As directly comparable figures were not usually available, data was processed as AH' and ACOz by integration of net C 0 2 uptake during phase I. Note was also made of C 0 2 exchange during phases I11 and I1 or IV. The contribution that respiratory C 0 2recycling made to nocturnal titratable acidity was calculated assuming that there is a strict stoichiometry of 2Ht :1 ma1 : C 0 2 and that malic acid is the only organic acid participating. Two expressions are used: % recycling =

(0.5

AH') - ACO;, 0.5 x AH'

X

absolute recycling 6H' = AH' - (2 x AC02)

(5)

(6) In view of the variation in units used to express data (see footnotes to Table 11), percentage recycling gives an easy comparison between all studies, but absolute recycling (6H') is a more precise indication of the relationship between the Ht derived from respiratory C 0 2 and total C 0 2 fixation. In some studies, units were not compatible, but an estimate of recycling has been included in parentheses, based on an assumption of leaf succulence. An important caveat should be that since most studies do not include a measure of succulence for each treatment, some of the variation in recycling could result from changes in leaf water content during the course of a series of experiments. The most striking feature of this compilation is the extent to which recycling of respiratory C 0 2 is a ubiquitous characterstic of constitutive CAM (Table 11). There are only a few examples of little or no recycling, perhaps reflecting optimal growth conditions, since it only occurs under one of the experimental or environmental conditions for each species [e.g. D. pilosilloides (Ong et al., 1986); S. medley-woodii (Eller and Ruess, 1986; Liittge, 1986); T. usneoides (Martinet al., 1981, 1986), Table 111. However, the comparative studies of single species under a range of environmental conditions reveal more about the regulation of carbon balance by recycling. The contribution of CAM to net carbon gain increases with age [e.g. 6I3C becomes less negative (Medina, 1984; see also Osmond, 1978; Ting, 1985; Winter, 1985)], but there is also ontogenetic variation in the proportion of recycling [e.g. S. medley-woodii (Ruess and Eller, 1985; Luttge, 1986; Table 11). There may also be seasonal variation as shown by T. usneoides (Martin et al., 198l), although in other bromeliads 6H' may alter but % recycling was similar in the wet and dry seasons [e.g. B. humilis (Lee et al., 1988); T. flexuosa (Griffiths et al., 1988)l. The ultimate state of recycling (i.e. loo%), occurs during CAM idling

66

H. GRIFFITHS

(Szarek et al., 1973; Osmond, 1978; Ting, 1985; Winter, 1985), and thus it would be natural to expect that there may be a progression, with recycling increasing with drought stress. This is borne out by many of the studies analysed for Table 11, whereby recycling increases in both percentage and absolute terms [e.g. H . carnosa (Rayder and Ting, 1983b); 0. inermis (Osmond et al., 1979); A . aquilega and A . nudicaulis (Griffiths et al., 1986)l. In some species, however, recycling increased in percentage terms but decreased in absolute terms, indicating that total malate accumulation is also affected by drought stress [ X . danguyi (Rayder and Ting, 1983a; see also de Luca et al., 1977); T. flexuosa (Griffiths et al., 1988); T. schiedana (Martin and Adams, 1987)l. Salt stress has a similar effect to drought stress, in that recycling increased when C. validus was subjected to 400 mM NaCl (Nobel et al., 1984). In terms of responses to PAR, the degree of recycling seems to depend on the original habitat preference of the plant. Naturally shaded plants tend to increase recycling under high PAR [e.g. P. longifolia (Winter et al., 1986a; Ong et al., 1986); T. usneoides (Martin et al., 1986)], whereas those usually growing under high PAR may increase recycling under low PAR [e.g. 0. ficus-indica (Nobel and Hartsock, 1986); A . nudicaulis (Griffiths et al., 1986); An. comosus (Sale and Neales, 1980)l. Bromelia humilis could well be better adapted to shaded understorey conditions, since the fully exposed plants are chlorotic and have a higher rate of recycling (Medina et al., 1986; Lee et al., 1988; Table 11). When measured under field conditions, recycling by epiphytic bromeliads in the genus Aechmea was found to be correlated with night temperature (Griffiths et al., 1986). In other studies summarized in Table 11, there is also an increase in recycling with night or growth temperature, particularly in the stem succulents [e.g. 0. ficus-indica (Nobel and Hartsock, 1984); 0. vulgaris and M . woodsii (Winter et al., 1986b)l. In leaf succulents, however, there is a variable response to night temperature: in the genus Kafanchoe,some studies indicate a relatively constant proportion of recycling with temperature [e.g. K . pinnata (Medina, 1982); K . daigremontiana (Medina and Osmond, 198l)], whereas in E . columbiana and K . daigremontiana there was a significant increase in recycling with temperature (Medina and Delgado, 1976; Winter et al., 1986b). Measurements of the increase in titratable acidity in C0,-free air were also made for these plants, with this novel approach to the quantification of absolute recycling first used by Medina and Osmond (1981). Although such manipulations could result in increased stomata1 conductance, there was good agreement with the calculated recycling (see Table 11). When the leaves of B. humilis were sealed with petroleum jelly during the dark period, the titratable acidity was 50-75% of that in normal air (H. Griffiths, unpublished data). It was suggested that the degree of recycling could be related to the occurrence of WSP, with leaf and stem succulents with separate WSP (e.g.

CRASSULACEAN ACID METABOLISM

67

Bromeliaceae and Cactaceae) being more prone to recycling than uniformly chlorenchymatous leaf succulents (Griffiths et a f . , 1986). Work on WSP respiration by Liittge and Ball (1987) indicates that this is not due to respiration by the WSP per se. Although Kalanchoe spp. seem generally to undertake a relatively constant proportion of recycling under a wide range of conditions, unlike other constitutive CAM plants (Table II), perhaps it is the performance of “perfect” CAM in Kafanchoe spp. which has led to the neglect of this important facet of CAM carbon balance. The origin of the author’s interest in this phenomenon arose as a result of field studies carried out on epiphytic bromeliads in Trinidad, 1983 (Griffiths et al., 1986). Extensive recycling (70-90%) was found in a variety of epiphytic bromeliads, where in one species the highest known AH’ was observed [474 mol H+ m-3 for A . nudicaufis (Smith et al., 1986b)l. Such a large discrepancy between AC02 and AH’ was not found to be anomalous in the Bromeliaceae, since even well-watered Ananas cornosus was found to derive 45% of malate from respired C 0 2 [data of Sale and Neales (1980); Table 11). This is remarkable, considering the enormous potential productivity of this crop (Sale and Neales, 1980; Bartholomew, 1982). It is now clear that while net and gross respiratory components may not have been considered relevant by plant physiologists or biochemists they should now be quantified since CAM plants cycle a large proportion of carbon reserves under both field and laboratory conditions. The energy utilized by these processes is not insignificant, and the possible reasons for this process are considered in Section IV. However, there is no doubt that CAM researchers should now be prepared to quantify the discrepancy between net C 0 2 uptake and acidification, as exemplified by the work of P. S. Nobel.

111. PLASTICITY OF METABOLIC RESPONSE: SHADES OF CAM While data presented in Table I1 show how environmental variables may alter the magnitude of the various phases of constitutive CAM, this section considers the extent to which CAM characteristics may be induced or repressed. There are few reports of constitutive CAM reverting completely to C3 [e.g. Agave deserti (Hartsock and Nobel, 1976)], and any such shift often seems to occur in response to extreme temperature regimes [e.g. Ananus cornosus (Neales et al., 1980)l. Although it has been suggested that A . C O ~ O S U Sis completely C3 under a 38/18 day-night temperature regime [Ting (1985), quoting Bartholomew (1982)], this is perhaps due to some discrepancy in the original presentation (Table I; Bartholomew, 1982). A much more common phenomenon is that of the complete induction or repression of CAM in C3-CAM intermediate plants. Indeed, many terminologies have been coined in order to describe the phenotypic plasticity of

68

H. GRIFFITHS

CAM expression. In this section, an attempt will be made to review the occurrence of CAM in C3-CAM intermediate plants and to demonstrate that these “shades” of CAM represent only slight modifications to the original CAM criteria, particularly if consideration is taken of the extent of respiratory C 0 2 recycling (Lee and Griffiths, 1987; Luttge, 1987; Teeri, 1982a,b).

A. CHARACTERISTICS OF C3-CAM INTERMEDIATES

One of the first attempts to categorize CAM used the terms “weak”, “full” and “super” CAM (Neales, 1975). In view of the potential modification of each phase in constitutive CAM, outlined in Section 1I.D. (see also Table II), the responses of any one plant may be encompassed by several of these terms. Inducible CAM has been defined by Winter (1985) as the “potential for expanded leaves . . . to attain CAM. . . in a short period compared to the leaf‘s total life cycle”. Other such descriptions include “facultative CAM” or “optional CAM” (Kluge and Ting, 1978; Osmond, 1978). The most thoroughly documented plant in this category is Mesembryanthemum crystallinum: work by Winter and co-workers has rigorously characterized the ecophysiological importance of the C3-CAM transition (Winter and von Willert, 1972; Winter and Luttge, 1976; Winter et al., 1978; Heun et al., 1981) as well as more general details of biochemical regulation of CAM (Winter, 1985). In summary, M . crystallinum induces the metabolic machinery of CAM in response to salt or drought stress, but also to reduced root temperature or anoxia (Winter, 1985). CAM may also be induced in K . blossfeldiana by photoperiod, where short days result in the irreversible induction of CAM and flowering (Brulfert and Queiroz, 1982; Stimmel, 1984). Photoperiod may also enhance the transition to a more pronounced form of CAM in Peperomia camptotricha (Sipes and Ting, 1985). Ting and co-workers have defined two categories of CAM where there is little or no C 0 2 uptake at night (Rayder and Ting, 1981, 1983a,b; Ting, 1985). These categories represent two extremes, since in one, constitutive CAM, plants close stomata completely during drought stress but still show a large die1 fluctuation in titratable acidity [CAM idling (Szarek et al., 1973; Ting, 1985; Brulfert et al., 1987)l. In the second, C 0 2 uptake is mainly during the day, but there is an increase in titratable acidity at night [CAM cycling (Ting and Rayder, 1982;Ting, 1985; Section III.D)]. The accumulation of organic acids is attributed to the recycling of respiratory C 0 2in each case. A further intermediate variant which has been proposed is that of latent CAM (Schuber and Kluge, 1981), whereby organic acid concentration remains constant throughout day and night, but at a higher level than that normally found in C3-plants.Although 14C02labelling suggested the opera-

CRASSULACEAN ACID METABOLISM

69

tion of C4 carboxylation in the latent CAM form of S. acre, there was no difference in the leaf water content between this and the induced CAM form (Schuber and Kluge, 1981). More recently, Lee and Griffiths (1987) have suggested that “latent CAM” may be a part of a progression between C3and CAM in S. telephium (see also Fig. 2). A further CAM variant is that occurring in submerged aquatic macrophytes: the determined efforts of Keeley and co-workers have now shown that the biochemical C02-concentrating mechanism in Zsoetes spp. (Lycopodiatae) and the other angiosperm “isoetids” fulfils all the CAM criteria (Keeley, 1981, 1983; Keeley and Bowes, 1982; for review see Keeley, 1987). Carbon dioxide supply to isoetids is also augmented by root uptake, with diffusion through internal lacunae to the leaves, and thus CAM operates as a biochemical C02-concentratingmechanism enhancing a physical C02-concentratingmechanism (Richardson et al., 1984; Raven, 1984; Raven et al., 1987; Section 1V.B). However, CAM is repressed in emergent leaves of these plants, which may also develop functional stomata (Keeley, 1983), and thus CAM operates in order to regulate C 0 2 supply in the aquatic environment, where diffusion is limiting (Richardson et al., 1984). It is apparent, therefore, that many of these plants can be classed as C,-CAM intermediates, with the recycling of respiratory C 0 2 being an important facet of isoetid carbon balance (Madsen, 1987; Robe and Griffiths, unpublished results; Section 1V.B). Interestingly, a limited form of CAM activity has also been found in a “terrestrial isoetid”, Stylites andicola. Lurking in boggy regions high in the Peruvian Andes, this plant was characterized by Keeley et al. (1984) as taking up C 0 2via the roots and showing CAM. CAM has also been found in plant roots: epiphytic “shootless” orchids take up C 0 2 and accumulate organic acids in the roots at night (Benzing et al., 1983; Hew et al., 1984; Cockburn et al., 1985; Winter et al., 1985). While this expression of CAM is clearly allied to constitutive CAM (Hew et al., 1984; Cockburn et al., 1985), any mechanism controlling C 0 2 diffusion through the orchid root velamen is obscure (Benzing et al., 1983; Cockburn et al., 1985; Winter et al., 1985). It is, however, possible to quantify the gas exchange and titratable acidity of Chiloschista usneoides by integrating the data of Cockburn et al. (1985) (see also Table 11). With a AH’ of 75 mmol (kg fresh wt)-’, the rate of recycling can be calculated as 86% [or 6H’ of 65 mmol (kg fresh wt)-’ in absolute terms, see Table 111, with net C 0 2release of 1.3 mmol (kg fresh wt)-’ during the day. Thus the plasticity of CAM responses encompasses many variations of physiological form and function: these have been recently reviewed in more detail by Cockburn (1985), who has also proposed that each variation be defined as a separate category of CAM. However, these variations in CAM are based on the same biochemical metabolism, and each type of C,-CAM intermediate can simply be characterized by the extent of respiratory C 0 2 recycling; this is considered in more detail in Section IV.

70

H. GRIFFITHS B. OCCURRENCE. DISTRIBUTION AND EVOLUTION

C3-CAM intermediates are found in most families which also show constitutive CAM, and representatives from most have been summarized in Table 111. In the pteridophytes, it is also possible to identify C3-CAM characteristics in the Polypodiaceae (Winter er al., 1983, 1986a; Sinclair, 1984). The evidence for CAM in Welwitschia has long provoked discussion, and has recently been extensively reviewed (von Willert, 1985; Winter, 1985). However, it seems that Welwitschia uses recycling to at least prevent respiratory C 0 2 losses (Winter and Schramm, 1986) with possibly both malate and citrate being utilized (von Willert et a f . ,1982). In evolutionary terms, C3-CAM intermediates are often found in those genera which are considered as displaying more primitive characteristics. In the Cactaceae, C3-CAM characteristics are demonstrated only by the Pereskioideae and leafy members of the Opuntioideae, the more advanced genera being constitutive CAM (Nobel and Hartsock, 1986). The excellent work by Teeri (1982a,b, 1984), relating variation of CAM expression in field and laboratory conditions, suggests that there is a gradation of CAM within the Crassulaceae, as identified by carbon isotope ratio and dependence on respiratory COz recycling. This suggests that the capacity for expression is related to both genetic and environmental factors: for example, crossing two species, identified as C3 and CAM in terms of their 613C ratios and leaf thicknesses, produced a plant with intermediate characteristics (Teeri, 1982a). Teeri (1982b) made another important point in terms of phenotypic variation in dplant.The same species grown in a phytotron, when compared to plants grown in a greenhouse, demonstrated plasticity of CAM expression which would not be normally revealed by sampling from a single experimental regime: the need for more integrated long-term studies is yet again emphasized (see also Section 1V.B). It was suggested that for the Crassulaceae, CAM could have been both developed and lost in evolutionary terms (Teeri, 1982b). Such analyses are also required in many other families, since many “chicken and egg” arguments appear to centre around the development (or loss) of CAM in conjunction with other characteristics, e.g. epiphytism in the Bromeliaceae (Benzing and Renfrow, 1971; Benzing, 1980; Medina, 1974; Medina et al., 1977; Smith et a f . ,1986a: for review see Griffiths, 1988). The progression of “shades of CAM” between tropical and temperate members of the same family also requires more rigorous analysis. In the Crassulaceae, for instance, those in the genus Kufunchoeseem to show little phenotypic plasticity, apart perhaps from the induction of CAM by the epiphytic K . unifrora (Schafer and Luttge, 1986; Table 111; Section 1II.C). However, there seems to be much more flexible expression of CAM in the genus Sedum (Kluge, 1977; Schuber and Kluge, 1981; Smith and Eickmeier, 1983; Teeri, 1982b; Teeri er a l . , 1986; Groenhof et al., 1986; Lee and

TABLE I11 Variation in C3-CAM intermediates: summary of integrated studies showing response of AH', CO, exchange during day and night, recycling of respiratory C02,6'3C and 6 0 to environmental perturbation

co2

succu-

Familyandspecies" Polypodiaceae Pyrrosia confluens (r, 11295)

lence" L

Welwitschiaceae Welwitschia L mirabilis (r, 112,5> W . mirabilis L (r, 1 4 5 )

AH'"

CO, Light

Recycling' 613C

Dark" Releaseb Uptakeb %

6H+

%d

%

Conditions

Reference

PAR (pmol m-'s-l)

0 9 15 24

40 42 35 27

+

+ + +

118 48 45 30

-25.3 -24.3 -24.2 -20.5

46 26 70 (46 24)

35 24 0

24

830 220 30

-24.7

+

6D

200 shade 401 Epiphytic 400 100 sun

Winter et al. (1983)

Ting and Burk (1983)

100

70

Field

-18.5

to

von Willert et al. (1982)

-19.5

A izoaceae Mesem bryanthemum crystallinum (r, 3 ) Cactaceae Pereskia aculeata (r, 2)

L

L

0

0

448 552

150 165

78 101

20 24

0

-

0

-

2708 664 1164

207 87

0 33 149 40 222

-30 -24 -26 to - 16

100 78 100 101 -27.@

-446

+H,O +400 mM NaCl NaCl+ H 2 0 Field (wet + dry)

Winter (1974) Winter and Liittge

+H,O -H20

Rayder and Ting

(1976)

Winter et al. (1978)

(1981)

Sternberg et al. (1984b) continued

TABLE 111-Continued C 0 2 Light succuFamilyandspecies" lence"

AH'O

Dark" Releaseb Uptakeb %

Cactaceae (continued) P. grandifiora (r, 2 )

L

35 50

Clusiaceae Clzuia rosea (r, 2)

L

18 60

L

57 62 202 350 0 73 73 94 76 855 330

13 32 0 175 0 0 22 38 25 150 120

39

11 0

Crassulaceae Kalanchoe unifIora (r,2,5> Sedum acre ( r , 3 ) Sedum telephium (3)

Sempervivum montanum (r, 3) Gesneriaceae Condonanthe crassifolia

(r,112,5)

L L

L

L

Recycling'

co2

44

0 0

-

-

59 11

100 100

6H+

613C Omd

-

+ -

+ + + + -

-

840 308 43 35 10 0 15 1042 310 43 4

%

(31) (0) 202 0 0 73 29 18 26 355 90

Conditions

Reference

+HzO -H20

35 50 -17.9

(54) (0) 100 0 0 loo 40 20 44 65 27

6D

I

gn: : :

Epiphytic

'"'"}Epiphytic -H,O -27.5 -27.0

+HiOSpring - H 2 0 Summer +H20 :C3

-25.06

+HZO: C&AM -H20: C3-CAM -H20: CAM -H20: CAM 300 p n o l m - 2 s 900 PAR

I

- 18.0

-24*

+13'

+H20}Epiphytic -HzO

Ting et al. (1985a)

Schafer and Luttge (1986) Schuber and Kluge (1981) Lee and Griffiths (1987) Griffiths (unpublished) Wagner and Larcher (1981) Guralnick et al. (1986) 'Tingetal. (1985a)

Piperacaceae Peperomia camptotricha (I,

112,5)

Portulaceae Portulacaria afra (I,

1/29 5 )

36 55 93 80 60 220 36 200

7 48 5 0

180 50 56 1 0

15 9 7

26 46 69

60

Young -27.7'

+23'

-27.5' -17.5'

+H,O - H 2 0 1week -H,O 2 weeks

-10'

August February April

Sipes and Ting (1985) Sternberg et al. (1984a)-

'

Guralnick et al. (1984) Sternberg et al. (1984b) Mooney et al. (1977) Martin and Zee (1983)

'

' Talinum calycinum (r, 3 )

Vitaceae cissus quadrangularis (r, 112,5) Bromeliaceae Guzmania monostachia ( I ) Nidularium innocenti (r, 3 )

Liliaceae Yucca gloriosa (r, 3 ) For notes, see Table 11.

400 180

0 104

1020 37

33 78

0 78

70 89 15 17

2 7

210

0

100 400 0 0

-27.8

+H,O -H,O 5 days

121 0

loo (50)

-25.3 -19.4 -17.8

Leaf Stem Old stem

Ting et al. (1983)

33

94

66

18

3

-26.76 -26.5' -31.5' -24'

Sun Sun Shaded 10°C, 16 h dark period

'Smith et al. (1986a) 'Smith et al. (1985)

-22.0

Field

Martin et al. (1982)

2

100 210

Griffithset al. (1986)

McWilliams (1970)

'Medina et al. (1977)

74

H. GRIFFITHS

Griffiths, 1987), which is in common with many other temperate succulent families such as the Sempervivoideae (Osmond et al., 1975; Wagner and Larcher, 1981; Earnshaw et al., 1985) and Portulacaceae (Martin and Zee, 1983; see also Ting, 1985). There also seems to be an example of convergent evolution in two families of predominantly neotropical distribution, the Piperaceae and Gesneriaceae (Sipes and Ting, 1985; Ting et al., 1985a; Guralnick et al., 1986). Many species in both families are epiphytic [up to 20% in the Piperaceae (Ting et al., 1985a)], and have similar leaf structure and distribution of WSP. Additionally, both families display a variant of CAM in which recycling of respiratory COz predominates [CAM cycling (Ting et al., 1985a; Guralnick et al., 1986; Table III)]. The C,-CAM transition has also been described for deciduous climbers in the Vitaceae [e.g. Cissus trifoliata (Olivares et al., 1984); C . quadrangularis (Ting et al., 1983; Virzo de Santo et al., 1983)], whereby both stems and leaves may show a varying degree of CAM depending on water status. Finally, a most interesting recent development has been the identification of CAM in hemi-epiphytic stranglers in the genus Clusia (Tinoco-Ojanguren and Vazquez-Yanes, 1983; Ting et al., 1985b). This is the first example of CAM in a woody dicotyledenous tree, and the regulation of organic acid metabolism (i.e. malic and citric acid) is currently receiving more attention (Luttge, 1988; Popp and Liittge, personal communication; Section 1V.C).

C. RESPIRATORY COZ RECYCLING BY C3-CAM INTERMEDIATES

Having demonstrated that the magnitude of AH' in constitutive CAM can be related to the extent of respiratory COz recycling under a range of environmental conditions, a similar analysis is made in this section for those plants usually identified as C,-CAM intermediates (Table 111;see also Table 11). The plasticity of the CAM response in these plants may entail considerable daytime COz uptake, without any ready separation into the light period phases of constitutive CAM. Data on COz uptake in the light have been integrated and are presented as the magnitude of net uptake and/or release (Table 111). Recycling, where possible, has been calculated as described previously (Section II.G, Table 11), and values of 613C and 6D have been included where applicable (see Section 1I.F). Although it was not possible to calculate recycling for Pyrrosia confluens, there is clearly considerable plasticity of response by sun and shade populations (Winter et al., 1983; Table 111). Daytime C 0 2 uptake was related to PAR, but the concomitant dark C 0 2 uptake in each of the treatments may not have been directly related to AH', indicating a contribution from recycling (Table 111). For the controversial Welwitschia, two contrasting studies have suggested that respiratory COz recycling is the major feature of

CRASSULACEAN ACID METABOLISM

75

any CAM-like adaptation. In laboratory studies AH' and AC02 were both measurable (Ting and Burke, 1983), but in the field a possible (malate citrate) fluctuation was accompanied by C 0 2 uptake only in daytime (von Willert et al., 1982; von Willert, 1985). Despite the many publications featuring M. crystallinum, it was with difficulty that comparable AH+ and ACOz could be obtained, by applying correction factors calculated from Winter and Luttge (1976) to data published by Winter (1974) for similarly treated plants (Table 111). In the C3 mode, H' does not fluctuate, and there is marked daytime C 0 2 uptake. When the plants are grown with 400 mol mP3 NaCl, there is a gradual decrease in daytime COz uptake (Winter, 1974; Winter and Luttge 1976) and increases in AH' and ACOz, with 33% recycling. On returning to control (salt-free) solution, CAM is retained with an increase in AH', ACOz and daytime C 0 2 uptake; recycling also increased in both percentage and absolute terms (Table 111). During CAM induction, 613C of new plant material shifts from -30 to -24%,, but studies in the field have perhaps shown that the CAM transition is more important, with Bplantvarying from -26 to -16%, as drought stress is prolonged (Winter et al., 1978). However, the respiratory C 0 2 contribution to AH' has not been measured for field-grown populations. In the Portulacaceae, there are also several species which have been described as inducible CAM plants (Ting and Hanscom, 1977; Guralnick et al., 1984a,b). However, as pointed out by Winter (1985), this does not represent true CAM induction but a changing proportion of light :dark C 0 2 uptake, with recycling probably contributing to the AH' (Guralnick et al., 1984a,b; Table 111). The characterization of CAM in Pereskia spp. has been made by Rayder and Ting (1981) and Diaz and Medina (1984). During progressive drought stress, AH' increases in field-grown plants of Pereskia guamacho (Diaz and Medina, 1984), and in laboratory experiments on two other species there was no net C 0 2 uptake at night (i.e. 100% recycling) by P. aculeata and P . grandiflora (see also Nobel and Hartsock, 1986). This is a unique response, since in all other C3-CAM intermediates examined, there is some net C 0 2 uptake at night under certain conditions (see Table 111). One recurrent theme in Table I11 is the number of epiphytes which are C3-CAM intermediates: Kalanchoe uniflora induces CAM in response to drought stress, although recycling ceases (Schafer and Luttge, 1986). Thus 50% of the AH' is derived from respiratory C 0 2 before stress, but then the more usual generic characteristics of Kalanchoe, with little or no recycling, are shown as plant water status decreases (Table 111; see also Table 11). This is in contrast to many of the other epiphytes and C3-CAM intermediates in general [with the exception of S. acre (Schuber and Kluge, 1981)], although rates of recycling could not be properly calculated for any of the Piperaceae or Gesneriacae (Table 111). In the Bromeliaceae, Guzmania monostachia

+

76

H. GRIFFITHS

(subfamily Tillandsioideae) was originally thought to be a classical inducible CAM plant (Medina et a f . , 1977), but field measurements showed little uptake of C 0 2during the dark period (corresponding to 94% recycling), and uptake of COz early in the light period (Griffiths et a f . , 1986; Table 111). Other field studies did correlate AH' with 613C (Smith et al., 1985; Table 111), and more work is required to fully evaluate the CAM status of this plant. Using combined data of McWilliams (1970) and Medina et al. (1977), it appears that Nidularium innocenti (subfamily Bromelioideae) is also a C,-CAM intermediate, with a small AC02 and AH' being accompanied by a C3-like 613C ratio (Table 111). Using similar criteria, Yucca gforiosa also appears to be a C3-CAM intermediate (Martin et a f . ,1981; Table 111). Sedum spp. and Sempervivum montanum also show a range of recycling characteristics, with recycling decreasing under high PAR in S. montanum (Wagner and Larcher, 1981). However, there is a decrease with stress in S. acre (Schuber and Kluge, 1981), and a whole range of values depending on the stage of CAM induction in S . telephium (Lee and Griffiths, 1987; Table 111). The nature of the regulation of +CAM transitions is discussed in more detail in the following section. Finally, it is worthwhile re-emphasizing that many of these C3-CAM intermediates seem to be related to the type or degree of succulence (see Section 1I.B). It is essential that future work takes consideration of succulence, both in terms of variation in plant water status during experimental procedures, and in terms of the respiratory C 0 2 recycling characteristics found in different families. D. PHYSIOLOGICAL CHARACTERISTICS OF THE C&AM

TRANSITION

In view of the phenotypic variation found in all CAM plants, there may perhaps be a simple explanation for the apparent diversity of C3-CAM intermediates, as exemplified by the proliferation of terminologies. Many studies have not used a sufficient range of environmental variables over a long enough period of plant growth and development to account for the possible range of CAM characteristics. There have only been a few integrated measurements of the full C,-CAM transition, collated in Table 111, and it is worthwhile highlighting the complete studies on Sedum acre (Kluge, 1977; Schuber and Kluge, 1981), Portulacaria afra (Guralnick etaf.,1984a,b; Guralnick and Ting, 1986) and M . crystaffinum(Winter, 1985). Figure 2 illustrates the range of physiological characteristics which occur during the complete +CAM transition by Sedum tefephium (Lee and Griffiths, 1987; Table 111). Plants were grown with an intermittent water supply so as to create C3-CAM intermediates; one group was then continuously watered, while the water supply was withdrawn from the other group. Measurements were made after 24 h and then again at 5-day intervals, and included C 0 2 uptake, AH' and xylem sap tension (pressure chamber, data in MPa).

77

CRASSULACEAN ACID METABOLISM

a 100 0 40 Dusk xylem 0.36 sap tension 0 0.50 (MPa) Recycling %

-

n

I

x

12

0

a

20 0.29

0

0

0.50

O0.50

12

12

0

44 a0.26

I

I

I

I

24

a 0

24

12

12

I

24

12

Fig. 2. The transition between C, and CAM by Sedum telephium in response to water supply. C3-CAMintermediate plants were watered (0)or drought-stressed(0), with measurement of C 0 2 uptake, titratable acidity and xylem sap tension being made at intervals (days 1,s and 10) following the imposition of the watering regimes. (a) day 1; (b) day 5; (c) day 10. Redrawn from Lee and Griffiths (1987).

During the initial 24-h sampling period, for both + and - water treatments, the AH' was 74 pmol H+ (g fresh wt)-' (Fig. 2a). Those undergoing continued drought showed net C 0 2uptake at night, while the watered plants had C3 gas exchange characteristics. After 5 days, the rewatered plants maintained a relatively constant level of titratable acidity, while those undergoing drought increased AH' to 94 pmol H' (g fresh wt)-' (Fig. 2b). After 10days, the background level of titratable acidity had fallen to 35 pmol H' (g fresh wt)-' (i.e. fully C,), but the plants undergoing drought were obviously stressed and AH' was reduced to 76 pmol H' (g fresh wt)-' (Fig. 2c). At each stage, the dusk values of xylem sap tension reflected improved water status in the rewatered plants, with the C3transition being represented by (0.364.26 MPa); in the CAM mode, xylem sap tension was constant at 0.5 MPa. As part of this progression from C3 to CAM in S. telephiurn it is possible to identify a number of the specific intermediate types described in Section III.A, including latent CAM (Fig. 2b, water), CAM cycling (Fig. 2a, - water) and perhaps a shift towards CAM idling (Fig. 2c, - water). How much simpler, therefore, to utilize the AH+ in conjunction with the proportion of respired C 0 2to describe the entire transition from C3to CAM (Lee and Griffiths, 1987). What, then, is the stimulus for the transition between C3and CAM? Plant

+

78

H. GRIFFITHS

water status is most simply implicated, and there was a change in xylem sap tension in S . telephium as CAM was repressed (Fig. 2; Lee and Griffiths, 1987). Ting and Rayder (1982) proposed that the mechanism may be controlled by stomata, and cite the high background levels of organic acids in P. afra as being poised for CAM (i.e. latent CAM?). In S. tefephium,there were similar high background levels of titratable acidity but also daytime C02uptake (Fig. 2b), and C 0 2uptake also occurred in the light following a large AH' (Fig. 2a). Daytime C02 uptake immediately following organic acid accumulation (i.e. effectively during phase 111) was observed in nearly all of the studies listed in Table 111, and the regulation of stomata1 aperture during deacidification urgently requires clarification, particularly as Nishio and Ting (1987) have suggested that daytime C4carboxylation may occur in Peperomia. The observation that root anoxia or low temperature induces CAM in M . crystallinurn (Winter, 1985) also points to a mesophyll-based stimulus. Ting and Rayder (1982) also suggested a role for ABA in CAM induction, but this has been discounted by Winter (1985) since ABAper se did not induce CAM in M . crystaflinum. A further suggestion, made by Lee and Griffiths (1987), was that the increasing background levels of organic acids could parallel osmotic adjustment in C3plants (Teeri et a f . , 1986). Thus as drought stress continues, the organic acids start to fluctuate diurnally, with the C 0 2being mainly derived from respiratory C 0 2 . This could also be related to the potential for water uptake during dewfall as x increases (Ruess and Eller, 1985; Liittge, 1987,1988; Smith et a f . ,1987; Section 1I.E). Investigations in our laboratory are currently directed towards characterizing this transition for S . tefephium.Preliminary observations in watercultured plants have shown that a AH' is induced by PEG within 24 h of changing the rooting media, and that levels of proline are three times higher in water culture + PEG, and six times higher in soil-grown plants with CAM, when compared to well-watered controls (Lee and Griffiths, unpublished results). At any event, it seems that any inductive stimulus is likely to be found in the mesophyll, but much more work is required to characterize the physiology and biochemistry of the C,-CAM transition before the inductive process can be identified.

IV. SIGNIFICANCE OF RESPIRATORY CO2 UTILIZATION DURING CAM Recycling has been shown above to be a recurring phenomenon during CAM. An attempt is now made to relate the significance of this apparently energetically futile process to the regulation of CAM under natural conditions in terrestrial and aquatic environments. Perhaps a close parallel can

CRASSULACEAN ACID METABOLISM

79

be seen between recycling in CAM and photorespiration in C3plants: both seem in teleological terms to have little purpose, but are ubiquitous nevertheless (Osmond et al., 1980). A. RECYCLING IN THE TERRESTRIAL ENVIRONMENT

To date, only Winter et al. (1986b) have attempted a systematic study of respiratory COz recycling, although it has always been quantified by P. S . Nobel. Using two stem succulents (0.vulgaris and M . woodsii) and a leaf succulent (K.daigrernontiana), COzuptake and acidification (also in COzfree air) were measured under a range of temperatures and imposed drought stress (Winter et al., 1986b). Some of the results have been summarized in Table I11 (see Section III.C), but in general terms they showed that recycling of respiratory COz conserved carbon at high temperatures, with -20% of carbon being lost at 30°C (Winter et al., 1986b). Intriguingly, under drought stress AH' was greater at high night temperatures, but net carbon balance was negative (Winter et al., 1986b). Results in Table IV are presented from a series of experiments on two species of epiphytic bromeliads in which distribution is clearly related to altitudinal zonation and rainfall in Trinidad (Griffiths, unpublished results). A . nudicaulis is one of the most widely distributed CAM epiphytes, throughout most of the rainfall zones, but at upper altitudes it occurs less frequently and A . fendleri predominates (Smith et al., 1986a). A . nudicaulis forms a tank from a tight rosette of leaves, as opposed to the looser rosette and less succulent leaves of A . fendleri. Field measurements in Trinidad indicated that these two species also differ markedly in terms of expression and regulation of CAM activity, and this had in part been correlated to the degree of succulence (Griffiths et al., 1986). Plants were grown in a humid glasshouse and then transferred to growth chambers. The regulation of CAM, gas exchange and water relations were studied under two light regimes (100 and 300 pmol m-'s-l PAR), and a range of night temperatures (12,18 and 25°C). Phase I COz uptake was significantly greater at higher PAR under the 25/25 regime, but under low PAR ACOzwas similar under both temperature regimes (Table IV). For both species, ACOz increased with temperature under high PAR, but was maximal at 18°C under low PAR. Our previous studies have shown that under natural conditions A . fendleri has a lower percentage of recycling than A . nudicaulis (Griffiths et al., 1986). This also holds for these laboratory-grown plants, although under some night temperature regimes (under low PAR) recycling is similar for the two species in percentage terms, but in absolute terms dH+ was generally greater in A . nudicaulis (Table IV).

TABLE IV Respiratory C 0 2 recycling by two epiphytic bromeliuds (Aechmea fendleri and A . nudicaulis) in response to variations in PAR and night temperature

Night temp. (“C)

Dusk-dawn titratable acidity AH+ (mol m-3)

Net C 0 2 uptake dark period (m mol kg-’)

250

12 18 25

191 249 254

100

12 18 25

250

PAR (pmol m-’s-’) Aechmea fendleri

Aechmea nudicaulis

100 ~

~~

~~

Data of Griffiths, unpublished.

Recycling of respired C 0 2

Net dark respiration (pmol O2m-2 s-l)

Calculated respiration

%

6H’

86.3 123.4 134.6

10 1 0

18 2 0

0.40 1.83 3.67

0.09 0.01

242 233 243

62.3 110.0 88.2

49 6 27

117 13 67

0.52 1.56 3.77

0.60 0.07 0.34

12 18 25

160 163 250

47.3 55.2 73.1

41 32 38

65 104

0.40 2.84 5.36

0.71 0.57 1.13

12 18 25

113 123 144

27.0 58.1 45.6

52 6 37

59 7 53

0.73 2.50 5.29

0.64 0.07 0.58

53

CRASSULACEAN ACID METABOLISM

81

At 25°C under high PAR, A . fendleri took up more C 0 2 than could be accounted by the AH’; at 18”C,there was nearly the exact 2 : 1stoichiometry of AH’ : C 0 2 uptake, and at 12°C only 10% of AH’ was derived from respired C 0 2 . Under low PAR, there was much more recycling by A . fendleri at 12 and 25°C (49% and 27% respectively), but 18°C seemed optimal with only 6% recycling. A . nudicaulis showed a similar response, with 3241% recycling under high PAR, but also near-optimal6Yo recycling at 18°C under low PAR. A theoretical rate of respiration can be calculated from the discrepancy between C 0 2uptake at night and the observed degree of acidification (Griffiths et al., 1986, 1988; Luttge and Ball, 1987;Lee et al., 1988). This is shown in Table I1 in comparison with the measured rate of respiration, the latter being determined with the Hansatech leaf electrode at each of the night temperatures utilized. In only three cases ( A . fendleri, 12”C,low PAR; A . nudicaulis, 12”C,high and low PAR) does the theoretical respiration rate approach that measured with the O2 electrode. This indicates that the rates of recycling of respired C 0 2 , derived indirectly from differences in C o t uptake and acidification at night, are not unreasonable in physiological terms and emphasizes the importance of recycling to plant carbon balance. This also validates the field data for many bromeliad species (Griffiths et al., 1986). Xylem sap tension did not significantly alter between treatments (data not shown), and although bromeliad stomata may respond directly to humidity (Lange and Medina, 1979; Griffiths et al., 1986), the variation in recycling shown in Table IV can be ascribed to a direct interaction between temperature, respiration and PAR supply. Osmond et al. (1980) were the first to suggest that the recycling of respired C 0 2 during CAM may have a role in preventing photoinhibition by maintaining photosystem stability. Fluorescence kinetics (variable and low temperature) demonstrated that photochemical stability was maintained after six months of drought stress in CAM-idling plants (Osmond, 1982). It has also been demonstrated that CAM plants show photoinhibition (Nobel and Hartsock, 1983; Martin et al., 1986; Table II), and more recently that 0. basilaris may be photoinhibited throughout the year in Death Valley, California (Adams et al., 1987). There are three characteristic phenotypic variations of Bromelia humilis found under natural conditions which have been described as “yellow exposed”, “green exposed” and “green shade” (Lee et al., 1988; see also Medina et al., 1986). Measurements of variable fluorescence in the field on transplanted populations of the three phenotypes have also demonstrated that photoinhibition may be a limiting factor in the exposed plants under natural conditions (Lee et al., 1988). It would be premature to suggest that recycling is only related to photoinhibition, particularly as many laboratory studies are carried out well below PAR saturation (see Tables 11,111 and IV). In A . fendleri and A . nudicaulis there is clearly an interaction between PAR and night temperature which

82

H. GRIFFITHS

results in reduced recycling at 18°C; the significance of these processes is as yet unresolved. In view of the 1: 1stoichiometry between AH' (estimated as Amal) and An (Smith and Luttge, 1985; Smith et al. , 1986b), it is unlikely to be due to any major changes in organic acid speciation (see Section 1V.C). What is clear, however, is that integrated studies which consider leaf energy status (Koster and Winter, 1985), quantum yield (Adams et al., 1986; Adams et al., 1987) and regulation of respiratory processes in all CAM phases (Luttge and Ball, 1987) are now required. B . RECYCLING IN THE AQUATIC ENVIRONMENT

The activity of CAM as a biochemical C02-concentrating mechanism has been shown to be a response to C 0 2 limitation in the aquatic habitat (for review see Keeley, 1987; Raven, 1984; Raven et al., 1987). Although the parallel between this form of CAM and that in terrestrial plants has already been drawn (Section III.A), there are further similarities in terms of the extent of respiratory C 0 2recycling (Richardson et al., 1984; Madsen, 1987). Madsen has shown that internal C 0 2 levels within the lacunae of the submerged aquatic Littorella unifroru are maintained well above bulk water equilibration concentrations, equivalent to 1-2% C 0 2in air. Table V shows the data for the regulation of gas exchange (C02and O2by both roots and shoots) in L . unifZora grown under two PAR regimes, 50 and TABLE V Recycling of respiratory C 0 2and net carbon balance of Littorella uniflora under two P A R regimes Recycling PAR (pmol m-2 s-')

AH' (pmol (g fresh wt)-')

ACOza

%

6H'

50

42

+8

60

26

300

117

+17

71

83

Dark respirationb 29 49

Balance of gas exchange by intact plants" Roots PAR (pmol m-2 s-l)

50 300

Shoots

co2

+12 +17

-3

+0.3

Whole plant co2

0 2

+8

+11

+ 17

+20

'Determined by changes in bulk medium concentration around intact plants (pmol (g fresh wt)-') in a 12-h dark period. + = net C 0 2or 02 uptake; - = net C 0 2 or O2release. bDeterminedby O2electrode with leaf slices (pmol (g fresh wt)-'). Data of Robe and Griffiths, unpublished.

CRASSULACEAN ACID METABOLISM

83

300 pmol m-2 s-l (Robe and Griffiths, unpublished). In terms of the regulation of CAM, AH’ and AC02 were greater under high PAR, and although recycling expressed as a percentage is similar, the absolute 6H’ is much greater under high PAR (60-71%, 26-83 pmol H+ (g fresh wt)-’), in parallel with the higher rate of net dark respiration (Table V). The overall carbon balance of the plant is determined by a preponderance of root over shoot C 0 2 uptake, with C 0 2 being lost from the leaves of the lower light treatment. There was, however, an excellent agreement between total C 0 2 uptake (as measured by infrared gas analysis) and O2 (as measured by O2 electrode) for the intact plant. A constant concentration of C 0 2 was maintained within the leaf lacunae throughout day and night, at 1.03 and 0.86 mol m-3 (low and high PAR, respectively; data not shown). The significance of this C02-concentrating mechanism can easily be understood for isoetids growing in “vernal pools” which undergo large diel fluctuations in 02,C 0 2 , pH and temperature (Keeley, 1981, 1983, 1987). These are in marked contrast to the relatively constant physicochemical conditions in large oligotrophic water bodies where isoetids are also commonly found, e.g. Loch Brandy, Tayside, Scotland (Richardson etal., 1984; and also Keeley, 1987). Just as the same criteria for CAM apply to plants in terrestrial and aquatic habitats, so perhaps does the rationale behind the extent of respiratory C 0 2 recycling. In any event, the unique plasticity of CAM shows remarkably uniform responses when analysed comparatively in this way, for a pathway which has evolved independently so many times throughout the plant kingdom. C. ORGANIC ACID SPECIATION: THE NEWCASTLE HYPOTHESIS REVISITED

One of the most widely quoted reviews originally defining CAM is that of Ranson and Thomas (1960), who characterized the biochemical regulation of the diel cycle (which we now recognize as the CAM phases) but perhaps failed to stress the ecophysiological adaptations now inherent in our understanding of CAM. Perhaps we could now explain the then puzzling discrepancy in malic acid : C 0 2uptake as resulting from the recycling of respiratory c02. One other common theme throughout the review of Ranson and Thomas concerned the role of citric acid and other organic acid fluctuations, but because they “behaved erratically” the authors went on to “focus attention primarily on . . . malate” (Ranson and Thomas, 1960), the role of citric acid being put aside in favour of calculations of malic acid stoichiometry and energetics. However, an example was given whereby citric acid could account for up to 25% of the AHf in Bryophyllum calycinum (Pucher et al., 1949). At the same time, Wolf (1960) also reviewed CAM, and cited his

84

H. GRIFFITHS

work where 4 6 8 0 % of AH' could be accounted for by citric acid in the same species (Wolf, 1939). Clusia rosea (Ting et al., 1985b), Welwitschia mirabilis (von Willert, 1985) and S. acre (Schuber and Kluge, 1981) have also shown the potential for citric acid accumulation during CAM. Luttge has suggested that citric acid has often been ignored because the strict stoichiometry between malate, C 0 2 and 2H' had been demonstrated for several species (see Luttge, 1988). Then, as now, Ranson and Thomas were unable to perceive any rationale behind the biochemistry or bioenergetics of citric acid accumulation. Luttge (1988) makes a detailed analysis of the energetic costs of malic versus citric acids, and points out that there is no advantage in terms of net C 0 2 gain if hexoses or glucans are used as carbon reserves via citrate synthase. In energetic terms, citric acid is more favourable because synthesis and storage produces net ATP, as compared to malic acid (Luttge, 1988; Section 1I.C). The only function for citric acid as an osmoticum [in terms of enhanced water uptake (Luttge, 1987, 1988; Ruess and Eller, 1985; von Willert and Brinckmann, 1985; Smith et al., 1987)] could be when it is derived from storage glucans rather than hexoses (Luttge, 1988). Citric acid could be important in terms of respiratory C 0 2recycling, which seems predominantly to conserve carbon during CAM. Unfortunately, few studies have measured the combined contribution of malic + citric acid to nocturnal acidification [see Ong et ul. (1986) where shikimic acid fluctuations rather than citric acid are concomitant with malic acid in ferns]. These deficiencies are amply demonstrated in Tables I1 and 111, whereby all calculations were based on the 2H': lmal: 1 C 0 2stoichiometry. Based on our current understanding of the regulation of citrate metabolism, it is only possible for citric acid to result in net C 0 2 fixation if acyl-CoA has been derived from fatty acids by P-oxidation. Current work at Newcastle has not only indicated that P-oxidation of fatty acids can occur in plant mitochondria (Thomas and Wood, 1986; Wood et al., 1986), but has also emphasized the role of carnitine as a transmembrane fatty acid carrier. Carnitine has been shown to be important in the regulation of fatty acid transfer in both mitochondria and chloroplasts (Wood et al., 1984), and the full implications for cellular biochemical control are still being elucidated. What, then, would result if carnitine was fed to K . crenatu leaves in the transpiration stream? Figure 3 shows the results from one of a series of experiments, whereby 20 mol mP3 D/L carnitine was added to the water supply of detached leaves of K . crenata for 24 h prior to measurement of titratable acidity during the dark period (Brown, Griffiths, Thomas and Wood, unpublished results). It can be seen that carnitine significantly enhanced the rate of acidification, AH' and deacidification (Fig. 3), but it should be noted that no such response was found when 10 mol m-3 D/L carnitine was supplied (data not shown). Although an investigation into the role of citric acid in CAM is currently

CRASSULACEAN ACID METABOLISM

85

I

24.00

06.00

12.00

Fig. 3 . Stimulation of acidification by carnitine in detached leaves of Kulunchoe crenata. Petioles were incubated for 24 h prior to sampling in 50 mol m-3 phosphate buffer (I = control) or in 20 rnol m-3 carnitine in 50 mol m-3 phosphate buffer (0 + carnitine). Unpublished results of Brown, Griffiths, Thomas and Wood.

being undertaken at Newcastle in terms of regulation by mitochondrial, chloroplastic and microbody reactions, we have at the moment no conventional explanation for the data shown in Fig. 3; it is presented simply to show that there is still much to learn about the regulation of organic acid synthesis and decarboxylation during CAM.

V. CAM: DEVELOPMENT OF INTEGRATED RESEARCH The recent review by Luttge (1987) extolled the virtues of CAM research as “exemplifying the need for integration in ecophysiological work”, and the synthesis presented in this chapter certainly bears out this view. An attempt has been made to re-appraise the regulation of CAM, as found in a wide range of terrestrial and aquatic habitats, and it is hoped that the initial “reductionist” aim of simplifying and unifying our interpretation of CAM has been achieved. It is also hoped that a number of avenues of future research interest have been identified: an essential feature of any future work will be no longer simply to examine CAM activity at a single level, but to compare environmental interactions with regard to both the phenotypic and genotypic plasticity in CAM. More than ever, there is a need for rigorous comparative approaches to as many components of CAM as possible, e.g. AH’, organic acid speciation, biochemical regulation of carboxylation and respiration, gas exchange and leaf-cell water relations (e.g. Osmond et al., 1982; Osmond, 1984,1987). With CAM thus unified in

86

H.GRIFFITHS

theory, but diverse in form and function, we may be able to evaluate more fully the many facets of this intriguing metabolic adaptation.

ACKNOWLEDGEMENTS Financial support from NERC, The Nuffield Foundation and The Royal Society is gratefully acknowledged. I would also like to thank the following colleagues who have stimulated discussion and have allowed access to unpublished data: J. Brown, Dr H. S. J. Lee, Professor U. Luttge, Dr C. Martin, W. Robe, Dr J. A. C. Smith, Dr D. R. Thomas and Dr C. Wood. I am also indebted to Lynn Wilson for processing the ms with alacrity and efficiency, and to M. Green, N. M. Griffiths and C. S. Hetherington for technical assistance.

REFERENCES Acevedo, E., Badillo, I. and Nobel, P. S. (1983). Plant Physiol. 72,775-780. Adams, W. W. 111, Noshida, K. and Osmond, C. B. (1986). Plant Physiol. 81, 297-300. Adams, W. W. 111, Smith, S. D. and Osmond, C. B. (1987). Oecologia 71,221-228. AndrC, M., Thomas, D. A., von Willert, D. J. and Gerbaud, A. (1979). Planta 147, 141-144. Bartholomew, D. P. (1982). In “Crassulacean Acid Metabolism” (I. P. Ting and M. Gibbs, eds), pp. 278-294. American Society of Plant Physiologists, Rockville. Beardall, J., Griffiths, H. and Raven, J. A. (1982). J . Exp. Bot. 33,729-737. Benzing, D. H. (1980). “The Biology of the Bromeliads.” Mad River Press Inc., Eureka, California. Benzing, D. H. and Renfrow, A. (1971). Bull. Torrey. Bot. Club 98,322-327. Benzing, D. H., Friedman, W. E., Peterson, G . and Renfrow, A. (1983). Am. J. Bot. 70,121-133. Black, C. C., Carnal, N. W. and Kenyon, W. H. (1982). In “Crassulacean Acid Metabolism” (I. P. Ting and M. Gibbs, eds), pp. 51-68. American Society of Plant Physiologists, Rockville. Boston, H. L. and Adams, M. S. (1986). Oecologia 68,615622. Brulfert, J. and Queiroz, 0. (1982). Planta 154,339-343. Brulfert, J., Kluge, M., Guerrier, D. and Queiroz, 0. (1987). Planta 170,92-98. Chen, S . 4 . and Black, C. C. (1983). Plant Physiol. 71,373-378. Cockburn, W. (1983). Plant Cell Environ. 6,275-279. Cockburn, W. (1985). New Phytol. 101,3-24. Cockburn, W.,Ting, I. P. andSternberg,L. 0.(1979). PlantPhysiol. 63,1029-1032. Cockburn, W., Goh, C. J. and Avadhani, P. N. (1985). Plant Physiol. 77,53-86. Deleens, E. and Garnier-Dardart, J. (1977). Planta 135,241-248. Deleens, E. and Queiroz, 0. (1984). Plant Cell Environ. 7,279-283. Deleens, E., Treichel, I. and O’Leary, M. H. (1985). Plant Physiol. 79,202-206. de Luca, P., Alfani, A. and Virzo de Santo, A. (1977). Bot. Gaz. 138,474-478. Diaz, M. and Medina, E. (1984). In “Eco-fisiologica de Plantas CAM” (E. Medina, ed.), pp. 98-113. CIET (IVIC-UNESCO), Caracas. Earnshaw, M. J . , Carver, K. A. and Lee, J. A. (1985). Oecologia 67,486-492.

CRASSULACEAN ACID METABOLISM

87

Earnshaw, M. J., Carver, K. A. and Charlton, W. A. (1987). P h t a 170,421432. Edwards, G. E. and Walker, D. A. (1983). “C3, C4: Mechanisms and Cellular and Environmental Regulation of Photosynthesis.” Blackwell, Oxford. Eller, B. M. and Ruess, B. R. (1986). J. Plant Physiol. 125,295-309. Evans, J. K., Sharkey, T. D., Berry, J. A. and Farquhar, G. D. (1986). Aust. J. Plant Physiol. 13,281-292. Farquhar, G. D. (1980). In “Carbon Dioxide and Climate: Australian Research” (G. I. Pearman, ed.), pp. 105-110. Australian Academyof Science, Canberra. Farquhar, G. D. and Richards, R. A. (1984). Aust. J . Plant Physiol. 11,539-552. Farquhar, G. D., O’Leary, M. H. and Berry, J. A. (1982). Aust. J. Plant Physiol. 9, 121-137. Fischer, A. and Kluge, M. (1984). Planta 160,121-128. Friemert, V . ,Kluge, M. and Smith, J. A. C. (1986). Physiol. Plant. 68,353-358. Fu, C. L. and Hew, C. S. (1982). Bot. Caz. 294-297. Griffiths, H. (1984). In “Eco-fisiologia de Plantas CAM (E. Medina, ed.), pp. 45174. CIET (IVIC-UNESCO), Caracas. Griffiths, H. (1988). In “Phylogeny and Physiology of Epiphytes” (U. Luttge, ed.). Springer, Berlin. (in press). Griffiths, H. and Smith, J. A. C. (1983). Oecologia 60,176-184. Griffiths, H., Luttge, U., Stimmel, K.-H., Crook, C. E., Griffiths, N. M. and Smith, J. A. C. (1986). Plant Cell Environ. 9,385-393. Griffiths, H., Smith, J. A. C., Luttge, U., Cram, W. J. C., Lee,H. S. J.,Medina, E., Popp, M., Schafer, C. and Stimmel, K.-H. (1988). New Phytol. (in press). Groenhof, A. C., Bryant, J. A. and Etherington, J. R. (1986). Ann. Bot. 57, 689-695. Guralnick, L. J. and Ting, I. P. (1986). Oecologia 70,85-91. Guralnick, L. J., Rorabaugh, P. H. and Ting, I. P. (1984a). Plant Physiol. 75, 454-457. Guralnick, L. J., Rorabough, P. H. and Ting, I. P. (1984b). Plant Physiol. 76, 643-646. Guralnick, L. J., Ting, I. P. and Lord, E. M. (1986). Am. J. Bot. 73,336-345. Hartsock, T. L. and Nobel, P. S. (1976). Nature 262,574-576. Heun, A.-M., Gorham, J., Luttge, U., Wyn-Jones, R. G. (1981). Oecologia 50, 66-72. Hew, C. S., Ng, Y.W., Wong, S. C., Yeoh, H. H. and Ho, K. K. (1984). Physiol. Plant. 60,154-158. Holtum, J. A. M., O’Leary, M. H. and Osmond, C. B. (1983). Plant Physiol. 71, 602-609. Holtum, J. A. M., Simmons, R. A., Roeske, C., Comins, H. and O’Leary, M. H. (1984). In “Advances in Photosynthesis Research” (C. Sybesma, ed.), Vol. 111, pp. 545-548. Dr W. Junk, Dordrecht. Keeley, J. E. (1981). A m . J . Bot. 66,1079-1084. Keeley, J. E. (1983). Oecologia 58,57-62. Keeley, J. E. (1987). In “The Ecology of Amphibious and Intertidal Plants” (R. M. Crawford, ed.). Cambridge University Press. Keeley, J. E. and Bowes, G. (1982). Plant Physiol. 20,1455-1458. Keeley, J. E., Osmond, C. B. and Raven, J. A. (1984). Nature 310,634-635. Kemp, P. R. and Gardetto, P. E. (1982). Oecologia 55,149-156. Kerby, N. W. and Raven, J. A. (1985). In “Advances in Botanical Research” (J. A. Callow and H. W. Woolhouse, eds.), pp. 71-123. Academic Press, London. Kluge, M. (1977). Oecologia 29,77-83. Kluge, M. and Ting, I. P. (1978). “Crassulacean Acid Metabolism.” Springer, Berlin.

88

H . GRIFFITHS

Kluge, M., Fischer, K. and Buchanan-Bollig, B. (1982). In “Crassulacean Acid Metabolism” (I. P. Ting and M. Gibbs, eds), pp. 31-50. American Society of Plant Physiologists, Rockville. Koster, S. and Winter, K. (1985). Plant Physiol. 79,520-524. Lange, 0 .L. and Medina, E. (1979). Oecologia 40,357-363. Lange, 0 . L. and Zuber, M. (1980). Flora 170,529-553. Lee, H. S. J. and Griffiths, H. (1987). J. exp. Bot. 38,834-841. Lee, H. S. J., Smith, J. A . C., Cram, W. J., Griffiths, H., Luttge, U., Medina, E., Popp, M., Schafer, C. and Stimmel, K.-H. (1988). New Phytol. (in press). Littlejohn, R. 0. and Ku, M. S. B. (1984). Plant Physiol. 74,1050-1054. Luttge, U. (1986). Planta 168,287-289. Liittge, U. (1987). New Phytol. 106,593-629. Luttge, U. (1988). Plant Cell Environ. (in press). Luttge, U. and Ball, E. (1979). J . Membr. Biol. 47,401-422. Luttge, U. and Bzll, E. (1980). Plant Cell Environ. 3, 195-200. Luttge, U. and Ball, E. (1987). Plant Physiol. Biochem. 25,3-10. Luttge, U. and Clarkson, D. (1985). Prog. Bot. 47,73-86. Luttge, U. and Nobel, P. S . (1984). Plant Physiol. 75,804-807. Luttge, U. andsmith, J. A. C. (1984a). J . Membr. Biol. 81,149-158. Liittge, U. and Smith, J. A. C. (1984b). In “Salinity Tolerance in Plants: Strategies for Crop Improvement” (R. C. Staples and G. H. Toenniessen, eds), pp. 125150. Wiley and Sons, New York. Luttge, U., Smith, J. A . C. and Marigo, G. (1982). In “Crassulacean Acid Metabolism” (I. P. Ting and M. Gibbs, eds), pp. 69-91. American Society of Plant Physiologists, Rockville. Luttge, U., Stimmel, K.-H., Smith, J. A. C. and Griffiths, H. (1986). Plant. Cell Environ. 9,377-383. Madsen, T. V. (1987). J . Exp. Bot. 38,367-377. Martin, C. E. and Adams, W. W. I11 (1987). Photosynth. Res. 237-244. Martin, C . E. and Zee, A . K. (1983). Plant Physiol. 73,718-723. Martin, C. E., Christiansen, N. L. and Strain, B. R. (1981). Oecologia 49,322-328. Martin, C. E., Lubbers, A. E. andTeeri, J. A. (1982). Bot. Gaz. 143,491-497. Martin, C. E., Eades, C. A. and Pitner, R. A. (1986). Plant Physiol. 80,23-26. McWilliams, E. (1970). Bot. Gaz. 131,285-290. Medina, E. (1974). Evolution 28,677-686. Medina, E. (1982). 2. Pflanzenphysiol. 107,251-258. Medina, E. (1984). In “Eco-fisiologia de Plantas CAM” (E. Medina, ed.), pp. 5272. CIET (IVIC-UNESCO), Caracas. Medina, E. and Delgado, M. (1976). Photosynthetica 10,155-163. Medina, E. and Osmond, C. B. (1981). Aust. J. Plant Physiol. 8,641-649. Medina, E. and Troughton, J. H. (1974). Plant Sci. Lett. 2,357-364. Medina, E., Delgado, M., Troughton, J. H. and Medina, J. D. (1977). Flora 166, 137-152. Medina, E., Olivares, E. and Diaz, M. (1986). Oecologia 70,441-446. Mook, W. G., Bommerson, J. C. and Staverman, W. H. (1974). Earth Plan. Sci. Lett. 22, 169-176. Mooney, H. A., Troughton, J. H. and Berry, J. A. (1977). Oecologia 30,295-306. Nalborczyk, E., La Croix, L. J. and Hill, R. D. (1975). Can. J. Bot. 53,1132-1138. Neales, T. F. (1975). In “Environmental and Biological Control of Photosynthesis” (R. Marcelle, ed.), pp. 299-310. Dr W. Junk, The Hague. Neales, T. F. and Hew, C. S. (1975). Planta 123,303-306. Neales, T. F., Sale, P. J. M. and Meyer, C. P. (1980). Aust. J . Plant Physiol. 7, 375-385.

CRASSULACEAN ACID METABOLISM

89

Nishio, J. N. and Ting, I. P. (1987). Plant Physiol. 84,600-604. Nobel, P. S. (1976). Plant Physiol. 58,576582. Nobel, P. S . (1982). In “Crassulacean Acid Metabolism” (I. P. Ting and M. Gibbs, eds), pp. 260-277. American Society of Plant Physiologists, Rockville. Nobel, P. S. (1983). Plant Cell Environ. 6,711-719. Nobel, P. S. and Hartsock, T. L. (1978). Plant Physiol. 61,510-514. Nobel, P. S . and Hartsock, T. L. (1983). Plant Physiol. 71,71-75. Nobel, P. S. and Hartsock, T. L. (1984). Physiol. Plant. 60,98-105. Nobel, P. S. and Hartsock, T. L. (1986). Plant Physiol. 80,913-917. Nobel, P. S., Luttge, U., Heuer, S. and Ball, E. (1984). Plant Physiol. 75,799-803. O’Leary, M. H. (1981). Phytochemistry 20,553-587. O’Leary, M. H. and Osmond, C. B. (1980). Plant Physiol. 66,931-934. O’Leary, M. H., Treichel, I. and Rooney, M. (1986). Plant Physiol. 80,578-582. Olivares, E., Urich, R., Montes, G., Coronel, I. and Herrera, A. (1984). Oecologia 61,358-362. Ong, B. L., Kluge, M. and Friemert, V. (1986). Plant Cell Environ. 9,547-557. Osmond, C. B. (1978). Ann. Rev. Plant Physiol. 29,379-414. Osmond, C. B. (1982). In “Crassulacean Acid Metabolism” (I. P. Ting and M. Gibbs, eds), pp. 112-127. American Society of Plant Physiologists, Rockville. Osmond, C. B. (1984). In “Advances in Photosynthesis Research” (C. Sybesma, ed.), Vol. 111, pp. 557-564. Dr W. Junk, Dordrecht. Osmond, C. B. (1987). New Phytol. 106 (supplement), 161-175. Osmond, C. B., Ziegler, H., Stichler, W. and Trimborn, W. (1975). Oecologia 18, 209-217. Osmond, C. B., Bender, M. M. and Burris, R. H. (1976). Aust. J . Plant Physiol. 3 , 787-799. Osmond, C. B., Nott, D. L. and Firth, P. M. (1979). Oecologia 40,331-350. Osmond, C. B., Winter, K. and Powles, S. B. (1980). In “Plant Responses to Water and High Temperature Stress” (N. C. Turner and D. J. Kramer, eds), pp. 139-154. Wiley, New York. Osmond, C. B., Winter, K. and Ziegler, H. (1982). In “Encyclopaedia of Plant Physiology”, Vol. 12B, “Physiological Plant Ecology 11” (0. L. Lange, P. S. Nobel, C. B. Osmond and H. Ziegler, eds), pp. 479-549. SpringerVerlag, Berlin. Pearman, G. I. (1980). In “Carbon Dioxide and Climate: Australian Research” (Pearman, G . I., ed.), pp. 11-20. Australian Academy of Science, Canberra. Pucher, G. W., Vickery, H. B., Abrahams, M. S. and Leavenworth, C. S. (1949). Plant Physiol. 24,610-620. Ranson, S . L. and Thomas, M. (1960). Ann. Rev. Plant Physiol. 11,81-100. Raven, J. A. (1984). “Energetics and Transport in Aquatic Plants.” A. R. Liss, New York. Raven, J. A. (1986). New Phytol. 104,25-77. Raven, J. A. (1987). In “Biochemistry of Plants” (D. D. Davies, ed.), Vol. 13, pp. 127-180. Academic Press, London. Raven, J. A., Beardall, J. and Griffiths, H. (1982). Oecologia 53,68-78. Raven, J. A., MacFarlane, J. J. and Griffiths, H. (1987). In “The Ecology of Amphibious and Intertidal Plants” (R. M. Crawford, ed.), pp. 129-149. Cambridge University Press. Rayder, L. and Ting, I. P. (1981). Plant Physiol. 69,139-142. Rayder, L. and Ting, I. P. (1983a). Plant Physiol. 72,606-610. Rayder, L. and Ting, I. P. (1983b). Photosynth. Res. 4,203-211.

90

H.GRIFFITHS

Richardson, K., Griffiths, H., Reed, M. L., Raven, J. A. and Griffiths, N. M. (1984). Oecologia 61,115-121. Ritz, D., Kluge, M. and Veith, H. J. (1986). Planta 167,284-291. Ruess, B. R. and Eller, B. M. (1985). Planta 166,57-66. Rundel, D. W., Rundel, J. A., Ziegler, H. and Stichler, W. (1979). Oecologia 38, 45-50. Rygol, J., Buchner, K.-H., Winter, K. and Zimmermann, U. (1986). Oecologia 69, 171-175. Sale, P. J. M. and Neales, T. F. (1980). Aust. J. Planf Physiol. 7,363-373. Schafer, C. and Liittge, U . (1986). Oecologia 71,127-132. Schuber, M. and Kluge, M. (1981). Oecologia 50,82-87. Sinclair, R. (1984). Plant Physiol. 77,5943. Sipes, D. L. and Ting, I. P. (1985). Plant Physiol. 77,59-63. Smith, F. A. and Raven, J. A. (1979). Ann. Rev. Plant Physiol. 30,289-311. Smith, J. A. C. (1984). In “Eco-fisiologia des Plantas CAM” (E. Medina, ed.), pp. 30-51, CIET (IVIC-UNESCO), Caracas. Smith, J. A. C. and Liittge, U. (1985). Planta 163,272-283. Smith, J. A. C. andNobel, P. S. (1986). J. Exp. Bot. 37,1044-1053. Smith, J. A. C., Uribe, E. G., Ball, E., Heuer, S. and Liittge, U. (1984a). Eur. J . Biochem. 141,415-420. Smith, J. A. C., Uribe, E. G., Ball, E. and Liittge, U. (1984b).Planta 162,299-304. Smith, J. A. C., Griffiths, H., Bassett, M. and Griffiths, N. M. (1985). Oecologia 67, 475-485. Smith, J. A. C., Griffiths, H. andLiittge, U. (1986a). Plant Cell Environ. 9,359-376. Smith, J. A. C., Griffiths, H., Liittge, U., Crook, C. E., Griffiths, N. M. and Stimmel, K.-H. (1986b). Plant Cell Environ. 9,395-410. Smith, J. A. C., Schulte, P. J. and Nobel, P. S. (1987). Plant Cell Environ. 10, 639-648. Smith, T. L. and Eickmeier, W. G. (1983). Oecologia 56,374-380. Spalding, M. H., Schmitt, M. R., Ku, J. B. and Edwards, G. E. (1979). Plant Physiol. 63,738-743. Sternberg, L., De Niro, M. J. and Ting, I. P. (1984a). Plant Physiol. 74,104-107. Sternberg, L., De Niro, M. J. and Johnson, H. B. (1984b). Plant Physiol. 74, 537-561. Steudle, E., Smith, J. A. C. and Luttge, U. (1980). Plant Physiol. 66,1155-1163. Stimmel, K.-H. (1984). In “Eco-fisiologia de Plantas CAM” (E. Medina, ed.), pp. 73-82. CIET (IVIC-UNESCO), Caracas. Szarek, S. R. and Troughton, J. H. (1976). Plant Physiol. 52,539-541. Szarek, S . R., Johnson, H. B. and Ting, I. P. (1973). Plant Physiol. 52,539-541. Teeri, J. A. (1982a). In “Biochemical Aspects of Evolutionary Biology” (M. H. Nitacki, ed.), pp. 93-130. University of Chicago Press, Chicago. Teeri, J. A. (1982b). In “Crassulacean Acid Metabolism” (I. P. Ting and M. Gibbs, eds), pp. 224-259. American Society of Plant Physiologists, Rockville. Teeri, J. A. (1984). Oecologia 64,68-73. Teeri, J. A., Tonsor, S. J . and Turner, M. (1981). Oecologia 50,367-369. Teeri, J. A., Turner, M. and Gurevitch, J. (1986). Plant Physiol. 81,678-680. Thomas, D. R. and Wood, C. (1986). Planta 168,261-266. Ting, I. P. (1985). Ann. Rev. Plant Physiol. 36,595-622. Ting, I. P. and Burk, J. A. (1983). Plant Sci. Lett. 32,279-285. Ting, I. P. and Gibbs, M. (1982). “Crassulacean Acid Metabolism.” American Society of Plant Physiologists, Rockville. Ting, I. P. and Hanscom, Z. (1977). Plant Physiol. 59,511-514.

CRASSULACEAN ACID METABOLISM

91

Ting, I. P. and Rayder, L. (1982). In “Crassulacean Acid Metabolism” (I. P. Ting and M. Gibbs, eds), pp. 193-207. American Society of Plant Physiologists, Rockville. Ting, I. P. and Sipes, D. (1985). In “Nitrogen Fixation and C0,Metabolism” (P. W. Ludden and J. E. Burris, eds), pp. 371-378. Elsevier, Holland. Ting, I. P., Sternberg, L. 0. and De Niro, M. J. (1983). Plant Physiol. 71,677-679. Ting, I. P., Bates, L., Sternberg, L. 0. and De Niro, M. J. (1985a). Plant Physiol. 78,246-249. Ting, I. P., Lord, E. M., Sternberg, L., Da, S. L. andDe Niro, M. J. (1985b). Science 229,969-971. Tinoco-Ojanguren, C. and Vazquez-Yanes, C. (1983). Bol. SOC.Bot. Mex. 45, 150-153. Tomlinson, P. B. (1969). In “Anatomy of the Monocotyledons” (C. R. Metcalfe, ed.), 111. Commelinales-Zingiberales. Oxford University Press, Oxford. Virzo de Santo, A., Alfani, A., Russo, G. and Fioretto, A. (1983). Bot. Gaz. 144, 342-346. Virzo de Santo, A., Ligrone, R . , Alfani, A., Fioretto, A. and Russo, G. (1984). Plant Cell Environ. 7,105-112. von Willert, D. J. (1985). In “Advances in Botanical Research” (J. A. Callow and H. W. Woolhouse, eds), pp. 157-191. von Willert, D. J. and Brinckmann, E. (1985). Ber. Deutsch Bot. Ges. Bd. 98, 455-464. von Willert, D. J., Eller, B. M., Brickmann, E . and Baasch, R. (1982). Oecologia 55,21-29. Wagner, J. and Larcher, W. (1981). Oecologia 50,88-93. Walker, D. B., Gysi, J., Sternberg, L. and De Niro, M. J. (1983). Science 220, 419-421. Winter, K. (1974). Ber. Deutsch Bot. Ges. Bd. 86,467-476. Winter, K. (1982). In “Crassulacean Acid Metabolism’’ (Ting, I. P. and Gibbs, M., eds), pp. 153-169. American Society of Plant Physiologists, Rockville. Winter, K. (1985). In “Photosynthetic Mechanisms and the Environment” (J. Baxter and N. R. Baker, eds), pp. 329-387. Elsevier, Amsterdam. Winter, K. and Luttge, U. (1976). In “Ecological Studies”, Vol. 19, “Water and Plant Life” (0.L. Lange, L. Kappen and E.-D. Schultze, eds), pp. 323-334. Springer-Verlag, Berlin. Winter, K. and Schramm, M. J. (1986). Plant Physiol. 82,173-178. Winter, K. and Tenhunen, J. D. (1982). Plant Physiol. 70,1718-1722. Winter, K. and von Willert, D. J. (1972). 2. Pflanzenphysiol. 267,166-170. Winter, K., Troughton, J. H., Evanari, M., Lauchli, A. and Luttge, U. (1976). Oecologia 25,125-143. Winter, K., Luttge, U., Winter, E. and Troughton, J. H. (1978). Oecologia 34, 225-237. Winter, K., Edwards, G. E. and Holtum, J. A. M. (1981). Plant Physiol. 68,355-357. Winter, K., Wallace, B. J., Stocker, G. C. and Roksandic, Z. (1983). Oecologia 57, 129-141. Winter, K., Medina, E., Garcia, V., Mayoral, M. L. and Muniz, R . (1985). J . Plant Physiol. 113,73-78. Winter, K., Osmond, C. B. and Hubick, K. T. (1986a). Oecologia 68,224-230. Winter, K., Schroppel-Maier, G. and Caldwell, M. M. (1986b). Plant Physiol. 81, 390-394. Wood, C., Noh, Hj Jalil, J., MacLaren, I . , Yong, B. C. S . , Ariffin, A., McNeil, P. H., Burgess, N. and Thomas, D. R. (1984). Planra 161,255-260.

92

H.GRIFFITHS

Wood, C., Burgess, N. and Thomas, D. R. (1986). Planta 167,54-57. Woodhouse, R. M.,Williams, J. G . and Nobel, P. S. (1980). Am. J . Bot. 67, 1179-1 185. Wolf, J. (1939). Plunta 29,450467. Wolf, J. (1960). in “Encyclopaedia of Plant Physiology 12” (Ruhland, W . , ed.), pp. 809-889. Springer, Berlin. Zeiger, E. (1983). Ann. Rev. Plant Physiol. 34,441475. Ziegler, H.,Osmond, C. B . , Stichler, W. and Trimborn, P. (1976). Planta 128, 85-92.

Potassium Transport in Roots

LEON V . KOCHIANa and WILLIAM J . LUCASb a

U.S.Plant. Soil and Nutrition Laboratory. USDA.ARS. Cornell University. Ithaca. New York. USA Department of Botany. University of California. Davis. California. USA

I. I1.

Introduction

. . . . . . . . . . . . . . . . .

Plasma Membrane Transport of K+ in Roots . . . . . . . A . Early Work: the Carrier-Kinetic A proach . . . . . . B . Are Root K+ Fluxes Coupled to H ? . . . . . . . . C . Uptake at High K+ Concentrations: the Linear Component D . Summary . . . . . . . . . . . . . . . .

P

I11.

Redox-coupledPlasmalemmaTransportof K+ . . A . Influence of Exogenous N A D H on K+Influx . B Membrane Transport and the Wound Response C. DevelopmentofanIntegratedNADHModel .

V.

VI .

.

.

.

.

.

94 94 99 120 128

. . . . . . . . . . . . . . .

129 130 131 132

. . . . . . . . .

136 137 140 143 145

Regulation of K+ Fluxes within the Plant . . . . . . . . . A . Allosteric Regulation of K+ Transport . . . . . . . . B . K+ Cycling within the Plant: an Integration of Regulatory . . . . . . . . . . . . . . . . Mechanisms

151 152

.

IV .

94

Radial K+ Transport to the Xylem . . A . Site of K+ Entry into the Symplasm B . Radial Pathway . . . . . . C . Lag Phase in Xylem Loading . . D . K+ Transport into the Xylem . .

Future Research and Prospects

. . . .

. . . .

. . . .

. . . .

. . . . .

. . . .

. . . .

. . . .

. . . .

. . . . . . . . . . .

Advances in Botanical Research Vol . 15

. . . .

163 169

Copyright 0 1988 Academic Press Limited All rights of reproduction in any form reserved .

ISBN 0-12-005915-0

93

94

L. V. KOCHIAN AND W. J. LUCAS

I. INTRODUCTION The involvement of Kf in a number of plant functions (enzyme activation, maintenance of turgor, protein synthesis, etc.), along with the relatively high permeability of plant cell membranes to K’, has been the basis for the study of K+ transport as a model system for plant ion transport. Numerous investigations have been conducted over the past 40 years, aimed at furthering our understanding of the mechanism(s), energetics, and cellular location of K+ transport systems in roots. Despite the surfeit of literature in these areas, it can be said with some confidence that the processes by which K+ ions are transported into and across the root are still far from being fully resolved. This is due, in part, to the complex nature of the root, with its various cell types and tissues. Additional complications arise from the fact that the individual cells are coupled (both electrically and physiologically) via plasmodesmata, and also simply from the complex nature of individual cells, with their multiple compartments and subcompartments. In the first part of this chapter, we will outline and summarize what is known (or at least thought to be known) concerning mechanistic aspects of K+ transport in roots. Once this background has been established, the regulation of these transport processes can be discussed in the context of the integration of these cellular processes at the organ and whole plant level.

11. PLASMA MEMBRANE TRANSPORT OF K+ IN ROOTS A. EARLY WORK: THE CARRIER-KINETIC APPROACH

Over the past half century, most of the studies that have been conducted concerning ion absorption by plants have generally utilized three basic types of plant material: the giant algae such as Chum, Nitellu, and Vuloniu, slices cut from storage tissue of beet, potato, and carrot, and either intact roots or roots excised from seedlings. The use of excised roots as research material can be traced to the classical paper of Hoagland and Broyer (1936). They found that the roots of barley seedlings grown in dilute salt solutions exhibited extremely high initial rates of ion accumulation. Because radioisotopes had not yet been introduced, the ability of these roots to maintain high rates of accumulation made them very useful experimental material. Hence, the now well-known “low-salt roots” characterized by low salt content, high sugar content, and a large capacity for ion transport, became widely used in research. It was during this era that many of the basic concepts of membrane transport were developed. Much of the work conducted with low-salt roots concerned the absorption of K+ and other alkali cations (for the reasons discussed above). Although the concept of the lipid bilayer nature of the plasmalemma had not yet been fully developed, many researchers were

95

POTASSIUM TRANSPORT IN ROOTS

beginning to realize that plant (and animal) cells were bounded by a non-aqueous layer that was relatively impermeable to electrolytes (Osterhout, 1931). Furthermore, the rapid absorption of Kf by plant cells led a number of investigators to hypothesize that special structures must exist that would facilitate K+ uptake across this outer boundary (Jacobsen and Overstreet, 1947; Jacobsen et al. , 1950; Osterhout, 1950,1952). The concept that, in the plasmalemma, specific carriers were involved in K+ uptake was developed further by Epstein and co-workers. When K+(86Rb+)uptake from dilute solutions into excised barley roots was studied, the kinetic profile approximated a rectangular hyperbola. This response was quite similar to that found in classical enzyme kinetic studies [Fig. 1 (see low concentration range)]. Epstein and Hagen (1952) were the first to apply Michaelis-Menten enzyme kinetics to ion transport. They postulated that specific alkali cation transport systems operated in a fashion analogous to substrate-specific enzymes. They went on to show, for the uptake of K+ and other alkali cations, that at higher external concentrations their observed kinetics deviated from classical Michaelis-Menten form (Leggett and Epstein, 1956; Epstein et af., 1963). Saturation was attained at low external K+ concentrations, but then at higher concentrations the curves appeared to reach a second level of saturation (Fig. 1). This biphasic pattern was labelled the “dual isotherm of uptake” by Epstein, and was hypothesized to be due to the operation of two separate classes of carriers in the plasmalemma. In the low K+ concentration range (<0.5 mM), mechanism I was hypothesized to be a K+ transport system with a high affinity for K+

25 I

I

K+concentration ( mM )

Fig. 1 . Rate of K+ absorption in barley roots as a function of K+concentration. Note that the concentration scale has been changed after 0.2 mM K’. Data from Epstein et al. (1963).

96

L. V . KOCHIAN AND W. J. LUCAS

(particularly with respect to Na’). In the higher concentration range, mechanism I1 was proposed to operate as a low-affinity, high-velocity transport mechanism that was less specific for K’. Further studies in the mechanism I1 range suggested that the kinetic profile was not a smooth curve, but was characterized by multiple inflections (Elzam et al., 1964; Epstein and Rains, 1965). This pioneering work by Epstein and his colleagues had a significant influence on the field of plant membrane transport. Subsequently, publications emerged from many laboratories indicating that the uptake of a number of inorganic and organic solutes in most plant tissues followed complex kinetics. These complex kinetics were, for the most part, interpreted to be the result of two or more transport systems. Subsequent kinetic analyses suggested that each transport system (for a particular solute) exhibited distinctly different Michaelis-Menten kinetics. The great volume of published work reflected the interest and importance associated with this field. A number of controversies arose concerning the kinetic aspects of K+ absorption (and the uptake of other solutes). Conflicting hypotheses were presented concerning such aspects as the number of phases and the cellular and subcellular localization of each phase (for a review see Laties, 1969; Epstein, 1976). The most direct challenge to Epstein’s “dual isotherm” hypothesis was introduced in the early 1970s by Nissen, who presented an alternative interpretation for multiphasic uptake kinetics (Nissen, 1971, 1973, 1974). Nissen’s approach was to analyse both his own and other researcher’s kinetic data for solute uptake (uptake versus solute concentration) by performing various kinetic transformations of the primary data. These transformations (usually Lineweaver-Burk reciprocal plots) yield linear relationships for uptake kinetics that followed Michaelis-Menten relationships. Because the kinetics for solute uptake in plants are almost always complex, the reciprocal plots, quite naturally, yielded non-linear transformations. Nissen has fitted the transformed data by means of a statistical computer program and has found that the best fit, generally, for uptake of Kf and any other solutes that he and others have studied, is a series of adjacent linear segments (e.g. Nissen and Nissen, 1983; Nissen, 1987). Nissen claims that this type of analysis is strong evidence for the operation of a single complex transport system located in the plasmalemma. As external K+ concentration is increased, this putative transport system is thought to undergo abrupt transitions at discrete external K+ concentrations. Within each phase, the transport system is thought to obey Michaelis-Menten kinetics, with the kinetic constants for each phase usually increasing in a fairly regular manner. This approach is somewhat controversial and has been subjected to some fairly strong criticisms (Walker, 1974; Wyn Jones, 1975; Borstlap, 1981a,b, 1983; Kochian and Lucas, 1982b).

97

POTASSIUM TRANSPORT IN ROOTS

1

-0

1

2

3

4

5

6

7

6

9

1

0

K+concentration ( mM) Fig. 2. Potassium (=Rb') influx as a function of Kf concentration in the experimental medium, obtained on corn root segments that were grown in 5 mM KCI + 0.2 mM CaS04 (high-salt status), Flux determinations were performed using the apparatus developed by Kochian and Lucas (1982a). Data points represent the mean k SE (points which lack error bars do so because the standard errors were smaller than symbols used). Linear regression performed on data points for K+ concentrations from 1.0 to 10 mM gave a value of 0.334 wmol (g fresh weight)-' h-' for the first-order rate coefficient of the linear component (the regression coefficient was 0.999).

Some of these criticisms, and an overall critique of the carrier-kinetic approach to Kf transport in roots, will be discussed in a later section. Most of the work of this era centred around the deduction of ion transport mechanisms from either the shapes of kinetic curves (influx versus concentration) or from the shapes of curves resulting from various mathematical transformations of the kinetic data. Many of these published data are of limited value because, in many cases, the kinetic profiles lack a sufficient number of data points for the concentration ranges tested. Also, data replication was often lacking. Kochian and Lucas (1982a) addressed this problem by developing an experimental apparatus which enabled them to generate large numbers of data points on the [ S ] axis, with enough replicates to obtain sufficiently precise values. As illustrated in Fig. 2, no abrupt discontinuities were observed for K+ transport into corn roots. Kinetic curves for Kf influx into roots of both low- and high-salt-grown corn seedlings were found to be smooth and nonsaturating; the curves approached linearity at concentrations above 1 mM and exhibited no tendency towards saturation at concentrations up to 50 mM. The kinetics for K+ transport could be resolved into saturable and nonsaturating, or first-order kinetic components. The saturable component could be specifically inhibited by application of sulphhydryl reagents (Fig. 3), whereas the nonsaturating component could be specifically inhibited by either the application of K+ channel-blocking

L. V. KOCHIAN AND W. J. LUCAS

98

1

--PI

"

0

2

4 6 Rb+concentration ( m M )

8

10

Fig. 3. Kinetic curves for 86Rb+influx into high-salt-grown corn root segments. In (a), the control curve (0)has been separated into its saturable and linear components. In (b), the influence of N-ethyl maleimide (NEM) exposures on %Rb+ influx is depicted. Corn root segments were pretreated with 0.3 mM NEM for 10 or 30 s, then washed for 10 min in 1 mM dithioerythritol (DTE) prior to %Rb+uptake. Data from Kochian and Lucas (1982a).

agents, or by substituting C1- in the uptake solution with other anions (Kochian et al., 1985).The discrepancy between Epstein's mechanism I1 and the first-order kinetic component observed in corn roots, will be addressed in Section 1I.C. Although the carrier-kinetic approach has provided important information on the mechanisms of ion transport, clearly there are limitations associated with this approach. In studying root ion transport, one is dealing with a system consisting of many cell types, each potentially in a different physiological state, with the problems of unstirred layers and diffusion limitation further confounding data interpretation. In studying the concentration dependence of uptake, the bulk solution ion concentration is generally taken as representing the substrate level available for transport, despite the knowledge that the cation and anion concentrations at the surface of the plasmalemma are probably considerably different from those in the bulk solution. This point is highlighted by the recent work of Newman et al. (1987), in which ion-selective microelectrodes were utilized to analyse the

POTASSIUM TRANSPORT IN ROOTS

99

electrochemical gradients for H+ and K+ in the unstirred layer near the surface of corn roots. Newman etal. (1987) found that in the presence of low , steady-state concentration at the root levels of K+ (less than 50 p ~ )the surface was about one-half that of the bulk solution, due to depletion arising from net K+ uptake into the root. This considerable difference may be magnified further within the cell walls and at the surface of the plasmalemma, where additional diffusion limitation and ionic effects may have a dramatic influence on cation and anion activities. These points are of particular significance when mechanistic models, such as those presented by Nissen, are based solely on small differences between and within uptake kinetic curves. ' ? B. ARE ROOT K+ FLUXES COUPLED TO H

It has long been accepted that K+ influx into roots is either coupled to the transport of H+ or, at the very least, influenced by H+ activities (or activity gradients) associated with the K+ transport system. Like most areas of plant membrane transport, the concept that K+ and H+ fluxes are coupled has been controversial. Over the past 40 years, various models have been presented to explain the mechanism(s) by which K+ uptake is coupled to H+ efflux. The earliest models proposed that H+ efflux was dependent upon K+ uptake, or more specifically, depended upon cation uptake that was in excess of anion absorption. The H+ excreted by the root had a compensatory role; that is, they were involved in maintaining charge balance across the plasmalemma. Organic acid metabolism and cytoplasmic pH regulation were also involved in this model (Ulrich, 1941;Jacobsen et al., 1950;Jackson and Adams, 1963;Torii and Laties, 1966; Hiatt, 1967a,b). More recent models have reversed the dependency between K+ and H+ fluxes. The influence of Mitchell's chemiosmotic hypothesis (Mitchell, 1970) has led some researchers to hypothesize that active H+ efflux is the primary event (and driving force) for the subsequent (and dependent) K+ uptake. Although most plant transport physiologists now generally accept that K+ uptake is dependent on H+ efflux, controversy still exists concerning the degree of this dependence. There appear to be two groups of adherents, with some researchers subscribing to a direct chemical coupling (Poole, 1974; Hanson, 1977; Leonard and Hanson, 1972; Lin and Hanson, 1976; Cheeseman and Hanson, 1979a,b; Cheeseman et al., 1980), and others arguing that the two fluxes are indirectly coupled, through the electrical component of the protonmotive force generated by the electrogenic, plasmalemma H+ATPase (e.g. Pitman et al., 1975; Marrk, 1977, 1979; Bellando et al., 1979; Bellando and Trotta, 1980). Glass and co-workers have severely questioned the existence of either a direct or an electrical coupling of the two fluxes, and have again suggested that the maintenance of charge balance may be the primary role for H+ fluxes in relation to K+ uptake (Glass and Siddiqi, 1982; Siddiqi and Glass,

100

L. V. KOCHIAN AND W. J. LUCAS

1984). However, recent studies with Neurospora (Rodriguez-Navarro and Ramos, 1986; Rodriguez-Navarro et al., 1986; Blatt and Slayman, 1987; Slayman et al., 1988) and Zea mays roots (Kochian and Lucas, 1987; Kochian et al., 1987, 1988; Newman et al., 1987) have provided evidence consistent with the hypothesis that a K+-Hf co-transport system may operate in the plasmalemma of fungal and higher plant cells. The evidence supporting the various models based on different modes of coupling between K+ and H+ fluxes will now be considered in more detail. 1. Unequal CationlAnion Uptake Charge Balanced by H' It has long been recognized that both intact and excised roots can maintain disparate rates of cation and anion absorption, depending on the composition of the salt solution bathing the roots. Even in early studies on root ion absorption, it was observed that, for example, excess cation absorption could occur from a solution containing K2S04,while excess anion uptake was often seen in solutions containing CaCl, or Ca(NO& (Brooks, 1929; Hoagland and Broyer, 1940). Uptake from KCl solutions usually results in a balance between cation and anion fluxes. During these early studies, it was recognized that the processes which transport net charge across the plasmalemma could not continue indefinitely, due to the axiom that electroneutrality must be maintained within living cells. (Additionally, we now realize that such processes could not continue in an uncompensated fashion due to the dielectric properties and limited capacitance of biological membranes.) Therefore, compensatory reactions must occur, in order to maintain charge balance both across membranes and within cells. Ulrich (1941) was the first to propose an interrelationship between unequal catiodanion uptake, cytoplasmic pH, and organic acid metabolism. He found that under conditions of excess cation uptake, organic acid content increased in proportion to the cation excess; the converse was found for uncompensated anion absorption. Ulrich hypothesized that under conditions of excess cation (usually K f ) absorption, H+ were exchanged for K+ ions. This could give rise to an alkalinization of the cell sap, stimulating the synthesis of organic acids which would help to buffer internal pH. Subsequent work from a number of laboratories extended these initial speculations. Several researchers dealt with the mechanism of K+ (and other cation) uptake in terms of a K+-H+ exchange and anion absorption as an anion-OH- antiport. Both processes would result in a trend towards electroneutrality (Jacobsen et al., 1950; Jackson and Adams, 1963). It was also amply established that during excess Kf uptake, the pH and buffering capacity of expressed root sap increased, as did dark COP fixation and organic acid (usually malate) synthesis (Jacobson and Ordin, 1954; Jacobson, 1955; Torii and Laties, 1966; Hiatt, 1967a; Hiatt and Hendricks, 1967; Jacoby and Laties, 1971). It is still not clearly understood how excess K+ absorption stimulates organic acid synthesis in the root. It has been proposed

POTASSIUM TRANSPORT IN ROOTS

101

that the K+-H+ exchange involved in charge balance should increase cytoplasmic pH, and such a pH shift has been suggested from pH measurements of total expressed root sap. This pH shift would “turn on” carboxylating enzymes in the cytoplasm (PEP carboxylase + malate dehydrogenase) and result in the synthesis of malic acid. Subsequent dissociation of the malic acid carboxyl groups would yield H+ that could both help maintain cytoplasmic pH and permit continued K+-H+ exchange (Hiatt, 1967a,b). This model has been promoted as a regulatory mechanism involved in the maintenance of cytoplasmic pH (Davies, 1973; Raven and Smith, 1974; Smith and Raven, 1976). Alternatively, other models have been proposed which involve alterations in the levels of substrate (C02 versus HCO;) available for organic acid synthesis (Jacoby and Laties, 1971), feedback regulation of PEP carboxylase by cytosolic malate levels (Osmond, 1976), and ionic effects on carbon metabolism (Osmond, 1976; Osmond and Greenway, 1972; Schnabl, 1987). The above discussion has dealt more directly with regulatory aspects of K+ (and general cation) uptake, and not specifically with mechanisms of K+ absorption. It was important to introduce this topic in this section because this area of research was instrumental in developing the concept that K+ and H+ fluxes were, in some fashion, correlated. For a more detailed consideration of the interactions of ion absorption, carbon metabolism, and cytoplasmic pH regulation, the reader is referred to reviews by Osmond (1976), Smith and Raven (1976,1979), and Davies (1979).

2. Direct Coupling: K + - H + Antiport It has been well documented that plant cells accumulate K+ in the cytoplasm and generally maintain electrochemical potential gradients for K+ across the plasmalemma (Lauchli and Pfluger, 1979;Leonard, 1984,1985). Numerous studies have demonstrated that K+ uptake into root tissue is extremely sensitive to metabolic inhibitors. Uptake is energy-dependent, and a close correlation with cellular levels of ATP (modified through the application of metabolic inhibitors) has been demonstrated (Petraglia and Poole, 1980). However, the problems of elucidating underlying mechanisms on the basis of such studies is all too obvious. Nevertheless, by combining electrophysiological, kinetic, and inhibitor studies, it has been possible to demonstrate that at least at low external K+ levels, uptake is a thermodynamically active process (see Pitman, 1976; Cheeseman and Hanson, 1980). Additionally, compelling evidence has been accumulated over the past 20 years that electrogenic transport processes operate across both the plasmalemma and tonoplast of lower and higher plant cells. Further, it is now well accepted that these transport systems are membrane-bound ATPases that are involved (at the very least) in the active transport of H+ (for reviews see Poole, 1978; Spanswick, 1981; Leonard, 1982; Serrano, 1984; Marrk and Ballarin-Denti, 1985;Sze, 1985). During this same period, the chemiosmotic

102

L. V. KOCHIAN AND W. J. LUCAS

Pathway 1

K+

ADP

+ K+pi>uL

Pathway 2

Fig. 4. Possible mechanisms for K+ transport into plant cells. The plasma membrane H+-ATPase produces a pH gradient and a membrane potential during ATP hydrolysis. K+ transport can be either driven by the electrical component of the membrane potential with movement occurring through a membrane protein (pathway 1) or by direct transport via a K+/H+-ATPase (pathway 2). It is also possible that both pathways can occur. Modified from Briskin (1986), with permission.

hypothesis, developed by Mitchell (1970) to explain energy transduction in mitochondria and chloroplasts, had alerted plant physiologists to the possibility that H+ gradients could be coupled to other ion fluxes. Because plant transport scientists already had observed associations between K+ uptake and H+ efflux, it was only natural that all of these factors would influence thinking about active K+ influx. Hence, a number of mechanistic models were developed that coupled Hf to Kf fluxes. As mentioned previously, two types of coupling were hypothesized, and these are detailed in Fig. 4. For a number of different root tissues it has been demonstrated that as external pH is increased, a concomitant increase in K+ influx occurs (e.g. Poole, 1966, 1974; Lin, 1979; Glass and Siddiqi, 1982). This apparent pH dependency for washed corn root segments and intact barley roots is shown in Fig. 5 . Poole (1974), working with slices from red beet roots, observed that as external pH was increased from 5.5 to 8.0, there were parallel stimulations of Hf efflux and K+ uptake, and a hyperpolarization of the Em. He was probably the first to suggest that the plasmalemma electrogenic proton pump of plant cells was actually a Kf-Hf exchange ATPase, facilitating K+ uptake in response to active Hf efflux.

103

POTASSIUM TRANSPORT IN ROOTS

3.0 n c

I

c

c

/

!- 2.5c

L

c

/

u)

Q

--

t m 2.00

1.5-

X

r E

+.-

=+

z”

O

O

Intact barley roots

-6

-5

/ /

E

5 -a

---“--3--

-4

1 Excised corn roots

1.0- /

.3

:.2

:? pH ofabsorption solution

Fig. 5 . Effect of external pH on K+(‘‘Rb’) influx into 4-h-washedcorn root segments (0) and intact barley roots (m). Where present, error bars represent +SE. Note the apparent pH insensitivity from pH 6-9, (Scale on right-hand side for intact barley roots.) 0: Data replotted from Lin (1979). .: Data replotted from Glass and Siddiqi (1982).

Hanson and co-workers, working with intact and excised corn root segments, have been among the more vocal proponents of a direct K+-H+ exchange mechanism (Hanson, 1977).Lin and Hanson (1976) demonstrated that the sulphhydryl-containing compound dithioerythritol (DTE) could “activate” a passive and stoichiometric (1: 1) K+-H+ exchange system. However, as they note, in their experiments this antiporter was involved in the release of K+ from corn root cells. They speculated that this passive antiport mechanism could be involved in K+ uptake under conditions where the thermodynamic gradients favoured K+ influx (K+ directed inwardly andor H+ out). However, this seems to be unlikely for the thermodynamic gradient for H+ would rarely be directed out of the cell. Lin and Hanson presented a model in which an active K+/H+-ATPase(for which no evidence was presented) was coupled to the observed DTE-induced passive K+-H+ exchange system as well as to an active anion uptake system (anion-OHantiport). An alternative function could exist for the passive K+-H+ antiport system, if it is physiologically significant. As the authors mention, this system could be involved in osmotic regulation and/or control of internal K+ under conditions of high K+ status. A precedence exists for such a situation. Active K+ uptake into the photosynthetic bacterium, Chromatiurn vinosum, is mediated via a K+-ATPase that exhibits a high affinity for K+, while a low-affinity K+-H+ antiport system (which mediates K+ efflux) is involved

104

L. V. KOCHIAN AND W. J. LUCAS

in osmotic balance and cytoplasmic pH regulation under conditions of high internal K+ (Davidson and Knaff, 1981,1982). In a subsequent paper investigating ion transport (K+ and Pi) in corn root segments, the differential effects of pH and several chemical modifiers on K+ and Pi transport were shown and used to support the existence of separate (and active) K+-H+ and Pi-OH- exchange systems (Lin, 1979). The apparent stimulation of K+ uptake at high pH values (Fig. 5 ) was correlated with increased values of H+efflux. These results were considered to be consistent with an active exchange system whose H+ efflux activity was increased in response to a reduction in the A& opposing H+ extrusion as the pH was increased. Lin found that the fungal toxin, fusicoccin, which has been shown to stimulate both H+ efflux and cation uptake, and to hyperpolarize the Em of plant cells (Marrk, 1979), rapidly stimulated K+ uptake while having little influence on phosphate transport. Additionally, the ATPase inhibitor diethylstilbestrol (DES) and the H+ ionophore carbonyl cyanide 4-(trifluoromethoxy)hydrazone (FCCP) both immediately inhibited Kf uptake while having an apparently less direct effect on Pi uptake. These results were taken by Lin asfurther evidence that K+ uptake occurs by an active, electrogenic K+-H+ exchange (even though no evidence for such a mechanism was presented in earlier papers). It should be noted that these results are quite circumstantial, and can be used to support other models for K+ transport, as will be discussed later. Cheeseman and Hanson (1979a,b) conducted a detailed study investigating the influence of anoxia and uncouplers on K+ influx and Emover a range of K+ concentrations. They analysed their data using the Goldman equations [used to describe passive ionic fluxes (Goldman, 1943; Hodgkin and Katz, 1949; Hodgkin and Huxley, 1952)] in order to determine the contributions of passive flux to the total K+ influx. They proposed the model illustrated in Fig. 6, which indicates that at low external K+ concentrations, active K+ influx is coupled to the H+-translocatingATPase. As K+ concentrations are increased in the mechanism I range, saturation of this carrier occurs and the increasing inward K+ current would have a depolarizing effect. As external K+ increases into the mechanism I1 range, K+ influx occurs via a passive electrophoretic uniport. However, in this concentration range, external K+ is thought to have a stimulatory effect on the ATPase. Cheeseman and Hanson originally suggested that both K+ transport functions were carried out by the same system, utilizing a single type of H+-ATPase. However, in a later publication (Cheeseman et al., 1980) they revised their model. After analysing the influence of ATPase inhibitors on Em and K+ influx, they hypothesized that at low K+ levels an electrogenic system dominates that is sensitive to ATPase inhibitors, and is associated with active K+ influx and H+ efflux. A second electrogenic system was proposed which operates at all external K+ concentrations and becomes dominant at higher K+ concentrations. This system drives passive K+ influx,

105

POTASSIUM TRANSPORT IN ROOTS Mechanirm I

EAT: [K+I,c 1mM

"1 outride

a I

plarmalrmma

Em

EK

EpL wllh IK'1, t

K+ influx c o u p l r d l o ATParr cytoplarm

H+ M r c h a n i r m I I [K+&> 1mM

Em

>EK

E, i w l t h IK'I,

t

K'influx via uniport

Fig. 6. Schematic model developed by Cheeseman and Hanson (1979b) to explain the role of the electrogenic H+-ATPase in Kf transport across the corn root plasmalemma when the external K+ concentration is within the mechanism I and I1 range. Note that within the mechanism I1 K+ concentration range, it is proposed that external K+ has a stimulatory effect on the rate at which H t are pumped out by the ATPase. E, is the electrogenic component of the Emcontributed by the H+-ATPase.

and is insensitive to ATPase inhibitors. Essentially, what Cheeseman and co-workers were proposing was a model that incorporates both types of coupling illustrated in Fig. 4. At low external K+ levels, active Kf uptake would occur via a K+-H+ exchange ATPase (pathway 2 of Fig. 4). At higher external Kf concentrations, passive K+ uptake would be mediated by a K+ channel, with the driving force arising primarily from the electrical potential difference developed across the plasmalemma by a H+-ATPase. Recently, evidence in support of this model was presented following a re-evaluation of the electrogenic nature of K+-H+ fluxes in corn roots (Thibaud et al., 1986). It should be noted that if Kf uptake is active at low K+ levels, then the K+ transport system must be coupled either directly to an ATPase (K+-ATPase or K+-H+ exchange ATPase), or to the transmembrane proton electrochemical gradient (PMF) via a Kf-H+ co-transport system. Indirect coupling (pathway 1, Fig. 4)can only be invoked for the passive K+ uniport. Early evidence for the existence of a Kf-transporting ATPase came from work which correlated Kf influx in corn, wheat, oat, and barley roots with K+-stimulated ATPase activity located in a microsomal membrane preparation isolated from these roots (Fisher and Hodges, 1969; Fisher et al., 1970). The putative plasmalemma ATPase was further characterized in subsequent studies (Hodges et al., 1972; Leonard and Hodges, 1973; Leonard and Hotchkiss, 1976). The enzyme showed an acidic pH optimum for activity, and a requirement for Mg2+,and was further stimulated by

106

L. V. KOCHIAN AND W. J. LUCAS

"0

10

20

30

40

50

20

40

60

80

1c

I

0

KCI concentration ( mM Fig. 7. Correlation between the kinetics of 42K+influx into excised oat roots (a) and the stimulation of a microsomal ATPase activity as a function of K+ present in the reaction medium (b). ATPase activity was determined by the amount of Pi released. Insets show 42K+ influx and Pi released over the concentration range of 0.01-0.35 mM KCI. Data from Leonard and Hodges (1973).

monovalent cations, particularly K+ and Rb'. Leonard and Hodges (1973) showed that the complex kinetics for K+ absorption, in oat roots, were quite similar to the kinetics observed for K+ stimulation of ATPase activity (Fig. 7). It seemed appropriate, therefore, to correlate K+ transport in plants with the H+/K+-ATPase of the gastric mucosa and the Na+/K+ATPase of animal cells, since these systems all exhibited cation-induced stimulation and transported K+ directly (Hodges, 1976; Cantley, 1981; Faller et al., 1982; Leonard, 1984; Briskin, 1986a). The correlation between plasma membrane-associated, K+-stimulated ATPase activity and K+ influx, and the similarity between the sequence for monovalent cation stimulation of ATPase activity (K' > NH: > Rb+ > Cs+ > Li') and the specificity of monovalent cation uptake into roots (Sze and Hodges, 1977), has been used as further evidence that this ATPase is involved in K+ uptake, putatively as a K+-H+ exchange system. It has also been demonstrated that microsomal membrane vesicles isolated from

POTASSIUM TRANSPORT IN ROOTS

107

tobacco callus, and believed to be plasmalemma in origin, exhibited a stimulation of K+-ATPase activity in the presence of nigericin (which facilitates electroneutral K+-H+ exchange), or valinomycin plus a protonophore (Sze, 1980). These results again suggest that the plasmalemma ATPase may be involved in K+-H+ exchange. Recent work with isolated plasma membranes from corn roots (Briskin and Leonard, 1982a,b), oat roots (Vara and Serrano, 1983), and red beet storage tissue (Briskin and Poole, 1983a,b), has demonstrated that the plasmalemma ATPase forms a covalently bound, phosphorylated intermediate during the course of ATP hydrolysis. Briskin and Leonard (1982a,b) showed that K+ mildly stimulated the breakdown of this phosphorylated intermediate, using a crude preparation from corn roots, and concluded that this stimulation is further evidence for a K+ transport function by the ATPase. An analogy was again drawn with the reaction mechanism for the Na+/K+-ATPaseof animal cells, which also exhibits a K+ stimulation of dephosphorylation, and transports K+ directly (Cantley, 1981). However, it should be noted that the Na+/K+-ATPase phosphoenzyme intermediate has a much higher turnover rate, and the influence of K+ on the rate of dephosphorylation is considerable. Similar K+ effects on the dephosphorylation step were obtained with the plasmalemma ATPase isolated from red beet (Briskin and Poole, 1983a,b; Briskin and Thornley, 1985;Briskin, 1986b). However, in these studies the authors appear to place less emphasis on ascribing a K+ transport function for the ATPase. In contrast to the above studies, Vara and Serrano (1983) could find no effect of K+ on the phosphorylation and dephosphorylation of the plasmalemma ATPase from oat roots. This result led Serrano (1984) to propose that the ATPase is not a K+-dependent enzyme involved in K+ transport. As he points out, this interpretation is supported by the observation that, in the studies of the corn root microsomal membranes, the turnover of the phosphoenzyme was quite slow, and the K+ stimulation of the phosphoenzyme breakdown was much too small to support the concept of a direct role for K+ in the reaction kinetic scheme for this enzyme. Recently, Spanswick and Anton (1988) presented a model describing the kinetic characteristics of a H+-translocatingATPase purified from isolated plasmalemma vesicles of tomato roots (Fig. 8). They observed that the kinetics for the ATPase were similar to those for other cation-transporting ATPases that form phosphorylated intermediates. Their model does not ascribe a K+ transport role to the ATPase. Instead, stimulation of the breakdown of the phosphorylated intermediate by cytoplasmic K+ appears to be sufficient to explain the K+-stimulation of ATPase activity, without invoking a K+ transport role for the ATPase. It was proposed that K+ transport occurs by a separate mechanism operating in conjunction with the H+-ATPase. There are additional criticisms that can be directed at the work published in support of a K+ transport function for the plant plasmalemma ATPase.

L. V. KOCHIAN AND W. J. LUCAS

108

t

K:Vi

-“-f E2-p

Fig. 8. A reaction scheme for the plasma membrane ATPase proposed by Spanswick and Anton (1988). The binding of H+ at the cytoplasmic face of the plasma membrane occurs via reaction 4 to give HE1, subsequent hydrolysis of ATP via reaction 1 produces HE, P and from this state the H+ is released to the outer surface of the plasma membrane, via reaction 2. Potassium is proposed to act from the cytoplasmic face of the membrane to accelerate the dephosphorylation of E2-P, to return the ATPase to the state in which it can again bind H+. (Vi stands for inorganic vanadate which acts as a competitive inhibitor on the E2complex.)

-

First, the degree of K+ stimulation of ATPase activity is quite low when compared with total ATPase activity, and the K+ concentrations often used to achieve this stimulation are excessively high (50 mM) when considering Kf concentrations likely to be present in the soil solution or root apoplasm. These results again suggest that the K+effect may be a general salt effect, or activation by cytoplasmic K+, and not evidence for Kf requirement by a K+ transporting ATPase. Additionally, K+ stimulation is observed only at pH values below 7 (Leonard and Hodges, 1973; Leonard and Hotchkiss, 1976). These considerations led Serrano (1984), in his quite comprehensive review of fungal and higher plant plasmalemma ATPases, to state, “Thus, the plasma membrane ATPase is not a potassium-dependent enzyme although it is specifically stimulated by this ion”. An alternative view can be taken which would fit both mechanistic views. The dominant transport protein in the plant plasmalemma could be the H+-translocatingATPase, which might not be expected to exhibit Kf-stimulated activity. However, a Kf-ATPase could also be present, at much lower levels. This could explain the small K+ stimulation of ATPase activity and phosphoenzyme turnover, which would be measured against a large background of activity due to the H+-ATPase. It is not unrealistic, in the light of work from animal and bacterial systems, to suggest that a number of

POTASSIUM TRANSPORT IN ROOTS

109

different ATPases might be contained in the various plant membranes, operating in response to different signals and participating in various cell functions. The resolution of this controversy may ultimately come from demonstrations of ATP-dependent K+ and H+ transport in purified and reconstituted systems (Briskin, 1986a). Much of the early work on plasma membrane ATPases shared the rather common problem of contamination by other membranes. For example, it was only recently that ATP-dependent transport was demonstrated with membrane fractions isolated from plant roots. Using microsomal membranes isolated from corn, several different groups have now been able to demonstrate the existence of an ATP-dependent proton pump (Sze and Churchill, 1981; Churchill and Sze, 1983; DuPont et al., 1982; Mettler et al., 1982; Stout and Cleland, 1982; Bennett and Spanswick, 1983). However, it was unclear at the time whether the membrane vesicles exhibiting H+ transport originated from the plasmalemma. It appears that much of the early examples of ATP-dependent H+ pumping were being conducted with tonoplast vesicles (Mettler et al., 1982; Serrano, 1984). It has now been amply demonstrated that at least two distinct ATPases were contained in these microsomal membrane preparations: a K+-stimulated, vanadate-sensitive, NO;-insensitive ATPase that is plasma membrane in origin, and a C1--stimulated, vanadate-insensitive, NO;inhibited ATPase that is located in the tonoplast (Marrb and Ballarin-Denti, 1985). ATP-dependent Hf pumping has been demonstrated with both types of membrane vesicles. Recent improvements in plant membrane fractionation techniques have allowed for the isolation of fairly pure membrane fractions. Hence, it is now possible to reconstitute partially purified preparations of plasmalemma ATPases into proteoliposomes (Vara and Serrano, 1982; O'Neill and Spanswick, 1984). However, an unequivocal demonstration of ATP-dependent K+transport is still lacking. When partially purified oat root (Vara and Serrano, 1982) and red beet storage tissue plasmalemma ATPases (O'Neill and Spanswick, 1984) were reconstituted into liposomes, only H+ transport was observed. Vara and Serrano demonstrated ATP-dependent Hf pumping with inverted liposomes containing K+-free solutions (Fig. 9), which suggested to them that the ATPase did not mediate K+-H+ exchange. Additionally, it can also be seen in Fig. 9 that introducing K+ produced an instantaneous stimulation of H+ transport, which would be expected if K+ was acting on the active site of the ATPase facing the external medium (cytoplasmic side of inverted vesicles). They interrupted this result as further evidence against a K+-H+ transport function, since a lag in the stimulation by Kf would be expected as K+ diffused into the vesicle and became available for transport. Because of the demonstrated H+-pumping capacity of these membranes, and a lack of K+ transport ability, the present consensus of most researchers

110

L. V. KOCHIAN AND W. J. LUCAS

ATP

KCI\

ATP K W \

,_

I

Fig. 9. Changes in 9-amino-6-chloro-2-methoxyacridine fluorescence upon energization of Kf-free proteoliposomes prepared from oat root plasma membrane ATPase. Assay medium contained either 25 mM MgS04 (A, B and C), or 25 mM Mg(N03)2(D). Tris-ATP (1.25 mM), K2S04(25 mM), KCI (50 mM), KN03 (50 mM), imidazole-HC1 (pH 6.5,20 mM), and gramicidin D (5 pg ml-') were added as indicated. The initial rate of quenching expressed as per cenrof total fluorescence min-' is indicated. Data from Vara and Serrano (1982).

in this area appears to be that the plant plasmalemma ATPase acts as an electrogenic Hf transport system rather than a K+/H+-ATPase. However, this is an area of research in which rapid progress is being made, and it may well be that this consensus will be modified in the near future. Very recently, evidence suggesting ATP-dependent K+ transport in plasmalemma vesicles isolated from red beet storage tissue has been presented (Giannini et af., 1987). In this study, inverted plasmalemma vesicles were loaded with 86Rb+-labelled K+ solutions by a freeze-thaw technique. Although these vesicles were leaky for K+, the addition of ATP stimulated Kf efflux over and above that observed in the absence of ATP. The ATP-dependent efflux was completely inhibited by vanadate, but only partially inhibited by carbonyl cyanide N-chlorophenylhydrazone (CCCP), which should have abolished the H+ electrochemical gradient. Giannini et al. suggest that K+ transport may be mediated by two systems; one system would be indirectly coupled to the H+ gradient, while the other would be directly coupled to the ATPase, as a K+-translocating ATPase. It should be noted that the results of Giannini et af. do not support the K+-H+ exchange hypothesis. As noted by Serrano (1984), H+ ionophores

POTASSIUM TRANSPORT IN ROOTS

111

should not influence the transport function of ATPases that directly transport K+, either as a Kf-ATPase or as a K+/H+-ATPase.However, since H+ ionophores dissipate both the chemical and electrical components of the protonmotive force generated by a Hf-ATPase, any indirect coupling of K+ transport to a H+-ATPase would be significantly inhibited. This would include K+ uniports driven by the electrical component of the PMF and H+-K+ co-transport. Evidence for the co-transport of K+ with H+ will be discussed in Section II.B.4. 3. Indirect Coupling: Electrophoretic K + Uniport A number of researchers have argued that K+ uptake is not directly coupled to the H+-translocatingATPase, but is associated indirectly with the activity of the H+ pump through a K+ transport system driven by the electrical component of the PMF (Pitman et al., 1975; Marrb, 1979 and references therein). Pitman and co-workers were the first to propose this type of electrophoretic coupling, following studies on the effects of fusicoccin (FC) on K+ and H+ fluxes in low-salt and salt-saturated barley roots (Pitman et al., 1975). Their studies were generally conducted at high external K+ levels (5 mM). In low-salt roots, the FC-stimulated H+ efflux was similar, whether the roots were exposed to 5 mM KCl or 5 mM NaC1. However, in saltsaturated roots, which have a greater passive permeability to K+ than Na+ (in relation to low-salt roots), H+ efflux was stimulated to a greater degree in KC1 solutions. Hence, they suggested that FC acted to stimulate passive K+ influx, through the increased electrical gradient created when H+ extrusion was enhanced by FC. Subsequent studies, which further support the hypothesis of an electrical coupling between the H+ efflux and K+ influx, are based on the application of FC to various plant tissues at K+ concentrations (>1 mM) where uptake is considered to be passive (Marr5, 1977, 1979). It has been consistently observed that under these conditions FC stimulates both H+ efflux and K+ uptake, and elicits a hyperpolarization of Em.A large body of circumstantial evidence has been accumulated, indicating that FC acts directly on the plasmalemma H+-ATPase to stimulate H+ translocation (MarrP, 1979). Since, in either the direct or indirect mode of coupling, a stimulation of H+ efflux should enhance K+ uptake, it is not possible to differentiate between the two models simply through the application of FC. Obviously, other approaches must be used in conjunction with FC experiments. Marrb (1979), in his review on FC, has summarized the observations in support of indirect (electrophoretic) coupling. They are as follows: 1. The stoichiometry of the FC-stimulated K+-H+ exchange is often quite close to 1: 1 (at least for corn roots and pea stem segments), particularly when corrections are made to account for H+ consumption during anion uptake. Although this observation can be used in support of either mode of

112

L. V. KOCHIAN AND W. J. LUCAS

coupling, it appears to support indirect coupling when considered in conjunction with the following observations. 2. Under the conditions in which FC-stimulated fluxes are generally studied, Kf is in passive equilibrium across the plasmalemma (Pitman etal., 1975; Cocucci et al., 1976). Hence, any stimulation of net Kf influx would involve an increase in passive uptake. 3. It has been demonstrated that lipophilic cations, such as tributylbenzylammonium and tetraphenylphosphonium, can substitute for K+ in its role in promoting FC-stimulated H+ efflux (Bellando et al., 1979; Bellando and Trotta, 1980). Presumably, these cations cross the plasmalemma nonspecifically, by permeating the lipid bilayer. When this occurs, it has been shown that Em is depolarized and H+ efflux is stimulated. Therefore, it has been argued that the apparent dependency of H+ efflux on K+ is due to a K+ influx-induced depolarization of Em,which would activate the electrogenic H+-translocating ATPase. Conversely, chemical modifiers that stimulate the proton pump (such as FC) would hyperpolarize the membrane potential and increase the driving force for passive, electrophoretic K+ influx. Interpretation of FC experiments is based on the premise that FC acts specifically on the plasma membrane ATPase to increase its activity. The rapidity of the FC effects on plasmalemma ion transport, and the similar inhibitions of FC-stimulated Kf and H+ fluxes, and ATPase activity, by known plasmalemma ATPase inhibitors, are often used in support of this premise (Marrb et al., 1974a,b). Additionally, [3H]-FC has been shown to bind specifically to a protein component of the plasmalemma-enriched fraction isolated from corn coleoptiles (Dohrmann et al., 1977), and FC is known to stimulate ATPase activity associated with plasmalemma-enriched membrane fractions (Beffagna et al., 1977). However, caution must be exercised when interpreting these data. Several groups have solubilized the putative FC-binding protein from plasmalemma-enriched membrane fractions of corn coleoptile (Pesci et al., 1979; Tognoli et al., 1979) and oat roots (Stout and Cleland, 1980). In each case, it has been shown that the FC-binding protein can be separated from the Kf-stimulated plasmalemma Hf-translocating ATPase. Although it has been suggested in each of the above studies that the FC-binding protein could be a subunit of a multisubunit ATPase, the possibility exists that FC may act at sites totally separate from the Hf-ATPase. In a recent electrophysiological study on Viciafaba guard cells, Blatt (1987) obtained evidence that FC may not act on the pump, but via an effect on the related co-transport processes. Additionally, his current-voltage data indicated that FC may act to block a Kf channel involved in K+ efflux from guard cells. Such a mode of action could also explain the FC effects on other plant cell membranes, and it may be necessary to conduct a complete re-evaluation of these FC data. The use of the electrophoretic coupling model to explain Kf uptake over all K+ concentrations is subject to a number of criticisms, some of which

113

POTASSIUM TRANSPORT IN ROOTS

have been rather eloquently summarized (Glass and Siddiqi, 1982; Siddiqi and Glass, 1984). They point out that K+ influx often greatly exceeds H+ efflux, an observation that we have also stressed (Kochian et al., 1987,1988; Newman et al., 1987). This is a feature which has been presented in work dealing with K+ and H+ fluxes, although it has often gone unmentioned in the texts of these papers (e.g. Poole, 1974; Pitman et al., 1975; Lado et al., 1976; Lin and Hanson, 1976). It is difficult to reconcile a model where K+ influx is energetically dependent on the membrane potential (which, in turn, is dependent primarily on H+ efflux), when the H+ fluxes are usually much smaller than the associated K+ fluxes. Sometimes this disparity has been explained on the basis that the measured, “apparent” net H+ efflux is considerably smaller than the true unidirectional efflux, due to processes that also consume H+. These processes might include anion-H+ co-transport, passive “leaks” that would tend to return H+ into the cell, and H+ captured by the cell wall. However, Glass and Siddiqi (1982) conducted a careful assessment of the contribution of these processes to the underestimation of “true” H+ efflux, and concluded that for their system (low-salt barley roots) the measured net H+ efflux is a good approximation of the unidirectional flux. Furthermore, the data presented in Fig. 10 indicate that H+ efflux may be dependent upon external K+, or K+ influx, per se. Clearly, as stressed by Glass and Siddiqi (1982), this is the converse of what would be expected if K+ uptake were dependent upon H+ extrusion. They also noted that at

-0

-0

E,

f

X

2 1

5 -

r

iii

+

I

+ Y

0

0 0.01

0.1

10

1

Potassium Concentration

(

rnol rn-3

)

Fig. 10. Potassium influx (0, A) and H+ efflux (0,A ) as a function of K’ (K,SO,) in the bathing medium. Experiments were conducted on two varieties of barley, var. Fergus (0, 0) andvar. Conquest ( A , A). Data from Glass and Siddiqi (1982).

114

L. V. KOCHIAN AND W. J. LUCAS

higher K+ levels (>1 mM), H+ efflux often levels off or declines slightly (see Fig. 10). This saturation is somewhat surprising, considering earlier published reports which specifically linked K+ and H+ fluxes at higher K+ concentrations (Pitman, 1970; Marr2, 1977). Because it was found that anion influx [SO:- in the case of Glass and Siddiqi (1982)l was greatly increased at higher substrate levels (>0.5 mM K2S04), these potentially anomalous results could be explained by a reduction in net H+ efflux due to the consumption of H+ during SO:- influx, if uptake were mediated by an anion-H+ co-transport. However, these data, and most of the results used in support of models coupling K+ and H+ fluxes, could be explained equally well by a H+ extrusion mechanism involved in the maintenance of charge balance (Glass and Siddiqi, 1982), as originally suggested by Ulrich (1941). The constraints of charge balance are most evident when low-salt roots are transferred from a CaS04medium to one containing various inorganic ions, whence they begin to increase their internal salt levels. The roots would then be in a transition between two regulatory states. Pitman (1970) proposed that the stimulated K+ uptake, H+ efflux, and excess cation absorption exhibited as salt status is increased, would be a reflection of this transitional stage. Siddiqi and Glass (1984) have extended this hypothesis, by suggesting that these transitional changes are the result of a concerted effort, by the root, to maintain electrical neutrality during the alteration of salt status. This attempt to maintain electroneutrality does not appear to be specific for H+ efflux. Under certain conditions, plants can use other ion transport processes in order to effect charge balance (Siddiqi and Glass, 1984). For example, both K+-stimulated Na+ efflux and NHi-stimulated K+ efflux have been observed in plant roots. Thus, the apparent coupling of K+ and H+ fluxes may reflect a more general feature of plants, i.e. that cation exchanges (H+, K+, Na', Ca2+etc.) may be needed to maintain charge balance during times when high rates of cation uptake occur (i.e. during the transition from low-salt to high-salt status). Finally, it seems unlikely that a single type of mechanism, or coupling mode, will suffice to account for K+ uptake in higher plants. This is particularly apparent when one considers the potentially wide range of soil K+ concentrations that a root may experience. There are many examples in the literature for plant, bacterial, and animal transport systems, where the uptake of a particular solute is mediated by two or more types of transport mechanisms. Since it appears that K+ uptake in higher plants is active at low K+ levels, and thermodynamically passive at higher concentrations, it seems plausible to speculate that two different K+ transport systems might exist. The existence of K+ channels has been known for many years in animal and bacterial membranes, and increasingly strong evidence for K+ channels in higher plants has been accumulating (Schroeder et al., 1984,1987; Bentrup et al., 1985; Kochian et al., 1985; Kolb et al., 1987). Hence, it is quite possible that K+ channels in the plasmalemma facilitate passive K+ uptake (or release) at higher (>0.5 mM) concentrations. The direction and magnitude

115

POTASSIUM TRANSPORT IN ROOTS

of these fluxes would depend upon the value of the membrane potential and on any voltage-dependent characteristics that these channels may possess. However, at low K+ concentrations, where active K+ uptake most probably occurs, a different type of transport system must operate. Uptake must either be directly coupled to an ATPase, as a K+-ATPase (or the less likely case of a K+/H+-ATPase),or active K+ uptake may be coupled to the H+ gradient, via a K+-H+ co-transport system. The K+-H+ co-transport hypothesis will be considered in the next section. Other Modes of Coupling: K + - H + Co-transport? Another approach that has been taken in order to simultaneously study K+ and H+ transport in roots involves the use of ion-selective microelectrodes for K+ and H+ (Kochian and Lucas, 1987; Kochian et al., 1988; Newman et al., 1987). An ion-selective microelectrode system was developed that could quantify and map the extracellular electrochemical potential gradients for K+, H+, and C1- along the roots of 4-day-old corn seedlings. From an analysis of the extracellular ion gradients, it is possible to simultaneously determine and monitor the net H+ and K+ fluxes associated with a few cells at the root surface (due to the small electrode tip diameter of 0.5-1.0 pm). Because this system provides a high degree of spatial (and temporal) resolution, it has proven to be a useful method for studying the coupling of K+ and H+ fluxes at the cellular level. A diagrammatic representation of the system used in these studies is shown in Fig. 11. Data collected with this system indicated that at any point along the root, large fluctuations in the fluxes (particularly H+ fluxes) occurred with time. On many occasions, H+ efflux was near zero while K+ influx was “normal”; the converse was also occasionally observed. Additionally, when repeated measurements were made at the same location on the root, both H+ efflux and K+ influx could vary significantly, and usually the variation of the two fluxes did not show any correlation. The data presented in Table I illustrate 4.

TABLE I Measurement of net H + and K+ influx in low-salt-grown corn roots K+ Influx

Time (min)

H+ Efflux (pmol (g fresh wt)-’ h-’)

(pmol (g fresh wt)-’ h-’)

0 3 8 12 20

1.87 0.76 0.00 0.00 1.12

2.92 3.18 2.52 2.89 2.27

Measurements were made with pH and K+ microelectrodes at a position 3.0 cm from the root apex at distances of 50 and 125 pm from the root surface. Bathing solution consisted of 0.1 mM K2S04 + 0.2 mM CaS04. Data taken from Kochian and Lucas (1987).

116

L. V. KOCHIAN AND W. J. LUCAS

€'=c E,v-€; 3

(- ,

+ c v-vo I,-I

6

f

Rootradius = 500pm

v, C O

( a )

E;

__

Ell+EKt

!I/

I

' Centre of corn roo1

I -

Portion of an intact root or root segment

1 5 p m j r M i c r o e l e c t r o d e tipslocated within15-20pm (b 1

_ -/ Cno,r////

root

'///////'Surface

of corn root

Fig. 11. Schematic representation of the ion-specific microelectrode system used by Newman et al. (1987) to measure the Kt and H+ electrochemical potential gradients which develop at the surface of intact corn roots. Individual voltage components are shown in (a): EL and E r represent the EMF of the half-cells for the reference and ion-selective electrode, respectively; E: is the EMF of the half-cell for the extracelluar electric potential electrode; E, is the EMF developed across the ion-exchange resin; V and V , are the extracellular potentials near the root surface and in the background bathing solutions, respectively. A scaled representation of the relative positions of Kt,Htand extracellular potential (EI) microelectrodes is shown in (b). Here the microelectrodes (0.5 pm tip diameter) ae shown at their closest position to the corn root surface.

this point. Here, five separate measurements of K+ influx and H+ efflux were made over a 20-min period, and it can be seen that H + efflux varied rather dramatically, while K+ uptake was relatively stable. From these studies no fixed stoichiometry was found for the two fluxes. Generally, net K+ uptake was significantly larger than H+ efflux; the K+ :H + flux ratio ranged from 3.6: 1 to 1:2.75. A similar lack of correlation between K+ and H+ fluxes was observed in barley roots (Glass and Siddiqi, 1982). Increasing the bathing medium pH resulted in a gradual, but small, increase in K+ influx (see Fig. 5 ) . However, H+ efflux did not exhibit a similar pH dependency, but rather appeared to vary independently of the pH value. In this situation, the K+:H+ flux stoichiometries varied from 1:1 to 12: 1, again with no apparent pH dependency. The response of the cortical cell membrane potential to changes in

117

POTASSIUM TRANSPORT IN ROOTS

7 -100 E Y

-0 -120.-

E -1400

c

0

0.2rnM CaS04

I-7Z-l 0.I rnM K2S04

a -1600)

-180-

n

E -200-

5

I

Time (min)

Fig. 12. Response of the cortical membrane potential of intact low-salt-grown corn roots to changes in the concentration of external K+ (as K2S04). (CaSO,, was present at 0.2 mM in all solutions.) Data from Newman et al. (1987).

external K+ was also investigated in these studies. Unlike the case in earlier studies (Cheeseman and Hanson, 1979a,b), the low-salt corn root membrane potential was extremely sensitive to external K+. Increasing the Kf concentration from 0.2 to 20 p~ caused a rapid 70 mV depolarization, from - 190 mV to - 120 mV (see Fig. 12). A further increase in K+ to 200 p~ only elicited a further 7 mV depolarization. Newman et al. (1987) made simultaneous measurements of Em and K+ fluxes in response to changes in external K+,using a three-microelectrode system (one to measure Em,and two K+ microelectrodes for K+ flux measurements). Both the kinetics (dependence on external K') for K+ influx and the depolarization of the membrane potential yielded similar (and quite low) apparent K m values in the 6-9 p~ range. These results indicated that low-salt corn roots possess a very high affinity K+ uptake system that is extremely electrogenic (Newman et al., 1987). In view of the highly electrogenic nature of this system, it is possible that other cations may be transported into the cell with Kf. Since these measurements were made in simple salt solutions (10 p~ K2S04 0.2 mM CaS04), H+ appears to be the most likely candidate for co-transport with K ' . In such a situation, active K+ influx would be achieved using the energy gained from the movement of H+ down their A&. Strong evidence for the existence of a K+-H+ co-transport system in fungi has been recently provided by work on Neurospora (Rodriguez-Navarro and Ramos, 1986; Rodriguez-Navarro et al., 1986;Blatt and Slayman, 1987; Slayman et al., 1988). Slayman and co-workers have described a high-affinity (K, = 1-10 p ~K+ ) uptake system, in K+-starved Neurosporu cells, that is highly electrogenic, as demonstrated by the response presented in Fig. 13. Measurements of internal and external K+, and Em,suggested that this uptake system may mediate active K+ uptake, and the system appears to be

+

118

L. V. KOCHIAN AND W. J. LUCAS ( 0

1

(b)

=-o

-I

100 urn Sdutlan

flow

~

IK'1,-

I

50~M

[K+l,

2

++

3 -304

5aM

,

4n - 2 1 8 +

-305

2014

1 2 0 9 + +

-307

5 0 ~ M

++

200g

Fig. 13. Effect of added extracellularKCon the membrane potential measured in low-K+ spherical cells of Neurospora crassa. (a) Diagram of the arrangement of the cell, the impaling electrode and a K+-floodingpipette for rapid introduction and removal of K+. (b) Depolarization of the Emwith 50 p~ K'. (c) Condensed record from a single cell, showing the Emresponse to four different K+ concentrations. Symbols I and I1 indicate cell impalement and electrode removal, respectively. Data from Rodriguez-Navarroet al. (1986).

coupled to the very active plasmalemma H+-ATPase of Neurospora. A net stoichiometry of one H+ out for one K+ in was demonstrated, which could be taken as evidence for K+-H+ exchange. However, current-voltage analysis conducted by Slayman's group indicated that the K+-associated inward current was twice that of the net K+ influx (see Fig. 14). Thus, one additional positive charge enters with every K+. In addition, the following points must be considered: (1) the H+-ATPaseoperates in parallel with the K+ uptake system; (2) almost every charge absorbed must be balanced by an extruded H+; and (3) only a single H+ is measured (released to the external solution) for every Kf taken up. Hence, the second charge coming in with the K+ must be a H+,which indicates that the high-affinity K+ uptake system operates as a K+-H+ symport (Rodriguez-Navarro et al., 1986; Slayman et al., 1988). Slayman and co-workers speculate that the data presented for higher plants could also be explained by this type of co-transport system coupled to the plasmalemma H+-ATPase. As discussed previously, it has been proposed that a similar co-transport system exists in the plasmalemma of corn root cells (Newman et al., 1987). Further support for this hypothesis comes from the observed similarities between the K+ uptake system seen in K+-starved Neurospora cells and that described above for low-salt corn roots (Kochian et al., 1987,1988; Newman et al., 1987). Both are very high affinity K+ uptake systems with almost identical K , values for K'. Furthermore, K+ uptake through both systems is

119

POTASSIUM TRANSPORT IN ROOTS

? C

2

A

30

0

40

80

120

External K + Concentration

200

160 (

p~

Fig. 14. Stoichiometry of the K+-Hf co-transport system of Neurosporu crassa. The smooth curves are Michaelis-Menten functions fitted to the two sets of data, using a common value of 14.9 ~ L MKt for the apparent K,. Separate values of V,,, are 15.3 k 1.0pmol cm-'s-' for the flux, and 30.1 f 1.6 pmol cm-2 s-' for the measured current. The stoichiometric ratio of the current to net K+ flux is very close to two. Data from Rodriguez-Navarroet af. (1986).

highly depolarizing (compare Figs 12 and 13). The similarities between the two systems, taken in conjunction with the lack of evidence, in low-salt corn roots, for a coupling between K+ influx and H+ efflux, strongly suggest that such a co-transport system could be operating in higher plants. On a kinetic basis, it can be reasoned that a K+-H+ co-transport system should reflect a sensitivity to changes in both extracellular and cytoplasmic pH values. Previous reports of increasing K+ influx, in response to decreasing H+ concentrations in the bathing medium, appear to be somewhat at variance with this prediction. However, in these reports, K+ influx was insensitive to external pH values from 6 to 9 (see Fig. 5 ) . To further investigate this hypothesis for K+ influx into corn roots, Kochian et al. (1987) used K+-selective microelectrodes and 86Rb+to measure K+ influx as a function of pH. Varying the external pH from 4 to 8 had no effect on either net K+ influx (measured with K+ microelectrodes) or unidirectional K+(86Rb+)influx. Additionally, Kochian et al. found that the K+-induced depolarizations of Em (with solution of 50 PM K+) also exhibited absolutely no pH dependence from pH 4 to 8. These results suggest that the K+-H+ co-transport system may have an extremely high affinity for H+. However, an alternative interpretation is that K+ influx is mediated by a different molecular mechanism, e.g. a K+-ATPase,which would not, apriori,show a sensitivity to external pH values.

120

L. V. KOCHIAN AND W. J. LUCAS

C. UPTAKE AT HIGH K+ CONCENTRATIONS: THE LINEAR COMPONENT

As discussed in Section II.A, studies on the kinetics of K+ uptake, for various plant tissues, have generally focused on the generation of complex and often discontinuous kinetic curves, via either the operation of multiple Michaelis-Menten transport systems (such as Epstein's mechanisms I and 11), or by complex, multisite carriers. However, there are numerous examples, particularly in association with studies involving animal and bacterial systems, of complex transport kinetics that appear to be due to the parallel operation of one or more saturable transport mechanisms plus a system that exhibits nonsaturating, or first-order uptake kinetics (linear component). For example, a linear transport component has been demonstrated for the uptake of amino acids and sugars in animal cells (Akedo and Christensen, 1962; Christensen and Liang, 1966; Munck and Schultz, 1969; Cohen, 1975, 1980; Debnam and Levin, 1975), and for the uptake of lactose (Maloney and Wilson, 1973), K+ (Rhoads et al., 1976; Epstein and Laimins, 1980) and amino acids (Wood, 1975; Iaccarino et al., 1978) in E. coli. In many of these studies, nonsaturating solute uptake was dismissed as a physiologically insignificant process, because it was considered to reflect passive diffusion across the lipid portion of the plasma membrane. However, in work on K+ uptake in E. coli, linear Kf uptake was hypothesized to be mediated by a transport protein, which was presumably a Kf channel (Rhoads etal., 1976; Epstein and Laimins, 1980). Furthermore, Christensen and Liang (1966) demonstrated that nonsaturating amino acid uptake in Ehrlich tumour cells was substrate-specific, and exhibited considerable sensitivity to pH and temperature. Thus, passive diffusion was discounted in favour of a more complex system involving a transport protein. In recent years, there has been an increasing interest in linear, nonsaturating solute uptake kinetics from studies involving plant tissues. First-order kinetics have been observed for the uptake of Fe2+in rice roots (Kannan, 1971), sucrose and 3-0-methylglucose in Ricinus cotyledons (Komor, 1977; Komor et al., 1977), sucrose in sugar beet leaf and petiole sections (Maynard and Lucas, 1982a,b), soybean cotyledons (Lichtner and Spanswick, 198l), Vicia leaves (Delrot and Bonnemain, 1981) and red beet vacuoles (Willenbrink and Doll, 1979), and sucrose, fructose, and glucose in Allium leaf discs (Wilson et al., 1985), and for amino acid transport in Lemna (Fischer and Luttge, 1980), and suspension-cultured tobacco cells (Blackman and McDaniel, 1978). Separation of the contribution made by the linear transport component from the saturable mechanisms has been achieved through the use of various inhibitors (Debnam and Levin, 1975; Polley and Hopkins, 1979; Maynard and Lucas, 1982b; Van Be1 et al., 1982). For K+ uptake into corn roots, Kochian and Lucas (1982a,b, 1983) have shown that the complex kinetics

121

POTASSIUM TRANSPORT IN ROOTS

could be resolved into saturable and first-order kinetic components, through the use of sulphhydryl modifiers. When corn root segments were subjected to a series of increasing N-ethyl maleimide (NEM) exposures (e.g. 0,10, and 30 s NEM exposures for the data illustrated in Fig. 3b), saturable K+ uptake was specifically inhibited and subsequently abolished, while linear uptake was relatively unaffected. Hence, it was suggested that the saturable and linear kinetic components represented separate K+ transport systems. Saturable K+ uptake would be analogous to Epstein's mechanism I; however, nonsaturating K+ uptake, which dominates uptake in the concentration range associated with Epstein's mechanism 11, appears to be distinctly different from the transport system described by Epstein and co-workers. Therefore, subsequent work was carried out in order to characterize this linear component for Kf uptake (Kochian and Lucas, 1984; Kochian et al., 1985); certain features of this transport system will now be discussed.

I . Anion Involvement in Nonsaturating K + Uptake Early work by Epstein et al. (1963) demonstrated that in barley roots K+ uptake by mechanism I1 was dramatically inhibited when C1- was replaced by SO:- in the uptake solution. Similarly, it was shown that linear K+uptake in corn roots was partially dependent on the presence of CI- in the uptake solution (Kochian et al., 1985). As shown in Fig. 15, replacing C1- in the

0

2

4

6

8

10

K+concentration (mM

Fig. 15. Influence of the accompanyinganion on K+(86Rb+)influx intocom root segments grown in 0.2 mM CaSO, (low-salt status). The first-order rate coefficients, k , in pmol (g fresh wt)-' h-' m K ' , for the linear component of K+ influx were as follows: 0.5 (Cl-), 0.21 (SO$-), and 0.24 (H2P0,). Data from Kochian et al. (1985).

122

L. V. KOCHIAN AND W. J. LUCAS

uptake solution with SO$-, H2PO; or NO; resulted in a 60% reduction in K+ uptake by the linear component in low-salt roots, while causing only a 15% reduction in the V,,, for the saturable system. In high-salt roots (grown on 5 mM KCl), a similar reduction in linear K+ uptake was seen, while saturable uptake was unaffected. This association between the linear component for K+ uptake and the presence of C1- was shown to be related to a coupling via the saturable C1-

Control

.

I , L

6 DIDS

-a

=

OCO”

o

. 2

4

6

8

10

8

10

CI-concentration (mM 1

0

2

4

6

K+concentration (mM)

Fig. 16. Effect of the anion transport inhibitor DIDS on the influx of wl- (a) and K+ (=Rb’) (b) into low-salt corn root segments. For K+ influx, k had values of 0.40 and 0.19 pmol (g fresh wt)-’ h-’ mM-’for the control and DIDS treatment, respectively. Data from Kochian etal. (1985).

POTASSIUM TRANSPORT IN ROOTS

123

acid influx process. Application of 1mM 4,4-diisothiocyano-2,2’-disulphonic stilbene (DIDS), which has been shown to be an anion transport inhibitor in red blood cells (Cabantchik et al., 1978), corn roots (Lin, 1981) and Chum (Keifer et al., 1982), abolished the saturable component for C1- influx into low-salt corn roots (Fig. 16a). In the presence of this inhibitor, the linear component of K+ uptake was suppressed to an identical degree to that seen when C1- was replaced in the uptake solution by other anions (compare Figs 15 and 16b). These results strongly suggest that the linear component of K+ influx in corn roots is, in some way, linked, at least partially, to saturable C1- uptake.

2. Involvement of K’ Channels? The effect of the quaternary ammonium salt, tetraethylammonium chloride (TEA), which has been shown to block K+ channels in excitable membranes, such as the plasma membrane of nerve fibres (Tasaki and Hagiwara, 1957; Armstrong, 1969), and in Chara (Keifer and Lucas, 1982), was studied on K+ uptake in low- and high-salt corn roots. In high-salt roots, TEA caused a dramatic (75%) and specific inhibition of the linear component of K+ influx (Fig. 17), which suggests that K+ channels may be involved. However, although low-salt roots possess a similar linear component for K+ influx, it was found to be insensitive to TEA, which seems at variance with the above interpretation. Uptake studies with [I4C]-TEAindicated that high-salt corn roots exhibited much higher rates of TEA accumulation than did low-salt roots, presumably via a transport system for quaternary ammonium salts

K+concentration (mM influx into high-salt-growncorn roots. Fig. 17. Influence of 10 mM TEA-CI on K+ (86Rb+) Roots were pretreated with a solution containing 10 mM TEA-C1,5 mM KCI, and 0.2 mM CaS04 for 30 min prior to K+ (“Rb’) uptake (10 mM TEA-CI was included in the uptake solutions). The first-order rate coefficients, k, for the linear component were 0.31 and 0.09 pmol (g fresh wt)-’ h-’ mM-’ for the control and TEA-C1, respectively. Data from Kochian et al. (1985).

124

L. V. KOCHIAN AND W. J. LUCAS

that has been demonstrated in higher plants (Michaelis et al., 1976). Hence, it was proposed that corn root tissue behaves in a similar manner to nerve axons, where it is necessary to inject the TEA into the axoplasm in order to block Kf channels (Tasaki and Hagiwara, 1957; Armstrong, 1969). Highsalt roots may be able to accumulate more TEA in the cytoplasm relative to low-salt roots. If this were the case, then K+ channels could be involved in nonsaturating K+ influx in both low- and high-salt roots; but only in high-salt roots would the cytoplasmic TEA concentration rise to a level high enough to block the putative Kf channels at the cytoplasmic face of the channel. In recent years, the application of the patch-clamp technique to plant cell membranes has provided direct evidence for the existence of K+ channels in higher plant cell membranes. For example, K+ channels have been demonstrated in the plasmalemma of guard cells (Schroeder et al., 1984, 1987), Samanea pulvinar cells (Moran et al., 1987) and corn root cells (Ketchum et al., 1987), and in the tonoplast of Chenopodium suspension cells (Bentrup et al., 1985) and barley mesophyll cells (Kolb et al., 1987). Therefore, it seems reasonable to speculate that one or more classes of K+ channels could operate in the plasmalemma of root cells, in order to facilitate passive K+ uptake at high external levels of K+. Although it is axiomatic that any transport system should have a finite transport capacity, it is possible that a system like an ion channel would not exhibit saturation kinetics under physiological conditions. As Cohen (1975) has pointed out, such a channel-mediated process should eventually saturate at high substrate concentrations. However, in his system (amino acid uptake into mouse brain slices), at high substrate levels, the medium changes from isotonic buffered saline to hypertonic buffered amino acid saline. Thus, any changes in transport kinetics may be due to changes in media composition. The same applies for K+ influx in corn roots. Potassium uptake has been studied from a range of concentrations up to 50 mM. At these high K+ levels, there was no significant change in nonsaturating K+ uptake (Kochian and Lucas, 1982a). Uptake was not studied from solutions of higher concentration, because it was felt that ionic and osmotic effects could alter membrane lipid/protein structure and make data interpretation tenuous.

3. Root Salt Status and Nonsaturating K + Uptake Uptake into low-salt barley roots, at low external K+ levels, is highly specific for K', while at higher K+ concentrations (>0.5 mM), Na+ can competitively inhibit K+ influx (Epstein et al., 1963). A similar response was also observed with low-salt corn roots. Inclusion of 3 mM NaCl in the uptake solution (K+ concentration was from 0.1 to 10 mM) caused a 50% inhibition of the linear component for K+ uptake, while saturable uptake was unchanged (Kochian et al., 1985). The interesting feature is that within the mechanism I1 concentration range, high-salt corn roots exhibit a much higher selectivity for K+ influx over other cations (Pitman, 1967, 1970; Pitman et al., 1968; Kochian et al., 1985). If we accept that the linear component represents the

POTASSIUM TRANSPORT IN ROOTS

125

operation of a Kf channel, these data indicate that some aspect of tissue salt status modifies the channel characteristics from relatively nonspecific to highly K+-specificin high-salt roots. It will be interesting to see whether this prediction can be confirmed using the patch-clamp technique.

4. Alternative Explanation f o r the Linear Component Sanders (1986) has developed a generalized model to explain biphasic or otherwise complex transport kinetics, based on the co-transport of a solute with a driver ion. For Kf influx into roots, a K+-H+ co-transport system would be driven by the inwardly directed A& generated by the H+-ATPase. In Sander’s reaction kinetic model, the assumption is made that both solute (K’)and driver ion (H’) can bind randomly to the carrier, and the limitation is made that the carrier can cross the membrane only as the fully loaded complex (influx) and can return to the outside free of ligand. The reaction kinetic scheme for this model is shown in Fig. 18. Numerical analysis

Fig. 18. Membrane transport modelled on the basis of a reaction kinetic scheme. (a) Reaction kinetic scheme for random binding of solute (S) and H+ to a membrane-bound carrier (X)which catalyses the transport of S across the membrane. Carrier is represented as transporting positive charge in the loaded form. (b). As in (a), but with loaded carrier being neutral and charge transfer occurring on the unloaded form of the carrier. (c) Generalized reaction kinetic scheme for the charged and uncharged models. Concentration (density) of carrier state “j” is designated as Ni, with rate constants (not shown) from carrier state i tostate j designated kii.Redrawn from Sanders (1986), with permission.

126

L. V. KOCHIAN AND W. J. LUCAS

simulation of this model demonstrated that it generates biphasic kinetics under conditions resembling those present in many uptake experiments. The utility of this model is that it makes a number of predictions that are experimentally testable. These predictions are (in terms of a K+-H+ cotransport system): 1. As [H'], is increased to saturating levels, the kinetics for K+ uptake should change from a complex biphasic to a monophasic profile. 2. At saturating [H'],, increasing levels of internal K+ should result in an uncompetitive inhibition of K+ influx.

Fig. 19. Comparison between the experimental data (W) obtained for 6-deoxyglucose influx into Chorella, as a function of bathing medium pH (Komor and Tanner, 1975), and the theoretical simulation of Sanders' (1986) reaction kinetic model. The solid lines represent a simulation of the charged carrier model, simplified for very negative Em and internal 6-deoxyglucose concentration set at zero. Data from Sanders (1986).

POTASSIUM TRANSPORT IN ROOTS

127

3. As [H’l0 is decreased, the relative contribution to the uptake isotherm made by the “apparent” low-affinity K+ transport should increase.

As Sanders (1986) noted, there are few published examples of data concerning pH effects on complex isotherms for solute uptake in plants that permit his predictions to be properly tested. However, such kinetic isotherms do exist for the H+-sugar co-transport system of Chlorella (Komor and Tanner, 1975). As shown in Fig. 19, these data give a reasonable fit to the first prediction of the Sanders model. However, if one then examines the pH dependency of the detailed kinetics for sucrose uptake into sugar beet leaves obtained by Maynard and Lucas (1982a), it is apparent that as the external pH is decreased from 9.0 to 4.0, both the saturable and nonsaturable components of sucrose uptake are stimulated. Clearly, these data do not fit the first prediction of the Sanders (1986) model. Furthermore, it is difficult to reconcile K+ uptake into high-salt corn roots with this Sanders model, since, in this system, it was possible to speciJically abolish saturable K+ uptake with NEM, while leaving nonsaturable uptake relatively unaffected (Fig. 3b). Conversely, it was possible to specifically inhibit the linear transport component either by replacing C1- in the uptake solution with other anions (Fig. 15), or by treating the roots with the K+ channel blocking agent, TEA (Fig. 17). At present it is not obvious how these perturbations, which appear to specifically resolve the complex kinetics for K+ uptake into separate kinetic components, can be explained by the Sanders reaction kinetic model. 5. Physiological Role for Nonsaturating K + Uptake? It has been noted that nonsaturating solute uptake often occurs over a concentration range which exceeds the levels normally experienced by roots in the soil, and so it is difficult to assign a physiological role for such a transport system. Reisenauer (1966) has pointed out that the majority of soil Kf is below 2 mM; yet, in corn roots, nonsaturating K+ uptake becomes significant above 1 mM external K’. What then would be the physiological relevance of a transport system that operates at substrate levels rarely experienced by the plant? The answer to this question may come from transport studies conducted on E. coli and Neurospora. In both organisms, it has been well documented that multiple carrier systems are often involved in the transport of a single solute. These organisms tend to combine constitutive, low-affinity,high-capacity transport systems with derepressible high-affinity systems; glucose and phosphate uptake into N . crassa are excellent examples of this strategy (Lowendorf et al., 1974; Scarborough, 1970). These systems give the organism the adaptive advantage of most effectively obtaining nutrients whose concentrations may vary considerably over a period of time. A root growing through the soil may often experience extremely low K+ levels. Therefore, a high-affinity system may be necessary in order for the

L. V. KOCHIAN AND W. J. LUCAS

128

plant to satisfy its requirements for this essential nutrient. However, if the growing root encounters a localized region of high soil K', the high-affinity system may not have the capacity to utilize excess K'. In such a situation, a low-affinity, high-velocity system would allow the plant to effectively utilize these pockets of high soil K+. Furthermore, as soil water content (potential) declines, the root system must respond in terms of osmotic adjustment. If only a high-affinity K+ transport system were present, the root would not be able to take advantage of the increase in the K+ concentration of the soil solution that would occur along with the decline in soil moisture. The presence of a high-velocity system may provide the plant with an advantage in terms of its response to water stress. In this way, the plant may maintain relatively efficient methods of dealing with its varying environment. D. SUMMARY

A summary of the mechanisms by which K+ may enter the symplasm of the root is given in Fig. 20. Uptake at low external K+ levels is facilitated by a

H+ f

H't-

K'Channel

-

-

- -

~

F e e d b a c k on c a t i o n specificity

Cell Wall

Fig. 20. Schematic representation of possible K+ transport systems operating at the plasmalemma to facilitate K+ uptake at both low and high external K+ concentrations. At low K+ levels, a high-affinity system is hypothesized to be either a K+-ATPase, or a H+-K+ co-transport system, coupled to the H+-ATPase. Both systems would be subject to feedback control by internal K+. Under high K+ levels, nonsaturating K+ uptake involves a K+channel transport. The degree of K+ specificity exhibited which is, in some way, coupled to anion (CI-) by this channel may be influenced by cytoplasmic K+ levels.

POTASSIUM TRANSPORT IN ROOTS

129

very high affinity transport system that is strongly depolarizing, and can transport K+ against its electrochemical potential gradient. Several possible models appear to fit the published data equally well. In our evaluation, the most likely candidates would either be a K+-translocating ATPase or, as in Neurospora, a Kf-Hf symport coupled to the H+-translocating ATPase. The possibility exists that both types of active transport could be functioning in parallel. This transport system would appear to be subject to kinetic and thermodynamic control, and would also be subject to feedback (allosteric) regulation from Kf contained in an internal compartment (see Section V. A). Potassium uptake at higher external levels would be due to a lower-affinity transport system that would transport K+ passively, down its electrochemical potential gradient. This system could be a K+ channel, or a less specific cation channel, which is coupled, in some way, to a saturable C1uptake system which may function as a H+-Cl- co-transport system (Sanders, 1980b; Jacoby and Rudich, 1980). This system does not appear to be subject to feedback regulation by internal K+ levels, but does exhibit changes in substrate specificity as internal salt levels are altered.

111. REDOX-COUPLED PLASMALEMMA TRANSPORT OF K+ The influence of exogenous electron donors and acceptors has been studied in a variety of plant cell types (Bienfait, 1985; Lin, 1985; Liittge and Clarkson, 1985; Moller and Lin, 1986). Although much of this work has concerned the role of plasmalemma electron transport (redox) systems in Fe” reduction at the root surface (Luttge and Clarkson, 1985; Mdler and Lin, 1986), several studies have addressed the more general topic of the involvement of plasma membrane redox systems in the energetics of solute transport into plant cells. Crane and co-workers (Craig and Crane, 1980, 1981; Misra et al., 1984) investigated the effects of exogenous NADH and ferricyanide on membrane transport and cell growth in suspension cells of carrot. They reported that carrot suspension cells can oxidize exogenous NADH with a concomitant increase in O2consumption (30% stimulation), and that this oxidation results in a stimulation (-60%) of K+ influx. Misra et al. (1984) postulated that a plasma membrane redox system operates in close association with the H+-translocating ATPase of this membrane to exert an influence on K+ transport. In experiments conducted on corn root cortical protoplasts, Lin (1982a,b, 1984) reported that addition of NADH tripled O2consumption, and caused a two- to three-fold increase in K+ influx, a marked stimulation of H+efflux and a moderate (20 mV) hyperpolarization of the E m . Although the extent of the NADH influence was less, Lin reported that similar results were obtained on excised corn root segments. Consistent with the results of earlier workers, Lin (1984) proposed a model in which the oxidation of

130

L. V. KOCHIAN A N D W. J. LUCAS

exogenous NADH would result in the establishment of a transplasma membrane H+ gradient. The relationship between this H+ gradient and K+ influx remains to be established (but see Rubinstein and Stern, 1986). A. INFLUENCE OF EXOGENOUS NADH ON K+ INFLUX

Almost all of the above-mentioned NADH studies were conducted at one particular K+ concentration, usually 0.2 or 1.0 mM. However, Lin (1984) showed that the addition of 1.5 mM NADH to corn root protoplasts had its main effect on K+ influx within the mechanism I range. In these protoplast experiments the apparent K,,, for K+ influx (0.3 mM) was not affected by NADH, but the V,,, underwent a four-fold stimulation. Uptake of K+ over the higher concentration range did not appear to be affected by exogenous NADH. Although in some cases a stimulatory effect of exogenous NADH has been observed on K+ uptake, a clear discrepancy exists between these reports and the findings of Kochian and Lucas (1985) and Thom and Maretzki (1985). Figure 21 illustrates the inhibitory effect that 1.5 mM NADH had on K+ influx into corn root segments (Kochian and Lucas, 1985). At low external K+ concentrations (C0.5 mM)) influx was inhibited by 80%, and a similar degree of inhibition of K+ influx into protoplasts, prepared from sugar cane suspension cells, was reported by Thom and Maretzki (1985). (Leucine, arginine and 3-0-methyl glucose influx into

, I" r

ol

Kt concentration ( mM 1

Fig. 21. Influence of 1.5 mM NADH on @Rb+ influx into high-salt-grown corn root segments. (A similar response was observed in NADH experiments conducted on low-salt corn roots.) Data from Kochian and Lucas (1985).

POTASSIUM TRANSPORT IN ROOTS

131

these protoplasts were similarly inhibited by NADH.) Interestingly, K+ influx into corn roots was also inhibited by the electron acceptor ferricyanide (Kochian and Lucas, 1985; Rubinstein and Stern, 1986). It has been suggested that these two chemicals elicit this inhibition by separate mechanisms, based on the different K+influx kinetics elicited after exposure to NADH or ferricyanide, and the different time courses for recovery from inhibition. B. MEMBRANE TRANSPORT AND THE WOUND RESPONSE

Kochian and Lucas (1985) proposed that exogenous NADH reacts with the outer surface of the plasmalemma to signal a tissue wound response. Hanson and co-workers have shown that corn roots are a particularly sensitive tissue, and that physical handling of the roots, imposition of cold shock etc., can elicit a wound response (Leonard and Hanson, 1972; Gronewald and Hanson, 1980; Chastain and Hanson, 1982; Zocchi and Hanson, 1982). This response is generally characterized by a reduction in K+ influx and a stimulation of efflux, a reduction in H+ efflux, and a depolarization of the Em.A recovery period of about 4 h is usually required before the “wounded” tissue returns to control levels of physiological functioning. It was shown that NADH does not affect K+ influx if the corn root segments have not yet recovered from excision-associated wounding (Kochian and Lucas, 1985). Additionally, if cycloheximideis included in the recovery medium, K+ influx does not return to the control level, and again NADH has no effect on K+ influx. These results indicate that activation of a wound response blocks the NADH effect. Kochian and Lucas (1985) demonstrated that in this state no NADH-stimulated O2consumption can be detected. Only in the recovered state could they measure a 30% stimulation of O2 uptake upon addition of 1.5 mM NADH. Interestingly, although addition of NADH perturbed K+ influx (and efflux), it continued to be oxidized by these perturbed root segments. As mentioned earlier, when the redox system in corn roots (or protoplasts) is stimulated by NADH, net apparent H+ efflux was reported to increase (Lin, 1984). However, Kochian and Lucas (1985) found that NADH elicited a significant decrease in H+ efflux in both washed corn root segments and intact roots (see also Lucas and Kochian, 1987). In some cases an alkalinization of the external medium was observed after addition of NADH, which implies net apparent H+ influx. A similar situation has also been reported for sugar cane protoplasts (Thorn and Maretzki, 1985). Transferring NADH-treated roots to fresh control solution (minus NADH) resulted in an almost immediate recovery of net apparent H+ efflux to pretreatment values (Kochian and Lucas, 1985). This imbalance between recovery of H+ efflux and K+ influx has also been observed with roots that

132

L. V. KOCHIAN AND W. J. LUCAS

are recovering from excision wounding (Gronewald and Hanson, 1982; Kochian and Lucas, 1985). Wounding the root elicits a depolarization of the Em,a response which is in direct contrast to the NADH-induced hyperpolarization of the potential reported by Lin (1982a). In recent studies conducted by Lucas and Kochian (1987), NADH was found to cause a significant depolarization of the corn root Em.In some cases the potential remained in this depolarized state, while in others it repolarized very slowly towards the pre-NADH resting potential. However, consistent with the response of H+ efflux discussed above, removal of NADH always resulted in a repolarization of the potential. Certainly, then, the NADH influence on K+ influx, H+ efflux, and the Em is consistent with the hypothesis that exogenous application of this redox reagent elicits some form of wound response in corn root tissue, while in protoplasts and suspension-cultured cells its application appears to stimulate these physiological processes. C. DEVELOPMENT OF AN INTEGRATED NADH MODEL

Although there is a clear conflict concerning the reported effects of exogenous NADH on plasmalemma transport of K+, there is little doubt that bona fide redox systems are located within this membrane (De Luca et al., 1984; Rubinstein et al., 1984; Buckhout and Hrubec, 1986; Bottger and Liithen, 1986; Macri and Vianello, 1986;Pupillo et al., 1986; Luster et al., 1987). The challenge is to develop a working model of the way in which these putative redox systems interact with the various transport systems functioning within the plasmalemma of plant cells. Such a model would have to account for these reported extremes in tissue response to NADH. In the present context, we will confine our attention to the effects of NADH on H+ and K+ fluxes and the membrane potential. Firstly, the inability of freshly cut, or wounded, corn roots to oxidize exogenous NADH must be explained in terms of the inability of NADH to further inhibit K+ influx, once the K+ transport system has received the wound response “signal(s)”. In analysing the NADH response we have, perforce, assumed that Kf transport occurs via a K+-H+ co-transport system. Zocchi el al. (1983) have shown that cold shock treatment of corn roots increased the phosphorylation of microsomal membrane proteins. These changes were correlated with a decrease in ATPase activity and, since the major changes in phosphorylated proteins occurred within the 92 to 100 kDa range, they concluded that one effect of injury is the blockage of H+ efflux via phosphorylation of the plasma membrane H+-ATPase. This hypothesis is consistent with the observed changes in adenine nucleotide content of corn roots during the injury recovery period (Zocchi et al., 1983). The critical step is the mechanism by which tissue injury is “sensed” by the cells and transduced into protein phosphorylation.

POTASSIUM TRANSPORT IN ROOTS

133

Several possible control systems appear to be tenable. Calcium influx has been shown to increase upon wounding (Zocchi and Hanson, 1982; De Quintero and Hanson, 1982), and the resultant change in the cytosolic Ca2+ level may activate a Ca2+-calmodulin system. One consequence of this activation could be the stimulation of a specific protein kinase. Alternatively, the turnover of phosphatidylinositol may function as a biochemical signal transduction system in roots in a manner analogous to its function in animal tissues (Berridge, 1983;Berridge and Irvine, 1984;Nishizuka, 1984). In animal cells, various external stimuli activate a phospholipase C on the inner plasma membrane surface, which then hydrolyses phosphatidylinositol 4,5-bisphosphate to release myoinositol 1,4,5-triphosphate (IP,) into the cytosol and diacylglyceride (DG) which remains within the membrane. IP3 can stimulate numerous biochemical and biophysical processes, including the release of internally sequestered Ca2+ and activation of certain kinase systems. Recent studies on plant tissues have demonstrated the presence and metabolic turnover of IP, (Boss and Massel, 1985; Morse et al., 1986; Sandelius and Sommarin, 1986; Rincon and Boss, 1987). These reports, along with the central role played by IP, in regulatory phenomena in animal tissues, provided the impetus to incorporate the IP3 cycle as a central component in our NADH “wound” model outlined in Fig. 22. When the alternative agonist membrane receptor (MR) has been stimulated by a wound “signal”, the membrane transmitter (MT) activates the signal transduction system (STS; phospholipase C), and this produces an increase in the level of IP3. The consequences of this increase in IP3are:

1. Activation of a protein kinase which results in phosphorylation of a protein subunit of the H+-ATPase near the Pi release site. This causes a reduction in H+ efflux; the effect will depend on the extent of phosphorylation and the dephosphorylation capability of the cell. 2. Ca2+ release, which may or may not exert its effects on cellular metabolism via Ca*+-calmodulin(CaCM). 3. Direct or indirect effects of IP3 on the gating (G) of ionic channels in both the plasma membrane and tonoplast. An important consequence of this wound activation of MT is that in this state the transmission system is incapable of accepting alternative stimuli. As indicated in Fig. 22, we suggest that an external NADH-oxidizing site can function to stimulate the MT system. Evidence for the presence of this NADH oxidation system comes from a reinterpretation of Lin’s trypsin data (Lin, 1982b; see also Buckhout and Hrubec, 1986). Although the SDSpolyacrylamide gel electrophoretogram of the TCA-precipitated protein obtained from the supernatant of trypsin-treated corn root protoplasts indicates a high level of endogenous protease activity, Lin (1982b) was able to show that the supernatant was capable of NADH oxidation. If endomem-

134

L. V. KOCHIAN AND W. J. LUCAS

EXOGENOUS

NADH

+2H+

I

I

+

H'

POTASSIUM TRANSPORT IN ROOTS

135

brane contamination (mitochondria and endoplasmic reticulum) was absent (but see Komor et al., 1987), this NADH oxidation may be attributed to the presence of a 42 kDa polypeptide released by mild trypsin action (Lin, 1982b, 1984). This protein appears to be able to oxidize NADH and reduce O2 (provided that O2is present); these properties are inconsistent with the NADH redox model proposed by Lin (1984). Note that in the model presented in Fig. 22, exogenous NADH will result in the following: 1. A stimulation of the MT and STS, with the consequences outlined above. 2. A continued stimulation of O2consumption, even though the cell has received a wound signal. 3. Proton consumption at the outer surface of the plasmalemma; depending on the degree to which the total H+-ATPase system is inhibited, this may give rise to alkalinization of the root surface. 4. Continual exogenous NADH oxidation is required to maintain the enhanced IP3 level; removal of the exogenous NADH results in a decline in IP3with the rate being dependent on the levels and/or activation states of the enzymes involved in its breakdown and resynthesis to phosphatidylinositol 4,5-biphosphate. These properties would account for all of our experimental observations on the perturbative influence of exogenous NADH on K+ fluxes in corn roots and corn root segments (Kochian and Lucas, 1985;Lucas and Kochian, 1987). How, then, is it possible to explain the NADH-mediated stimulation of K+ influx observed in corn root segments, corn root protoplasts (Lin, 1982a, 1984) and carrot suspension cells (Misra el al., 1984)? Based on recent reports of plasmalemma-bound electron transport systems, we have included two forms of NADH redox systems in our model (Fig. 22). We propose that these systems would function only when the protoplast, cell or tissue is not in a “wound” state. The protoplast is the easiest system to

Fig. 22. Model summarizing the various effects of NADH on K+ influx into corn roots (or protoplasts). Two transplasmalemma NADH redox systems are shown. The Lin (1984) model (top) utilizes endogenous or exogenous NADH and transports both electrons and H+. The Rubinstein and Stern (1986) model transfers only electrons across the plasmalemma. In this model, generation of the H+ in the cytoplasm, by the oxidation of NAD(P)H, is compensated for by transport out of the cell by the H+-ATPase. In both models, stimulation of K+ uptake would occur by H+-K+ co-transport. We propose that the NADH-induced “wound response” exhibited by corn, is mediated by the coupling of a membrane-bound signal transduction system (STS) to the IPS cycle. Two agonist membrane receptors (MR) are postulated, one being an NADH oxidase [the 44 kDa protein from Lin (1984)], and the other being an alternative receptor mediating other types of wound response. Binding to the MR activates the STS (phospholipase C) via a membrane transmitter (MT). The resultant release of IP3 acts via the activation of protein kinases andor the release of Ca2+, to modify transport proteins (H+ATPase, cation channels) at the plasmalemma and tonoplast. Redrawn from Lucas and Kochian (1988).

136

L. V. KOCHIAN AND W. J. LUCAS

rationalize, because lysis during preparation could release proteases that actually remove the NADH MR protein. Buckhout and Hrubec (1986) also reported that washing their isolated plasmalemma preparation (with or without salt) resulted in a partial loss of NADH reductase as well as oxidase activity. In this state, the protoplasts may oxidize NADH (either exogenously supplied or via the cytosol) to cause an increase in the A& across the plasmalemma by direct transport of H+ (Lin, 1984; Bottger and Luthen, 1986) or by coupling to the H+-ATPase (Rubinstein and Stern, 1986). The IP, recycling system may also vary from cell to cell, or from one tissue to another, as a consequence of the physiological prehistory of the plant system. Since lithium acts as an inhibitor of IP, conversion to inositol and hence its resynthesis to phosphatidylinositol 4,5-bisphosphate, growing cultures in the presence of lithium may reduce the sensitivity of the IP, regulatory system, thereby allowing the NADH redox system to function. Introduction of myoinositol should re-establish the wound response. A further test of the IP,-regulation/wounding hypothesis may be achieved using FC. Chastain and Hanson (1982) have shown that in wounded corn roots FC can override the endogenous control mechanism that inhibits the plasmalemma H+-ATPase. It will be of interest to see whether FC can negate the perturbative influence that exogenous NADH has on K+ influx into roots. Purification of these putative redox systems of the plasmalemma should greatly assist the elucidation of their function. Questions of sidedness, direction of electron transport, and natural substrates (donors and acceptors) are all important issues that need to be resolved. Reconstitution of these proteins into liposomes may provide an unambiguous answer to the question of whether or not they are important in energizing K+ transport into roots.

IV. RADIAL K+ TRANSPORT TO THE XYLEM A symplasmic pathway for K+ movement, from the epidermis to the xylem parenchyma, was first proposed by Crafts and Broyer (1938). In their hypothesis, the cortex was the site where nutrients, like K’, were taken up into the symplasm by active, 02-dependent processes. Movement to the stele was via plasmodesmata, and release into the xylem vessels was by diffusion, or leakage, since the compact tissue of the stele was thought to be low in oxygen. Some 38 years later, Anderson (1976), in his review on solute transport across the root, concluded that, “In the stele, the most common proposal is that the parenchyma either leaks or secretes solutes which then diffuse or are swept along into the vessels.” Figure 23 is a diagrammatic representation of the pathway discussed by Anderson (1976). Based on Anderson’s reviews, it

POTASSIUM TRANSPORT IN ROOTS Epidermis

Cortex

~

Endodermis

near- unity 0 value

w

137

e

low d value

7

External

-

APOPl and water fluxes

probably turgor drlven

Arrows show @ fluxes

Apo$asmic barrier

Fig. 23. Diagrammatic sketch of the root in cross-sectional view showing what Anderson (1976) considered to be the “usually accepted mechanisms of ion and water through-putto the xylem vessels.” Stippled areas represent the cytosol which, in this Anderson model, also extends in the xylem vessel. From Anderson (1976).

would appear that over the period from 1938 to 1976 the only improvement to the Crafts and Broyer hypothesis was the concept that solutes might be secreted into the xylem vessels. However, the mechanistic basis for this “secretion” remained poorly defined. Note also that Fig. 23 suggests that the xylem vessels may be symplasmically connected to the surrounding parenchyma and contain parietal cytoplasm. For those who are quick to be critical of this concept, we should point out that the controversy of whether living xylem vessels are engaged in the release of K+ to the translocation stream is still unresolved (Lauchli et al., 1978; McCully et al., 1987). Clearly, the complexity of the root has made it difficult to obtain a rigorous test of the Crafts and Broyer hypothesis. However, experiments of the past decade have provided some new insights. A. SITE OF K+ ENTRY INTO THE SYMPLASM

1. Epidermis or Cortex? Most current textbooks on ion transport and plant physiology discuss K+ movement into the cortex in terms of two parallel pathways, namely the apoplasmic and symplasmic routes. However, there is still some question as to the relative importance of each component (Anderson, 1976; Lauchli, 1976; Van Iren and Boers-van der Sluijs, 1980; Kochian and Lucas, 1983). Using a theoretical model for Rb’ uptake into barley roots, Bange (1973) found that the model only generated reasonable profiles when transport was

138

L. V. KOCHIAN AND W. J. LUCAS

restricted to the epidermis. The studies of Van Iren and Boers-van der Sluijs (1980) offered experimental support for this hypothesis. These workers investigated K+ uptake by applying the assumption that plasmolysis would sever plasmodesmata, thereby symplasmically isolating each cortical and epidermal cell of a barley root. Autoradiographic localization of 86Rbf, accumulated by the root following plasmolysis, was generally limited to the root periphery. These results suggested that when roots are exposed to low K+ concentrations (<1 mM), transport into the symplasm occurs at the epidermis; i.e. the apoplasmic pathway and subsequent uptake by the cortical cells is of little significance. The criticism which must be levelled at this work is that it may be erroneous to base a hypothesis concerning “normal” transport function on data obtained from a system where transport physiology may have been dramatically changed by plasmolytic treatment. Furthermore, the work of Anderson and Reilly (1968) concerning fluid exudation from the xylem of corn roots indicated that the cortical cells do have the capacity to absorb ions. They found that surgical removal of the epidermis and outer cortex of excised corn roots did not prevent significant fluxes of ions and H 2 0 to the xylem. Kochian and Lucas (1983) examined this issue by using radioactively labelled sulphhydryl reagents, [ 3H]N-ethylmaleimide (NEM) and [*03Hglpchloromercuribenzenesulphonic acid (PCMBS). A brief (30-60 s) exposure of corn roots to either NEM or PCMBS dramatically reduced K+ influx into the root, without affecting root respiration. Autoradiographic localization studies revealed that sulphhydryl binding occurred almost exclusively in the cells of the root periphery (see Fig. 24). However, protoplasts isolated from corn root cortical tissue exhibited significant “Rb+ influx, the kinetics of which were identical in shape to those obtained on corn roots (Kochian and Lucas, 1982a, 1983). These results suggest that although cortical cells possess the capacity to absorb ions at the low concentrations likely to be present in normal soils, Kf influx into the root symplasm most probably occurs at the root periphery.

2. Aerenchyma Induction The role of the cortex in K+ uptake and radial movement into the stele has been investigated by the induction of aerenchymatous tissue (Drew et al., 1980; Drew and Saker, 1986;Deacon etal., 1986). Drew et al. (1980) showed that an 02-stresstreatment of young corn roots induced aerenchyma in the developing, adventitious roots (Fig. 25). Microscopic examination of this root tissue revealed extensive breakdown of cells within the mid-cortex, while the epidermis, endodermis and stele remain unaffected. A comparison of this aerenchymatous tissue with control roots revealed that the volume of the cortex was reduced by 36% and that there was an 80% reduction in the symplasmic pathway available for K+ movement from the epidermis to the stele. Thus, if the cortex were involved in Kf uptake, the flux into these

POTASSIUM TRANSPORT IN ROOTS

139

60

so

5 8

30 a NEM Trootmont

40

a a

3

10

0

Fig. 24. Microautoradiographic localization of [3H]N-ethyl maleimide (NEM) following a 3fJ-9exposure of corn root segments to 0.3 mM NEM. Numerous cross-sections were analysed

for grain distribution with an Imanco Quantimet 720 Image Analysing Computer. These data have been averaged and presented in the histogram below the micrograph. Data from Kochian and Lucas (1983).

Fig. 25. Scanning electronmicrograph illustrating the aerenchyma that develops in nodal, adventitious roots of corn grown in non-aerated culture solution. Root sections were cut from the region 8-10 cm from the root tip and lyophilized before being viewed in the electron microscope. Symbols are as follows: C, cortical air spaces; W, wall residues of collapsed cells; I, intact cells linking inner and outer cortex. (Bar represents 500 pm.) From Drew etal. (1980), with permission.

140

L. V. KOCHIAN AND W. J. LUCAS

aerenchymatous roots should be reduced relative to control roots. Drew and Saker (1986) investigated the ion-transporting capability of such corn roots and found that, with respect to K', PO$- and C1-, aerenchymatous roots were at least as effective as control roots. These results obtained using aerenchymatous roots are consistent with the hypothesis that ion uptake, at low concentrations, occurs primarily at the epidermis. In view of this lack of effect of a reduction in the symplasmic pathway, it would also appear that the radial flux to the stele is rate-limited by transport into the symplasm (see Anderson, 1976). However, a note of caution is needed in the interpretation of these flux data. The experimental sys;tem used by Drew et al. (1980) and Drew and Saker (1986) imposed a relatively thick unstirred layer at the surface of both control and aerenchymatous roots. It is possible that the equivalent performance of the two root types was due to a limitation in the supply of nutrients to the actual root surface. This possibility should be investigated using a well-mixed bathing medium. Alternatively, flux analysis using the technique of Newman et al. (1987) might be informative. B. RADIAL PATHWAY

1 . Symplasmic Route Although the symplasmic pathway is the currently accepted route for K+ movement from the epidermis to the stele, there have been only a few studies on plasmodesmatal structure, frequency and function in root tissues. Spanswick (1972) was the first to demonstrate that corn root cortical cells are electrically connected via plasmodesmata. Clarkson and Robards (1975) showed that plasmodesmatal connections exist between cortical, endoderma1 and stelar parenchyma cells in the mature, transporting region of the root. The elegant work by Overall et al. (1982) and Overall and Gunning (1982) provided important information on the substructure of the plasmodesma in Azolla roots but, unfortunately, this work focused on the root apex. Consequently, their findings with respect to symplasmic coupling cannot be extended to the mature region of terrestrial root systems. Plasmodesmatal frequency distribution studies have been conducted on the root apex of corn (Juniper and Barlow, 1969), and this work indicated that as the cells matured, the number of plasmodesmata per unit area decreased. However, this study also focused on developing tissue. Perhaps the most interesting study, from the viewpoint of radial transport, is the electrical systems analysis of Lipidium sativum roots reported by Behrens and Gradmann (1985). By employing a combination of microelectrodes located at various positions along the root, it was possible to detect electrical anisotropy within the root, there being a preference for longitudinal over transverse (or radial) coupling. These results suggest that further studies are

141

POTASSIUM TRANSPORT IN ROOTS

needed on the spatial distribution of plasmodesmata within the cortical tissue of mature roots. It seems almost inconceivable that the concept of symplasmic K+ transport is so well entrenched, in view of the almost complete absence of supporting cytological data obtained on mature root tissue.

2. Compartmentation and the Endoplasmic Reticulum? In the Robards (1971,1975) model of the plasmodesma, there is continuity between the endoplasmic reticulum (ER) of neighbouring cells, via the desmotubule. Lauchli and others have extended this concept of an ER continuum in terms of providing a specialized K+ transport compartment in roots (see Fig. 26). Much of the justification for this special compartment comes from X-ray microprobe analysis and cytohistochemical precipitation data. Lauchli et al. (1971) analysed the radial distribution of K+ within the roots of 8-day-old corn seedlings. Roots were pretreated for 4 h in 0.2 mM KCI plus 0.5 mM CaS04,before the terminal 1cm was removed and subsequent 1 mm sections cut and immediately frozen in liquid N2. Following freezesubstitution and embedding in Spurr's resin, anhydrous sections were cut (1-2 pm) and analysed for K+ distribution using X-ray microprobe. The reported line scan indicated a high level of K+ in the epidermis and within the cells of the stele; the xylem parenchyma appeared to contain the highest levels of K+. Clearly, there are problems in evaluating these X-ray microprobe data, the question of artefacts generated during tissue preparation being the most serious. Of course, there is also the problem of spatial resolution, with the possibility of contamination from the vacuole and, to a lesser extent, the cell walls. To circumvent these problems, Lauchli et al. (1977) used the bulktissue frozen-hydrated specimen protocol developed by Lauchli (1975). Barley root segments, obtained from a position 1-2 cm from the apex of soil-grown tissues, were frozen in liquid N2 and then fractured under liquid Xylem

Vessc1

Fig. 26. Pathways of ion (K') transport in roots, involving the endoplasmic reticulum (ER) as a special compartmentof the symplasmicroute. Symbols: C, cytoplasm; C,, Casparian strip; V, vacuole. From Lauchli (1976), with permission.

142

L. V. KOCHIAN AND W. J. LUCAS

N2. This tissue was then quickly transferred to an SEM cryostage (held at - 15OOC). A comparison of the X-ray spectra collected from epidermal and xylem parenchyma cells again revealed that K+ was higher in the xylem parenchyma. However, the K+ level in the metaxylem vessels was suspiciously high, perhaps indicating the presence of developing vessel elements (McCully et al., 1987). Although one might have more confidence in the K+ distribution determined on frozen-hydrated tissue, one must not lose sight of the fact that barley roots are still quite thick and so ice crystal artefacts cannot be discounted, especially deep within the root where the rate of cooling would be the slowest. Spatial resolution on bulk frozen-hydrated tissue is also a serious limitation, especially when the SEM is operated at 10 keV. Since the cytosolic compartment in the barley xylem parenchyma cells was less than 0.5 pm [see Fig. 5 of Lauchli et al. (1977)], it is highly unlikely that the system could resolve K+within the cytosol. Thus, provided that ice crystal artefacts are not seriously affecting the K+ concentrations, the above data establish that the K+ level in the vacuole of the xylem parenchyma is higher than elsewhere in the root. The problem of spatial resolution on frozen-hydrated root tissue was further investigated by Echlin et al. (1981) and Pitman et al. (1981). In these studies, and especially in the latter, it is clear that the SEM system was operating at the limit of resolution. Additionally, the images showed clear evidence of ice crystal artefacts and so X-ray microprobe data on such tissue must be interpreted with caution. Finally, Pitman et al. realized that their data on relative distributions of K+ and Na’ were of a preliminary nature with respect to assigning valid concentrations. The problem of chemical activities versus “bulk” concentrations remains a formidable problem. Even with the above-mentioned limitations of the X-ray microprobe technique, data collected on root tissues by several groups show that variations in K+, Na+ and C1- distributions, among individual cells of the root, do not correlate with their spatial position in the root (Lauchli et al., 1977; Echlin et al., 1981; Pitman et al., 1981; Chino, 1981; Harvey, 1985). As stressed by Pitman et al. (1981), these spatial inhomogeneities in ionic distribution draw into question one of the most basic assumptions underlying radioisotope flux analysis, namely, a homogeneous distribution of the label. Finally, at present, it is our assessment that statements made about K+ concentration gradients across the root must be interpreted in terms of vacuolar changes. It is anticipated that X-ray microprobe studies, conducted with more advanced systems capable of higher spatial resolution and lower accelerating voltages, will provide the much needed information on cytosolic versus vacuolar concentrations at all sites across the root. Without such data, it will be difficult to make progress in elucidating the role played by the xylem parenchyma in K+ transport into the xylem. A final comment on the putative role of the ER in K+ transport is

POTASSIUM TRANSPORT IN ROOTS

143

necessary. Although no unique model of the plasmodesma is accepted at present, the model proposed by Overall et al. (1982) (see also Lopez-Saez et a f . , 1966) appears to be consistent with most studies on symplasmic permeability (Erwee and Goodwin, 1983, 1984; Erwee et a f . , 1985; Madore et a f . , 1986; Terry and Robards, 1987). In this model, the “desmotubule” is presented as a solid phospholipid rod, not an open tubule as in the Robards model. Movement through the plasmodesma is considered to occur via the “cytoplasmic annulus” and not via the ER-associated desmotubule. Hence, the role of the E R in intercellular K+ transport should be viewed as being entirely speculative and perhaps at variance with the current view of the structure of the plasmodesma.

C . LAG PHASE IN XYLEM LOADING

An extensive literature exists on the application of compartmental flux analysis to K+ movement across the root. Theoretical aspects of this work have been reviewed by Walker and Pitman (1976). For a broad description of compartmental flux analysis, as it pertains to ion transport across the root, the reader is referred to Jeschke (1980) and Pitman (1982). For the present, we will focus on the phenomenon of the lag phase in K+ transport into the xylem, to illustrate the complex nature of the processes controlling K+ movement into the various compartments of the root (see also Section V). When low-salt-grown plants are transferred from their CaS04medium to a solution containing K+ (or “Rb’), there is immediate uptake into the symplasm of the root.,However, as shown by the data presented in Fig. 27, transport into the xylem translocation stream is delayed, or lags behind root uptake by a period of 1-4 h (Hooymans, 1976; Bange, 1977; Glass, 1978; Van Iren e t a f . ,1981). Van Iren etal. (1981) showed that this lag was not due to a time-dependent penetration of K+ into the stele. Also, the extent of the lag phase was independent of both the amount of K+ in the bathing medium and the extent of Kf uptake by the root (Fig. 27). However, it would appear that the abrupt transition from root accumulation to transport to the xylem is influenced by the rate at which the root is taking K+ into the symplasm (see Fig. 27c). In subsequent studies, Bange (1977,1979) has shown that during this lag phase, %Rb+accumulation into the vacuole starts slowly, even though the cytoplasmic “compartment” seems to fill quite rapidly. Consequently, the difference between the lack of K+ accumulation in the barley roots bathed in 5 and 10 ,UM K+ (Fig. 27a,b) and the accumulation found in 1.0 mM K+ (Fig. 27c), must reflect the operation of a regulatory mechanism that in some way senses the K+ requirement for xylem translocation. When this requirement is exceeded, vacuolar transport of K+ appears to be activated [but see Glass (1978) for an alternative explanation]. At present no informa-

144

L. V. KOCHIAN AND W. J. LUCAS

.-

(a 1 n

5pM Kt

-

c

I

I

I

1 2 3 4 5 6 7 8

0

I root

1 2 3 4 5 6 7 8

12 34 5 6 7 8

Time (hours) Fig. 27. Influence of extracellular K+ levels on Kf uptake into intact barley roots (0)and subsequent translocation to the shoot (0) as a function of time after roots were introduced to K2S04.Experimental solutions contained 5 mM CaCI2,0.1 mM Ca(HC03)2and 5, 10 or 1000 ~ L K+ M in (a), (b) and (c), respectively. Note the pronounced lag in the supplyof K+ to the shoot. Data from Hooymans (1976).

tion exists as to the nature of this putative sensing mechanism (see Section V for a further discussion on regulation of K+ transport within the root). Hooymans (1976) argued that the insensitivity of the lag phase to both external and bulk root K+ concentration supported the hypothesis that a specialized compartment or “organelle” was involved in K+ transport to the xylem. Glass (1978) offered a seemingly more plausible hypothesis that the lag phase reflected an inductive period associated with establishing the conditions necessary for K+ transport into the xylem. Induction by high symplasmic Kf concentration (Glass, 1978) would be inconsistent with Hooymans’ (1976) data; consequently, a threshold K+concentration may be involved in the induction process. Van Iren et al. (1981) attempted to test this induction hypothesis by giving low-salt-grown barley plants a brief exposure to K+. They argued that this brief treatment would allow sufficient K+ to enter the symplast to activate the induction process. If this were correct, the lag would be either absent or considerably shortened when these pretreated plants were resupplied with K’. In experiments where a 5-min K+pretreatment was employed there was no observable reduction in the lag period. It is unfortunate that Van Iren et al. did not investigate the effect of longer pretreatment periods.

POTASSIUM TRANSPORT IN ROOTS

145

Another approach to studying the nature of this lag phase would be to employ p-fluorophenylalanine (FPA), an analogue of the amino acid phenylalanine, which produces abnormal or defective proteins. A review of the effect of this compound on transport across the root is given by Pitman (1977). When high-salt barley roots are treated with FPA, transport into the root remains unaffected, whereas transport into the xylem becomes inhibited (Schaefer et al., 1975). The response to FPA is quite rapid, which indicates that the essential “transport” protein(s) involved in the release of K+ or C1- to the xylem turns over quite rapidly. Low-salt-grown roots could be pretreated with FPA, for various times, prior to introducing K+ to commence the induction phenomenon. Roots pretreated in FPA should not commence transport into the xylem, but upon FPA removal transport should commence almost immediately if a threshold level of cytoplasmic K+ is involved. Studies by Jeschke (1984) also implicate a role for the phloem in the duration of the lag period. On a speculative note, it may be that the induction of transport to the xylem may be affected by the level of K+ recirculating to the root via the phloem. In any event, the abrupt transition to xylem loading (Fig. 27) is at variance with loading via K+ leakage (see Fig. 23), and at first sight seems inconsistent with K+ release into the translocation stream by late-maturing xylem elements (McCully et a l . , 1987). D. K+ TRANSPORT INTO THE XYLEM

Roots grown in a full nutrient medium (high-salt status) take up K+ and transport a major part of this influx to the shoot. By contrast, in low-saltgrown roots a definite lag period exists between K+ entering the root and its transfer to the xylem (see Fig. 27). Numerous studies have contributed to this type of phenomenological characterization of the transport of mineral nutrients from the soil solution to the shoot (Pitman, 1977,1982). Although it was necessary to utilize a flux compartmental analysis approach in many of these studies, the data could be evaluated without the assignment (or knowledge) of specific transport mechanisms. The active or passive nature of these specificfluxes (to the symplasm and the xylem vessels) was deduced from studies involving various metabolic inhibitors and chemical agents, as well as from frustratingly difficult attempts at determining the electrochemical potential gradients for the various ions (Lauchli, 1976; Bowling, 1981; Pitman, 1982). Although it may come as a surprise to the reader, it was not until 1978 that Mitchell’s chemiosmotic hypothesis was applied to solute transport at both boundaries of the symplasm; i.e. at the plasmalemma of the epidermal and xylem parenchyma cells. Hanson (1978) developed a “speculative” model in which he placed two “opposing” H+-translocating ATPases in the plasma

146

L. V. KOCHIAN AND W. J. LUCAS

membranes of the outer cortex and the stelar parenchyma. The impermeable nature of the Casparian strip of the endodermis separated the apoplasmic solutions of the cortex and stele, with the symplasm operating as a “bridging” osmotic unit. Transport of cations and anions was proposed to occur by various uniport and antiport systems. It is worth stressing that K+ transport into the xylem was hypothesized to take place by a K+-H+ antiporter with the flux being dependent on the A&. An important prediction of the Hanson (1978) hypothesis was that the transroot potential, measured between the apoplasm of the xylem and the bathing medium, should have two opposing (or “antagonistic”) electrogtnic components. Although many electrophysiological studies had been conducted on the nature of the transroot potential, no study indicated the presence of an electrogenic component at the xylem parenchyma plasma membrane (Shone, 1968,1969; Davis and Higinbotham, 1969; Ansari and Bowling, 1972; Bowling and Ansari, 1972). In general, these studies suggested that the transroot potential (measured from the xylem with respect to the bathing medium) could be interpreted based on the transmembrane potential difference equivalent to that of a single cell. Although their experiments were conducted on an excised hypocotyl-root system from Vigna sesquipedalis, Okamoto et al. (1978, 1979) were able to demonstrate the existence of opposing electrogenic pumps between the surface of the hypocotyl and the xylem vessels. De Boer et al. (1983) constructed a similar experimental system to investigate the transroot potential of two Plantago species. The elegant system developed by De Boer et al. (1983) is illustrated in Fig. 28 (the reader is directed to the original paper for important technical details). An investigation of the response of excised Planfago roots to changes in O2partial pressure revealed three classes of transroot potentials. One class, having transroot potentials of approximately -100 mV, was comparable to the earlier reports of Davis and Higinbotham (1969) and Shone (1968, 1969). The second and third class of roots had transroot potentials from 0 to -20 mV and -50 to -70 mV, respectively, with the latter class being the less frequently observed of the three classes. The response of the second class of Planfago roots to changes in the O2 partial pressure is illustrated in Fig. 29. Anoxia, and recovery from anoxia, elicited a transient change in the transroot potential (Fig. 29a), while exposure of the root to 10% oxygen resulted in a shift in the transroot potential from -20 mV to -60 mV (Fig. 29b). An electrical analogue of Hanson’s (1978) model is presented in Fig. 29c. As discussed by De Boer et al. (1983), this model can readily account for the observed responses in the transroot potential when the excised Plantago root undergoes changes in oxygen status. If the two opposing electrogenic components are functioning and almost equal, their net effect on the transroot potential will be small, and the value of the potential will depend

Fig. 28. Experimental system used by De Boer et uf.(1983) to simultaneously measure the transroot potential(TRP), ionic activitiesfor H', K+and Na', and the xylem sap flow rate (Jv) on an excised Pfuntugo root. Symbols: A, amplifier;A', precision intrumentationamplifier; SOS, slotted optical switch. From De Boer et al. (1983), with permission.

148

L. V . KOCHIAN AND W. J. LUCAS

-80

I 40

-60

'

-,

30 20

-40

E

10

-

oe 40

4

30 -40

20

-20

10

0

50

Time ( C )

100 150 minutes 1

1

200

(

Potential Profile

I Caaparian strip

Fig. 29. Response of the transroot potential of Plantago to changes in the O2tension of the medium bathing the roots. (a) Roots brought under anoxia and then returned to 30% 02. (b) Oxygen tension reduced from 30% to 10% and then subsequently returned to 30%. (c) Schematic representation of the electrophysiological organization of the Planfago root. The Em of the epidermaUcortical cells ( E l ) and the xylem parenchyma cells (Ez) result from the combined EMF of the passive diffusion potential (EDland E D 2 , respectively) and the electrogenic pump (Epl and Ee2, respectively). The hyperpolarization produced by the electrogenic pumps (EHypIand EHm) will depend upon the EMF and conductances ( g ) of the pump and passive diffusion systems. The transroot potential is given by: TRP = El - E2 = (EHYPI + EDI)- (EHY~L + ED^). Data from De Boer et al. (1983).

POTASSIUM TRANSPORT IN ROOTS

149

upon the difference between the diffusion potentials generated at the boundary of the cortex and the xylem parenchyma. When the O2level in the bathing solution is lowered from 30 to 10% (Fig. 29b), the O2supply to the inner regions of the root will be reduced, causing a decline in the available ATP level, and so the activity of the electrogenic pump on the xylem parenchyma plasmalemma will be decreased. The outer region of the root will still have an adequate O2supply, and so the continued operation of the outwardly directed H+-ATPase will cause the transroot potential to shift to a more negative value (Fig. 29b). Anoxia eventually reduces the activity of both H+-ATPases, but the differing time courses of the decline in ATP give rise to the observed transient change in the transroot potential (Fig. 29a). The electrophysiological data obtained on Pluntugo roots (De Boer et al., 1983; De Boer and Prins, 1984, 1985) and the related measurements conducted on Vigna hypocotyls (Okamoto et ul., 1978, 1979) provide support for the most important aspect of the Hanson model for ion transport into the xylem. Further support is also offered by the xylem perfusion studies of Clarkson et al. (1984) and Clarkson and Hanson (1986). Using a special xylem perfusion system, these workers introduced unbuffered solutions of different pH values into excised roots of AZZium cepa. Irrespective of whether the initial pH value was acidic (pH 3.9) or basic (pH 9.3), the pH value of the perfusate leaving the root was adjusted to within a range of pH 5.5-6.5. Furthermore, the extent of H+ efflux into the xylem perfusate could be increased greatly by buffering the medium at alkaline pH values. It is also interesting to note that malate has been reported to be a normal constituent of the xylem sap in some roots (Butz and Long, 1979); the pK,(2) of malate is -5.3. These pH shift data were interpreted in terms of pH regulation of the xylem sap, in situ; a process which would be essential if K+ transport into the xylem were to occur via a K+-Hf antiporter as suggested by Hanson (1978). Figure 30 represents a synthesis of current data on K+ transport into the symplasm of the root and its efflux across the xylem parenchyma to the xylem vessel. Unfortunately, the complexity of the root, and especially that of the stele, has made it extremely difficult to conduct the necessary biophysical experiments to gain further data in support of the Hanson model. Although various attempts have been made to determine electrochemical gradients across the cells of the root (Dunlop and Bowling, 1971; Bowling, 1972; Dunlop, 1973,1982;Davis and Higinbotham, 1976), these studies generally involved making successive electrode impalements through cells of the epidermis, cortex and stele. As stressed by Anderson and Higinbotham (1975), great caution should be used when evaluating these data. Thus, with the spatial resolution of the X-ray microprobe system(s) used to examine K+ distribution across the root being such that only vacuolar estimates can be deemed reliable, and the microelectrode studies on the stelar tissue being

150

L. V. KOCHIAN AND W. J. LUCAS

Fig. 30. Schematic representation of the transport processes thought to be involved in the acquisition of K+ from the soil solution and in its subsequent transport from the symplasmic compartmentof the root into the xylem vessels. This model has been developed on the basis of the chemiosmotic scheme originally proposed by Hanson (1978) and later refined by Clarkson and Hanson (1986). (Open “arrowed” circles on tonoplast represent regulation of K+ movement between the cytoplasm, vacuole and radial transport “pools”, and rectangles represent channels.)

somewhat dubious, precious little information is currently available concerning the Kf gradients across the root. The only reliabledata concerns the K+ concentrations within the xylem vessels, and even here one must be concerned with extrapolation from in vitro measurements to the intact root. As a consequence of these limitations, we can only make general comments on the thermodynamic nature of the Kf fluxes at the two boundaries of the root symplasm. Provided that the Donnan charges within the epidermal cell wall do not result in a significant increase in the K+ concentration at the outer surface of this plasma membrane, the typical gradients illustrated in Fig. 30 indicate that net Kf influx into the symplasm is clearly an active process. Our comments concerning the situation at the xylem parenchyma plasmalemma cannot be made with the same degree of certainty. If the K+ concentration in the cytoplasm is within the range expected for normal physiological functioning (250mM), and the steady-state level in the xylem vessels is approximately 5 mM, the thermodynamic status of K+ efflux will depend very much on the magnitude of the potential across this membrane. If the value is similar to that deduced for the Plantugo root system, then K+ efflux would necessarily occur by an active process. Although the K+-H+ antiport hypothesis is still purely speculative, if K+ were to be transported by such a mechanism, with a 1: 1stoichiometry, the Em would not influence the thermodynamics of the situation, but it could have a significant influence on Kf efflux through its effect on the kinetic parameters of the putative transport system.

POTASSIUM TRANSPORT IN ROOTS

151

It is perhaps indicative of the complexity of the root that hypotheses like the Crafts and Broyer (1938) and Hylmo (1953) models have been so difficult to disprove. It is interesting to note that the Hylmo model was recently resurrected by McCully et al. (1987). The Hylmo model proposes that ion accumulation in the root involves uptake into the vacuoles of living xylem vessels, which upon maturation release their contents into the translocation stream. Although this model received some degree of support in the mid-1970s (Davis and Higinbotham, 1976), Lauchli et al. (1978) provided irrefutable evidence that ion transport occurs across the stele into mature xylem vessels. However, in the recent report by McCully et al. (1987), X-ray microprobe was used on bulk, frozen-hydrated tissue, to examine the K+ concentrations within the late-maturing xylem vessels of field-grown corn plants. Surprisingly, K+ values in the range of 150-400 mM were detected; the microprobe system was “calibrated” against known KC1 standards that were vacuum-infiltrated into the mature xylem vessels. Clearly, these levels of K+ could only be established by transport across the tonoplast of these maturing vessels. In view of these findings, McCully et al. claim that since the living, late-maturing xylem elements contain the highest K+ concentrations of any cells in the root, they “represent the end point of potassium accumulation from the soil before it is released as a slug of concentrated solution to the transpiration stream in the mature vessel”. In view of all the available data, it seems to us that a combination of the two routes may not be completely unrealistic.

V. REGULATION OF K+ FLUXES WITHIN THE PLANT In contrast to the situation in animal cells where homeostasis is mediated primarily at the tissue or organ level, plants must be able to regulate their metabolism at the cellular level. This is particularly true for the cells of a root growing in the soil, which can be exposed to a wide range of environmental conditions. Glass and Siddiqi (1984) suggest that two primary priorities for the regulation of inorganic nutrients in the plant are the maintenance of ionic composition of the cytoplasm and nutrient allocation to the shoot. These priorities are particularly applicable to K+.Although the levels of K+ do not appear to be precisely regulated, the cytoplasmic values seem to be constrained within certain limits. In glycophytic plants, cytoplasmic K+ concentrations range from 100 to 200 mM (Leigh and Wyn Jones, 1984). It has been suggested that maintenance of K+ within these limits is important because of its roles in activating a number of cytoplasmic enzymes and in protein synthesis (Lauchli and Pfluger, 1979; Wyn Jones et al.,1979). Unfortunately, direct measurements of cytoplasmicK+ have been difficult to accurately obtain. However, it has been well documented that “bulk” K+ concentrations are generally maintained at fairly constant levels, even in the

152

L. V. KOCHIAN AND W. J. LUCAS

face of widely varying Kf supply to the root. For example, Asher and Ozanne (1967) studied the growth and K+ content of 14 plant species grown at external K+ concentrations held constant at levels varying from 1 to 1000 p ~ They . found that at 24 PM K+, eight of the species reached maximum yield, while at 95 p~ K', the remaining species reached maximum yield, and increases in external K+ above 95 p~ had little further effect on K+ content or root :shoot ratios. Although our knowledge of ion transport regulation is far from complete, most of the work that has been conducted has been directed at the cellular level. A large portion of this work has focused on the regulation of K+ uptake across the plasmalemma of higher plant roots, undoubtedly because of the important metabolic and osmotic roles played by this ion. An examination of the pathway of K+ movement from the root surface to the shoot indicates a number of potentially important regulatory sites. These include the plasmalemma of epidermal cells, the tonoplast of epidermal and cortical cells, the plasmodesmata hypothesized to be involved in maintaining symplasmic continuity along this pathway, and the xylem parenchyma, which may form the site of K+ transfer into the xylem translocation stream. The location where ions are transported into the xylem vessels appears to be an important focal point for root-shoot interaction, in terms of regulatory signals moving between the shoot and root. We will advance the concept that there is a hierarchy of regulatory mechanisms operating to regulate K+ within the plant which extend from the subcellular to the whole plant level. These various levels of regulation are thought to establish a feedback system which influences K+ uptake across the plasmalemma of root epidermal cells. We will deal with regulation of K+ transport at the level of the plasmalemma. Regulation of transport across the root will then be examined in terms of the importance of the shoot in the regulation of root Kf fluxes. A. ALLOSTERIC REGULATION OF K+ TRANSPORT

A number of studies that have involved either bulk tissue K+ measurements, or indirect measurements of cytoplasmic K+ (i.e. compartmental analysis), have suggested that cytoplasmic K+ is maintained at fairly constant levels, while vacuolar K+ concentrations can vary more dramatically (e.g. Glass and Siddiqi, 1984; Leigh and Wyn Jones, 1984; Memon et al., 1985). Hence, a number of researchers have sought to uncover the regulatory mechanisms involved in the maintenance of cytoplasmic K+ levels. Certainly the most widely discussed and studied model involves allosteric regulation of the plasmalemma K+ transport system by cytoplasmic Kf concentration. I . Feedback by Internal K + A negative correlation between K+ uptake and internal K+ concentration ([KIi) has been reported for a number of plant tissues (Johansen et al., 1970;

153

POTASSIUM TRANSPORT IN ROOTS

Young and Sims, 1972; Glass, 1975, 1976, 1977; Jensen and Pettersson, 1978; Cram, 1980; Kochian and Lucas, 1982a). In many of these studies, K+ influx was correlated with bulk tissue K+ levels. However, in the work of Young and Sims (1972) and Glass (1977), the adjustment of K+ influx and [K], following either an increase or decrease in the external Kf concentration was quite rapid and correlated with known half-times for the turnover of cytoplasmic K+. These results provide circumstantial evidence in support of the hypothesis that a regulatory feedback loop exists between cytoplasmic K+ levels and K+ influx across the plasmalemma. The K+ transport protein(s) may possess both catalytic (transport) and regulatory (allosteric) sites (Cram, 1976; Glass, 1975, 1976, 1977, 1978; Jensen and Pettersson, 1978; Pettersson and Jensen, 1978, 1979). Binding of cytoplasmic K+ to internal allosteric sites on the transporter may induce a conformational change in the protein which would result in a reduction of K+ influx. Further evidence in support of the allosteric regulation hypothesis for K+ influx into roots has been provided by the work of Glass and co-workers (for reviews see Glass, 1983; Glass and Siddiqi, 1984). Glass (1975) observed that when low-salt barley roots were being exposed to high exogenous K+ levels, in order to increase [K], , Kf influx was inversely related to the square of [K],, a result previously reported for Lemna (Young and Sims, 1972). Subsequently, the kinetics for Kf influx at low external K+ concentrations ( ~ 0 . mM) 5 were obtained on barley root tissue in which different [K], levels had been established (Glass, 1976). As shown in Fig. 31a, as [K], increased from 26.3 pmol g-' to 130.5 pmol g-', the V,,, for the saturable component of K+ influx decreased, and the related K,,, increased. Replotting these data

( b ) \O-

r

- 3 c

+-

E

Y

1 0 40

Extemal K+ConcentraUon (mM)

80

120

Root K+ Content ( pnd g-' 1

Fig. 31. Evidence for allosteric control of K + influx into barley roots. (a) Influence of pretreatment in 50 mM KCI upon the influx kinetics obtained on low-salt-grown barley roots. Symbols: 0, 0 h; 0 ,3 h; W, 6 h; A, 12 h duration in 50 mM KCI. (b) Relationship between influx values from solutions containing various KCI concentrations as a function of [Kli. Symbols(mMKCI):0,0.32;0,0.16; 0,0.08; W,0.04; A,0.02; A,O.Ol. DatatakenfromGlass (1976).

154

L. V. KOCHIAN AND W. J. LUCAS

as a function of [K], indicated that a sigmoidal relationship exists between K' influx and [K], (Fig. 31b). Glass (1976) noted that these observations are characteristic of an enzyme exhibiting allosteric regulation. He applied a transformation of the Hill equation to these kinetic data to determine the number of allosteric binding sites [a description of the use of the Hill equation can be found in Segel (19791. Linear Hill plots were obtained from which Glass deduced that four allosteric sites were operational on each transport protein. Glass has quite cautiously suggested (due to the dangers inherent in extrapolating from simple enzyme kinetics to a complex tissue such as a root) that in low-salt barley roots the K+ carrier possesses one binding site for external K' and four allosteric binding sites for [K],. Similar results have been subsequently obtained for K+ influx into the roots of a number of different plants (Jensen and Pettersson, 1978; Pettersson and Jensen, 1979). In low-salt-grown roots, the reduction in net K' influx, by internal K + , was not due to a change in K+ efflux (Glass and Siddiqi, 1984). This is not surprising, since it is well established that in low-salt roots net K+ uptake is equivalent to unidirectional K + influx, indicating that K' efflux is quite small (Johansen et al., 1970; Glass and Siddiqi, 1982; Newman et al., 1987). It had been previously reported that as excised barley roots were loaded with KCl, K+ influx remained constant but efflux increased (Jackson and Edwards, 1966). However, the fluxes in this study were measured in 10 mM KCI, which is in the range of passive K+ uptake that we suggest is mediated by K+ channels. In this concentration range, K+ influx does not respond significantly to [K], (Glass and Dunlop, 1978). It would appear that only the high-affinity K' uptake system (K'-H+ symport or K+-ATPase?) is subject to feedback control by internal K + (see Fig. 20). 2. Involvement of Protein Synthesis andlor Degradation? The response of the K+ transport system to changes in [KIi can be quite rapid, which indicates a direct control on the transport system, as opposed to regulation via synthesis or degradation of transport carriers, or regulatory proteins (Glass, 1977; Glass and Siddiqi, 1984). Glass (1977) found that the reduction in K+ influx, in response to increasing [K],, showed no lag and appeared to be independent of RNA, DNA and protein synthesis. Pettersson and Jensen (1979), on the other hand, have hypothesized that a combination of both allosteric regulation and carrier synthesis occurs either simultaneously or sequentially during the period of adjustment following a perturbative treatment. Unfortunately, for higher plants, techniques have not been available that allow for the demonstration of transport regulation through carrier synthesiddegradation. As Glass and Siddiqi (1984) point out, even though direct evidence is lacking, a realistic model for transport regulation would involve short-term adjustments involving rapid responses of transport systems to

POTASSIUM TRANSPORT IN ROOTS

155

changes in internal ion levels, in conjunction with long-term adaptations (hours or days) involving control of carrier synthesis and degradation. In this context, recent work by Fernando etal. (1987) indicated that K+ deprivation in barley roots resulted in an increase of both a 34 and a 44 kDa plasmalemma protein, as well as K+ uptake. However, the increase in K+ uptake occurred before any changes in the plasmalemma protein complement could be detected. Fernando et al. (1987) speculated that allosteric regulation occurs prior to a stimulation in the synthesis of specific plasmalemma proteins (in response to Kf deprivation). Clearly, further studies in this area are needed to confirm this hypothesis.

3. Regulation of Tonoplast K + Transport? The complex structure of the root makes it difficult to measure tonoplast fluxes and vacuolar content. In general, only approximate values can be obtained through radioisotope compartmental analysis studies. Glass and Siddiqi (1984) have analysed this literature and conclude that as the root is loaded with K+, both plasmalemma and tonoplast K+ fluxes are reduced, with the greater reduction being observed at the tonoplast. Leigh and Wyn Jones (1984) have also proposed that a coupling exists between the cytoplasmic and vacuolar K+ pools. They developed a model which attempts to explain the relationship between growth and tissue Kf levels. The model is based on the obsecvations that: (1) K+ is not replaceable in the cytoplasm; (2) K+ salts in the vacuole that are involved in turgor generation can be replaced by other solutes like Na’, Mg2+,Ca2+and organic molecules, during periods of K+ deficiency; and (3) K+ is often unequally distributed between the vacuole and cytoplasm. They propose that as the Kf concentration in a plant tissue declines (during periods of K+ deficiency), cytoplasmic Kf concentrations are initially maintained while vacuolar Kf is depleted and replaced by other solutes. A minimum vacuolar K+ level (usually around 20 mM) is envisaged; when the vacuolar K+ reaches this minimum, cytoplasmic K+ begins to decline and metabolism and growth become negatively affected. Support for the Leigh and Wyn Jones (1984) model was provided by the K+-use efficiency study of Memon et al. (1985). In this work three barley varieties were examined and one was found to exhibit “typical” K+ deficiency symptoms when grown at 10 PM K’. The important point is that this same variety was capable of accumulating K+ to higher levels, in the root and shoot, than the other varieties used in this study. Using a compartmental analysis approach, it was shown that the variety which exhibited K+ deficiency symptoms retained more of its K+ in the vacuole. This failure to transport vacuolar K+ to the cytoplasm resulted in a depletion of cytoplasmic K+,which then caused the onset of K+ deficiency symptoms. These studies suggest that the relative levels of cytoplasmic and vacuolar K+ must, in some fashion, be integrated. Furthermore, these concepts can be extended to indicate that the transport systems at both membranes must

156

L.V. KOCHIAN AND W. J. LUCAS

share some common controls. Observations have been made of membrane perturbations acting specifically at the outer face of the plasmalemma (applications of PCMBS, NADH, ferricyanide) that have resulted in significant alterations in tonoplast K+ fluxes (Kochian and Lucas, 1982a, 1985). In these studies, it was suggested that signalling mechanisms operate between the two membranes. The mechanism(s) of K+ transport at the tonoplast is poorly developed, and information concerning regulation of transport across the tonoplast is essentially nonexistent. Recent studies with tonoplast vesicles purified from roots and storage tissue have shown that the tonoplast contains two H+translocating enzymes, an ATPase and a pyrophosphatase (Blumwald, 1987). Both function to generate an electrochemical gradient for H+ across the tonoplast. This gradient is dominated by the activity component, which is considered to be approximately two pH units (the vacuole being acidic), while a small electrical component is also thought to exist. A number of H+-associated antiport mechanisms have recently been described for the tonoplast. Examples include Ca2+ uptake into the vacuole of oat roots (Schumaker and Sze, 1985) and Na+ uptake into the vacuole of red beet storage tissue (Blumwald and Poole, 1985). Evidence consistent with a K+-Hf antiport for tonoplast vesicles from red beet was recently presented (Sarafian and Poole, 1987). In terms of mechanisms for the transport of K+ into and out of the vacuole, one could speculate that K+ uptake may be mediated by a K+-Hf antiport, while K+ efflux from the vacuole could be facilitated by gated K+ channels, where the driving force for K+ efflux would be derived from the small (interior-positive) membrane potential. Of course these speculations do not describe potential regulatory processes for the tonoplast K+ transport systems, or offer explanations for how the plasmalemma and tonoplast might “communicate”. Answers to these questions must be sought in future studies. Criticisms of the Allosteric Model The major criticism of the allosteric regulation model is based on our inability, due to technical limitations, to make accurate measurements of cytoplasmic K+ concentrations. As Sanders (1980a) has pointed out, it is impossible to demonstrate a direct interaction between cytoplasmic K+ and the plasmalemma, since changes in root tissue content reflect primarily those occurring within the vacuole. Furthermore, according to the model of Leigh and Wyn Jones (1984), changes in vacuolar K+ content may not necessarily reflect what is occurring in the cytoplasm. Conversely, alterations in the cytoplasmic concentration of an ion can occur without corresponding changes in vacuolar concentration, as has been observed for C1- in perfused Chum cells (Sanders, 1980a). Consequently, caution must be exercised when attempting to generate models dealing with ionic levels of cytoplasmic compartments from data based on average (or bulk) tissue content. 4.

POTASSIUM TRANSPORT IN ROOTS

157

This cautionary note is supported by the occasional observation that at times Kf fluxes can be altered even though no change in [KIiwas recorded. An excellent example of this type can be found in the article by Pettersson and Jensen (1978). These workers attempted to alter the internal K+ content of barley roots (grown in full nutrient solution that contained either 0.1 or 10 mM K+) by placing the low-Kf seedlings in 10 mM KCl solution and the high-K+ seedlings in 0.1 mM KN03 solution; these pretreatments were applied for up to 80 min. Pettersson and Jensen found that K+ fluxes either decreased or increased significantly, while root Kf content did not change (Fig. 32). The authors argued that this was evidence for allosteric regulation,

Time of pretreatment in 0.1mM K t min Fig. 32. Investigation of the apparent relationship between [KIi and Rb+(8hRb+)influx into barley roots. (a) Roots grown in the presence of 0.1 mM K+ in the cultivation medium and then transferred to a 10 mM K+ pretreatment solution for the times indicated. (b) As in (a), except 10 mM K+ in cultivation medium and 0.1 mM K" in pretreatment solution. Rb+(86Rbf) uptake (0)was measured from a complete nutrient medium containing 1.0 mM Rb'. At each pretreatment time internal Kf ([KIi) was also determined (A).Data from Pettersson and Jensen (1978).

158

L. V. KOCHIAN AND W. J. LUCAS

because it indicated (to them) that the cytoplasm was filling rapidly and thus influencing allosteric sites, while bulk K+ concentration, which represented the vacuole, remained constant. Unfortunately, although these events could be occurring, there is no possible way to extrapolate from the data to the conclusions drawn by Pettersson and Jensen. Obviously, precise measurements of cytoplasmic K+ concentrations must be made before such extrapolations can be accepted. Certainly those who have championed the allosteric model are aware of this criticism, and some have attempted to deal with it via indirect measurements of cytoplasmic K+. However, direct measurements are required. Hopefully, recent technological advances in ion-selective microelectrodes may enable routine (?) measurements of cytoplasmic K+ (Felle and Bertl, 1986). Furthermore, the ability to measure K+ influx while making instantaneous changes in the K+ concentration at the cytoplasmic face of the plasmalemma, as has been done for perfused Chum cells, would be a powerful technique in the study of allosteric regulation. The use of the patch-clamp technique to study plasmalemma (and tonoplast) K+ currents, while controlling the ionic environment on each side of a membrane, in either the “patch” or “whole cell” (“whole vacuole”) mode, may be the answer to this technical problem. However, because the patch-clamp technique, of necessity, involves the use of protoplasts and isolated vacuoles, the regulatory aspects of transport involved in the integration of cells and subcellular compartments at the tissue level will be lost. Many of the studies concerning both regulatory and mechanistic aspects of K+ uptake have utilized low-salt-grown roots. As we previously indicated, low-salt roots are in a transitional stage between two regulatory states. Glass and Siddiqi (1984) have defined uptake experiments in which plants are grown under one set of conditions and then exposed to uptake media that differ from the growth solutions, as a perturbation technique. The use of low-salt roots for the generation of kinetic curves for K+ influx is a prime example of a perturbation technique. As they have noted, this technique has been useful for the elucidation of transport mechanisms, but its utility as an approach for studying transport regulation is limited, primarily because this technique is incapable of documenting long-term adaptations. Glass and Siddiqi (1984) illustrate their point with a comparison of two kinetic curves for K+ influx (over identical external K+ concentrations) for barley plants grown under different growth regimes. The fluxes measured via the perturbation technique were obtained by using intact plants that had been grown in a balanced inorganic nutrient solution containing 40 PM K’. A kinetic profile for K+ influx was determined by transferring these plants to new solutions containing from 1 to 100 PM K+ (10 min uptake). Steady-state K+ influx, on the other hand, was determined by growing barley plants under conditions of continuous K+ supply at the same concentrations used to obtain the K+ fluxes. For example, plants grown in 10 PM K+ solution were

159

POTASSIUM TRANSPORT IN ROOTS

/

Steady- state flux

~oco-o-o-o-o-o-o-o-o.

20

40

60

80

100

Potassium concentration ( p m )

Fig. 33. Comparison between “steady-state’’and “perturbation”kinetic profiles obtained for Kt influx. Steady-state experiments were conducted on barley roots that were grown at the Perturbation experiments were conducted concentrationused in the K+ influx experiment (0). on barley plants that were grown in 40 /AM K+ and then transferred to the indicated K+ concentrationsto determine the “perturbationflux” (A).Data from Glass and Siddiqi (1984).

used only for flux determinations from an uptake solution containing 10 p~ K’. As shown in Fig. 33, the two kinetic profiles are dramatically different. The curve obtained using the perturbation technique yielded a V,,, and K , of 16.8 pmol g-’ h-’ and 26 p ~respectively, , while the steady-state fluxes were considerably smaller and were independent of external concentrations above about 10 p~ K+. The differences in the two kinetic curves graphically illustrate the influence that growth conditions can have on K+ fluxes. Although their point is well taken, there are some inconsistencies in the data presented by Glass and Siddiqi (1984). At an external K+ concentration of 40 p ~one , would expect that the fluxes obtained from both techniques would be quite similar, since the seedlings used for this particular concentration were grown at that concentration. However, the data in Fig. 33 indicate that the perturbation flux at 40 p~ K+ is approximately 10 times the steady-state flux. If one examines the original paper from which these data originate, the apparent discrepancy is explained. Siddiqi and Glass (1983) obtained the data for the perturbation fluxes from 8-day-old barley seedlings, while the steady-state fluxes were obtained with seedlings grown for 18 days. Therefore, a significant portion of the reduction seen in the steadystate fluxes was due to an increase in [KIi. For example, the root K+ content of the 8-day-old seedlings grown in 40 p~ K+ solution and used for the perturbation flux determinltions was 13.7 pmol g-’, while the roots used for steady-state fluxes that were grown in 10-50 )(LMKf solutions for 18 days contained between 35 to 68 pmol K+ g-’. Although we still agree with the ideas that were proposed, a better choice of data could have been made in order to illustrate the concept.

160

L. V. KOCHIAN AND W. J. LUCAS

Another caveat concerning the use of low-salt roots for studies on ion transport regulation was made by Cheeseman (1985) in his study on the influence of Na' on K+ uptake in both low-salt corn seedlings and corn seedlings grown in full nutrient solution. In an earlier study on excised roots from 3-day-old low-salt corn roots, Cheeseman (1982) found that inclusion of Na+, over a concentration range from 0.02 to 20 mM, had no effect on K+ uptake from solutions containing 2 PM to 20 mM K'. However, when Cheeseman (1985) conducted his experiments using the roots of 8-day-old corn seedlings that had been grown on a full nutrient solution (containing 1 mM K') he found that the inclusion of Na+ in the uptake solution caused a marked stimulation of K+ influx. When the kinetics for K' uptake were determined from solutions containing from 0.05 to 2 mM KCl and Na' at 45 times the K' concentration, influx values quite similar to those for low-salt roots were obtained. On the basis of these results, Cheeseman proposed that transport regulation in full-nutrient-grown plants may be qualitatively different from that operating in low-salt roots. This idea is similar to those expressed earlier by Pitman (1970) and Siddiqi and Glass (1984) concerning transitions between different regulatory modes as a low-salt root alters its salt status. It should be noted that with respect to this Na' response, there are some differences in terms of K' uptake between the low-salt seedlings used by Cheeseman (1982) and those used by other laboratories. The lack of inhibition of K+ influx, at high external K' concentrations (>1 mM), by Na' in Cheeseman's investigation is rather surprising, considering the measurements made by Epstein et al. (1963). In this work conducted on low-salt barley roots, mechanism I1 K' uptake was inhibited by Na'. Additionally, Kochian et al. (1985) have demonstrated, for low-salt corn roots, that the inclusion of 3 mM Na+ in the uptake solution inhibits the linear component for K+ uptake by approximately 50%. The possibility exists that, in addition to the existence of qualitative differences between Cheeseman's low-salt and full-nutrient-grown roots, there may be some qualitative differences between his low-salt roots and those grown by other laboratories. It is clear that in studies involving regulation of root ion transport, the various growth conditions used to produce experimental material may have a significant influence on the results obtained and the models generated from these data. Therefore, particularly when work is conducted with the relatively non-physiological low-salt roots, special care must be taken when evaluating the data. However, this does not mean that earlier work with low-salt roots is invalid, or that they are not still a useful research material. In fact, the use of nutrient-starved roots may be necessary in order to unmask transport mechanisms that would otherwise not be detected in nutrient-rich roots. For example, the high-affinity K' uptake system recently observed in low-salt corn roots (Newman et a f . ,1987) could not be detected in high-salt-grown roots. The same is true for the high-affinity K' transport

POTASSIUM TRANSPORT IN ROOTS

161

system operating in Neurospora. Earlier K+ uptake studies by Slayman and Tatum (1965) were carried out with K+-replete cells, and were conducted in the absence of extracellular Ca2+.In these cells, K+ uptake occurred via a low-affinity,high-velocity transport system. However, if K+-starved Neurospora cells were grown (analogous to low-salt roots), and K+ uptake studied in the presence of Ca2+, the high-affinity K+ transport system (K+-H+ symporter) was revealed (Rodriguez-Navarro and Ramos, 1986;RodriguezNavarro et al., 1986). 5. Alternatives to the Allosteric Model Although most of the root work concerned with feedback control of K+ uptake by [KIi has centred on allosteric models, other models can also account for regulation of ion transport in plants. At the cellular level, control of K+ uptake by cytoplasmic K+ could also be explained on the basis of a reaction kinetic model (see Fig. 18). Here, feedback inhibition by high cytoplasmic K+ levels can be accounted for in terms of a reduction in K+ dissociation from the cytoplasmic face of the carrier; i.e. in an analogous way to that proposed for the H+-Cl- co-transport system of Chara (Sanders, 1980a,b; Sanders and Hansen, 1981). Unfortunately, due to technical limitations, it has not yet been possible to conduct the appropriate experiments, in roots, that would enable a differentiation between these two models. Recently, a regulatory model was presented which suggested that under conditions of increased shoot demand, allosteric regulation of K+ transport at the root epidermal and cortical cell level is superseded by a more complex regulatory mechanism that focuses on ion transport to the xylem as a key point for root-shoot interactions. Drew et al. (1984) and Drew and Saker (1984) investigated the effects of K+ deficiency (along with PO!- and Cl-) on K+ uptake and translocation. In their initial study, they determined the kinetic parameters for K+ uptake into barley roots by monitoring rates of depletion of labelled ions from dilute K+ solutions (Claassen and Barber, 1974). By this means, they were able to measure K+ fluxes over a full range of concentrations (as the roots depleted external K+ from the initial value of 100 PM) on the same sets of plants. With 14-day-oldplants, they found that 1 day of K+ starvation caused the K , to decline dramatically (from 53 to 11 p ~ )while , the V,,, remained unaffected. Further K+ starvation (4 or 7 days) had no influence on the K,, while the V,,, increased two-fold. Bulk K+ content of the roots and shoots declined significantly after each period of starvation (1,4 and 7 days). Drew et al. (1984) proposed that the initial rapid decrease in K , reflected allosteric regulation of the carriers by decreasing [KIi. For the starvation periods longer than 1day, they speculated that the increase in V,,, was the result of protein synthesis which increased the number of available transport proteins with a high affinity for K'.

162

L. V. KOCHIAN AND W. J. LUCAS

In the accompanying work, Drew and Saker (1984) conducted dividedroot experiments with young barley plants. These plants were grown such that one seminal root was segregated from the remaining four or five seminal roots. This single root was continuously supplied with full nutrient solution (including K'), while the remaining roots were supplied with the same solution (controls), or with the same solution minus K+ for either 2 or 4 days. Using 42K+in the compartment containing the single root, net uptake of 42K+ into the root was monitored, as well as 42K+translocation to both the shoot and the K+-starved roots. Additionally, bulk Kf concentration of the single absorbing root, the other roots, and the shoot were determined at the end of the experimental period. The results of this study are presented in Table 11. Following K+ starvation of the other roots, for either 2 or 4 days, there was a marked stimulation in K+ uptake into the single fed root, as well as dramatic stimulations of K+ translocation to the shoot and other roots (compared with controls). However, the bulk K+ concentration of the fed root remained unchanged following either 2 or 4 days of K+ starvation of the other roots. The bulk K+ concentrations of the other roots declined significantly after both 2 and 4 days of starvation, while the shoots exhibited a significant decline in K+ content only after 4 days. Drew and Saker (1984) concluded that the marked stimulation of K+ uptake and translocation to the shoot that occurred during the period of K+

TABLE I1 Potassium uptake and internal ion concentration in barley plants with roots divided between K+-free (- K ) and K+-containing ( + K ) nutrient solution

Control (single root +K, other roots + K) Net uptake of 42K-labelledK+ by single root (pmol (g dry wt root)-' day-') Single root Shoots Other roots Total translocated Tissue K+ concentration ( w o l (g dry w9-l) Single root Shoots Other roots

Potassium-deficient (single root +K, other roots -K) 2 day

'

4 day

452' 481" 53" 533"

634 1280' 188' 1470'

826' 1830' 252' 2090"

1730" 1550" 1590"

1700" 1430" 970'

1600' 1280' 500'

Values across the table followed by a different letter are significantly different from each other (P < 0.05, N = 8). Data taken from Drew and Saker (1984).

POTASSIUM TRANSPORT IN ROOTS

163

deprivation, without any detectable change in Kf concentration within the absorbing root, was incompatible with the operation of an allosterically regulated transport mechanism. They suggested that allosteric regulation had been overriden by an alternative mechanism involving transport across the root; i.e. K+ uptake into the root was regulated by transport to the xylem. The implications of such a model are that allosteric regulation will be the primary regulatory mechanism when K+ flux to the xylem is not significant. An example of such a situation would be during the lag in K+ transport to the shoot which is characteristic of low-salt roots undergoing the transition to salt saturation. Furthermore, such a model suggests the integration of regulatory mechanisms (and signals) over the entire plant, because it indicates that increased nutrient demand on a root would influence the uptake characteristics of that root. This type of transport regulation would involve a hierarchy of regulatory mechanisms ranging from control at the cellular or membrane level, which might include allosteric and/or kinetic control, to transport across the root, which might involve a feedback loop from the xylem parenchyma to the epidermis. At yet a higher level, signals from the shoot might influence root ion transport processes via the Kf transport systemsinvolved in releasing K+ into the xylem (presumably at the plasmalemma of the xylem parenchyma). Siddiqi and Glass (1987) have challenged the concept that shoot demand directly controls root K+ levels. Various combinations of experimental protocols were used in order to manipulate barley root and shoot Kf levels. Usually, the trends in changing root and shoot Kf levels were similar; i.e. root and shoot K+ increased or decreased in concert. However, some of their experimental conditions resulted in situations where alterations in the K+ content of shoots and roots did not correlate. In these cases, the kinetic parameters for Kf influx were always correlated with root K'. Based on such analyses, Siddiqi and Glass proposed that although other types of transport regulation may operate, in addition to allosteric control, negative feedback from root K+ content is the primary effector in terms of regulating Kf uptake into the root. B. Kt CYCLING WITHIN THE PLANT: AN INTEGRATION OF REGULATORY MECHANISMS

When one considers the complexities involved in nutrient cycling within the plant, particularly for a highly mobile and important nutrient such as K', it seems reasonable to hypothesize a multicomponent regulatory system for the uptake and translocation of K'. This would particularly apply to the fully autotrophic plant growing in the field, which would encounter a wide range of different environmental variables for the root versus the shoot, and would also impose the constraints of shoot and root growth upon the regulation of

164

L. V. KOCHIAN AND W. J. LUCAS

K+ uptake and partitioning. Hence, a regulatory model which simply involves the control of K+ influx across the plasmalemma of the epidermis/ cortex appears to lack the level of integration necessary to achieve a balance of K+ between the shoot and root (Drew and Saker, 1984; Jeschke et al., 1985; Cheeseman and Wickens, 1986a,b). A number of reports have shown that alterations in root K+ uptake are not negatively correlated with root K+ content (e.g. Jensen and Pettersson, 1978; Pettersson and Jensen, 1978;Wild etal., 1979; Drew and Saker, 1984; Jeschke, 1984;Pettersson, 1986;Cheeseman and Wickens, 1986a,b). Generally, non-allosteric regulation has been observed when environmental conditions have been manipulated in order to change factors such as relative growth rate (RGR) or shoot: root ratio. As Siddiqi and Glass (1987) have argued, it is possible that an alteration of factors such as environmental conditions or shoot parameters indirectly influences regulation, by changing the quantitative relationship between root [KIi and K+ influx. That is, the magnitude of the change in K+ influx in response to a particular incremental change in [KIi could be altered in response, for example, to a change in shoot growth. This explanation would still be based on the premise that only root K+ levels directly control K+ transport in the plant. Alternatively, a model that relates the needs of the shoot to the root, via signals directed to the site of K+ transfer into the xylem, with subsequent changes in transport activity then feeding back (ultimately) to K+ uptake at the root surface, would appear to be better suited to the needs of the entire plant. Such a model would still be ultimately based on feedback control of root K+ uptake by root [KIi, but in a more complex manner. Unfortunately, few studies have addressed the relationship between growth and K+ uptake specifically in terms of transport regulation. Pitman (1972) looked at the influence of shoot growth rate (modified through variations in the photoperiod) on root K+ uptake and translocation to the shoot. He found that under conditions of low shoot growth, K+ uptake and translocation were decreased in relation to the situation observed under conditions of rapid shoot growth. These results suggested that “shoot demand” for K+ could override regulatory processes operating in the root. Pitman suggested that phytohormones, such as ABA and cytokinins, could be participating in feedback control of K+ transport. Although hormones are often considered as likely candidates for the signal between the shoot and root, the available data appears contradictory. Thus, we can only conclude that hormones may play a role in the regulation of ion transport within the plant and that this area of research awaits further clarification (for review, see van Steveninck, 1976). Jeschke (1982) has examined the effect of shoot :root ratios on K+ and Na+ fluxes in intact barley seedlings. The shoot : root ratio was increased by excising all but one seminal root. This resulted in an increase in plasmalemma K+ influx, net K+ uptake, transport to the xylem, and cytoplasmic

POTASSIUM TRANSPORT IN ROOTS

165

K+. The positive correlation that was observed between K+ influx and cytoplasmic K+ levels again suggests a situation where allosteric regulation has been overriden. Based on this work, Jeschke proposed that increasing the shoot :root ratio (which would be analogous to increasing shoot demand) altered the feedback control of root ion fluxes by the shoot, possibly through the participation of hormonal signals. In a subsequent study, Jeschke (1984) studied the influence of increased transpiration on K+ fluxes in barley seedlings in which the root system had been reduced. It was found that accelerated water flow stimulated K+ uptake and translocation to the shoot, while root K+ content decreased. These observations appear to be consistent with the allosteric model. However, compartmental analysis performed under these conditions indicated that vacuolar K+ declined, while cytoplasmic K+ remained constant, or increased slightly. These results further underscore the need for measurements of cytoplasmic K'. In terms of transport regulation, Jeschke has commented on the marked similarities between the effects of increased transpiration and increased shoot :root ratio on K+ fluxes and compartmentalization within the plant. He speculated that water flow and K+ recycling within the plant could play a role in regulatory interactions between the shoot and root. This would be particularly true for seedlings where the root system has been reduced to a single root. In this situation very high rates of phloem translocation are maintained from the shoot to the root (Passioura and Ashford, 1974). The participation of recycled K+ as a signal between the shoot and root is discussed below. Cheeseman and Wickens (1986a,b) conducted a detailed study on the relationships between Kf uptake, tissue K+ content, root : shoot ratios, and growth in intact, vegetative Spergularia marina plants grown under full nutrient conditions. Potassium uptake into S. marina was negatively correlated with root weight (or root :shoot ratio), but positively correlated with root K+ content. Again, these results are incompatible with the concept that allosteric regulation of uptake functions as the prime regulatory mechanism. In all of the studies relating ion transport to growth, the focus has been placed, quite logically, on the needs of the shoot. However, there are situations where root nutrient demands may change the way a regulatory network might operate. It is well documented that nutrient deficiency often results in an increase in root growth and, therefore, a decrease in shoot :root ratio (Clarkson and Hanson, 1980). Any regulatory scheme for K+ transport in the whole plant must take into account growth factors of both the shoot and root. In terms of an integrated regulatory scheme for K+ transport throughout the plant, one logical choice for the signal between the root and shoot would be K+ recirculated from the shoot to the root, via the phloem. At least three important sites need to be considered when developing a regulatory model

166

L. V. KOCHIAN AND W. J. LUCAS

for K+ transport. These include the site of entry for K+ (the root surface), the site of transfer to the shoot (the xylem parenchyma), and the site of K+ recirculation (the shoot). The critical “fulcrum” for such a model would be the xylem parenchyma, which would be the point of signal reception from the shoot, and also the site from which root K+ uptake would be modulated, presumably through the feedback control that exists between root [KIi and plasmalemma K+ influx. As Drew and Saker (1984) have proposed, K+ recirculated to the root via the phloem could be the signal that would affect xylem K+ transport, and, ultimately, root K+ uptake. Potassium recirculation within the plant, via the phloem, has been well documented (Greenway and Pitman, 1965; Pate, 1975; Armstrong and Kirkby, 1979; Jeschke et al., 1983). Although cytological studies within the root, detailing the relationship between various cells of the phloem and the xylem parenchyma, are lacking (and much needed), the close proximity of the xylem parenchyma to the phloem does make it possible that K+ released from the phloem could influence K+ transport by these cells. Recently, elegant studies on ion recirculation (K+, Mg2+,Ca2+,and Na’) throughout

Fig. 34. Model of K+ transport and accumulation in nodulated Lupinus albus plants. Rates of ion flows (figure at arrows) and of accumulation during growth (figures in boxes) are given in pmol (g fresh wt)-’ h-’. Arrow thicknesses relate to net uptake (1) into the root of 100%. Numbers represent: accumulation in the root (2) and shoot (6); (3) “direct” accumulation from the external medium; transport via xylem (4) and phloem ( 5 ) ; phloem to xylem transfer in the root (7). Data from Jeschke et al. (1985).

POTASSIUM TRANSPORT IN ROOTS

167

the plant have been conducted by Jeschke and co-workers on both nodulated white lupin and barley (Jeschke et al., 1985,1987;Wolf and Jeschke, 1987). Initially, Jeschke and co-workers chose white lupin because of its ability to bleed from both the xylem and phloem. Xylem and phloem sap was collected, and rates of ion flow and cycling between the shoot and root were calculated from the measured carbon :ion weight ratios in the xylem and phloem streams. These determinations were based on an empirical model previously developed for carbon and nitrogen partitioning in lupin (Pate et al., 1979).The model for K+ circulation in lupin is shown in Fig. 34. The data indicate that 52% of the absorbed K+ is recirculated back to the root, and that 76% of the phloem-borne K+ returning to the root re-enters the xylem stream and is transported again to the shoot! Clearly, since a large portion of the K+ moving in the xylem transpiration stream is supplied by the phloem, the concept gains strength that K+ transfer, from the phloem, influences the ionic environment of the root xylem parenchyma. An even more detailed study of K+ and Na+ recirculation in barley has been developed to take into account the interactions between older and younger leaves (Wolf and Jeschke, 1987). The Wolf and Jeschke model is presented in Fig. 35, and from this the complexities involved in K+ partitioning within the plant are clearly evident. A number of points are brought out by this model:

1. The older leaves were the major sinks for xylem-borne K+. Subsequently, most of this imported K+ was re-exported out of the older leaves via the phloem. 2. Most of the K+ supplied to the young leaves came frnm the older leaves, via the phloem. However, the oldest leaves retranslocated their K+ primarily to the root, with a smaller proportion going to the younger leaves. 3. A large portion of the shoot K+ was recirculated to the root, and subsequently returned to the shoot via the xylem. It was estimated that 43% of the K+ absorbed by the root was recirculated back from the shoot to the root. From these models, it is apparent that the K+ flows are strongly influenced by photoassimilate partitioning, as evidenced by the large proportion of K+ that is conducted in the phloem. When the details of such complicated flow schemes are examined in the context of K+ transport regulation within the whole plant, it becomes apparent that regulatory models based on the control of a single K+ transport system (i.e. allosteric regulation of root K+ uptake) are inadequate to account for the control that would be necessary to direct these flows. Instead, it should be recognized that feedback control of K+ uptake, at the cellular level, is probably the most critical component of a multicomponent, hierarchical regulatory system that would integrate K+ fluxes at the cellular, organ, and whole plant level. It should be made clear that, at this time, such a speculative regulatory model is only a conceptual

168

L. V. KOCHIAN AND W. J. LUCAS

Fig. 35. Rates of K+ flow between single leaves and the root of barley plants. Figures represent ion movements and increments in individual organs [pmol plant-' (5 days)-'] for the period 14-19 days after germination (four-leaf stage). Arrow thicknesses are drawn proportional to transport rates, with striped arrow indicating flows in the xylem and solid arrows flows in the phloem. Squares in leaves or roots represent ion increments due to growth andor to increases in tissue ion concentrations; areas are drawn in proportion to ion increments. The indicated rate of phloem to xylem transfer within the root represents a minimum estimate. Data from Wolf and Jeschke (1987).

framework from which future work can be directed. Certainly almost nothing is known, for example, about how alterations in the activity of K+ transport systems in the plasmalemma of the xylem parenchyma might feed back upon K+ uptake systems in the root epidermal plasmalemma. Obviously, as Cheeseman and Wickens (1986a,b) noted earlier, future research must be directed at the elucidation of the physiological basis for the signals and transduction systems operating within the plant, and also the mechanisms by which the plant integrates these components into an interactive and coherent system.

POTASSIUM TRANSPORT IN ROOTS

169

VI. FUTURE RESEARCH AND PROSPECTS We began this chapter by noting the lack of understanding and knowledge concerning the mechanisms and regulatory aspects of K+ transport in roots. It can certainly be said that progress in this area has not been rapid, primarily because of the complexities of the processes being studied. However, our examination of the literature indicates that over the past 10 years significant insight has been gained into both the mechanisms of K+ transport across the root and the control processes involved. Progress has come primarily through the application of a range of recent technological advances. In order to maintain this progress, we must not only continue to take advantage of new technologies as they become available, but we must also apply them in an integrated manner. In terms of mechanisms of plasmalemma and tonoplast K+ transport, a combination of biophysical, biochemical and molecular techniques should prove extremely fruitful. The application of patch-clamp studies (both whole cell and vacuole, and excised patch) will enable the investigation of the coupling and stoichiometry of K+ and Hf fluxes. In this way, it should be possible to study the putative Kf-Hf co-transport system in the most rigorous of ways. Additionally, we are pleased to note that the study of Kf channels with the patch-clamp technique has already begun and should yield important information on mechanisms and control of passive K+ transport systems. Recent improvements in cell fractionation and purification techniques should also enable researchers to obtain purified preparations of transport proteins (at least the more ubiquitous ones) which can be used for further reconstitution and transport studies. This approach should complement patch-clamp and traditional electrophysiological techniques in discriminating between Kf-ATPases, K+-H+ co-transport systems, and KfH+ exchange ATPases. Furthermore, the combination of membrane purification techniques with immunology and molecular genetics is enabling researchers to obtain the gene sequences for some transport proteins. Further progress in this exciting field will allow for the molecular characterization of various transport proteins. By integrating these new techniques into existing programmes, plant transport physiologists should be able to study the different Kf transport systems at the mechanistic/molecular level. The future prospects for research into Kf transport regulation, at the cellular, root, and whole plant level, are very exciting. A t the cellular level, the patch-clamp technique should again be very useful. This approach will enable researchers to study K+ transport while carefully controlling the ionic environment on both sides of the plasmalemma (or tonoplast). Hence studies into kinetic and allosteric control can be conducted. Additionally, this technique can be used to study the interactions between membraneassociated redox systems, K+ transport, and the control of growth. Recent

170

L. V. KOCHIAN AND W. J. LUCAS

advances in ion-selective microelectrode technology should enable researchers to use K+ and Hf-selective microelectrodes to obtain direct measurements of cytoplasmic and vacuolar pH and K+ activities, and to monitor changes in these activities as the root absorbs and translocates K+. Finally, the application of high-resolution SEM X-ray microprobe analysis should prove to be a powerful tool in the investigation of K+ compartmentalization within the various cells of the root, especially during alterations in the status of this ion. The integration of K+ transport across the root with root-shoot interactions is an important area of research that is gaining the attention of a number of plant physiologists. Studies must be focused on the roles of the xylem parenchyma in K+ transport to the shoot and Kf recirculation within the plant. Furthermore, possible modes of coupling between the xylem parenchyma and root epidermis, in terms of regulating K+ transport across the root, should be investigated. This will require the integration of cell cytology, biophysics, biochemistry and whole plant physiology. In view of the models indicating that Kf recirculation from the shoot to the root may play an important role in regulating K+ transport in the plant, it is obvious that an extensive cytological study is needed to establish whether symplasmic continuity exists between the phloem and xylem tissues of the stele. Additionally, the use of detopped roots for the study of both transroot electrical potentials and ion transport to the xylem, using combinations of various nutritional regimes and chemical probes directed at the xylem parenchyma, are needed to further characterize K+ transport across the root. Finally, further whole plant studies quantifying Kf flows within the plant, particularly under conditions where Kf concentrations and flow in the phloem are artificially modified, should prove of particular value. In conclusion, it is apparent that as we gain further understanding and insights concerning K+ transport within the plant, we also gain a growing appreciation of the complexities involved in this highly regulated physiological system. In the future, it will often be necessary to dissect the system into its cellular, subcellular and molecular components, even down to the gene level. However, we should take every opportunity to integrate these new findings to yield a highly cohesive approach to the study of K+ transport in the plant. We feel that only in this way will we be able to build upon the knowledge that has been yielded up so grudgingly by our beloved plant!

ACKNOWLEDGEMENTS Work on Kf transport conducted in our laboratories has been supported by grants from the United States National Science Foundation (W.J.L.) and the United States Department of Agricultural Research Service (L.V.K.). We also wish to thank Lyn Noah and Cheryl Redlich for their help in typing

POTASSIUM TRANSPORT IN ROOTS

171

t h e manuscript a n d t h e many scientists who provided us with reprintdpreprints of their work. W e offer special thanks to Drs Glass, Drew and Jeschke for valuable discussion. Finally, we would like t o thank D i a n a Lucas for her patience and encouragement throughout t h e preparation of this manuscript.

REFERENCES Akedo, H. and Christensen, H. N . (1962). J. Biol. Chem. 237,118-122. Anderson, W. P. (1976). In “Encyclopedia of Plant Physiology” (U. Luttge and M. G. Pitman, eds), New Series, Vol. 2B, pp. 129-156. Springer-Verlag, Berlin. Anderson, W. P. and Higinbotham, N. (1975). J. Exp. Bot. 26,533-535. Anderson, W. P. and Reilly, E . J. (1968). J. Exp. Bot. 19,19-30. Ansari, A. Q. and Bowling, D. J. F. (1972). J . Exp. Bot. 23,641-650. Armstrong, C . M. (1969). J. Gen. Physiol. 54,553-575. Armstrong, M. J. and Kirkby, E. A. (1979). PBemant Physiol. 63,1143-1148. Asher, C. J. and Ozanne, P. G . (1967). SoilSci. 103,155-161. Bange, G. G. J. (1973). Acta Bot. Neerl. 22,529-542. Bange, G. G . J. (1977). Acta Bot. Neerl. 2 6 , 5 3 4 2 . Bange, G. G. J. (1979). Physiol. Plant. 46,179-183. Beffagna, N . , Cocucci, S. and Marrb, E. (1977). Plant Sci. Lett. 8,91-98. Behrens, H. M. and Gradmann, D. (1985). Planta 163,453462. Bellando, M. and Trotta, A. (1980). Plant Sci. Lett. 17,467472. Bellando, M., Trotta, A., Bonetti, A., Colombo, R., Lado, P. and Mar&, E. (1979). Plant Cell Environ. 2 , 3 9 4 7 . Bennett, A. B. and Spanswick, R. M. (1983). J. Membr. Biol. 71,95107. Bentrup, F. W., Hoffmann, B., Gogarten-Boekels, M. and Gogarten, J. P. (1985). 2. Naturforsch 40,886-890. Berridge, M. J. (1983). Biochem. J . 212,849-858. Berridge, M. J. and Irvine, R. F. (1984). Nature312,315-321. Bienfait, H. F. (1985). J. Bioenerg. Biomembr. 17,73-83. Blackman, M. S. and McDaniel, C. N. (1978). Plant Sci. Lett. 13,27-34. Blatt, M. R. (1987). “Abstracts of XIV International Botanical Congress”, 1-16-9, Berlin, West Germany. Blatt, M. R. and Slayman, C. L. (1987). Proc. NatlAcad. Sci. USA 84,2737-2741. Blumwald, E. (1987). Physiol. Plant. 69,731-734. Blumwald, E . and Poole, R. J. (1985). Plant Physiol. 78,163-167. Borstlap, A. C. (1981a). Plant Cell Environ. 4,189-195. Borstlap, A. C. (1981b). Naturwiss. 6 8 , 4 1 4 3 . Borstlap, A. C. (1983). Plant Cell Environ. 6,407-416. Boss, W. F. and Massel, M. (1985). Biochenr. Biophys. Res. Commun. 132, 1018-1023. Bottger, M. and Luthen, H. (1986). J. Exp. Bot. 37,666-675. Bowling, D. J. F. (1972). Planta 108,147-151. Bowling, D. J. F. (1981). Physiol. Plant. 53,392-397. Bowling, D. J. F. and Ansari, A. Q. (1972). J. Exp. Bot. 23,241-246. Briskin, D. P. (1986a). Physiol. Plant. 68, 159-163. Briskin, D. P. (1986b). Arch. Biochem. Biophys. 248,106-115. Briskin, D. P. andLeonard, R. T. (1982a). Proc. NatlAcad. Sci. USA 79,6922-6926. Briskin, D. P. and Leonard, R. T. (1982b). Plant Physiol. 70,1459-1464. Briskin, D. P. and Poole, R. J. (1983a). Plant Physiol. 71,507-512.

172

L. V. KOCHIAN AND W. J. LUCAS

Briskin, D. P. and Poole, R. J. (1983b). Plant Physiol. 71,969-971. Briskin, D. P. and Thornley, W. R. (1985). Phytochemistry 24,2797-2802. Brooks, S. C. (1929). Protoplasma 8,389-412. Buckhout, T. J . and Hrubec, T. C. (1986). Protoplasma 135,144-154. Butz, R. G . and Long, R. C. (1979). Plant Physiol. 64,684-689. Cabantchik, Z. I., Knauf, P. A . and Rothstein, A . (1978). Biochim. Biophys. Acta 515,239-302. Cantley, L. C. (1981). Curr. Top. Bioenerg. 11,201-237. Chastain, C. J. and Hanson, J. B. (1982). Plant Sci. Lett. 24,97-104. Cheeseman, J. M. (1982). J. Membr. Biol. 70,157-164. Cheeseman, J. M. (1985). Physiol. Plant. 64,243-246. Cheeseman, J. M. and Hanson, J. B. (1979a). Plant Physiol. 63,1-4. Cheeseman, J. M. and Hanson, J. B. (1979b). Plant Physiol. 64,842-845. Cheeseman, J . M. andHanson, J. B. (1980). PlantSci. Lett. 18,81-84. Cheeseman, J. M. and Wickens, L. K. (1986a). Physiol. Plant. 67,7-14. Cheeseman, J. M. and Wickens, L. K. (1986b). Physiol. Plant. 67,15-22. Cheeseman, J. M., LaFayette, P. R., Gronewald, J. W. and Hanson, J. B. (1980). Plant Physiol. 65,1139-1145. Chino, M. (1981). Soil Sci. PlantNutr. 27,487-503. Christensen, H. N. and Liang, M. (1966). Biochim. Biophys. Acta 112,524531. Churchill, K. A. and Sze, H. (1983). Plant Physiol. 71,610-617. Claassen, N. and Barber, S. A . (1974). Plant Physiol. 54,564-568. Clarkson, D. T. and Hanson, J. B. (1980). Ann. Rev. Plant Physiol. 31,239-298. Clarkson, D. T. and Hanson, J . B. (1986). J . Exp. Bot. 37,1136-1150. Clarkson, D. T. and Robards, A . W. (1975). In “The Development and Function of Roots” (J. G. Torrey and D. T. Clarkson, eds), pp. 425-436. Academic Press, London. Clarkson, D. T., Williams, L. and Hanson, J . B. (1984). Planta 162,361-369. Cocucci, M., Marrk, E., Ballarin-Denti, A. and Scacchi, A . (1976). Plant Sci. Lett. 6,143-156. Cohen, S. R. (1975). J. Membr. Biol. 22,53-72. Cohen, S. R. (1980). J. Membr. Biol. 52,95-105. Crafts, A. S. and Broyer, T. C. (1938). A m . J. Bot. 25,529-535. Craig, T. A. and Crane, F. L. (1980). Proc. Indiana Acad. Sci. 90, 150-155. Craig, T. A . and Crane, F. L. (1981). Proc. Indiana Acad. Sci. 91, 150-154. Cram, W. J. (1976). In “Encyclopedia of Plant Physiology” (U. Liittge and M. G. Pitman, eds), New Series, Vol. 2A, pp. 284-316. Springer-Verlag, Berlin. Cram, W. J. (1980). Aust. J . Plant Physiol. 7 , 4 1 4 9 . Davidson,V. L. andKnaff, D. B. (1981). Photobiochem. Photobiophys. 3,167-174. Davidson, V. L. and Knaff, D. B. (1982). Arch. Biochem. Biophys. 213,358-362. Davies, D. D. (1973). Symp. SOC.Exp. Biol. 27,513-529. Davies, D. D. (1979). Ann. Rev. Plant Physiol. 30, 131-158. Davis, R. F. and Higinbotham, b’. (1369). Plant Physiol. 44, 1383-1392. Davis, R. F. and Higinbotha , N. (1976). Plant Physiol. 57,129-136. Deacon, J. W., Drew, M. C. and Darling, A . (1986). Ann. Bot. 58,719-727. Debnam, E. S. andLevin, R. J. (1975). J . Physiol. 246,181-196. De Boer, A. H. and Prins, H. B. A . (1984). Plant Cell Physiol. 25,643-655. De Boer, A. H. and Prins, H. B. A . (1985). Plant Cell Environ. 8,579-586. De Boer, A. H., Prins, H. B. A . and Zanstra, P. E. (1983). Planta 157,259-266. Delrot, S. and Bonnemain, J. L. (1981). Plant Physiol. 67,560-564. De Luca, L., Bader, U., Hertel, R. and Pupillo, P. (1984). Plant Sci. Lett. 36,93-98. De Quintero, M. R. and Hanson, J. B. (1982). Plant Physiol. 76,403-408. Dohrmann, U., Hertel, R., Pesci, P . , Cocucci, S. M., Marre, E., Randazzo, G. and Ballio, A. (1977). Plant Sci. Lett. 9,291-299.

POTASSIUM TRANSPORT IN ROOTS

173

Drew, M. C. and Saker, L. R. (1984). Planta 160,500-507. Drew, M. C. andSaker, L. R. (1986). 1. Exp. Bot. 37.22-33. Drew, M. C., Chamel, A., Garrec, J.-P. and Fourcy, A. (1980). Plant Physiol. 65, 506-511. Drew, M. C., Saker, L. R., Barber, S. A. and Jenkins, W. (1984). Planta 160, 49w99. Dunlop, J. (1973). J . Exp. Bot. 24,995-1002. Dunlop, J. (1982). J . Exp. Bot. 33,910-918. Dunlop, J. andBowling, D. J. F. (1971). J . Exp. Bot. 22,445-452. DuPont, F. M., Giorgi, D. L. and Spanswick, R. M. (1982). Plant Physiol. 70, 1694-1699. Echlin, P., Lai, C. and Hayes, T. L. (1981). Scan. Electron Microsc. 2,489-498. Elzam, 0 . E., Rains, D. W. and Epstein, E . (1964). Biochem. Biophys. Res. Commun. 15,273-276. Epstein, E. (1976). In “Encyclopedia of Plant Physiology” (U. Luttge and M. G. Pitman, eds), New Series, Vol. 2B, pp. 70-94. Springer-Verlag, Berlin. Epstein, E. and Hagen, C. E. (1952). Plant Physiol. 27,457474, Epstein, E. and Rains, D. W. (1965). Proc. NatlAcad. Sci. USA 53,1320-1324. Epstein, E., Rains, D. W. and Elzam, 0. E. (1963). Proc. Natl Acad. Sci. USA 49, 684-692. Epstein, W. and Laimins, L. (1980). TIBS 5,21-23. Erwee, M. G. and Goodwin, P. B. (1983). Planta 158,320-328. Erwee, M. G. and Goodwin, P. B. (1984). Protoplasma 122, 162-168. Erwee, M. G., Goodwin, P. B. and Van Bel, A. J. E. (1985). Plant Cell Environ. 8, 173-178. Faller, L., Jackson, R., Malinowska, D., Mukidjam, E., Rabon, E., Saccomani, G., Sachs, G. and Smolka, A. (1982). Ann N Y A c a d . Sci. 402,146-163. Felle, H. and Bertl, A. (1986). J. Exp. Bot. 37,1416-1428. Fernando, M., Siddiqi, M. Y. and Glass, A. D. M. (1987). Plant Physiol. 83, S-164. Fischer, E . and Luttge, U. (1980). Plant Physiol. 65, 1004-1008. Fisher, J. and Hodges, T. K. (1969). Plant Physiol. 44,385-395. Fisher, J., Hansen, D. and Hodges, T. K. (1970). Plant Physiol. 46,812-814. Giannini, J. L., Gildensoph, L. H., Ruiz-Cristin, J . andBriskin, D. P. (1987). Plant Physiol. 83, S-55. Glass, A. D. M. (1975). Plant Physiol. 56,377-380. Glass, A. D. M. (1976). Plant Physiol. 58,33-37. Glass, A. D. M. (1977). Aust. J . Plant Physiol. 4,313-318. Glass, A. D. M. (1978). Can. J. Bot. 56,1759-1764. Glass, A. D. M. (1983). Ann. Rev. Plant Physiol. 34,311-326. Glass, A. D. M. and Dunlop, J. (1978). Planta 141,117-119. Glass, A. D. M. and Siddiqi, M. Y. (1982). Plant Cell Environ. 5,385-393. Glass, A. D. M. and Siddiqi, M. Y. (1984). In “Advances in Plant Nutrition” (P. B. Tinker and A . Lauchli, eds), Vol. 1, pp. 103-147. Praeger Publishers, New York. Goldman, D. E. (1943). J . Gen. Physiol. 27,37-60. Greenway, H. and Pitman, M. G . (1965). Aust. J. Biol. Sci. 18,235-247. Gronewald, J. W. and Hanson, J. B. (1980). Plant Sci. Lett. 18,143-150. Gronewald, J. W. and Hanson, J. B. (1982). Plant Physiol. 69,1252-1256. Hanson, J. B. (1977). In “Water Relations in Membrane Transport in Plants and Animals” (A. M. Jungreis, T. K. Hodges, A. Kleinzeller and S. G. Schultz, eds), pp. 277-290. Academic Press, New York. Hanson , J . B . (1978). Plant Physiol. 62,402-405. Harvey, D. M. R. (1985). Planta 165,242-248. Hiatt, A. J. (1967a). Plant Physiol. 42,294-298.

174

L. V. KOCHIAN AND W. J . LUCAS

Hiatt, A. J. (1967b). Z . Pflanzenphysiol. 56,223-245. Hiatt, A. J. and Hendricks, S. B. (1967). Z . Pflanzenphysiol. 56,220-232. Hoagland, D. R. and Broyer, T. C. (1936). Plant Physiol. 11,471-507. Hoagland, D. R. andBroyer, T. C. (1940). A m . J. Bot. 27,173-185. Hodges, T. K. (1976). In “Encyclopedia of Plant Physiology” (U. Luttge and M. G. Pitman, eds), New Series, Vol. 2A, pp. 260-283. Springer-Verlag, Berlin. Hodges, T. K., Leonard, R. T., Bracker, C. E . and Keenan, T. W. (1972). Proc. Natl Acad. Sci. USA 69,3307-3311. Hodgkin, A. C. and Huxley, A. F. (1952). J . Physiol. 117,500-544. Hodgkin, A. C. and Katz, B. (1949). J. Physiol. 108,37-77. Hooymans, J. J. M. (1976). Z . Pflanzenphysiol. 79,182-186. Hylmo, B . (1953). Physiol. Plant. 6,333-405. Iaccarino, M., Guardiola, J. and DeFelice, M. (1978). J . Membr. Sci. 3,287-302. Jackson, P. C. and Adams, H. R. (1963). J . Gen. Physiol. 46,369-386. Jackson, P. C. and Edwards, D. G. (1966). J . Gen. Physiol. 50,225-241. Jacobsen, L. and Overstreet, R. (1947). A m . J . Bot. 34,415420. Jacobsen, L., Overstreet, R., King, H. and Handley, R. (1950). Plant Physiol. 25, 639-647. Jacobson, L. (1955). Plant Physiol. 30,264-269. Jacobson, L. and Ordin, L. (1954). Plant Physiol. 29,70-75. Jacoby, B. and Laties, G. G. (1971). Plant Physiol. 47,525-531. Jacoby, B. and Rudich, B. (1980). In “Plant Membrane Transport: Current Conceptual Issues’’ (R. M. Spanswick, W. J. Lucas and J. Dainty, eds), pp. 497498. ElsevierR’Jorth-Holland Biomedical Press, Amsterdam. Jensen, P. and Pettersson, S. (1978). Physiol. Plant. 42,207-213. Jeschke, W. D. (1980). I n “Plant Membrane Transport: Current Conceptual Issues” (R. M. Spanswick, W. J. Lucas and J. Dainty, eds), pp. 17-31. ElsevierR’JorthHolland Biomedical Press, Amsterdam. Jeschke, W. D. (1982). J . Exp. Bot. 33,601-618. Jeschke, W. D. (1984). J . Plant Physiol. 117,267-285. Jeschke, W. D., Stelter, W., Reising, B. and Behl, R. (1983). J . Exp. Bot. 34, 964-979. Jeschke, W. D., Atkins, C. A. and Pate, J. S . (1985).J. PlantPhysiof. 117,319-330. Jeschke, W. D., Pate, J. S . and Atkins, C. A. (1987). J . Plant Physiof. 128,77-93. Johansen, C., Edwards, D. G. and Lonergan, J. F. (1970). Plant Physiol. 45, 601-603. Juniper, B. E. and Barlow, P. W. (1969). Planta 89,352-360. Kannan, S. (1971). Planta 96,262-270. Keifer, D. W. and Lucas, W. J. (1982). Plant Physiof. 69,781-788. Keifer, D. W . , Franceschi, V. R. and Lucas, W. J. (1982). Plant Physiol. 70, 1327-1334. Ketchum, K. A., Cooper, E., Shrier, A. and Poole, R. J. (1987). Plant Physiol. 83, s-112. Kochian, L. V. and Lucas, W. J. (1982a). Plant Physiol. 70,1723-1731. Kochian, L. V. and Lucas, W. J. (1982b). What’s New in Plant Physiol. 1 3 , 4 5 4 8 . Kochian, L. V. and Lucas, W. J. (1983). Plant Physiol. 73,208-215. Kochian, L. V. and Lucas, W. J. (1984). In “Membrane Transport in Plants” (W. J. Cram, R. Rybova and K. Sigler, eds), pp. 410-411. Academia, Prague. Kochian, L. V. and Lucas, W. J. (1985). Plant Physiol. 77,429-436. Kochian, L. V. and Lucas, W. J. (1987). In “Redox Functions of the Eukaryotic Plasmalemma” (J. Ramirez, ed.), pp. 155-173. CSIC, Madrid. Kochian, L. V., Xin-Zhi, J. and Lucas, W. J. (1985). Plant Physiol. 79,771-776.

POTASSIUM TRANSPORT IN ROOTS

175

Kochian, L. V., Shaff, J. E. and Lucas, W. J. (1987). Plant Physiol. 83, S-111. Kochian, L. V., Newman, I. A. and Lucas, W. J. (1988). In “Proceedings of 7th International Workshop on Plant Membrane Transport”, Australian Academy of Science, Canberra. (In press.) Kolb, H.-A., Kohler, K. and Martinoia, E. (1987). J . Membr. Biol. 95, 163-169. Komor, E. (1977). Planta 137,119-131. Komor, E. and Tanner, W. (1975). Planta 123,195-198. Komor, E., Rotter, M. and Tanner, W. (1977). Plant Sci. Lett. 9,153-162. Komor, E., Thom, M. and Maretzki, A. (1987). Planta 170,34-43. Lado, P., DeMichelis, M. I., Cerana, R. and Mar&, E. (1976). Plant Sci. Lett. 6, 5-20. Laties, G. G. (1969). Ann. Rev. Plant Physiol. 20,89-116. Lauchli, A. (1975). J. Microscopie Biol. Cell. 22,4331140. Lauchli, A. (1976). In “Transport and Transfer Processes in Plants” (I. F. Wardlaw and J. B. Passioura, eds), pp. 101-112. Academic Press, New York. Lauchli, A. and Pfluger, R. (1979). In “Potassium Research-Review and Trends”, pp. 11 1-163. International Potash Institute, Worblaufen, Bern/Switzerland. Lauchli, A., Spurr, A. R. and Epstein, E . (1971). Plant Physiol. 48,118-124. Lauchli, A., Kramer, D., Sluiter, E. and Gullasch, J. (1977). Colloque du CNRS Echanges Ioniques Transmembranaires chez les Vegetaux 258,469-476. Lauchli, A., Pitman, M. G., Luttge, U., Kramer, D. and Ball, E. (1978). Plant Cell Environ. 1,217-223. Leggett, J. E. and Epstein, E. (1956). Plant Physiol. 31,222-226. Leigh, R. A. and Wyn Jones, R. G. (1984). New Phytol. 97,l-13. Leonard, R. T. (1982). In “Membranes and Transport” (A. N. Martonosi, ed.), Vol. 2, pp. 633-637. Plenum Press, New York. Leonard, R. T. (1984). In “Advances in Plant Nutrition” (P. B. Tinker and A . Lauchli (eds), Vol. 1, pp. 209-240. Praeger Scientific, New York. Leonard, R. T. (1985). In “Potassium in Agriculture” (R. 0. Munson, ed.), pp. 327-335. Am. SOC.Agron., Madison, WI. Leonard, R. T. and Hanson, J. B. (1972). Plant Physiol. 49,430-435. Leonard, R. T. and Hodges, T. K. (1973). Plant Physiol. 52,6-12. Leonard, R. T. and Hotchkiss, C. W. (1976). Plant Physiol. 58,331-335. Lichtner, F. T. and Spanswick, R. M. (1981). Plant Physiol. 68,693-698. Lin, W. (1979). Plant Physiol. 63,952-955. Lin, W. (1980). Plant Physiol. 66,550-554. Lin, W. (1981). Plant Physiol. 68,435438. Lin, W. (1982a). Proc. NatlAcad. Sci. USA 79,3773-3776. Lin, W. (1982b). Plant Physiol. 70,326-328. Lin, W. (1984). Plant Physiol. 74,219-222. Lin, W. (1985). Physiol. Plant. 65, 102-108. Lin, W. and Hanson, J. B . (1976). Plant Physiol. 58,276-282. Lopez-Saez, J. F., Giminez-Martin, G. and Risueno, M. C. (1966). Protoplasma61, 81-84. Lowendorf, H. S., Slayman, C. L. and Slayman, C. W. (1974). Biochim. Biophys. Acta 373,369-382. Lucas, W. J. and Kochian, L. V. (1987). In “Redox Functions of the Eukaryotic Plasmalemma” (J. M. Ramirez, ed.), pp. 176-193. CSIC, Madrid. Luster, D. G., Hrubec, T. C. and Buckhout, T. J. (1987). Plant Physiol. 83, S-54. Luttge, U. and Clarkson, D. T. (1985). Prog. Bot. 47,73-86. Macri, F. and Vianello, A. (1986). Plant Sci. 43,25-30. Madore, M. A., Oross, J. W. and Lucas, W. J. (1986). Plant Physiol. 82,432442. Maloney, P. C. and Wilson, T. H. (1973). Biochim. Biophys. Acta 330,196-205.

176

L. V. KOCHIAN AND W. J . LUCAS

Marre, E. (1977). In “Regulation of Cell Membrane Activities in Plants” (E. Marre and 0. Ciferri, eds), pp. 185-202. Elsevier/North Holland, Amsterdam. Marre, E . (1979). Ann. Rev. Plant Physiol. 30,273-288. Marrb, E. and Ballarin-Denti, A. (1985). J. Bioenerg. Biomembr. 17, 1-21. Mar&, E., Colombo, R., Lado, P. and Rasi-Caldogno, F. (1974a). Plant Sci. Lett. 2, 139-150. Marrb, E . , Lado, P., Ferroni, A. and Ballarin-Denti, A. B. (1974b). Plant Sci. Lett. 2,257-265. Maynard, J. M. and Lucas, W. J. (1982a). Plant Physiol. 69,734-739. Maynard, J. M. and Lucas, W. J. (1982b). Plant Physiol. 70,1436-1443. McCully, M. E., Canny, M. J. and Van Steveninck, R. F. M. (1987). Physiol. Plant. 69,7340. Memon, A. R., Saccomani, M. and Glass, A. D. M. (1985). J. Exp. Bot. 36, 1860-1876. Mettler, I. J., Mandala, S. andTaiz, L. (1982). Plant Physiol. 70, 1738-1742. Michaelis, G., Ehwald, R. and Goring, H. (1976). Biochem. Physiol. Pflanzen 169, 241-247. Misra, P. C., Craig, T. A. and Crane, F. L. (1984). J . Bioenerg. Biomembr. 16, 143-152. Mitchell, P. (1970). Symp. SOC. Gen. Microbiol. 20, 121-166. Mdler, I. M. and Lin, W. (1986). Ann. Rev. Plant Physiol. 37,309-334. Moran, N., Iwasa, K., Ehrenstein, G., Mishke, C., Bare, C. and Satter, R. (1987). Plant Physiol. 83, S-112. Morse, M. J., Crain, R. C. and Satter, R. L. (1986). Plant Physiol. 80, S-92. Munck, B. G. and Schultz, S. G . (1969). J. Gen. Physiol. 53,157-182. Newman, I. A., Kochian, L. V., Grusak, M. A. and Lucas, W. J. (1987). Plant Physiol. 84,1177-1184. Nishizuka, Y. (1984). Nature 308,693-698. Nissen, P. (1971). Physiol. Plant. 24,315-324. Nissen, P. (1973). Physiol. Plant. 28,113-120. Nissen, P. (1974). Ann. Rev. Plant Physiol. 25,53-79. Nissen, P. (1987). Plant Cell Environ. 10,475-485. Nissen, P. and Nissen, 8.(1983). Physiol. Plant. 57,47-56. Okamoto, H., Ichino, K. and Katou, K. (1978). Plant Cell Environ. 1,279-284. Okamoto, H., Katou, K. and Ichino, K. (1979). Plant Cell Physiol. 20, 103-114. O’Neill, S. D. and Spanswick, R. M. (1984). J. Membr. Biol. 79,231-243. Osmond, C. B. (1976). In “Encyclopedia of Plant Physiology” (U. Liittge and M. G. Pitman, eds), New Series, Vol. 2A, pp. 347-372. Springer-Verlag, Berlin. Osmond, C. B. and Greenway, H. (1972). Plant Physiol. 49,260-263. Osterhout, W. J. V. (1931). Biol. Rev. 6,369-411. Osterhout, W. J. V. (1950). Biol. Bull. 90,308-309. Osterhout, W. J. V. (1952). J. Gen. Physiol. 35,579-594. Overall, R. L. and Gunning, B. E. S. (1982). Protoplasma 111,151-160. Overall, R. L., Wolfe, J. and Gunning, B. E. S. (1982). Protoplasma 111,134-150. Passioura, J. B. and Ashford, A. E. (1974). Aust. J . Plant Physiol. 1,521-527. Pate, J. S. (1975). In “Encyclopedia of Plant Physiology” (M. H. Zimmermann and J. A. Milburn, eds), New Series, Vol. 1, pp. 451-473. Springer-Verlag, Berlin. Pate, J. S . , Layzell, D. B. and McNeil, D. L. (1979). Plant Physiol. 63,730-738. Pesci, P., Tognoli, L., Beffagna, N. and Marre, E. (1979). Plant Sci. Lett. 15, 313-322. Petraglia, T. and Poole, R. J. (1980). Plant Physiol. 65,967-972. Pettersson, S. (1986). Physiol. Plant. 66,122-128. Pettersson, S. and Jensen, P. (1978). Physiol. Plant. 44,110-114.

POTASSIUM TRANSPORT IN ROOTS

177

Pettersson, S. and Jensen, P. (1979).Physiol. Plant. 45,83-87. Pitman, M. G. (1967). Nature 216,1343-1344. Pitman, M. G. (1970).Plant Physiol. 45,787-790. Pitman, M. G. (1972).Aust. J. Biol. Sci. 25,243-257. Pitman, M. G. (1976).In “Encyclopedia of Plant Physiology” (U. Luttge and M. G. Pitman, eds), New Series, Vol. 2B, pp. 95-128. Springer-Verlag, New York. Pitman, M. G. (1977).Ann. Rev. Plant Physiol. 28,71-88. Pitman, M. G. (1982). Q. Rev. Biophys. 15,481-554. Pitman, M. G., Courtice, A. C. and Lee, B. (1968).Aust. J. Biol. Sci. 21,871-881. Pitman, M. G., Schaefer, N. and Wildes, R. A. (1975).Planta 126,61-73. Pitman, M. G., Lluchli, A. and Stelzer, R. (1981). Plant Physiol. 68,673-679. Polley, L. D. and Hopkins, J. W. (1979). Plant Physiol. 64,374378. Poole, R. J. (1966).J. Gen. Physiol. 49,551-563. Poole, R. J. (1974). Can. J. Bot. 52,1023-1028. Poole, R. J. (1978).Ann. Rev. Plant Physiol. 29,437460. Pupillo, P., Valenti, V., De Luca, L. and Hertel, R. (1986). Plant Physiol. 80, 384-389. Raven, J. A. and Smith, F. A. (1974). Can. J. Bot. 52,135-148, Reisenauer, H. M . (1966). In “Environmental Biology” (P. L. Altman and D. S. Dittmer, eds), pp. 507-508. Federation of American Societies for Experimental Biology, Bethesda, MD. Rhoads, D. B., Waters, F. B.and Epstein, W .(1976).J. Gen. Physiol. 67,325-341. Rincon, M . and Boss, W. F. (1987).Plant Physiol. 83,395-398. Robards, A. W. (1971).Protoplasma 72,315-323. Robards, A. W. (1975).Ann. Rev. Plant Physiol. 26,13-29. Rodriguez-Navarro, A. and Ramos, J. (1986).Biochim. Biophys. Acta 857,229-237. Rodriguez-Navarro, A., Blatt, M. R. and Slayman, C. L. (1986).J. Gen. Physiol. 87,649-674. Rubinstein, B. and Stern, A. I. (1986).Plant Physiol. 80,805-811. Rubinstein, B., Stern, A. I. and Stout, R. G. (1984).PlantPhysiol. 76,386-391. Sandelius, A. and Sommarin, M . (1986).FEBS Lett. 201,282-286. Sanders, D. (1980a).J. Membr. Biol. 52,5140. Sanders, D. (1980b).J. Membr. Biol. 53,129-141. Sanders, D. (1986).J. Membr. Biol. 90,67-87. Sanders, D. and Hansen, U.-P. (1981).J. Membr. Biol. 58,139-153. Sarafian, V. and Poole, R. J . (1987). Plant Physiol. 83, S-55. Scarborough, G. A. (1970).J. Biol. Chem. 245,3985-3987. Schaefer, N., Wildes, R. A. and Pitman, M. G. (1975). Aust. J. Plant Physiol. 2, 61-73. Schnabl, H. (1987). “Abstracts of XIV International Botanical Congress,” 1-16-1, Berlin, West Germany. Schroeder, J. I., Hedrich, R. and Fernandez, J. M . (1984). Nature312,361-362. Schroeder, J . I., Raschke, K. and Neher, E. (1987). Proc. NatlAcad. Sci. USA 84, 4108-4112. Schumaker, K. S. and Sze, H. (1985). Plant Physiol. 79,1111-1117. Segel, I. H. (1975). “Enzyme Kinetics.” John Wiley and Sons, New York. Serrano, R. (1984). Curr. Top. Cell Regul. 23,87-126. Shone, M . G. T. (1968).J . Exp. Bot. 19,468-485. Shone, M. G. T. (1969).J . Exp. Bot. 20,698-716. Siddiqi, M. Y . and Glass, A. D. M . (1983). Can. J. Bot. 61,1551-1558. Siddiqi, M . Y. and Glass, A. D. M . (1984). Plant Sci. Lett. 33, 103-114. Siddiqi, M. Y. and Glass, A. D. M . (1987).J. Exp. Bot. 38,935-947. Slayman, C. L. andTatum, E. L. (1965). Biochirn. Biophys. Acta 102,149-160.

178

L. V. KOCHIAN AND W.J. LUCAS

Slayman, C. L., Rodriguez-Navarro, A . and Blatt, M. R. (1988). In “Proceedings of 7th International Workshop on Plant Membrane Transport”. Australian Academy of Science, Canberra. (In press.) Smith, F. A. and Raven, J. A. (1976). In “Encylopedia of Plant Physiology” (U. Liittge and M. G . Pitman, eds), New Series, Vol. 2A, pp. 317-346. SpringerVerlag, Berlin. Smith, F. A. and Raven, J. A. (1979). Ann. Rev. Plant Physiol. 30,289-311. Spanswick, R. M. (1972). Planta 102,215-227. Spanswick, R. M. (1981). Ann. Rev. Plant Physiol. 32,267-289. Spanswick, R. M. and Anton, G. E. (1988). In “Proceedings of 7th International Workshop on Plant Membrane Transport”. Australian Academy of Science, Canberra. (In press.) Stout, R. G. and Cleland, R. E. (1980). Plant Physiol. 66,353-359. Stout, R. G. and Cleland, R. E. (1982). Plant Physiol. 69,798-803. Sze, H. (1980). Proc. NatlAcad. Sci. USA 77,5904-5908. Sze, H. (1985). Ann. Rev. Plant Physiol. 36,175-208. Sze, H. and Churchill, K. A. (1981). Proc. Nut1 Acad. Sci. USA 78,5578-5582. Sze, H. and Hodges, T. K. (1977). Plant Physiol. 59,641-646. Tasaki, I. and Hagiwara, S. (1957). J. Gen. Physiol. 40,859-885. Terry, B. R. and Robards, A. W. (1987). Planta 171,145-157. Thibaud, T.-B., Soler, A. and Grignon, C. (1986). Plant Physiol. 81,847-853. Thom, M. and Maretzki, A. (1985). Plant Physiol. 77,873-876. Tognoli, L . , Beffagna, N., Pesci, P. and Marrb, E. (1979). Plant Sci. Lett. 16,l-14. Torii, K. and Laties, G. G . (1966). Plant Cell Physiol. 7,395-403. Ulrich, A. (1941). A m . J. Bot. 28,526-537. Van Bel, A. J. E., Borstlap, A. C., Van Pinxteren-Bazuine, A. and Ammerlaan, A. (1982). Planta 155,335-341. Van Iren, F. and Boers-van der Sluijs, P. (1980). Planta 138,130-137. Van Iren, F., Joolen, M. L., Gerritsen, A. F. C., Noordervliet, M. A. W. and Boers-van der Sluijs, F. P. (1981). Physiol. Plant. 52,lS-22. Van Steveninck, R. F. M. (1976). In “Encyclopedia of Plant Physiology” (U. Liittge and M. G. Pitman, eds), New Series, Vol. 2A, pp. 307-342. Springer-Verlag, Berlin. Vara, F. and Serrano, R. (1982). J. Biol. Chem. 257,12826-12830. Vara, F. and Serrano, R. (1983). J. Biol. Chem. 258,5334-5336. Walker, N. A. (1974). In “Membrane Transport in Plants” (U. Zimmerman and J. Dainty, eds), p. 348. Springer-Verlag, Berlin. Walker, N. A. and Pitman, M. G. (1976). In “Encyclopedia of Plant Physiology” (U. Liittge and M. G. Pitman, eds), New Series, Vol. 2A, pp. 93-126. SpringerVerlag, Berlin. Wild, A., Woodhouse, P. J. and Hopper, M. J. (1979). J. Exp. Bot. 30,697-704. Willenbrink, J. and Doll, S. (1979). Planta 147,159-162. Wilson, C., Oross, J. W. and Lucas, W. J. (1985). Planta 164,227-240. Wood, J. M. (1975). J . Biol. Chem. 250,4477-4485. Wolf, 0 .and Jeschke, W. D. (1987). J. Plant Physiol. 128,371-386. Wyn Jones, R. G. (1975). In “IonTransport in Plant Cells and Tissues” (D. A. Baker and J. L. Hall, eds), pp. 193-229. American Elsevier, New York. Wyn Jones, R. G., Brady, C. J. and Speirs, J. (1979). In “Recent Advances in the Biochemistry of Cereals” (D. L. Laidman and R. G. Wyn Jones, eds), pp. 63-103. Academic Press, London. Young, M. and Sims, A. P. (1972). J. Exp. Bot. 23,958-969. Zocchi, G. and Hanson, J. B. (1982). Plant Physiol. 70,318-319. Zocchi, G., Rodgers, S. A . and Hanson, J. B. (1983). Plant Sci. Lett. 31,215-221.

Sporogenesis in Conifers

ROGER I. PENNELL

John Innes Institute and AFRC Institute of Plant Science Research, Colney Lane, Norwich NR4 7UH, UK

. . . . . . . . . . . . . . . . . .

179

11.

Sporogenesis in the Pollen-bearing Cone . . . . . . . . . A. The Archaesporium and Differentiation within the Sporangium B. Sporogenous Cells and Tapetum . . . . . . . . . . C. Meiosis . . . . . . . . . . . . . . . . . . D. ExinePatterningandtheFreeSporePeriod . . . . . .

181 181 182 183 185

111.

Megasporogenesis . . . . . . . . . . . . . . . . A. The Origin of the Reproductive Cell Lineage within the Ovule B. Mitochondria, Plastids and Planes of Division within the Megaspore Mother Cell . . . . . . . . . . . . . . . C. Megaspore Viability . . . . . . . . . . . . . .

190 190

I.

Introduction

191 193

I. INTRODUCTION Sexual reproduction is essential for the long-term vigour of all organisms. Although plants are capable of a variety of sexual methods of reproduction (e.g. apomixis, apogamy and vivipary), only sexual methods bring together genes from two parents in a blend which confers genetic uniqueness upon the progeny. The random reassortment of parental genes affords the chance that a proportion of the progeny will be pre-adapted to thrive in changing environments. Sexual reproduction therefore offers the means by which plants may spread globally and compete and succeed in conditions which might otherwise be exclusive. Copyright 01988 Academic Press Limited All rights of reproduction in any form reserved.

Advances in Botanical Research Vol. 15 ISBN 0-12-005915-0

179

180

R. I. PENNELL

In land plants it is the spore that allows sexual reproduction, since in all instances, both male and female, the spore is the ultimate product of meiosis and hence contains a haploid complement of chromosomes. Seed plants bear spores which develop heterotrophically into unisexual gametophytes capable of bearing either sperms or eggs, as distinct from the hermaphroditic gametophytes of many pteridophytes. They are thus termed “heterosporous”, and the origin of heterospory itself today forms the basis of active research. In the majority of heterosporous plants (those that also form seeds), the spore that gives rise to the male gametophyte (the microspore; the term does not imply volume relationship to its female counterpart) is liberated from the parent sporophyte, while that giving rise to the female gametophyte (the megaspore) is retained within a specialized part of the parent plant which subsequently functions as a part of the seed. The “pollen grains” of the yews (Tuxus spp.) are more correctly microspores since the single nucleus contained within each is one of a tetrad of four produced during meiosis. The term “pollen grain” is accurate only when post-meiotic divisions of the haploid nucleus take place, giving rise to the bi- and tricellular grains characteristic of Pinus and many other conifers and of the flowering plants. Although these plants do not liberate microspores as such, their pollen grains are nevertheless the product of the same sequence of events, known as microsporogenesis. Together with comparable events which take place in the ovular tissues (megasporangium or carpel), the process of sporogenesis is a lengthy one and involves a series of now well-documented stages. The cell lineage which ultimately gives rise to the spore appears within the plant well before meiosis occurs. In Tuxus buccatu a period of approximately four months separates the two events (Pennell and Bell, 1985), although in flowering plants only a few days may pass between the appearance of the reproductive cell lineage and the reductive division of meiosis. The first cell which can with certainty be identified as the forerunner of the meiocyte is generally termed the archaesporial cell, and collectively a mass of these cells forms the archaesporium. In a microsporangium there are several hundred sporogenous cells which consequently yield very many pollen grains, while in a megasporangium the single tetrad is formed from only one. The term “sporogenous cell” may then be applied to any destined to give rise to a meiotic cell once the archaesporium has differentiated, the remainder forming the tapetum. The sporogenous cells in both microsporangia and megasporangia are, like those of the archaesporium, meristematic, and undergo a finite number of divisions. The final division gives rise to a spore mother cell or meiocyte within which meiosis takes place. This brings to a close the diploid phase of the life cycle, but the haploid spores subsequently acquire elaborate surface architecture. These subjects will be developed at more length in the following pages. It is noteworthy that the conifers are generally difficult subjects for study since

SPOROGENESIS IN CONIFERS

181

the reproductive cells of interest are often inaccessible to experimental manipulation and difficult or impossible to isolate. For these reasons the most profitable approach has been an in situ one, using a variety of microscopical techniques. The Taxales have been regarded by many authors as a group closely allied to (or a part of) the Coniferales, and the behaviour of Taxus is here regarded as an example of this group as a whole.

11. SPOROGENESIS IN THE POLLEN-BEARING CONE A. THE ARCHAESPORIUM AND DIFFERENTIATION WITHIN THE SPORANGIUM

The origin of the cell which gives rise to the archaesporium is in all instances unclear. It is likely to be a hypodermal cell of the primordial microsporangium. Like meristematic cells generally, it gives rise to a number of others which divide rapidly, quickly filling the microsporangium with a mass of angular cells up to 10 p m in width (Moitra and Bhatnagar, 1982). The cytoplasms of these cells are interconnected by numerous plasmatic channels (a feature shared with the archaesporial cells of flowering plants) ,sometimes to such an extent that they become largely confluent. Nevertheless, the development of the archaesporial cells is characteristically asynchronous, and for this reason what is known of their metabolism comes from in situ studies following sectioning and staining. The archaesporial cells display many features characteristic of cells undergoing rapid metabolic reorganization, including fluxes in numbers of ribosomes (Pennell and Bell, 198S), extensive lytic activity (Pennell and Bell, 1985), and cyclical changes in the state of differentiation of the plastids (Moitra and Bhatnagar, 1982; Pennell and Bell, 1985). There are also changes in the affinity of the plastid envelopes for osmium (Dickinson and Bell, 1976; Pennell and Bell, 1985), which may represent increasing saturation of membrane lipids and accompanying enhanced permeability to metabolites (Bell, 1983), and changes in the structure and composition of the primary walls which precede the appearance within them of callose (Moitra and Bhatnagar, 1982). The transition from archaesporium to sporogenous tissue occurs in Pinus banksiana only a few days before meiosis takes place in the microspore mother cells (Dickinson and Bell, 1976), but in Taxus 4-5 weeks separate the two events (Pennell and Bell, 1985). The archaesporium differentiates into the sporogenous cells and tapetum in Taxus (Pennell and Bell, 1985) and Pinus banksiana (Dickinson and Bell, 1976), but the tapetum appears to arise from the sporangial wall cells in Pinus sylvestris (Walles and Rowley, 1982). The finding that in Taxus the cells destined to become tapetal contain in their inner radial walls thickenings that are unique to this site (Pennell and

182

R. I. PENNELL

Bell, 1985) indicates that their metabolism is different from those elsewhere in the microsporangium at a very early stage, even though the fine structures of their cytoplasms are identical. The tapetum and sporogenous cells develop along independent lines soon after such thickenings appear within these walls, but the stimulus which initiates their developmental separation is not yet known. Radial gradients within the sporangium have been invoked as being responsible for the dichotomy (Dickinson and Bell, 1976), and the unique thickenings in the inner tangential walls may be active in preventing such gradients from traversing this barrier. The absence of symplastic connections between the peripheral layer of cells of the Tuxus archaesporium would therefore lead to the physiological isolation of these cells and might even promote their subsequent degeneration. B. SPOROGENOUS CELLS AND TAPETUM

In gymnosperms generally, the sporogenous cells develop in a loculus enclosed by one or two layers of tapetal cells. Tapeta such as this, which remain in a peripheral position throughout microsporogenesis, are termed “secretory”. Although invasive tapeta are common in the flowering plants (Maheshwari, 1950), the partially invasive tapetum of Aruucuriu (Hodcent, 1965) and of Lurix (Mikulska et ul., 1969) remain unusual features within the gymnosperms. The development of the tapetum takes place at a rate that coincides with the development of the sporogenous cells within, so that by the time meiosis is completed the tapetum itself is almost wholly disorganized. Structural changes in the tapetum at this time include an increase in numbers of dictyosomes and vesicles, and in endoplasmic reticulum, which gradually becomes organized into whorls containing up to 12 gyres of cisternae (Pennell and Bell, 1985). In Lilium these membranes are assembled at the expense of “storage” lipid (Reznickova and Dickinson, 1982), and this is probably also the case in Tuxus. Meanwhile, the sporogenous cells in Tuxus undergo changes in basophilia that are the result of fluxes in ribosomes. Lytic vesicles appear in all and are likely to be responsible for the diminution in ribosome numbers that takes place before the cells enter meiosis (Pennell and Bell, 1985). These changes are difficult to quantify since, like those which accompany the development of the archaesporium, they take place at different rates in different cells. However, the technique of microfluorometry has now been applied for the first time to the microsporangia of Tuxus (Pennell and Bell, 1985), and has given support to the notion that marked changes in cytoplasmic RNA do occur in the developing sporangium. This technique involves staining high molecular weight nucleic acids with a dye which fluoresces when excited by far-blue light. The fluorochrome Acridine Orange will, at pH 4.0, form a complex with RNA

SPOROGENESIS IN CONIFERS

183

that fluoresces orange-red and with DNA that fluoresces green-blue. When the DNA has been removed by pretreatment of the tissue with DNase, the fluorescence of the RNA alone can be measured by a single photocell. Since the binding of the dye to the nucleic acid is stoichiometric (Bucknall and Sutcliffe, 1965; Darzynkiewicz et al., 1979), its intensity can be taken as indicative of the amount of RNA in the section of the cell and, by extrapolation, in the cell as a whole. In Tuxus the trend during the development of the sporogenous cells appears to be towards a general increase in RNA (Pennell and Bell, 1985), this being attributable to a proliferation of ribosomes. The significance of this is still unclear, but it is conceivable that it takes place in preparation for meiosis, when very special and specific proteins associated with the synaptonemal complex are assembled (Moens et ul., 1987). Other proteins are certain to include some that control the entry into prophase and the progression through meiosis, and, as during mitosis, those of the cytoskeleton which determine planes of division and cytokinesis (Wicket a l . , 1981). The rapid and timely assembly of these proteins may therefore necessitate the appearance within the sporogenous cells of additional ribosomes upon which their messenger RNAs may be translated. Once completed, the controlled elimination of ribosomes by the directed action of proteases and nucleases could then give rise to the fall of RNA levels observed immediately prior to meiotic prophase (Pennell and Bell, 1985). C. MEIOSIS

Meiosis within pollen mother cells has been best studied in flowering plants, and common features have appeared that may be significant controlling events in the life cycle (Dickinson and Heslop Harrison, 1977). One of the most important of these is the undoubted elimination of a major proportion of ribosomes during prophase (Mackenzie et al., 1967) and their subsequent restoration by the disintegration in the cytoplasm of “nucleoloids”, possibly assembled in the nucleus (Williams et al., 1973). This “ribosome cycle” is accompanied by the transient dedifferentiation of both the mitochondria and plastids (Dickinson and Heslop Harrison, 1977) and the formation of “nuclear vacuoles” within the prophase nuclei (Sheffield et a l . , 1979). These events have been interpreted as crucial in the reorganization of a diploid cell (a sporogenous cell) into one whose genome is haploid (the spore), and in which expression of that part of the genome which is concerned specifically with sporophytic growth must be replaced by that part concerned specifically with gametophytes (Dickinson and Heslop Harrison, 1977). The physiological isolation of the meiocytes by cell walls rich in callose [strictly those capable of forming a fluorescent complex with Aniline Blue (Mangin, l889), now proven to involve a variety of mixed P-linked glucans (Smith and McCully, 1978)], has been implicated as being significant in the change of phase (Heslop Harrison and Mackenzie, 1967).

184

R. I. PENNELL

Meiosis in conifers has, in the absence of evidence to the contrary, been thought comparable. Indeed, the process in both angiosperms and gymnosperms shares some common features, including the partial dedifferentiation of the mitochondria and plastids and the appearance of vacuoles within the nuclei (Sheffield et al., 1979). Meiosis in many conifers differs significantly in the duration of prophase, however. In Lilium (which may be regarded as representative of the situation in floweringplants generally) meiosis lasts for just 2-3 days (Dickinson, personal communication), while in Juniperus (Singh et al., 1983) and Taxus (Pennell and Bell, 1987a) it lasts for approximately five weeks, and in Pseudotsuga, Tsuga, Thuja and Larix for almost two months (Owens and Molder, 1971). In all instances it is the prophase of meiosis which is extended, the remaining stages of meiosis I and I1 occurring within a few days. Not only does this set the conifers apart from the flowering plants, but it also makes them favourable subjects for the study of events regulating meiosis. Advantage has recently been taken of this, and a new technique applied to the pollen mother cells of Taxus has provided novel information for an entirely new concept of controlling mechanisms (Pennell and Bell, 1986a, 1987a). The technique is also based upon the preferential staining with Acridine Orange of RNA in de-embedded sections of microspore mother cells. However, to obtain greater resolution than afforded by microfluorometry per se, the photographed images of the cells have been enlarged sufficiently to allow investigation of cytoplasm and nucleus separately, and these enlargements have been used as the subject of study. The density of reduced halide on the film, like the intensity of the fluorescence itself, is a direct measure of RNA within the section. When each compartment of the image of the cell is scanned separately in a microdensitometer, the original intensity of fluorescence can be judged from the restriction of the scanning beam, and quantified. Once cells from each of the crucial stages of meiosis have been examined in this way, the data coming from cytoplasm, nucleus excluding nucleolus, and nucleolus alone, may be plotted separately. When applied to the pollen mother cells of Taxus, it is evident that there is only slight diminution in cytoplasmic RNA during prophase (Pennell and Bell, 1986a, 1987a) and this is matched by reciprocal changes within the nucleus (excluding the nucleolus). Throughout prophase, RNA levels in the nucleolus remain constant. The significance of these findings may be great. It is conceivable that in Taxus, and possibly in the conifers generally, it is the absence of widespread degeneration of cytoplasmic RNA that is responsible for the protracted nature of prophase. The controlled hydrolysis of RNA that occurs during prophase in flowering plants may liberate oligoribonucleotides into the cytoplasm which affect the rate of expression of nuclear genes and therefore the progression through meiosis. The absence of such putative signalling molecules from the Taxus meiocytes may thus explain the slow passage through prophase. Whatever the case, it now seems clear that the “nuclear vacuoles” that have in the past been

SPOROGENESIS IN CONIFERS

185

invoked as the means by which information-carrying molecules move between nucleus and cytoplasm are induced only where the chromosomes exert traction on the nuclear envelope during contraction (Pennell and Bell, 1985, 1986b). Once prophase is completed the remainder of male meiosis proceeds in gymnosperms in much the same way as in angiosperms and heterosporous pteridophytes. Meiosis I gives rise to pairs of cells (diads), each of which quickly divides again into two spores. In Pinus these divisions take place without the formation of partitioning walls within the tetrad (Willemse, 1971; Moitra, 1980), but in Taxus a primary wall does develop during telophase I which separates the cells of the diad (Pennell and Bell, 1986b). It therefore seems likely that the sequence of partitioning walls within the pollen mother cell carries with it systematic significance, but it has no evident functional effect. What is perhaps more important functionally is the appearance of invaginations within the nuclei of post-meiotic cells of Pinus banksiana (Dickinson and Bell, 1970) and Podocarpus macrophylla (Aldrich and Vasil, 1970). These tube-like infoldings of the nuclear envelope have rightly been thought significant both because of the presence of highly ordered pore-like structures at their openings (Dickinson and Bell, 1972a), and by virtue of the accumulation within them of osmiophilic material (Dickinson and Bell, 1970), possibly RNA (Dickinson and Potter, 1975). When taken in the context of the change of phase which occurs during meiosis, these invaginations have been cited as specialized structures by which information-carrying molecules are provided en masse to the cytoplasm of the spore (Dickinson and Bell, 1972a). It has been thought conceivable, therefore, that these molecules are in some way specific to and crucial for the cytoplasm of the spore, and prime it for gametophytic growth (Dickinson and Bell, 1972a). Such a mechanism would find a parallel in the angiosperms, where “nucleoloids” become dispersed in the spore cytoplasm at the end of meiosis (Williams et al., 1973). In Taxus, however, there are neither nuclear invaginations nor nucleoloids, and there is no evidence of nucleocytoplasmic interaction of this or any other kind at the end of meiosis. This of course may be related to the limited flux of RNA between cytoplasm and nucleus during prophase described earlier.

D . EXINE PATTERNING AND THE FREE SPORE PERIOD

Once meiosis is completed a complex series of events occurs within the cytoplasm of the spores, and these are ultimately responsible for the patterning of the sporoderm. Exine formation in the pollen grain of Pinus banksiana has been studied in depth. The pollen grain bears two bladder-like

186

R. I. PENNELL

sacci distally and a punctate sexine perforated by a single colpus. The reticulate pattern of the sexine is established while the spores remain enclosed in the callose wall, and pattern formation begins when a number of large vesicles make contact and fuse radially at the plasma membrane of the young spore (Dickinson, 1971). The vesicles arise from what appear to be enlarged dictyosomes, and fuse en masse with the plasma membrane (Dickinson, 1971). The outer radial faces of the vesicles then become diffuse and electron-opaque, although the inner faces, which represent the new plasma membrane of the cell, continue to stain normally. The outer surfaces of the vesicles act as pattern initials, and their subsequent thickening and elaboration with sporopollenin give rise to the surface topography of the grain. Similar events take place close to the plasma membranes of the young spores of Abies concolor (Kurmann, 1986), Tsuga canadensis (Kurmann, 1984a) and Taxodium distichum (Kurmann, 1985), and in each instance development within the tetrad proceeds until elements of both nexine and sexine are in place. Detailed information is also available for the flowering plants Cosmos and Lilium (Dickinson, 1976), and is again comparable although in the angiosperms generally the architecture of the sexine is considerably more complex than that of Pinus. The patterning of the Pinus pollen grain does not take place at that part of the plasma membrane which becomes “masked” by cisternae of endoplasmic reticulum, and this does not develop into the layered exine characteristic of other parts of the grain (Dickinson, 1976). In Taxodium, the granular layer of the sexine is absent from the distal surface at the region of the germinal papilla, and the number of layers in the nexine is reduced (Kurmann, 1985), but in Tsuga there is no evidence of a pre-formed germinal aperture in the sporoderm (Kurmann, 1984b). The situation in Tuxus is simpler. The contribution of the gametophyte (spore) to the pattern of the exine takes the form of a number of small osmiophilic globules (indistinguishable in the electron microscope from sporopollenin) that emerges from the protoplast and form a confluent layer (Pennell and Bell, 1986b), or layers (Rohr, 1977), at its surface. This thin smear of sporopollenin encapsulates the microspore during the early free spore period, and serves as a surface upon which tapetally synthesized sporopollenin is placed (Pennell and Bell, 1986b). The second phase of accretion then takes place as orbicules enter the loculus from the degenerating tapetal cells, and appears to involve no more than the transfer of sporopollenin from individual orbicules onto areas of exposed gametophytic sporopollenin. The sporopollenin of the orbicules appears to have greater affinity for that layered upon the spore surface by the gametophyte than for either the lipidic core of the orbicule or orbicular sporopollenin elsewhere. In consequence, orbicular sporopollenin is transferred en masse to the surface of the maturing microspore, first as a series of concentric layers and then as discrete globules which confer the punctate pattern upon the sporoderm. The number of globules attached to the microspore surface

SPOROGENESIS IN CONIFERS

187

Fig. 1. Near-mature Taxus microspore, fixed inside a loculus of the microsporangium before dehiscence. The nexine (NX) consists of several distinct layers of sporopollenin and the sexine (SX) of a single layer of globules. The intine (I) is barely visible. N = nucleus, P = plastid. Scale bar = 1 pm.

increases until a complete layer is present, and no areas of layered sporopollenin remain available for transfer (Fig. 1).Similar events complete exine patterning in Tarodium (Kurmann, 1985). As in Tsuga, there is no evidence in Tuxus G f a germinal pore within the spore wall. Following the terminology of Erdtman (1966), the laminae of sporopollenin represents the nexine (of which several laminae can be resolved, presumably representing individual waves of synthesis from the tapetum), and the globules the sexine. It is notable that in Tuxus the entire sexine and much of the nexine are added to the microspores once the callose wall is dissipated, but in the members of the Pinaceae a large proportion of nexine and sometimes sexine is contributed by the spores themselves. Although the structural basis of patterning is now well-established for, amongst others, Abies, Pinus, Taxodium and Taxus, as well as for several angiosperms, there is still little information upon the means by which the subcellular events responsible for it are controlled. In Lilium it has now been demonstrated that neither the meiotic spindles nor the radial microtubules in the cytoplasm of the young spore control the positioning of pattern determinants in the plasma membrane of the young pollen grain (Sheldon

188

R. I. PENNELL

Fig. 2. Junction between three tapetal cells of Taxus. The radial cell wall (CW) contains numerous lipidic globules (L). The plasma membranes (PM) of the cells are associated with osmiophilic globules of sporopollenin (S), and similar globules are also visible inside the protoplast upon small vesicles (V). Two such vesicles appear to have traversed the plasma membranes and are entering the cell wall (arrows). Scale bar = 1pm.

and Dickinson, 1986), as the microtubules of root (Lloyd and Wells, 1985) and seed hairs (Quader et al., 1986) appear to control the orientation of cellulose microfibrils in the primary cell wall. It seems more likely that patterning determinants are present ab initio within the plasma membrane of the young pollen mother cells (Sheldon and Dickinson, 1983), and these become reorganized into a pattern which predicts that which develops subsequently in the pollen grain wall (Sheldon and Dickinson, 1986). The control of positioning of the dictyosomes which add to the pattern initials in Pinus (Dickinson, 1971), Tsuga and Taxodium (Kurmann, 1984b), however, seems likely to be involved with the cytoskeleton of the spore. In contrast, the means by which the pattern becomes impregnated with sporopollenin from the tapetum appears clear. The vehicle responsible for the transport of sporopollenin from the tapetum into the loculus of the sporangium is the orbicule. The lipidic centre of each orbicule first appears as a spherical globule within a radial tapetal wall (Pennell and Bell, 1986b) (Fig. 2). The osmiophilic material which subsequently encloses each of these “pro-orbicular cores” appears concomitantly within the tapetal protoplasts close to the radial walls (Fig. 2), in association with small vesicles (Pennell and Bell, 1986b). This substance, judged from its affinity for osmium to be sporopollenin, also moves into the radial walls of the tapetum, apparently by forming a transient association with the plasma membranes at these sites.

SPOROGENESIS IN CONIFERS

189

Fig. 3. Localization by lead capture of acid phosphatases in the tapetum of Taxus. Both the plasma membrane (PM) and the membrane of a small vesicle (V) appear to contain the enzyme. C = cytoplasm. Scale bar = 1 pm.

Several globules of sporopollenin then rapidly aggregate to form a sheet which enwraps a single pro-orbicular core (Pennell and Bell, 1986b). The composite body so formed is clearly identifiable as an orbicule. Many orbicules are released from the tapetal walls as the protoplasts degenerate and undergo hypertrophy. The formation of the orbicules appears therefore to owe much to the affinity of the sporopollenin sheets both for lipid and for sporopollenin elsewhere. Similarly, the transfer of sporopollenin from the orbicule to the developing exine seems to come about as a result of the differing affinity of orbicular sporopollenin for the lipidic centre of the orbicule and for the sporopollenin in place upon the surface of the microspore. Indeed, once the transfer has taken place, the pro-orbicular cores devoid of a sporopollenin encasement may be observed close by the microspores. The degeneration of the tapetum proceeds apace with the formation upon the microspores of the sporoderm. The complex membranous figures which all but fill the mature tapetal cells of Tuxus develop as the fine structure of the cells becomes disorganized, and may arise as a result of their increased physiological isolation by the peritapetal membrane. Such an investment has been observed in the sporangia of many conifers (Moitra and Bhatnagar,

190

R. I. PENNELL

1982) and appears to be composed of saturated lipid (Dickinson and Bell, 1972b). Soon after the peritapetal membrane is formed in the microsporangia of Taxus acid phosphatases appear in the plasma membranes of the tapetal cells (Pennell and Bell, 1986b) (Fig. 3). These enzymes may lead to the generalized perturbation of the plasma membranes by dephosphorylating membrane phospholipids and phosphate-containing proteins. However, what is perhaps of more interest in the study of the development of the microspores is that these enzymes (and presumably other hydrolases which accompany them) are already present and active within the tapetal plasma membranes when the microspore mother cells enter meiosis, some eight weeks before the tissue disintegrates.

111. MEGASPOROGENESIS In comparison with the development of the microsporangium, the events which take place within the ovules of conifers are poorly explored by modern techniques. This is due both to the morphology of the female cone (or fertile axis in the Taxales) and, as in all seed plants, the presence of just a single meiocyte within the ovule. Indeed, so unyielding are the megaspores and their antecedents to experimental methods that all that is known about them has been reasoned from structural studies and comparison with megasporogenesis in heterosporous pteridophytes and flowering plants whose sporangia are more accessible. Nevertheless, it is clear that sporogenesis within the ovule differs from microsporogenesis in several important ways, most strikingly in the development of the tetrad and survival of the spores. A. THE ORIGIN OF THE REPRODUCTIVE CELL LINEAGE

WITHIN THE O W L E

Little is known about differentiation within the ovule in conifers. In Tuxus, whose ovules are borne singly on compressed fertile axes, the nucellus is initially homogeneous. The nucellus begins to differentiate about 12 weeks after the appearance upon the shoot apex of the primordium of the fertile axis, when a mass of basophilic cells appears in the middle of the nucellus. In the electron microscope they are seen to contain many ribosomes and conspicuous globules of lipid (Pennell and Bell, 1987b). Since it is within this mass of cells that the megaspore mother cell develops, they have been thought to participate in the nutrition of the meiocyte and the female gametophyte which succeeds it when meiosis is completed. There is, however, no experimental evidence in support of this view, and only structural comparison of the cells with those of the tapetum in the microsporangium suggests that both behave similarly. The antecedent of the megaspore mother cell has not been identified with

SPOROGENESIS IN CONIFERS

191

certainty. That of Taxus may develop from a hypodermal cell inside the nucellus (Dupler, 1920). B . MITOCHONDRIA, PLASTIDS AND PLANES OF DIVISION WITHIN THE MEGASPORE MOTHER CELL

Meiosis in the megaspore mother cells of many conifers sees the unidirectional movement of the mitochondria and plastids into a single haploid cell. This was first observed at the beginning of the century as the appearance of a densely staining mass below the nuclei of the meiocytes of Larix, Taxus, Thuja, Taxodium (Coker, 1904) and Juniperus (Ottley, 1909). Since then the movement of amyloplasts within the megaspore mother cells of Pinus has been found to be similar (Willemse, 1968), and indeed in the zooidozamous Encephafartos (de Sloover, 1961) and Ginkgo (Stewart and Gifford, 1967) the distributions of both plastids and mitochondria become polarized in this way. Consequently, the partitioning of the megaspore mother cell during telophase I by a transverse wall gives rise to a pair of cells of which only one-that lying chalazally--contains either mitochondria or plastids, or both (Fig. 4). The organelles behave in a similar way during meiosis 11, so that in Taxus at least (the only genus to be examined thoroughly in this respect) the chalazal spore comes to contain the entire complement of organelles, the other three members of the tetrad remaining bereft of them (Pennell and Bell, 1987b). The mechanism responsible for the directed motion on the mitochondria and plastids is still unresolved. The belief that gravity is the motive force (Volkman and Sievers, 1975) seems untenable when chalazal accumulations are known to occur in the inclined or inverted ovules of both the cone-bearing conifers and taxads. More plausible is the possibility that the motion of the mitochondria and plastids is generated by their intimate association with the cytoskeleton. Comparable phenomena have been described in vivo (Inolie, 1981) and in vitro (Vale e t a f . , 1985b), and there now seems little doubt that microtubules are capable of directing the movement of membrane-bound organelles, possibly in association with the mechanochemical protein kinesin (Vale et a f . ,1985a). Although there is no direct evidence for such a mechanism in the megaspore mother cells of conifers (or zooidogamous gymnosperms), it is difficult to conceive of any hypothesis that is not in some way related to cytoplasmic structures akin to microtubules. The directed movement of the organelles may occur in all conifers, but the form of the tetrad of megaspores differs. In most, linear or T-shaped tetrads are formed (Maheshwari, 1950), but both isobilateral and tetrahedral tetrads occur in Sequoia (Looby and Doyle, 1942). In T-shaped tetrads the plane of division of the micropylar diad cell is perpendicular to its long axis and this is due to a similarly displaced spindle. The plane of division of plant

192

R. I. PENNELL

Fig. 4. Diad produced by meiosis I within the ovule of T a u s . The chalaza1 diad cell (lower) contains numerous mitochondria and plastids, but the micropylar diad cell (upper), while rich in lipid (L), contains none of these organelles. The bounding wall of the diad is thickened adjacent to the plasma membrane in the same way as that of the prophase meiocyte (arrow, inset). Scale bar = 5 p m (0.5 p m inset).

SPOROGENESIS IN CONIFERS

193

cells generally is now known to be predicted by a cortical “pre-prophase band” (Wick and Duniec, 1983,1984), and hence it follows that it is the plane of assembly of this whorl of microtubules which is ultimately responsible for the opposed planes of division of the two diad cells during meiosis 11. Whatever the force responsible, the organization of the spores within the tetrad appears to be of little phylogenetic or developmental importance. C. MEGASPORE VIABILITY

Megasporogenesis in conifers is completed once the tetrad is fully formed and each spore within it is wholly enclosed by bounding and partitioning walls. However, in conifers only the chalazal megaspore develops further (eventually giving rise to the female gametophyte), the remaining spores ultimately degenerating. The promotion of the chalazal megaspore has been attributed to the presence within it of the entire complement of organelles of the mother cell from which it is derived, this conferring in some way “fitness” upon the cell. Although in the conifers the promotion of the chalazal megaspore is in all known instances prefaced by the directed movement into it of mitochondria and plastids, this is not the case in other plants. In Myosurus (Woodcock and Bell, 1968) and Epipactis (Bednara et al., 1981) the embryo sac develops from a single spore, but all the members of the tetrad receive more or less equal numbers of organelles during meiosis. Similarly, in the heterosporous fern Marsilea, megaspore viability cannot be related to any visible differences between the cytoplasms of the four spores of a tetrad (Bell, 1981). In Zeu the distribution of mitochondria and plastids between the four megaspores is only partially polarized, and again only one develops into an embryo sac (Russell, 1979). Clearly, the enriched endowment of a megaspore with mitochondria and plastids cannot be a general explanation of its preferred success, and must merely complement other factors. An alternative suggestion has been proposed by Lintilhac (1974a,b), who believes that the functional megaspore lies at a site within the nucellus where inwardly directed radial forces (established by repeated cell divisions) are opposed, giving rise to an area of “null pressure”. Since this is the only region within the ovule where there would be little or no resistance to expansion, the megaspores lying within would be able to enlarge whilst those elsewhere would not. This hypothesis is, however, difficult to reconcile with the situation in Taxus, where a pair of diametrically opposed vascular bundles in the chalazal region of the integument would be likely to displace the tectonic symmetry of the ovule as expounded by Lintilhac. Similarly, the presence of bisporic ovules in many conifers suggests that a single region of “null pressure” does not exist, or that two are present side by side. Indeed, the location of the functional megaspore within the tetrahedral tetrad of Mur-

194

R. I. PENNELL

silea appears to bear no relation to its orientation within the radially symmetrical megasporangium (Bell, pers. comm.), and in this instance a tectonic explanation of megaspore viability is wholly unsatisfactory. The regular promotion of one member of a tetrad of four spores cannot, in fact, be readily explained by any contemporary model. This has led Bell (1981) to suggest that of the four genotypes present amongst the megaspores of Marsilea, only one is viable in the environment of the megasporangium, and this is invariably promoted at the expense of the remainder. The four genotypes would arise from recombination between two heterozygous loci which regularly experience crossing over. This hypothesis can only be applied at the present time to megaspore mother cells within which the formation of partitioning walls is delayed until both divisions of meiosis are completed. This is the case in Marsilea and the pteridophytes generally. In the linear and T-shaped tetrads of the conifers the genotype that becomes that of the functional megaspore is in all instances contained by telophase I in the chalaza1 diad cell, and is isolated from those that yield the pair of micropylar spores. For a Mendelian explanation to serve the spermatophytes it would therefore be necessary to invoke directed segregation of alleles during meiosis I, and this in itself is at variance with current thoughts upon genetic segregation preceding sexual reproduction.

ACKNOWLEDGEMENTS The author would like to express thanks to Professor P. R. Bell for helpful comments upon the manuscript.

REFERENCES Aldridge, H. C. and Vasil, I . K. (1970). J . Ultrastruct. Res. 32,307-315. Bednara, J., Kuras, M. and Rodkiewicz (1981). Acta SOC. Bot. Pol. 50, 127-130. Bell, P. R. (1981) J . Cell Sci. 51, 109-119. Bell, P. R. (1983). Eur. J . Cell Biol. 30,279-282. Bucknall, R. A . and Sutcliffe, J. F. (1965). J . Exp. Bot. 16,423-432. Coker, W. G. (1904). Bot. Gaz. 38,206-213. Darzynkiewicz, Z., Evenson, D . P . , Staiano-Coico, L., Sharpless, T. P. and Malamed, M. L. (1979). J. Cell Physiol. 100,425-438. de Sloover, J. L. (1961). La Cellule62,105-116. Dickinson, H. G. (1971). In “Sporopollenin” (J. Brooks et al., eds), pp. 31-68. Academic Press, London and New York. Dickinson, H. G. (1976). In “The Evolutionary Significance of the Exine” (K. Ferguson and J. Muller, eds.), pp. 67-90. Linnean Society Symposium, Series 1. Dickinson, H. G. and Bell, P. R. (1970). J . Ultrastruct. Res. 33,356-359. Dickinson, H. G. and Bell, P. R. (1972a). Dev. Biol. 27,425-429. Dickinson, H. G. and Bell, P. R. (1972b). Plants 107,205-215. Dickinson, H. G. and Bell, P. R. (1976). Ann. Bot. 40,103-113.

SPOROGENESIS IN CONIFERS

195

Dickinson, H. G. andHeslopHarrison, J. (1977). Philos. Trans. R. SOC.London. B 277,327-342. Dickinson, H. G. and Potter, U. (1975). Pluntu 122,99-104. Dupler, A. W. (1920). Bot. Guz. 69,492-521. Erdtman, G. (1966). Grunu Pulynol. 6,318-323. Heslop Harrison, J. and Mackenzie, A. (1967). J. Cell Sci. 2,387400. Hodcent, E. (1965). Bull. SOC. Bot. France 112,121-127. Inout, S . (1981). J. Cell Biol. 89,346-356. Kurmann, M. H. (1984a). “Abst. 6th Int. Palynol. Conf.”, Calgary, Canada. Kurmann, M. H. (1984b). “Abst. 2nd Int. Org. Palaeobot. Conf.”, Edmonton, Canada. Kurmann, M. H. (1985). Pulynology 9,246 (abstract). Kurmann, M. H. (1986). A m . J. Bot. 73,632 (abstract). Lintilhac, P. M. (1974a). A m . J. Bot. 61,135-140. Lintilhac, P. M. (1974b). A m . J. Bot. 61,230-237. Lloyd, C. W. and Wells, B. (1985). J. Cell Sci. 75,225-238. Looby, W. J. and Doyle, J. (1942). Proc. R. Dublin SOC. 23,35-54. Mackenzie, A., Heslop Harrison, J. and Dickinson H. G. (1967). Nature 215, 997-999. Maheshwari, P. (1950). “An Introduction to the Embryology of Angiosperm.” McGraw-Hill, New York. Mangin, L. (1889). Bull. SOC.Bot. France 33,512-517. Mikulska, E., Zolincrowicz, W. and Walek-Czernecka, A. (1969). Actu SOC. Bot. Pol. 38,291-301. Moens, P. B., Heyting, C., Dietrich, A. J. J., Raarnsdonk, W. van and Chen., Q. (1987). J. Cell. Biol. 105,93-103. Moitra, A. (1980). Ph.D. Dissertation, University of Delhi. Moitra, A. andBhatnagar, S. P. (1982). Gum. Res. 5,71-112. Ottley, A. M. (1909). Bot. Guz. 48,3146. Owens, J. N. and Molder, M. (1971). Can. J. Bot. 49,2061-2064. Pennell, R. I. andBell, P. R. (1985). Ann. Bot. 56,415-427. Pennell, R. I. andBell, P. R. (1986a). ActuSoc. Bot. Pol. 55,ll-16. Pennell, R. I. and Bell, P. R. (1986b). Ann. Bot. 57,545-555. Pennell, R. I. and Bell, P. R. (1987a). A m . J. Bot. 74,444-450. Pennell, R. I. and Bell, P. R. (1987b). Ann. Bot. 59,693-704. Quader, H., Deichgraber, G. and Schnepf, E. (1986). Pluntu 168,l-10. Reznickova, S . A. and Dickinson, H. G. (1982). Plantu 155,400408. Rohr, R. (1977). Cytologiu 42,157-167. Russell, S . D. (1979). Can. J. Bot. 57,1093-1110. Sheffield, E., Cawood, A. H., Bell, P. R. and Dickinson, H. G. (1979). Pluntu 146, 597-601. Sheldon, J. M. and Dickinson, H. G. (1983). J. Cell Sci. 63,191-208. Sheldon, J. M. and Dickinson, H. G. (1986). Pluntu 168, 11-23. Singh, H., Owens, J. N. and Dietrich, H. F. (1983). A m . J. Bot. 70,1272-1280. Smith, M. M. and McCully, M. E. (1978). Protoplusmu 95,229-254. Stewart, K. D. and Gifford, Jr., E. M. (1967). A m . J. Bot. 54,375-383. Vale, R. D., Reese, T. S. and Sheetz, M. P. (1985a). Cell 41,39-50. Vale, R. D., Schnapp, B. J., Reese, T. S. and Sheetz, M. P. (1985b). Cell 40, 449454. Volkman, D. and Severs, A. (1975). Pluntu 127,ll-19. Walles, B. and Rowley, J. R. (1982). Nord. J. Bot. 2,53-70. Wick, S . M. and Duniec, J. (1983). J. Cell Biol. 97,235-243. Wick, S . M. and Duniec, J. (1984). Protoplusmu 122,45-55.

196

R. I. PENNELL

Wick, S. M., Seagull, R. W . , Osborn, M., Weber, K. and Gunning, B . E. S. (1981). J . Cell Biol. 89,685-690. Willemse, M. T. M. (1968). Actu Bot. Neerl. 17,330-331. Willemse, M. T. M . (1971). Actu Bot. Neerl. 20,411-427. Williams, E., Heslop Harrison, J. and Dickinson, H. G . (1973). Protoplusmu 77, 79-83. Woodcock, C. L. F. and Bell, P. R. (1968). J. Ultrastruct. Res. 22,546-563.

AUTHOR INDEX

Barlow, P. W., 11, 39, 140, 174 Bartholomew, D. P., 44, 67, 86 Bates, L., 91 Baylor, D. A., 41 Beardall, J., 57, 86, 89 Beckerle, M. C., 39 Bednara, J., 193,194 Beffagna, N., 112,171, 176, 178 Behl, R., 174 Behrens, H. M., 9, 11, 29, 38, 40, 140,

A

Abrahams, M. S., 89 Acevedo, E., 52, 53,86 Adams, H. R., 99, 100,174 Adams, M. S . , 46,86 Adams, W. W., 61,66, 81, 82,86,88 Akedo, H., 120, 171 Aldrich, H. C., 185,194 Aldridge, B . B ., 33,38 Alfani, A., 86, 91 Ammerlann, A., 178 Anderson, W. P., 136, 137, 138, 140,

171

Bell, P. R., 180, 181, 182, 183, 184, 185, 186, 188, 189, 190, 191, 193,

149,171

AndrC, M., 64,86 Anikeeva, I., 39 Ansari, A. Q., 146,171 Anton, G. E., 107, 108,178 Ariffin, A., 91 Armstrong, C. M., 123, 124,171 Armstrong, M. J., 166, 171 Asher, C. J., 152, 171 Ashford, A. E., 165,176 Atkins, C. A., 174 Audus, L. J., 2, 7 , 19, 20, 32,38,39 Avadhani, P. N., 86 Axelrod, D., 40

194,194,195,196

Bellando, M., 99, 112, 171 Bender, M. M., 89 Bennett, A. B., 8, 41, 109, 171 Benolken, R. M., 8,38 Bentrup, F. W., 114, 124, 171 Benzing, D. H., 69, 70,86 Bergfeld, R., 39 Berlin, J. R., 41 Berridge, M. J., 133, 171 Berry, J. A., 87, 88 Berry, L. J., 9,38 Bertl, A., 158, 173 Bhatnagar, S. P., 181, 189,195 Bienfait, H. F., 129, 171 Bjorkman, T., 1-41 Black, C. C., 51, 54,86 Blackman, M. S., 120,171 Blatt, M. R., 100, 112, 117, 171, 177,

B Baasch, R., 91 Bader, U., 172 Badillo, I., 86 Bajer, A., 39 Ball, E., 49, 50, 51, 66, 81, 82, 88,89,

178

Blumwald, E., 156, 171 Boers-van cjer Sluijs, F. P., 137, 138,

90, 175

Ballarin-Denti, A., 101, 109, 172, 176 Ballio, A., 172 Bandurski, R. S . , 9, 25,38 Bange, G. G. J., 137, 143,171 Barber, S. A., 161, 172, 173 Barclay, G. F., 28,38 Bare, C., 176

178

Bomrnerson, J. C., 88 Bonetti, A., 171 Bonnemain, J. L., 120, 172 Borstlap, A. C., 96, 171, 178 Boss, W. F., 133, 171

197

198

AUTHOR INDEX

Boston, H. L., 46,86 Bottger, M., 132, 136,171 Bowes, G., 69,87 Bowling, D. J. F., 145, 146, 149,171, 173 Bracker, C. E., 174 Brady, C. J., 178 Brinckmann, 44,84,91 Briskin, D. P., 102, 106, 107, 109,171, 172,173 Brooks, S . C., 100,172 Brown, 84, 85 Brownlee, C., 28,38 Broyer, T. C., 94, 100, 136, 151,172, 174 Brulfert, J., 52, 68,86 Bruner, L. J., 33,38 Bryant, J. A., 87 Buchanan-Bollig, B., 88 Buchner, K.-H., 90 Buck, C., 39 Buckhout, T. J., 40,132, 133,136,172, 175 Bucknall, R. A., 183,194 Burdon-Sanderson, J., 8,38 Burgess, N., 91, 92 Burk, J. A., 71,75, 90 Burnett, J. H., 16,38 Burridge, K., 39 Burris, R. H., 89 Butz, R. G., 149, 172 Buxbaum, R. E., 22,38

Chino, M., 142,172 Christensen, H. N., 120,171,172 Christiansen, N. L., 88 Churchill, K. A., 109,172, 178 Claassen, N., 161, 172 Clarkson, D., 50,88 Clarkson, D. T., 129, 140, 149, 150, 165,172,175 Cleland, R. E., 10,33,38,109,112,178 Clifford, P. E., 28,38 Cockburn, W., 44, 45,48, 50, 69,86 Cocucci, S . M., 112,171, I72 Cohen, S . R., 120, 124,172 Coker, W. G., 191,194 Colombo, R.,171,176 Comins, H., 87 Cooper, E., 174 Corey, D. P., 26,38 Coronel, I., 89 Cosgrove, D., 33,38 Cosgrove, D. J., 33,38 Coster, H. G. L., 33,38 Courtice, A. C., 177 Courtice, G. R. M., 39 Cove, D. J., 39 Crafts, A. S . , 136, 151,172 Craig, T. A., 129, 172,176 Crain, R. C., 176 Cram, W. J. C., 87,88, 153, 172 Crane, F. L., 129,172,176 Crook, C. E., 90

C Cabantchik, Z. I., 123,172 Caldwell, M. M., 91 Cannell, M. B., 41 Canny, M. J., 176 Cantley, L. C., 106, 107,172 Carnal, N. W., 86 Carver, K. A., 86,87 Caspar, T., 16, 17,38 Cawood, A. H., 195 Cerana, R., 175 Chabot, J. F., 38 Chamel, A., 173 Chandra, S . , 28,38 Charlton, W. A., 87 Chastain, C. J., 131, 136, 172 Cheeseman, J. M., 99, 101, 104, 105, 117, 160, 164, 165, 168,172 Chen, S.-S., 54,86

Da, S . L., 91 Darling, A , , 172 Darwin, F., 7,38 Darzynkiewicz, Z., 194 Datta, N., 40 Dauvalder, M., 15, 40 Davidson, V. L., 104,172 Davies, D. D., 101, 172 Davis, R. F., 146, 151, 172 Dayanadan, 8,39 De Boer, A. H., 146,147,148,149,172 De Luca, L., 132,172,177 de Luca, P., 66,86 De Niro, M. J., 90, 91 De Quintero, M. R., 133,172 de Sloover, J. L., 191,194 Deacon, J. W., 138,172 Debnam, E. S . , 120,172 DeFelice, M., 174

D

AUTHOR INDEX

DeGreef, J. A., 41 Deichgraber, G., 195 Deleens, E., 57,63,86 Delgado, M., 59, 66, 88 Delrot, S., 120, 172 DeMichelis, M. I., 40, 175 Dennerll, T.,38 Derksen, J., 30,38 Diaz, M., 46,75,86,88 Dickinson, H. G., 181, 182, 183, 185, 186, 188, 190,194,195,196 Dieter, P., 14, 15, 28,38, 41 Dietrich, A. J. J., 195 Dohrmann, V., 112,172 Dolk, H. E., 8, 38 Doll, S., 120, I78 Doyle, J., 191, 295 Drew, M. C., 138, 139, 140, 161, 162, 164, 166,172, 173 Duggan, K., 39 Duniec, J., 193, 195,196 Dunlop, J., 149, 154, 173 Dupler, A. W., 191,195 DuPont, F. M., 109,173

E Eades, C. A., 88 Earnshaw, M. J., 49, 74, 86,87 Echlin, P., 142,173 Edwards, D. G., 154,174 Edwards, G. E., 44,87, 90, 91 Edwards, K. L., 20, 33,34,38 Ehrenstein, G., 176 Ehwald, R., 176 Eickmeier, W. G., 70, 90 Einstein, A., 27, 38 Eller, B. M., 54, 58,65,78,84, 87, 90, 91 Elson, E. L., 40 Elzam, 0. E., 96,173 Epstein, E., 95, 96, 121, 124,160,173, 175 Epstein, W., 120, 173,177 Erdtman, G., 187,195 Erwee, M. G., 143,173 Etherington, J. R., 87 Evanari, M., 91 Evans, J. K., 56, 63, 87 Evans, M. L., 9, 10, 15,39, 40, 41 Evenson, D. P., 194 Evers, J., 9,38

199

F Falke, L., 33,38 Faller, L., 106,173 Farquhar, G. D., 55, 56, 87 Fechheimer, M., 40 Fein, A., 13, 40 Felle, H., 158, 173 Fernandez, J. M., I77 Fernando, M., 155,173 Ferroni, A., 176 Fioretto, A., 91 Firth, P. M., 89 Fischer, A., 48, 87 Fischer, E., 120, 173 Fischer, K., 88 Fisher, J., 105,173 Fleurat-Lessard, P., 11, 40 Flewelling, R. F., 39 Fondren, W. M., 40 Fourcy, A., 173 FOX,A. P., 41 Franceschi, V. R., I74 French, A., 29,30,39 Friedman, W. E., 86 Friedrich, U., 23,27,31,39 Friemert, V., 48,53,87,89 Fu, C. L., 62,87 G Galston, A. W., 21, 39 Garcia, V., 91 Gardetto, P. E., 62, 87 Garnier-Dadart, J., 63,86 Garrec, J.-P., 173 Gaynor, J. J., 21,39 Geiger, B., 28, 39 Gerbaud, A., 86 Gerritsen, A. F. C., 178 Giannini, J. L., 110, 173 Gibbs, M., 45, 49, 86, 90 Gifford, E. M. Jr, 191, 195 Gildensoph, L. H., 173 Giminez-Martin, G., 175 Giorgi, D. L., I73 Glass, A. D. M., 99, 102, 103, 113, 114, 116, 143, 144, 151, 152, 153, 154, 155, 158, 159, 160,163, 164, 173,176,177 Gogarten-Boekels, M., 171 Gogarton, J. P., 171 Goh, C. J., 86 Goldman, D. E., 104,173

200

AUTHOR INDEX

Goodwin, P. B., 143,173 Gordon, S. A., 40 Gorham, J., 87 Goring, H., 176 Gradmann, D., 9,38, 40, 140,171 Grahm, L., 9, 11,39 Greenway, H., 101, 166,173, 176 Griffiths, H., 43-92 Griffiths, H. J., 20, 39 Griffiths, N. M . , 90 Grignon, C., 178 Groenhof, A. C., 70,87 Gronewald, J. W., 131, 132,172, 173 Grusak, M. A , , 176 Guardiola, J., 174 Guerrier, D., 86 Guharay, F., 33,39 Gullasch, J., 175 Gunning, B. E. S., 140, 176,196 Guralnick, L. J., 72,73, 74, 75, 76,87 Gurevitch, J., 90 Gysi, J., 91

H Haberlandt, G., 16, 39 Hagen, C. E., 95,173 Hagiwara, S., 123, 124, 178 Handley, R., 174 Hanscom, Z., 75, 90 Hansen, U.-P., 161, 177 Hanson, J. B., 99, 101, 102, 103, 104, 105. 113. 117. 131. 132. 133. 136. 14-51 146, 149; 150; 165; 172: 173; 175, 178 Harris, G. G., 7, 39 Hart, J. W., 7, 39 Hartsock, T. L., 44,46,50,52,59, 66, 67,70,75, 81,87,89 Harvey, D. M. R., 142, 173 Hasenstein, K. H., 9, 10, 39 Hayes, T. L., 173 Haynes, L. W., 41 Heathcote, D. G., 16, 27, 29, 30,39 Hedrich, R., 177 Heidemann, S. R., 38 Hendricks, S. B., 100,174 Hensel, W., 2, 30,39, 40 Herrera, A., 89 Hertel, R., 23, 27, 31,39, 172, 177 Hertz, C. H., 9,39 Heslop Harrison, J., 183, 195, I96 Heuer, S., 89, 90

Heun, A . - M . , 68, 87 Hew, C. S., 62,69,87,88 Heyting, C., I95 Hiatt, A. J., 99, 100, 101,174 Higinbotham, N., 146, 149, 151, 171, 172 Hill, R. D., 88 Hille, B., 31, 39 Hillman, S. K., 21,39 Hinchman, R. R., 17,27, 40 Ho, K. K., 87 Hoagland, D. R., 94, 100, 174 Hodcent, E., 182,195 Hodges, T. K., 105, 106, 108,173,174, 178 Hodgkin, A. C., 104,174 Hoffmann, B., 171 Holtum, J. A. M., 57, 63, 64,87, 91 Honig, B. H., 28,39 Hooymans, J. J. M., 143, 144,174 Hopkins, J. W., 120,177 Hopper, M. J., 178 Honvitz, A., 28, 39 Hotchkiss, C. W., 105, 108, 275 Hoyt, R. C., 9,38 Hrubec, T. C., 132, 133, 136,172, 175 Hubbell, W. L., 39 Hubick, K. T., 91 Hudspeth, A. J., 13, 14, 18,26,38,39 Huxley, A . F., 104,174 Hylmo, B., 151,174 I Iaccarino, M., 120, 174 Ichino, K., 176 InouC, S . , 191, 195 Irvine, R. F., 133,171 Iversen, T.-H., 16, 29, 32, 39, 40 Iwasa, K., 176

J

Jackson, M. B., 11,39 Jackson, P. C., 99, 154, I74 Jackson, R., 173 Jacobsen, L., 95, 99, 100, 174 Jacobson, L.,100,174 Jacobson, S. L., 8,38 Jacoby, B., 100, 101, 129,174 Jaffe, L. F., 9, 39 Jalil, J’., 91 Jenkins, G. I., 16,39 Jenkins, W., 173

AUTHOR INDEX

Jensen, P., 153, 154, 157, 158, 164, 174, 176,177 Jeschke, W. D., 143, 145, 164, 165, 166, 167, 168,174, I78 Johansen, C., 152, I74 Johnson, H. B . , 90 Johnsson, A . , 6, 7,39 Joolen, M. L., I78 Juniper, B . E., 29, 30,39, 140,174 K Kacmarek, L. K., 41 Kakimoto, T., 30,39 Kannan, S., 120, 174 Katou, K., I76 Katz, B . , 104, 174 Kaufman, P. B., 8 , 3 9 Keeley, J. E., 44, 46, 69, 82, 83, 87 Keenan, T. W., 174 Keifer, D. W., 123, I74 Keith, C. H . , 28,39 Kemp, P. R., 62, 87 Kenyon, W. H . , 86 Kerby, N. W., 57, 87 Ketchum, K. A., 124, 174 King, H . , I74 Kirkby, E. A . , 166, I71 Kluge, M., 44, 46, 48, 49, 52,53, 68, 69, 70, 72, 75, 76, 84, 86, 87, 88, 89, 90 Knaff, D. B . , 104, 172 Knauf, P. A., 172 Kochian, L. V . , 93-178 Kohler, K., I75 Kolb, H . - A . , 114, 124, 175 Komor, E., 120, 126, 127, 135, I75 Koppel, D. E., 40 Koster, S . , 82, 88 Kostina, L., 35,39 Kramer, D., 175 Ku, J. B . , 90 Ku, M. S. B . , 48, 53, 88 Kuras, M., I94 Kurmann, M. H., 186, 187, 188, I95 Kutschera, U., 7,39

L La Croix, L. J., 88 Lado, P., 113,171, 175, I76 LaFayette, P. R., 172 Lai, C., 173 Laimins, L., 120,173

201

Lange, 0. L., 51, 53, 81,88 Larcher, W., 72, 74, 76, 91 Larsen, P . , 7, 27, 32,39 Laties, G. G., 96,99, 100,101,175, I78 Lauchli, A . , 91, 101, 137, 141, 142, 145, 151,175, I77 Laurinavichyus, R. S., 39 Lawton, J. R., 28,39 Layzell, D. B., 176 Leavenworth, C. S . , 89 Lederer, W. J., 41 Lee, H. S. J., 44,45,48,51,53,55,61, 63, 65, 66,68, 69, 70,72,76, 77, 78, 81,87,88 Lee, J . A . , 86 Lee, J. S . , 9, 10, 11,30,39 Leggett, J. E., 95, 175 Leigh, R. A . , 151, 152, 155, 156, I75 Leonard, R. T., 99, 101, 105, 106, 107, 108, 131, 171,174, 175 Leopold, A . C., 9, 11, 15,25, 29, 30, 38, 40 Levin, R . J., 120, I72 Liang, M., 120, 172 Lichtner, F. T., 120, I75 Lin, C.-T., 15,39 Lin, W . , 99, 102, 103, 104, 113, 123, 129, 130, 131, 132, 133, 134, 135, 136, 175, 176 Linthilhac, P. M., 193, I95 Littlejohn, R. O., 48, 53, 88 Lloyd, C. W . , 188,195 Lohmann, K. J., 9, 32, 39 Lonergan, J. F., I74 Long, R. C., 149,172 Looby, W. J., 191,195 Lopez-Saez, J. F., 143, I75 Lord, E. M., 87, 91 Lowendorf, H. S., 127, 175 Lubbers, A . E., 88 Lucas, W. J., 93-178 Lund, E. J., 9,38 Lundborg, T., 30,39 Luster, D. J., 132, 175 Luthen, H . , 132, 136,171 Luttge, U., 44, 45, 48, 49, 50, 51, 52, 53, 54, 58,64,65, 67, 68,70, 71, 72, 74, 75, 78, 81, 82, 84, 85,87, 88, 89, 90,91, 120, 129,173, 175

M McClelen, C. E., 17, 40

202

AUTHOR INDEX

McCully, M. E., 137, 142, 145, 151, 183, 176,195 McDaniel, C. N., 120, 171 Macdonald, I. R., 7 , 3 9 McFadden, J. J., 40 Mackenzie, A., 183,195 MacLaren, I., 91 MacLean-Fletcher, S. D., 22, 23,39 McNeil, D. L., 176 McNeil, P. H., 91 Macri, F., 132, 175 McWilliams, E., 61, 73, 76, 88 Madore, M. A., 143,175 Madsen, T. V., 69, 82,88 Maheshwari, P., 182, 191,195 Malamed, M. L., 194 Malinowska, D., 173 Maloney, P. C., 120,175 Mandala, S., 176 Mangin, L., 183,195 Marban, E., 41 Marcum, H., 40 Maretzki, A., 130, 131,175,178 Marigo, G., 88 Marme, D., 41 Marre, E., 99, 101, 104, 109, 111, 114, 171,172,175,176, 178 Martin, C. E., 52,61,65,66,73,74, 76, 81,88 Martinoia, E., 175 Massel, M., 133,171 Mathews, G., 41 Maxfield, F. R., 39 Maynard, J. M., 120, 127,176 Mayoral, M. L., 91 Medina, E., 46, 51, 57, 59,61,63, 65, 66,70,73,75,76,81,86,87,88,91

Medina, J. D., 88 Memon, A. R., 152, 155,176 Merkis, A. I., 35,39 Mertens, R., 8, 9, 11, 39 Methfessel, C., 14, 39 Mettler, I. J., 109, 176 Meyer, C. P., 88 Michaelis, G., 124, 176 Mikulska, E., 182,195 Mishke, C., 176 Misler, S., 38 Misra, P. C., 129, 135,176 Mitchell, P., 99, 102, 176 Moens, P. B., 183,195 Moitra, A., 181, 185, 189,195

Molder, M., 184,195 Mdler, I. M., 30,39, 129, 176 Montes, G., 89 Mook, W. G., 57,88 Mooney, H. A., 73,88 Moore, R., 9, 17,35, 40 Moran, N., 124,176 Morrison, G. F., 38 Morse, M. J., 133, 176 Mukidjam, E., 173 Mulkey, T. J., 39 Munck, B. G., 120, I76 Muniz, R., 91

N Nakatani, K., 41 Nalborczyk, E., 57, 63,88 Neales, I. F., 44,60,62, 66, 67, 68,88, 90 Neher, E., 177 Newman, I. A., 98, 99, 100, 113, 115, 117, 118, 140, 154, 160, 175,176 Ng, Y. W., 87 Niedergang-Kamien, E., 11, 40 Nishio, J. N., 78, 89 Nishizuka, Y., 133, 176 Nissen, O., 96,176 Nissen, P., 176 Nobel, P. S., 44,46, 50,52,53, 54, 55, 59, 61, 66, 67,70, 75, 79, 81, 86, 87, 88, 89, 90, 92 Noh, Hj., 91 Noordervliet, M. A. W., 178 Nott, D. L., 89 Nuccitelli, R., 9, 39 0 Okamoto, H., 146, 149, 176 O’Leary, M. H., 55, 57, 63,86, 87, 89 Olivares, E., 74, 88,89 Olsen, G. M., 29, 40 O’Neill, S. D., 109, 176 Ong, B. L., 58, 66, 84, 89 Oostendorp, T., 38 Ordin, L., 100,174 Oross, J. W., 175,178 Osborn, M., 196 Osborne, D . J., 12, 41 Osmond, C. B., 44, 45, 46,49,50, 51, 52, 57, 59,63, 65, 66, 68, 74, 79, 81,85,86,87,88, 89, 91, 92, 101, 176

AUTHOR INDEX

Osterhout, W. J. V., 95, 176 Ottley, A. M., 191, 195 Overall, R. L., 140, 143, 176 Overstreet, R., 95, 174 Owens, J. N., 184,195 Ozanne, P. J., 152, 171 P Parthasarathy, M. V., 23,27,30,40,41 Passioura, J. B., 165,176 Pate, J. S., 167, 176 Payne, R., 13, 40 Pearman, G. T., 55,87,89 Pennell, R. I., 179-196 Perbal, G., 25, 30, 31, 40 Pesci, P., 112,172,176,178 Peterson, G., 86 Petraglia, T., 101, I76 Pettersson, S., 153, 154, 157, 158, 164, 174,176, 177 Pfeffer, W., 36, 40 Pfliiger, R., 101, 151,175 Pickard, B. G., 8,9, 11, 16, 17,20, 27, 32, 33, 34, 37,38, 40 Pilet, P. E., 30, 40 Pitman, M. G., 99, 101, 111, 112, 113, 114, 124, 142, 143, 145, 160, 164, 166, 173,175, 177, 178 Pitner, R. A . , 88 Pollard, T. D., 22, 23, 39 Polley, L. D., 120,177 Poole, R. J., 99, 101, 102, 107, 113, 156,171,174, 176, 177 Poovaiah, B. W., 15, 40 Popp, M., 74,87, 88 Potter, U., 185,195 Powles, S. B., 89 Prins, H. B. A., 149, 172 Pucher, G. W., 89 Pugliarello, M. C., 40 Pupillo, P., 132,172, 177

Q

Quader, H., 188, 195 Queiroz, O., 63, 68,86

R Raamsdonk, W., 195 Rabon, E., I73 Rains, D. W., 96, 173 Ramos, J., 100, 117, 161,177 Randazzo, G . , 172

203

Ranson, S. L., 45,48, 49, 83, 84,89 Raschke, K., 177 Rasi-Caldogno, F., 11, 40,176 Ratan, R., 39 Raven, J. A , , 50,55, 57,69,82,86,87, 89, 90, 101,177, I78 Rayder, L., 46, 58,60,66,68, 71, 75, 78, 89, 91 Rayle, D. L., 10, 38 Reddy, A. S. M., 40 Reed, M. L., 90 Reese, T. S., 195 Reilly, E. J., 138,171 Reinecke, D., 38 Reisenauer, H. M., 127,177 Reising, B . , I74 Renfrow, A., 70,86 Reznickova, S. A., 182, 195 Rhoads, D. B., 120,177 Richards, R. A , , 56, 87 Richardson, K., 46, 69, 82, 83, 90 Rincon, M., 133,177 Risueno, M. C., 175 Ritz, D., 49, 64, 90 Robards, A. W., 140, 141, 143, 172, 177,178 Robe, 69, 82, 83 Roblin, G., 11, 40 Rockwell, M. A., 23, 40 Rodgers, S. A., I78 Rodkiewicz, I94 Rodriguez-Navarro, A., 100, 117, 118, 119, 161,177,178 Roeske, C., 87 Rohr, R., 186, 195 Roksandic, Z., 91 Rooney, M., 89 Rorabaugh, P. A., 40 Rorabough, P. H., 87 Rothstein, A., 172 Rotter, M., 175 Roux, S. J., 9, 10, 15, 40 Rowley, J. R., 181,195 Rubery, P. A., 11, 40 Rubinstein, B., 130, 131, 132, 136,177 Rudich, B., 129, I74 Ruess, B. R., 54,58, 65,78, 84,87, 90 Ruiz-Cristin, J., 173 Rundel, D. W., 59,90 Russell, S. D., 193,195 Russo, G., 91 Rygol, J., 53, 54, 90

204

AUTHOR INDEX

S Saccomani, G., 173 Saccomani, M., 176 Sachs, F., 33,39 Sachs, G., I73 Sack, F. D., 22, 25,26, 28, 29, 30, 40 Saker, L. R., 138, 140, 161, 162, 164, 166, 173 Sakmann, B., 14,39 Sale, P. J. M., 44, 60, 66, 67, 88, 90 Salisbury, F. B., 10, 35, 40 Sandelius, A , , 133, 177 Sanders, D., 125, 126, 127, 129, 156, 161,177 Sarafian, V., 156,177 Satter, R., 176 Scacchi, A., 172 Scarborough, G. A., 127,177 Schaefer, N., 145,177 Schafer, C., 70, 75, 87, 90 Schafer, C., 72,88 Schlessinger, J., 40 Schmitt, M. R., 90 Schnabl, H., 101,177 Schnapp, B. J., 195 Schneider, C. L., 7, 41 Schnepf, E., 195 Schopfer, P., 39 Schramm, M. J., 70, 91 Schrank, A. R., 9, 40 Schroder, H., 21,40 Schroeder, J. I., 114, 124, 177 Schroppel-Maier, G., 91 Schroter, K., 21, 31, 40 Schuber, M., 68, 69, 70,72,76, 84,90 Schulte, P. J., 90 Schultz, S. G., 120, 176 Schulze, A . , 38 Schumaker, K. S., 156, 177 Seagull, R. W., 196 Segel, I. H., 154,177 Serrano, R., 101, 107, 108, 109, 110, 177, I78 Shaff, J. E., 175 Sharkey, T. D., 87 Sharpless, T. P., I94 Shaw, S., 8, 40 Sheetz, M. P., 195 Sheffield, E., 183, 184, 195 Shelanski, M. L., 39 Sheldon, J. M., 187, 188, 195 Sheldrake, A. R., 11, 40

Shen-Miller, J., 7, 17, 27, 40 Shibaoka, H., 30,39 Shone, M. G. T., 146,177 Shrier, A., 174 Shvyagzhdene, D. V., 39 Siddiqi, M. Y., 99, 102, 103, 113, 114, 116, 151, 152, 153, 154, 155, 158, 159, 163, 164, 173,177 Sievers, A., 11,21,28, 29, 30, 31,38, 40, 41, 191, 195 Simmons, R. A., 87 Sims, A. P., 153,178 Sinclair, R., 70, 90 Singh, H., 184, 195 Sipes, D. L., 68, 73, 74, 90, 91 Slayman, C. L., 100, 117, 118, 161, 171,175, 177, 178 Slayman, C. W., 175 Sliwinski, J. E., 35, 40 Slocum, R. D., 9, 10, 40 Sluiter, E., 175 Smith, F. A., 50, 90, 101, 177, 178 Smith, J. A. C., 44, 46,47, 49,50, 51, 52,53, 54,55,60, 61, 67, 70, 73, 76, 78, 82, 84, 87,88, 90 Smith, M. M., 183, 195 Smith, S. D., 86 Smith, T. L., 70, 90 Smolka, A , , 173 Soler, A., 178 Somerville, C., 38 Sommarin, M., 133,177 Song, G.-X., 39 Spalding, M. H., 48, 53, 90 Spanswick, R. M., 101, 107, 108, 109, 120, 171,173, 175,176,178 Speirs, J., 178 Spruyt, E., 41 Spurr, A. R., 175 Staiano-Coico, L., 194 Starzak, M. E., 25, 30, 40 Staverman, W. H., 88 Steive, H., 13, 40 Stelter, W., 174 Stelzer, R., 177 Steponkus, P. L., 33, 34, 41 Sterberg, L., 91 Stern, A. I., 130, 131, 136,177 Sternberg, L. O., 58,60,62,64,71,73, 86, 90, 91 Steudle, E., 33, 53, 54, 38, 90 Stewart, K. D., 191, 195

AUTHOR INDEX

205

Uribe, E. G., 90 Stichler, W., 89, 90, 92 Urich, R., 89 Stimmel, K . - H . , 68, 87,88, 90 Stinemetz, C. L., 15, 40, 41 Stocker, G. C., 91 V Stout, R. G., 109, 112,178 Vale, R. D., 191, 195 Strain, B. R., 88 Valenti, V., I77 Strong, J. A . , 16, 41 Valentin, P., 41 Sun, D., 39 Van Bel, A. J. E., 120, 173,178 Sutcliffe, J . F., 183,194 Van and Chen, Q., 195 Suyemoto, M. M., 40 VanIren, F., 137, 138, 143, 144, 178 Szarek, S. R., 45, 52,62,63,66, 68, 90 Van Pinxteren-Bazuine, A., 178 Sze, H., 101, 109, 156, 172, 177, 178 Van Steveninck, R. F. M., 176,178 Vara, F., 107, 109, 110, 178 Vasil, I. K . , 185, 194 T Vaulina, E., 39 Taiz, L., 176 Vazquez-Yanes, C., 74, 91 Tanada, T., 9, 41 Veith, H. J., 90 Tanner, W., 126, 127, 175 Verbelen, J.-P., 7, 16, 20, 41 Tasaki, I . , 123, 124, I78 Vianello, A , , 132,175 Tatum, E. L., 161,177 Vickery, H. B . , 89 Taylor, D. L., 40 Vinten-Johansen, C., 9, 41 Teeri, J. A., 55,63,68,70,78,88, 90 Virzo de Santo, A., 74, 86, 91 Tenhunen, J. D., 53, 91 Volkmann, D., 11, 28, 29, 30, 41, 191, Terry, B. R., 143, I78 195 Thibaud, T.-B., 105,178 Thimann, K. V., 7, 16, 17, 32, 40, 41 Thorn, M., 130, 131,175,178 W Thomas, D. A., 86 Wagner, G., 15, 18, 41 Thomas, D. R., 64, 84, 85, 90, 91, 92 Wagner, J., 72, 74,76, 91 Thomas, M., 45,48, 49, 83,84,89 Walek-Czernecka, A , , 195 Thornley, W. R., 107,172 Walker, D. A., 44,87 Ting, I. P., 44, 45,46, 49, 50, 52, 53, Walker, D. B., 63, 91 58, 60,64, 65,66, 67,68, 71,72, Walker, N. A . , 96, 143, 178 73, 74, 75, 76, 78, 84,86, 87, 89, Wallace, B. J., 91 90, 91 Walles, B., 181, 195 Tinoco-Ojanguren, C., 74, 91 Wang, C.-L., 40 Tognoli, L., 112, 176,178 Waters, F. B., 177 Tomlinson, P. B., 49, 91 Wayne, R. O., 40 Tonser, S. J., 90 Webb, W. W., 40 Torii, K . , 99, 100, 178 Weber, K . , 196 Trass, J. A., 38 Weiler, E. W., 8, 9, 11,39 Treichel, I., 86, 89 Weir, W. G., 41 Trewavas, A . , 11, 41 Weisenseel, M. H . , 38 Trimborn, P., 92 Weiss, S . , 38 Trimborn, W., 89 Wells, B., 188, I95 Trotta, A . , 99, 112,171 White, R., 40 Troughton, J. H . , 46,61,62,63,88,90, Wick, S. M., 183, 193,195, 196 91 Wickens, L. K . , 164, 165, 168, 172 Tsein, R. W., 41 Wild, A . , 178 Turner, M., 90 Wildes, R. A . , 177 Wilkins, M. B., 8, 9, 11, 21, 39, 40, 41 U Willemse, 185, 191,196 Ulrich, A . , 99, 100,114, 178 Willenbrink, J., 120, 178

206

AUTHOR INDEX

Willert, D. J. von, 44, 46, 68, 70, 71, 75,84,86, 91 Williams, E., 183, 185, 196 Williams, J. G., 92 Williams, L . , 172 Williams, S. E., 8, 41 Willows, A. 0. D., 9, 32,39 Wilson, C., 120, 178 Wilson, T. H., 120,175 Winter, E., 91 Winter, K., 44, 45, 46, 48, 49, 50, 52, 53,58,59,60, 62,63,65, 66, 68, 69,70,71,74,75,78,79, 82,88, 89, 90, 91 Witztum, A., 23, 27,41 Wolf, J., 83, 84, 92 Wolf, O., 167, 168,178 Wolfe, J., 33, 34, 41,176 Wong, S. C., 87 Wood, C., 64, 84, 85, 90,91, 92 Wood, J. M., 120,178 Wood, J . W., 28,38 Woodcock, C. L. F., 193,196 Woodcock, A. E. R., 9, 11,41 Woodhouse, P. J . , 178

Woodhouse, R. M., 61, 92 Wright, M., 12, 41 WU, J.-Y., 39 Wyn Jones, R. G., 87,96, 151, 152, 155, 156,175,178

X Xin-Zhi, J . , 174

Y Yau, K.-M., 13,41 Yeoh, H. H., 87 Yong, B. C. S., 91 Young, M., 153,178 Z

Zanstra, P. E., 172 Zee, A. K., 73,74,88 Zeiger, E., 50, 92 Ziegler, H., 64,89, 90, 92 Zimmermann, U., 33,38, 90 Zocchi, G., 131, 132, 133,178 Zolincrowicz, W., 195 Zuber, M., 53,88

SUBJECT INDEX

A Abscisic acid and potassium transport, 164 Acid metabolism, see Crassulacean acid metabolism Action potentials, 8-9 and gravitropism, 9 Activation energy of gravity perception, 5 Aerenchyma induction and potassium uptake, 138-140 Amyloplasts, see Statolith gravity sensors Anions, and linear potassium uptake, 121-123 ATP and proton pump, 109 ATPase in protodpotassium transport, 105-1 11, 117-1 18 Auxin and gravitropism polar transport, 11-12 translocation of anion, 9

B Bromeliads, carbon dioxide recycling in, 79-81 C Calcium ions and eye light adaptation, 13 in gravity perception, 14-15, 37 translocation, 9, 10-11 Calmodulin and gravity, 15-16 Carbon dioxide recycling and acid metabolism citric vs. malic acids, 83-85 and environment aquatic, 82-83 terrestrial, 79-82 and intermediate CAM variants, 74-76 regulation, 64-67

207

Carbon isotope studies in acid metabolism, 55-64 Carnitine and citrate acid metabolism, 84-85 Chloride ion and potassium uptake, 121-123 Citrate metabolism vs. malate, 83-85 carnitine and, 84-85 Cochlear, 13-14 Conifers, sporogenesis in, 179-196 heterospory, 180 megasporogenesis, 190-194 megaspore viability, 193-194 meiosis, 191-193 reproductive cell origin, 190-191 in pollen-bearing cone, 181-190 archaesporium development, 181-1 82 exine patterning, 185-190 meiosis, 184-185 sporogenous cellsltapetum, 182-183 process of, 180 Crassulacean acid metabolism (CAM), 43-92 carbon dioxide recycling, 78-85 aquatic environment, 82-83 citric vs. malic acids, 83-85 regulation, 64-67 terrestrial environment, 79-82 die1 cycle, 46-50 organic acid fluctuation, 49-50 phases in, 46-49 and succulence, 49 intermediate variants, 67-69 carbon dioxide recycling, 74-76 occurrencelevolution, 70-74 physiology of transmission, 76-78 isotope ratio analysis, 55-64 modification, environmental, 51-53 and photosynthetic radiation, 52

208

SUBJECT INDEX

CAM, modification-Cont. and temperature, 52-53 and water stress, 52 occurrence/distribution, 45-46 plant water relations, 53-55 regulation, biochemical, 50-51 malic acid accumulation, 51 phosphoenolpyruvate carboxylase in, 50-51 research integration, 85-86 interest, 4 4 4 5 Cytoskeleton displacement and gravity, 27-28 shear and statolith motion, 21-24

D Die1 cycle in Crassulacean acid metabolism, 46-50 and organic acid fluctuation, 49-50 phases in, 46-49 and succulence, 49 Diethylstilbestrol and ion flux in roots, 104 Drought and Crassulacean acid metabolism. 52

E Ear, signal transduction in, 13-14 Electrophoresis in gravitropism, 9-10 Endoplasmic reticulum compartmentation and potassium transport, 141-143 displacement and gravity, 28-29 Epiphytes, carbon dioxide recycling in, 75,75-76. 79-81 Exine patterning in sporogenesis, 185-1 90 mechanism, 187-190 the process, 185-187 Eyes, signal transduction in, 13

F Fusicoccin and ion flux in roots, 104, 111-1 12 G Gravitropism, 141 definitions, 2 perception, 12-37 multiple systems, 16-17 non-statolith, 32-36

signal transduction, 1-2, 12-16 statolith sensors, 16, 18-32, 37 susception, 3-7 mechanism of sensing, 3-4 and thermal noise, 4-7 transmission, 7-12 chemical, 10-12 electrical, 8-10 Growth regulators auxin anion translocation, 9 pumping in gravitropism, 11-12 see also Phytohormones and potassium transport

H Heterospory in conifers, 180 Hormones and potassium transport, 164 Hydrogen ion/potassium transport in roots, 99-119 charge balance maintenance, 100-101 co-transport, 115-119 and hydrogen ATPase, 117-118 hydrogen/potassium independence, 115-116 membrane potential and potassium flux, 116-117 and pH, 119 dependence, inter-ion, 99 direct coupling, 101-111 active system, 104 and ATPase activity, 105-111 electrogenic system, 104-105 passive system, 103 and pH, 102 indirect coupling, 111-115 ‘apparent’/‘true’ proton flux, 113 charge balance constraints, 114 and fusicoccin stimulation, 111-112 and potassium channels, 114-115 see also Potassium channels; Proton Pump I ‘Idling’ in Crassulacean acid metabolism, 52 Ions anions, and potassium uptake, 121-123 channels, tension sensitive, and gravity, 33-34 pumping in gravitropism, 10-11

SUBJECT INDEX

see also Electrophoresis in gravitropism Isoetids, carbon dioxide recycling, 83 Isotope studies in Crassulacean acid metabolism, 55-64 lipid metabolism, 63-64

L Ligand binding in statolith location, 30 Lipid metabolism, isotope ratio studies of, 63-64 Littorella unifiora, carbon dioxide recycling, 82-83 M Macrophytes, aquatic, acid metabolism in, 69 Malic acid, see Crassulacean acid metabolism Meiosis in conifers, 180 female, 191-193 male, 184-185 in flowering plants, 183 Membrane tension and gravity sensing, 33-35 Mesembryanthemum crystallinum, acid metabolism in, 68 and carbon dioxide recycling, 75 Microfluorometry and conifer sporogenesis, 182-183 Mitochondria sedimentation, 20 Mitosis and spaceflight, 35 N NADH and potassium transport exogenous, and potassium ion influx, 130-131 integrated model, 132-136 and wound response, 131-132 0 Orchids, Crassulacean acid metabolism in, 69 Osmotic pressure and Crassulacean acid metabolism, 53-54 Oxygen isotope studies of respiration, 64

P Pereskia spp., carbon dioxide recycling in, 75

209

pH and protonlpotassium transport co-transport, 119 direct coupling, 102 Phloem, potassium recirculation in, 165-167 Phosphoenol pyruvate carboxylase in acid metabolism, 50-51 Phosphoinositides in gravitropism, 15-16 Photoinhibition and carbon dioxide recycling, 81 Phytohormones and potassium transport, 164 Plasmalemma in root potassium transport, 94-129 carrier-kinetic approach, 94-99 dual isotherm hypothesis, 95-96 limitations, 98-99 ‘low-salt’ roots, 94-95 and hydrogen ion transport, 99-119 charge balance maintenance, 100-101 co-transport, 115-119 dependence, inter-ion, 99 direct coupling, 101-1 11 indirect coupling, 11 1-1 15 linear uptake and anions, 121-123 physiological role, 127-128 and potassium channels, 123-124 reaction kinetic model, 125-127 and root salt status, 124-125 redox-coupled, 129-136 and exogenous NADH, 130-131 integrated NADH model, 132-136 and wound response and NADH, 131-132 Plasmodesmata and root potassium uptake, 140-141 Pollen, see Meiosis Potassium channels, 114-115 and linear uptake, 123-124 and medium concentration, 124 and tetraethylammonium chloride, 123-124 Potassium transport in roots, 93-178 plasmalemma in, 94-129 carrier-kinetic approach, 94-99 and hydrogen ion transport, 99-119 linear uptake component, 120-128 redox-coupled, 129-136

210

SUBJECT INDEX

Potassium transport, redox coupledCont. and exogenous NADH, 130-131 integrated NADH model, 132-136 and wound response, 131-132 regulation, 151-152 ‘bulk’ levels, 151-152 hierarchy of, 152 reaction kinetic model, 161 sites, 152 by transport to xylem, 161-163 regulation, allosteric, 152-161 carrier synthesisldegradation, 154-155 criticism of, 156-161 feedback, 152-154 tonoplast fluxes, 155-156 regulation, multicomponent, 163-168 and environmental factors, 164 feedback control in, 164 and growth factors, 165 phloem recirculation, 165-167 and transpiration, 165 research directions, 169-170 to xylem, 136-151 electrogenic mechanisms, 145-151 lag phase in, 143-145 radial pathway, 140-143 site of entry, 137-140 Proton pump, 10 ATP dependent, 109 Protoplasm weight and gravity sensing, 32

R Rhodopsin activation, 5, 13 Ribosome cycle in meiosis, 183 RNA in conifer sporogenesis meiosis, 184 cell development, 183 Roots, gravity sensing by, 36 see also Potassium transport in roots S

Sedum spp., carbon dioxide recycling in, 76 Sedum telephium, Crassulacean acid metabolism in, 76-78 Sempervivum spp., carbon dioxide recycling in, 76 Signal transduction, 1-2, 12-16

in ears, 13-14 in eyes, 13 in gravitropism, calcium ions in, 9, 14-15, 37 phosphoinositides in, 15-16 Spaceflight and mitosis, 35 Starch depletion and gravity sensing, 16 see also Statolith gravity sensors Statolith gravity sensors, 16, 18-32, 37 action, 21 displacement, 26-29 cytoskeleton stretching, 27-28 of endoplasmic reticulum, 28-29 identifying intracellular, 19-21 intramembrane, 18-19 motion, 21-25 cytoskeletal shear, 21-24 electrical field, 25 kinetic energy, 25 in multiple systems, 17 position, 29-32 chemicaYelectrostatic interactions, 31 and electrostatic attraction, 30 and ligand binding, 30-31 Susception in gravitropism, 3-7 mechanism of sensing, 3 4 and thermal motion, 4-7

T Temperature and Crassulacean acid metabolism, 52-53 Thermal motion and gravity sensing, 4-7 activation energy, 5 presentation time, 6-7 threshold response, 7 Tonoplast in potassium regulation, 155-156 Transmission of gravity perception, 7-12 chemical, 10-12 growth regulator pumping, 11-12 ion pumping, 10-11 electrical, 8-10 action potentials, 8-9 electrophoresis, 9-10 gradient. 9 Transpiration and potassium flux, 165 Turgor pressure sensing, 32-33

SUBJECT INDEX

V Vacuoles, see Tonoplast in potassium regulation W Water stress and Crassulacean acid metabolism, 52 Wound response, membrane transport and, 131-133

X X-ray microprobe for potassium in roots, 141-142

211

Xylem, root, potassium transport to, 136-151 electrogenic mechanisms, 145-151 lag phase in, 143-145 and potassium regulation, 161-163 radial pathway, 140-143 symplasmic, 140-141 and endoplasmic reticulum cornpartentation, 141-143 site of entry, 137-140 and aerenchyma induction, 138-140 epidermislcortex, 137-138.

This Page Intentionally Left Blank