Proc. Nati. Acad. Sci. USAVol. 87, pp. 9305-9309, December 1990Botany

Repetitive increases in cytosolic Ca2+ of guard cells by abscisic acidactivation of nonselective Ca2+ permeable channels

(stomata/ion channel/receptor-operated Ca2+ channel/fura-2)

JULIAN I. SCHROEDER* AND SUSUMU HAGIWARAtDepartment of Physiology, Jerry Lewis Neuromuscular Research Center, University of California at Los Angeles School of Medicine, Los Angeles, CA 90024

Communicated by Bernard 0. Phinney, August 29, 1990 (received for review June 28, 1990)

ABSTRACT Many signal-transduction processes in higherplant cells have been suggested to be triggered by signal-induced opening of Ca21 channels in' the plasma membrane.However, direct evidence for activation of plasma-membraneCa21 channels by physiological signals in higher plants has notyet been obtained. In this context, several lines of evidencesuggest that Ca2+ flux into the cytosol of guard cells is a majorfactor in the induction of stomatal closing by abscisic acid(ABA). ABA closes stomatal pores, thereby reducing transpi-rational loss of water by plants under drought conditions. Todirectly investigate initial events in ABA-induced signal trans-duction in guard cells, we devised an experimental approachthat allows simultaneous photometric measurements of cyto-solic Ca2+ and patch-clamp recordings of ion currents acrossthe plasma membrane of single Vicia faba guard cells. Usingthis approach, we found that the resting cytosolic Ca21 con-centration was 0.19 ± 0.09 FM (n = 19). In responsive guardcells, external exposure to ABA produced transient repetitiveincreases in the cytosolic free Ca2+ concentration. These Ca21transients were accompanied by concomitantly occurring in-creases in an inward-directed ion current. Depolarization of themembrane terminated both repetitive elevations in cytosolicCa2+ and inward-directed ion currents, suggesting that ABA-mediated Ca2+ transients were produced by passive influx ofCa2+ from the extracellular space through Ca2+-permeablechannels. Detailed voltage-clamp measurements revealed thatABA-activated ion currents could be reversed by depolariza-tions more positive than -10 mV. Interestingly, reversalpotentials of ABA-induced currents show that these currentsare not highly Ca2+-selective, thereby permitting permeation ofboth Ca2 and K+. These results provide direct evidence forABA activation of Ca2+-permeable ion channels in the plasmamembrane of guard cells. ABA-activated ion channels allowrepetitive elevations in the cytosolic Ca2+ concentration, which,in turn, can modulate cellular responses promoting stomatalclosure.

Elucidation of the molecular mechanisms responsible forstimulus-dependent activation of Ca2l fluxes is central tounderstanding the initial events in signal transduction inhigher plant cells (1). Control of stomatal pore movements bythe physiological growth regulator abscisic acid (ABA) pro-vides an opportune system for the investigation of mecha-nisms underlying Ca2"-dependent signal transduction (2-6).

Stomatal pores permit diffusion of CO2 into leaves forphotosynthetic carbon fixation and the transpiration of watervapor to the atmosphere. This exchange of gases is regulatedby movements ofguard cell pairs that surround each stomatalpore. When water stress develops, the plant growth regulatorABA triggers stomatal closing synergistically with Ca2+ (2,3). Recent detailed investigations of several plant species

have shown that stomatal closing in response to ABA followsa nonuniform behavior such that stomatal pores in distinctsmall patches of the leaf surface area- close in response toABA, whereas stomata in other small areas of the leafepidermis remain open (7-11). This nonuniform ABA re-sponse has been suggested to play a key role in ABA-dependent regulation of carbon fixation (7-11).

Stomatal movements are mediated by changes in the ioncontent of guard cells, which depend on large ion fluxesacross guard-cell membranes (5, 6). Stomatal opening re-quires organic-anion synthesis and K+ uptake (5). Inward-conducting Ca2l-regulated K+ channels represent a majorpathway for K+ uptake (4, 12). Stomatal closing is producedby release of K+ and anions from guard cells (13, 14).ABA-mediated closing of stomata proceeds in a Ca2+-dependent manner (2, 3). The modulation of voltage-dependent anion channels (4, 15) by cytosolic Ca2+ (4) andthe resulting activation of outward-conducting K+ channels(12) have been suggested to provide a molecular basis for theCa2' dependence of ABA-induced stomatal closure (4).Recent research has indicated that ABA can elevate the

cytosolic free Ca2+ concentration ([Ca2+]cyt) in guard cells ofCommelina communis (16). The mechanisms by which[Ca2+]cyt is elevated remain unknown. In the present studywe have attempted to determine the effects of ABA on[Ca2+]cyt in Vicia faba guard cells by adopting a method ofinvestigation that allows the simultaneous monitoring of[Ca2 ]cyt and ion currents across the plasma membrane(plasmalemma) ofguard cells. This approach permits distinc-tion between signal-dependent activation of Ca2+ channels inthe plasma membrane and release of Ca2+ from intracellularorganelles.

METHODSCell Isolation. V. faba plants were grown in a controlled

environment growth chamber (Conviron E15) at 20°C with a12-hr light, 12-hr dark day/night cycle and the use of fluo-rescent and incandescent illumination at a photon fluencerate of 150 ,uE m 2 s- (where E = 1 mol of photons). Guardcell protoplasts were isolated from 2- to 3-week-old V. fabaplants by 60- to 75-min incubation in 1.7% Cellulase OnozukaRS (Yakult Honsha, Tokyo), 1.7% Cellulysin (CalBiochem),0.026% Pectolyase Y-23 (Seishin Pharmaceutical, Tokyo)following a described procedure (17).

Patch Clamp and Solutions. The tight-seal whole-cell con-figuration of the patch-clamp technique (18) was applied toisolated guard-cell protoplasts as described (12, 17). Duringrecordings, cells were bathed in solutions that contained 2

Abbreviations: ABA, abscisic acid; [Ca2+]cyt, cytosolic free Ca2+concentration.*To whom reprint requests should be sent at present address:Department of Biology C-016, University of California at SanDiego, La Jolla, CA 92093.tDeceased April 1, 1989.

9305

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

June

7, 2

021

9306 Botany: Schroeder and Hagiwara

mM MgCl2, 1 mM CaCI2, 5 mM Mes [2-(N-morpholino)ethanesulfonic acid], 1 mM KOH (pH 6.15) and either 10mMpotassium glutamate (17) or 10 mM KCI as indicated. Thecytosol of single guard cells was perfused with a pipettesolution that included 2 mM MgCl2, 5 mM Hepes, 0.1 mMGTP, 1 mM MgATP, pH adjusted to 7.2 with KOH, andeither 100 mM potassium glutamate (17) or 80 mM potassiumglutamate with 20 mM KCl as specified. The Ca2+ indicatordye fura-2 (100 ,uM) (Molecular Probes) or EGTA (100 ,uM)was added to the pipette solution as indicated. All solutionswere adjusted to final osmolalities of 500 mmol-kg-1 byaddition of D-mannitol or D-sorbitol and verification with avapor pressure osmometer (Wescor 5100B). Liquid junctionpotentials were accounted for as described (17). Patch-clamped guard cells were externally perfused with ABA(0.5-5 ,.uM) (98% (+) cis-trans isomer; Sigma, A 4906) eitherby bath perfusion or by short-term local application throughejection from a micropipette positioned in the vicinity of thecell under investigation. Experiments were at 20-22°C.

Photometric Apparatus for Single-Cell [Ca2+],yt Measure-ments. Experiments were performed with a Zeiss IM-35inverted microscope. For epifluorescence excitation, lightfrom a mercury arc lamp (100 W) passed through twoalternating UV interference filters (350 nm DF10 and 380 nmDF10; Omega Optics) mounted on an air turbine-drivenrotating wheel (Biomedical Instrumentation Group, Univer-sity of Pennsylvania). UV light was deflected by a dichroicmirror (Zeiss FT 425) into the microscope objective (NikonFluor 40 x 1.3 oil immersion), which was focused on afura-2-loaded guard cell. Fluorescence from the cell wasobserved through a barrier filter (Zeiss 470 LP) and thenreached the photometer (Thorn EMI RFI/QL-30) by passingthrough a filter combination that allowed broad band trans-mission from 470 to 540 nm (Schott BG 39 and Ditric 540 SPfilters). Cells were viewed during experiments by illumina-tion through a red filter (Schott RG 630) that did not cause anymeasurable increase (

Proc. Natl. Acad. Sci. USA 87 (1990) 9307

[Ca2+]

1 pM

0

2 min ABA

FIG. 2. ABA-induced Ca2+ transient in a V. faba guard cell. Theguard cell was externally exposed to 1 ,uM ABA. Membrane potentialof the cell was held at -40 mV. Intracellular solution contained 100mM potassium glutamate, and extracellular solution included 10 mMpotassium glutamate.

Using whole-cell patch-clamp techniques in conjunctionwith measurements of [Ca2+],y, allows the simultaneousmonitoring of ABA effects on both transmembrane ion cur-rents and [Ca2+]cyt. This approach permits direct identifica-tion of the mechanisms by which [Ca2+]cyt elevations aretriggered. Ca2+ release from intracellular stores may increase[Ca2+]cyt without a concomitantly occurring plasma-membrane Ca2+ current, whereas activation of plasma-membrane Ca2+ channels would increase both [Ca2+]cyt andinward Ca2+ currents.

In the present study, the effects of ABA on [Ca2+J]cytmembrane current, or membrane potential were recorded in37% of the cells studied (n = 62). When studying stomatalmovements in epidermal strips, ABA treatment was ob-served to close approximately one-third of the stomata (J.I. S.and P. Zanirato, unpublished results). In the present studyresponsive guard cells showed characteristic effects as de-scribed below.When guard cells were externally perfused with 1 ,uM

ABA, a rapid and transient increase in [Ca2+]cyt was recorded(Fig. 2). Peak amplitudes of the initial Ca2+ transient ob-served after ABA application ranged from 0.5 to 5 AM.Simultaneous Effects of ABA on Transmembrane Ion Cur-

rents and [Ca2+Jcyt. To study the mechanisms by whichtransient rises in [Ca2+]cyt were mediated, ion currents acrossthe plasma membrane and [Ca2+Jcyt were simultaneouslymeasured. When guard cells were continuously exposed toABA by bath perfusion and the membrane potential (V.m) washeld at -54 mV, repetitive rises in [Ca2+]cyt were observed(Fig. 3). With each rise in [Ca2+Jcyt (Fig. 3; lower trace), aconcomitant rise in inward ion current across the guard-cellplasma membrane was seen (Fig. 3, upper trace). These data

suggest that Ca2l influx through ABA-activated ion channelsmay be responsible for ABA-mediated increases in [Ca2+]cyt.To test the suggestion that rises in [Ca2J]cyt were mediated

by passive ion-channel-mediated flux into the cell, the mem-brane was depolarized to 0 mV (Fig. 3). Depolarizations to 0mV eliminated transient elevations in [Ca2+]cy, and inducedsmall outward currents (Fig. 3). After depolarization to 0 mVfor several minutes, the membrane was repolarized to -54mV. Upon repolarization to -54 mV immediate continuationof the occurrence of transients in [Ca2+]cyt was observed.Each transient Ca2+ elevation was accompanied by a tran-sient increase in an inward ion current (Fig. 3).Subsequently the membrane was polarized for periods

lasting from 1.5 min to 30 sec to +5 mV, followed by -54 mV,+5 mV, and -54 mV (Fig. 3). Each depolarization to +5 mVled to termination of transient elevations in [Ca2+]cy, as wellas cessation of inward currents, whereas each hyperpolar-ization resulted in transient [Ca2+]cyt elevations accompaniedby concomitantly occurring inward currents showing thatCa2+ influx across the plasma membrane contributed toABA-induced Ca2+ transients (Fig. 3).

Nonselective Current Activation by ABA. The Ca2+ equi-librium potential with 1 mM Ca2+ in the external medium and

+20 current

pA 1 IV KJ-20 -

ii

2+i /12 MM [Ca0 --------- ---,)-----------------------------------------------------------------

fABA 1 5 min

-54

membrane potential |Vm=54 V OmV -54mV +5 I-54mV

FIG. 3. Simultaneous recordings ofABA-induced [Ca2+]cyt elevations (lowertrace) and concomitantly occurring inwardion currents (downward deflections in up-per trace) in a V.faba guard cell. The guardcell was exposed continuously to 5 ,AMABA by bath perfusion. ABA-inducedrises in [Ca2+]cyt and inward ion currentsdepended on the imposed membrane po-tential (VIm, indicated at bottom) as de-scribed in text. Internal solution included100 mM potassium glutamate, and externalsolution included 10 mM potassiumglutamate.

Botany: Schroeder and Hagiwara

Dow

nloa

ded

by g

uest

on

June

7, 2

021

9308 Botany: Schroeder and Hagiwara

0.5 pM ABA20 PA- losI

-3 mvFIG. 4. ABA-activated ion currents re-

corded in whole-cell mode of patch-clamptechnique. Downward deflections corre-spond to ABA-activated inward ion cur-rents, and upward deflections are ABA-induced outward currents. ABA was ap-plied for several seconds as indicated bythe bar on the top, from a perfusion pipettelocated in the vicinity of the guard cell.Membrane potentials are indicated at theupper left of current traces. Internal solu-tion included 100 AtM EGTA, 80 mM po-tassium glutamate, and 20 mM KCI; ex-ternal solution included 10 mM KCL.

+50 mV) and the K+ equilibrium potential (EK+ = -52 mV).Furthermore, the reversal potential of ABA-activated cur-rents was more negative than equilibrium potentials of allother ions in the pipette and bath solutions (See Methods;EfreeMg2+ = -1 mV; Ecl- = +10 mV; EH+ = +61 mV).Equilibrium potentials were calculated after corrections forionic activities (27).

DISCUSSION

The reversal potential ofABA-activated currents (Figs. 4 and5) and current-mediated Ca2' elevations (Fig. 3) showed thatCa2+ ions as well as K+ ions were permeable to ABA-activated channels. Similar reversal potentials of ABA-activated currents were found when using other recordingsolutions, confirming the conclusion that ABA-induced ele-vations in [Ca2+]cy were mediated by Ca2+ influx throughnonselective ion channels.

Ca2+-channel activation by physiological stimuli such asplant hormones, light, and fungal elicitors has been suggestedto play a primary role in the initiation of signal-transductionprocesses of higher plant cells (for review, see ref. 1). Thefinding that ABA activates Ca2+-permeable currents in guardcells provides direct evidence for signal-activated, Ca2+-permeable channels in the plasma membrane ofa higher plantcell.Mechanism of ABA-Induced Channel Activation. ABA-

dependent Ca2+-permeable channels were activated in a

20-

0.

c O

C)m

c -20(D

A

A

-50 -40 -30

A

-20 -10 0 10 20 30 40Membrane potential, mV

FIG. 5. Peak amplitudes of ABA-activated ion currents recordedfrom the guard cell of Fig. 4 as function of membrane potential; thereversal potential was interpolated to -11 mV.

repetitive manner during continuous ABA applications (Fig.3). This repetitive activation pattern during continuous ABAapplication is unlike the activation of transmitter-receptorion channels, when the receptor and the ion channel consti-tute one molecular entity. In general, such receptor-ionchannel complexes in animal cells show activation followedby desensitization at high ligand concentrations. However,recovery of receptor-ion channel proteins from desensitiza-tion typically requires removal ofthe ligand from the externalmedium (28). Therefore, the repetitive activation pattern ofABA-activated currents during continuous ABA applicationsuggests that intermediate signaling events may be requiredbetween ABA exposure at the plasma membrane and ion-channel opening.

In animal cells, receptor-operated Ca2l channels havebeen postulated to explain Ca2' influx into a variety of cellsthat do not possess voltage-gated ion channels for electricalexcitability, as well as into electrically excitable cells (29).However, direct evidence for the existence of receptor-operated Ca2l channels remains sparse to date (29). In onedetailed study of peritoneal mast cells, second-messenger-dependent Ca2' influx was found in 26% of the investigatedcells, indicating that receptor-activated Ca2+ influx mayunderlie complex activation and Ca2+-translocation mecha-nisms (30).Although elevations in [Ca2+]cyt of guard cells correlate

with ABA-activated inward currents, the possibility cannotbe excluded that release of Ca2+ from intracellular organellesprovides an additional contribution to the observed rises in[Ca2+],yt. During prolonged depolarizations, no increases in[Ca21]cyt were observed (Fig. 3), suggesting that Ca2' releasefrom intracellular organelles may require hyperpolarizationand/or Ca2' influx across the plasma membrane. Ca2+-induced Ca2' release from intracellular organelles, as foundin animal cells (31), could lead to additional elevation in[Ca2+]cy1 under these conditions. Our results deviate from thehypothesis that Ca2+ release from intracellular organelles isthe initial mechanism of ABA-induced stomatal closing,which was derived from injection of synthetic caged inositolphosphates into guard cells (32, 33). Further investigationswill be needed to assess the possible contribution of Ca2+release from intracellular organelles to ABA-mediated risesin [Ca2]+cyt.ABA Responses of Guard Cells. Our data correlate with

recent findings by McAinsh et al. (16), which showed ABA-

-23 mV

Proc. Natl. Acad. Sci. USA 87 (1990)

Dow

nloa

ded

by g

uest

on

June

7, 2

021

Proc. Natl. Acad. Sci. USA 87 (1990) 9309

dependent elevation of [Ca2"],y, in guard cells from C.communes. The more repetitive nature of Ca2" elevationsobserved in the present study as well as subsequent extrusionof Ca2" may account for the difficulty in resolving ABA-induced changes in 45Ca2" fluxes in leaf epidermi (for detaileddiscussion, see ref. 34). Small 45Ca2" flux changes in ABA-treated epidermi (34) may also be attributed to the recentfindings of several groups that show that stomata respond toABA treatment by closing in small domains and by remainingopen in other small domains of the leaf epidermis (7-11). Thisdifferential ABA response has been suggested as the basis forthe physiologically observed regulation of photosyntheticcarbon fixation by ABA (7-11). The nonuniform occurrenceof ABA-induced stomatal closure in leaves (7-11) may also beresponsible for our findings that approximately one-third ofthe stomata in the epidermis of V. faba leaves closed afterexposure to ABA and 37% of the guard cells studied by patchclamping responded to ABA. Physiological modification ofthe primary ABA signaling mechanisms may account for theobservation of nonuniform closure of stomata by ABA. Thecorrelation between physiological ABA responses in leaves(7-11) and results on ABA-activated Ca2"-permeable chan-nels from the present study calls for further investigation todetermine whether modification of initial events in ABA-mediated signal transduction contribute to physiological ef-fects of ABA on photosynthetic carbon fixation in leaves(7-11).

Effects of [Ca2J1cyt on Guard-Cell Ion Transport. Increasesin [Ca2J]cy, have been suggested to play a key role in inducingstomatal closing by ABA (refs. 2-4 and 16; for reviews, seerefs. 5, 6, and 35). ABA-induced stomatal closing is mediatedby release of K+ and anions from guard cells (13, 14).Increases in [Ca2+]CYJ to the micromolar level have beenshown to inhibit inward rectifying K+ channels and to acti-vate voltage-dependent anion channels (4). The resultingefflux of anions through anion-selective channels (15) could,in turn, depolarize guard cells sufficiently (4) to activateoutward rectifying K+ channels (12, 17). In addition ABA-dependent enhancement of K+ currents has been reported(36). Simultaneous activation of anion channels and K+channels would permit ion efflux across the plasma mem-brane required for stomatal closing (4-6).

Equivocal comparison of steady-state anion currents (15)with non-steady-state anion currents (4) has led to the sug-gestion that two types of anion channels prevail in the plasmamembrane of guard cells with different voltage dependencies(15). However, direct comparison of steady-state currentsshows that the voltage dependencies of Ca2'-activated anioncurrents (4) and single-anion channel currents (15) are similar(37). In the present study ABA-activated Ca2'-permeablecurrents could be investigated in the absence of anion chan-nels due to the inactivation and wash-out of these voltage-dependent anion channels (4, 37).

Previous results have suggested that other second messen-gers, in addition to [Ca2+]CYt, may be important for regulatingvoltage-dependent anion channels (4, 37). Additional second-messenger processes may be triggered by ABA, as the ABAactivation of Ca2+-permeable channels appears to occur bymeans of intermediate coupling mechanisms (see above).Whole-cell patch-clamp experiments executed in conjunctionwith [Ca2+]cyt measurements, as used here, may allow furtherinsight into initial signal-transduction events of ABA-mediated stomatal closure.

We thank H. Fang for careful handling of protoplast preparationsand Dr. J. Kourie for comments on the manuscript. We thank Drs.R. Penner and J. Vergara for help and advice during initial Ca2+

measurements and Dr. E. Tobin for the use of growth chambers. Thisresearch was supported by National Institutes of Health GrantNS09012-21 to the late S. Hagiwara and by Alexander von Humboldtand National Institutes of Health Fellowships to J.1.S.

1. Hepler, P. K. & Wayne, R. (1985) Annu. Rev. Plant Physiol.36, 397-439.

2. De Silva, D. L. R., Cox, R. C., Hetherington, A. M. & Mans-field, T. A. (1985) New Phytol. 101, 555-563.

3. Schwartz, A., Ilan, N. & Grantz, D. A. (1988) Plant Physiol.87, 583-587.

4. Schroeder, J. 1. & Hagiwara, S. (1989) Nature (London) 338,427-430.

5. MacRobbie, E. A. C. (1988) Bot. Acta 101, 140-148.6. Schroeder, J. 1. & Hedrich, R. (1989) Trends Biochem. Sci. 14,

187-192.7. Farquhar, G. D., Hubrick, K. T., Terashima, I., Condon,

A. G. & Richards, R. A. (1987) in Progress in PhotosynthesisResearch, ed. Biggins, J. (Nishoff, Amsterdam), Vol. 4, pp.209-212.

8. Downton, W. J. S., Loveys, B. R. & Grant, W. J. R. (1988)New Phytol. 108, 263-266.

9. Terashima, I., Wong, S. C., Osmond, C. B. & Farquhar, G. D.(1989) Plant Cell Physiol. 29, 385-394.

10. Daley, P. F., Raschke, K., Ball, F. & Berry, J. A. (1989) PlantPhysiol. 90, 1233-1238.

11. Smith, S., Weyers, J. D. B. & Berry, J. W. G. (1989) Plant CellEnviron. 12, 653-660.

12. Schroeder, J. I., Raschke, K. & Neher, E. (1987) Proc. Nati.Acad. Sci. USA 84, 4108-4112.

13. Raschke, K. (1979) in Encyclopedia of Plant Physiology, eds.Haupt, W. & Feinleib, M. F. (Springer, Berlin), Vol. 7, pp.384-441.

14. MacRobbie, E. A. C. (1981) J. Exp. Bot. 32, 563-572.15. Keller, B. U., Hedrich, R. & Raschke, K. (1989) Nature

(London) 341, 450-453.16. McAinsh, M. R., Brownlee, C. & Hetherington, A. M. (1990)

Nature (London) 343, 186-188.17. Schroeder, J. 1. (1988) J. Gen. Physiol. 92, 667-683.18. Hamill, 0. P., Marty, A., Neher, E., Sakmann, B. & Sigworth,

F. J. (1981) Pflugers Arch. 391, 85-100.19. Grynkiewicz, G., Poenie, M. & Tsien, R. Y. (1985) J. Biol.

Chem. 260, 3440-3450.20. Neher, E. (1988) J. Physiol. (London) 395, 193-214.21. Toeplitz, B. K. (1979) J. Am. Chem. Soc. 101, 3344-3349.22. Williamson, R. E. & Ashley, C. C. (1982) Nature (London)

296, 647-651.23. Hepler, P. K. & Callaham, D. A. (1987) J. Cell Biol. 105,

2137-2143.24. Miller, A. H. & Sanders, D. (1987) Nature (London) 326,

397-400.25. Bush, D. S. & Jones, R. L. (1987) Cell Calcium 8, 455-472.26. Felle, H. (1988) Planta 174, 495-499.27. Robinson, R. A. & Stokes, R. H. (1955) in Electrolyte Solu-

tions (N.Y. Acad. Sci., New York), pp. 480-499.28. Hille, B. (1984) Ionic Channels of Excitable Membranes (Sin-

auer, Sunderland, MA).29. Rink, T. J. (1988) Nature (London) 334, 649-650.30. Penner, R., Matthews, G. & Neher, E. (1988) Nature (London)

334, 499-504.31. Berridge, M. J. & Irvine, R. F. (1989) Nature (London) 341,

197-205.32. Gilroy, S., Read, N. D. & Trewavas, A. J. (1990) Nature

(London) 346, 769-771.33. Blatt, M. R., Thiel, G. & Trentham, D. R. (1990) Nature

(London) 346, 766-769.34. MacRobbie, E. A. C. (1989) Planta 178, 231-241.35. Mansfield, T. A., Hetherington, A. M. & Atkinson, C. J.

(1990) Annu. Rev. Plant Physiol. Plant Mol. Biol. 41, 55-75.36. Blatt, M. R. (1990) Planta 180, 445-455.37. Schroeder, J. I. & Hagiwara, S. (1990) in Calcium in Plant

Growth and Development, ASPP Symposium Series, eds.Leonard, R. T. & Hepler, P. K. (ASPP, Beltsville, MD), Vol.4, pp. 144-150.

Botany: Schroeder and Hagiwara

Dow

nloa

ded

by g

uest

on

June

7, 2

021