Proc. Natl. Acad. Sci. USAVol. 91, pp. 12639-12643, December 1994Developmental Biology

Increased internal Ca2+ mediates neural induction in theamphibian embryo

(intracellular calcium/calcium imaging/Pleurodeles walt/protein kinase C)

MARC MOREAU*, CATHERINE LECLERC, LYDIE GUALANDRIS-PARISOT, AND ANNE-MARIE DUPRATCentre de Biologie du D6veloppement, Unitd mixte de Recherche 9925, Centre National de la Recherche Scientifique/Universitd Paul Sabatier, 118, route deNarbonne F-31062 Toulouse Cedex, France

Communicated by John B. Gurdon, July 25, 1994

ABSTRACT The molecular mechanism of neural induc-tion is still unknown and the identity of the natural inducerremains elusive. It has been suggested that both the proteinkinase C and cAMP signal transduction pathways may beinvolved in mediating its action. Here we provide evidence thatCa2+ is implicated in the process of transduction of theneuralizing signal. We find that an increase in intraceDlularCa2+ concentration [Ca2+1J occurs during neural inductionprovoked in vitro by the lectin Con A in Pleurodeles walWlembryo. We demonstrate that specific L-type Ca2+ channelagonists also trigger neural induction. Conversely, noninduc-ing lectins do not raise [Ca2+]J. Ryanodine and caffeine triggerneural induction. An increase in [Ca2+]J was also observed aftertreatment with the phorbol 12-myristate 13-acetate, which hasbeen reported to be inductive. The [Ca2+]J increase triggered byphorbol ester and Con A was abolished by staurosporine andby L-type Ca2+ channel antagonists. Our findings demonstratethat the [Ca2+]i increase occurs via L-type Ca2+ channels. Wesuggest an amplification of this increase by a Ca2+-inducedCa2+ release mechanism which involves intracellular ryano-dine-sensitive stores. We propose that Ca2+-dependent pro-cesses controlled by protein kinase C are implicated in theregulation of gene expression in response to neural induction.

On the other hand, induction of the ectoderm toward theneural pathway can be provoked simply by modifying theextracellular concentration ofdivalent cations (16, 17). Theseobservations and the availability of microspectrofluorimetrictechniques prompted us to examine the role of Ca2+ intransduction of the neuralizing signal. We triggered in vitroneural induction on excised dorsal ectoderm explants of theurodele amphibian Pleurodeles waltl with the lectin Con A.This lectin is an inducer for both anurans and urodeles, actingat the plasma membrane level. So far, it is the only lectinknown to have inducing activity (7, 8, 18).Here we demonstrate that neural induction triggered by

Con A involved the activation of L-type Ca2+ channels. Inour system L-type Ca2+ channels are upmodulated by PKC,suggesting that inductive effects ofphorbol esters, previouslyobserved on Xenopus ectoderm (9, 10), may be partly ex-plained by the activation of L-type Ca2+ channels. In addi-tion, an increase in intracellular Ca2+ concentration ([Ca2+],)obtained by direct stimulation ofinternal stores by caffeine orryanodine triggers neural induction. We conclude that anincrease in [Ca2+]i is sufficient to induce the competentectoderm toward the neural pathway.

In amphibians, neural induction takes place during gastrula-tion, as a consequence of an interaction between the chor-damesoderm (inductive tissue) and the ectoderm (targettissue) (1, 2). Although numerous unrelated substances canact as inducers (2, 3), the natural inducer remains unidenti-fied. Recently an endogeneous soluble protein, noggin, hasbeen shown to have neural inducing activity in Xenopus (4).It has been also suggested that the inhibition of the signaltransduced by the activin type II receptor leads to neural-ization (5). Follistatin, a natural antagonist of the activinreceptor, can induce neural tissue in vivo (6). Further, theinducing signal from the chordamesoderm is recognized atthe level of the plasma membrane of the ectoderm (3, 7, 8).Work with Xenopus embryos has led to the suggestion thatactivation of protein kinase C (PKC) by phorbol esters leadsto neural induction (9-12). An increase in cAMP-dependentprotein kinase (PKA) activity during neural induction hasalso been observed (11), suggesting a cross talk between PKAand PKC. Additional support for a role of a PKC pathway inneural induction is provided by phorbol ester induction of theneural-specific src+ mRNA in dorsal competent ectoderm ofXenopus embryo (13, 14).

Recently, a direct examination of ionic signaling (15) hasrevealed that dorsal ectoderm cells of Xenopus undergo anincrease in internal pH in response to the neural inducingsignal, suggesting that intracellular alkalinization may par-ticipate in gene expression associated with neural induction.

MATERIALS AND METHODSChemicals. Bis(2-aminophenoxy)ethane-N,N,N',N'-

tetraacetic acid (BAPTA) AM and fluo-3AM (MolecularProbes); nifedipine, nimodipine, staurosporine, and ionomy-cin (Sigma); and phorbol 12-myristate 13-acetate (PMA) and4a-phorbol 12-myristate 13-acetate (4a-PMA) (Research Bio-chemicals, Natick, MA) were prepared as stock solutions indimethyl sulfoxide and used at appropriate concentrations.(S)-(-)-Bay K 8644 (Calbiochem), isradipine (Research Bio-chemicals), and Ryanodine (Sigma) were dissolved in abso-lute ethanol. Control experiments have shown no biologicaleffect ofthese solvents. Con A (IBF, Villeneuve-la-Garenne,France) and succinyl-Con A (Sigma) were dissolved to con-stitute stock solutions of 2 mg/ml in Holfreter medium(containing 0.88 mM Ca2+) and used at 300 ug/ml (7, 8).Caffeine (Sigma) was dissolved in distilled water.

Explant Preparation. Dejellied embryos from P. waltl weremaintained in Holfreter medium (60 mM NaCl/0.67 mMKCl/0.88 mM CaCl2/5 mM Tris, pH 8.5). All stages refer tothose of Gallien and Durocher (19). Explants of competent(stage 8a) ectoderm from the dorsal side and noncompetent(stage 6) ectoderm of P. waltl (8) were kept flat in a goldfolding microscopy grid (diameter, 3.5 mm; fields, 200 ,am x200 pm; Pelanne Instruments, Paris) which permitted direct

Abbreviations: AM, acetoxymethyl ester; BAPTA, bis(2-aminophe-noxy)ethane-N,N,N',N'-tetraacetic acid; [Ca2+]i, intracellular Ca2+concentration; GFAP, glial fibrillary acidic protein; PMA, phorbol12-myristate 13-acetate; PKA, cAMP-dependent protein kinase;PKC, protein kinase C.*To whom reprint requests should be addressed.

12639

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.

12640 Developmental Biology: Moreau et al.

observation of a field of about 30-50 cells and 10-20 cells forstage 8 and stage 6 explants, respectively.

[Ca2+1J Measurements. Measurements of [Ca2+]i were per-formed by emission microspectrofluorimetry. Cells wereloaded for 60 min at 20°C with 5 ,uM fluo-3 acetoxymethylester (AM) (Molecular Probes) under gentle agitation. Afterwashing, the preparation was placed on the stage of aninverted microscope (Diaphot, Nikon) and observed with aNikon objective (x40; n.a., 0.83) illuminated by a 75-Wxenon lamp, attenuated with neutral density filters of ade-quate values. Excitation light was 490 nm with a 525-nmbarrier filter. Fluorescence was detected by a silicon inten-sified target camera. Images were captured at intervals of 5or 10 sec with 8-bit resolution (256 levels) and processed withthe Argus 50 processing image system (Hamamatsu Photo-nics, Hamamatsu, Japan). Time courses of Ca2+ signals inselected regions were analyzed with the Argus 50 software.Because fluo-3 does not permit the use of ratio measurementsto determine absolute free Ca2+ levels, data are presented inarbitrary units as percentage of fluorescence variation withrespect to the resting level. However, comparisons of fluo-rescence signal within each experiment provide a relativeindication of free Ca2+ levels (20).Immunocytochemistry. Neurons and glial cells were char-

acterized with antibodies directed against specific antigenicmarkers of neurofilaments or the NC1 epitope, which inPleurodeles cultures is expressed on the membrane of dif-ferentiated neurons (21), and glial fibrillary acidic protein(GFAP) (22). Immunocytochemical experiments were car-ried out as described (21). Direct observation by confocalmicroscopy (LSM, Zeiss) allowed us to visualize neural cells.

RESULTSIn P. waltl the neural competence of the ectoderm starts atstage 7 and this tissue can no longer be induced after stage 12(18, 23). In response to ConA (300 ug/ml) added to theexternal medium, [Ca2+]i rose in stage 8a ectoderm explantsas revealed by fluo-3 (see Fig. 2A). This rise could bedetected <20 min after addition of ConA, and its maximumcorresponded to 10-25% (mean ± SEM, 15.2 ± 4.3%; n = 28)ofthe resting [Ca2e]i (see ref. 24 for details ofmeasurements);the maximum amplitude was reached 25-60 min after addi-tion of Con A. The entire phase of increase lasted about 90min, but less sustained increases could sometimes be ob-served. In some cases [Ca2e]i decreased gradually from 15 to60 min after the initial increase. In such cases the fluores-cence stabilized at about 10% above the resting value. Inparallel control experiments, Con A treatment of ectodermsexcised from the same batches of embryos clearly elicitedneuralization as judged from the presence of numerousneurons and glial cells differentiating after 5-6 days ofcultureat 20°C (Fig. 1 A and B). No (Ca2+]i increase was observedwhen Con A was omitted. A strict correlation between anearly increase in [Ca2+]i and neural induction was observedin the following experimental situations. (i) No variation in[Ca2+]i was detected following treatment with the noninduc-ing lectin succinyl-Con A at 300 pg/ml (n = 5) (Fig. 2A Inset).(ii) Noncompetent ectoderm treated with Con A at 300 pg/ml

1.2

1.1

1.0

1.2

1.1

01.0

C1.2r

1.1 [1.0

C1.2

1.1

1.0

0

FIG. 1. Cultured competent ectodermal cells after various treat-ments. The assays for neuralization were performed on embryosfrom the same batch as those used for [Ca2+]i measurements. Theobservations were performed after 5 days ofculture, as described (8).Preparations were visualized by epifluorescence confocal micros-copy. (Left) Immunocytochemical staining using a monoclonal an-tibody (NC1) directed against gangliosides specific for the neuronalmembrane (21). (Right) Glial cells identified with GFAP antibodies(22). (A and B) Dorsal ectoderm cells (stage 8) treated with Con A(300 pg/ml). Note the presence of neurons (n) with neurites (neur)and areas of glial cells (gc). (C and D) Ectoderm treated for 3 hr with20 fiM Bay K 8644. (E and F) Effect of treatment for 30 min with 10mM caffeine. In this experiment numerous neurons (n) and largeareas of glial cells (gc) have differentiated. (Bar = 50 pm.)

20 40 60 80

Time, min

FIG. 2. Neural induction is correlated with the activation ofL-type Ca2+ channels. (A) Recording of fluorescence (F) changes ofthe Ca2+ indicator fluo-3, revealing elevation of [Ca2+]i in Pleuro-deles ectoderm cells. (A) Response of competent tissue (stage 8a) toCon A (300 jug/ml). (Inset) Sibling ectoderm treated with succinyl-ConA (300 p,g/ml). (B) Noncompetent ectoderm (stage 6) stimulatedby Con A (300 pg/ml). (C) Effect of pretreatment for 30 min with 10,uM isradipine on the increase in [Ca2+]i triggered by Con A (600pg/ml). The increase in [Ca2+]i is totally inhibited. (Inset) Controlexperiment showing the [Ca2+]j increase triggered by Con A (600pg/ml) (embryos from the same batch). (D) Action of 20 j.M Bay K8644 on [Ca2+]i in competent ectoderm.

A1.2-

ConY I1 1 1.0

0 20 40

0 20 40 60 8c

B

ConA

I~1

10l 20 30

C + isradipine1.2 1

ConA 1.0 < .

lI 0 15 30

10 20 3c

D

BayK* 1

Proc. Natl. Acad. Sci. USA 91 (1994)

Proc. Natl. Acad. Sci. USA 91 (1994) 12641

did not develop an increase in [Ca2W]i above 2% of the initialvalue (n = 3) (Fig. 2B). (iii) Neural induction in response toCon A did not occur when competent ectoderm was loadedwith the Ca2+ chelator BAPTA (0.4 uM BAPTA AM). Invitro induction resulting from association of the blastoporallip with dorsal ectoderm was also inhibited when ectodermalcells were loaded with BAPTA (data not shown).The next question we addressed was the origin of the

increased [Ca2+]1i. The major route for Ca2+ entry into cellsinvolves voltage-operated Ca2+ channels. Fixation sites ofCon A are localized on the plasma membrane of the respon-sive ectoderm (25), and Con A binds to the a2 subunit of theL-type Ca2+ channel (26, 27). To investigate the possibleaction of Con A on these channels, the use of a Ca2+-freemedium was excluded since it triggers a complete dissocia-tion of ectoderm cells and, indirectly, neural induction (17).However specific L-type Ca2+ channel antagonists are avail-able. We tested the effect of 10 ,uM isradipine on [Ca2+Ji inectoderm after treatment with Con A. Isradipine totallyinhibited the Con A stimulation of [Ca2+]1i (n = 7) (Fig. 2C),as well as neural induction, since only epidermal cells dif-ferentiated in culture. Nifedipine (100 ,uM), another specificinhibitor of L-type Ca2+ channels, also produced the sameeffects. Conversely, a transitory increase in [Ca2+]i of 10-30min was observed after addition ofthe L-channel agonist BayK 8644 (10 or 20 ,uM) to the external medium (mean ± SEM,12.2 ± 5.1%; n = 3) (Fig. 2D). Further, this transitoryincrease in [Ca2W]i was sufficient to trigger neural inductionin all cases, as revealed by the development in culturedexplants of neuron and glial cells (Fig. 1 C and D).The above results demonstrate that Con A acts via L-type

Ca2+ channels. We sought to determine whether neuralinduction could also be triggered by Ca2+ freed from intra-cellular stores. To answer this question, we used drugs thatliberate sequestered Ca2+ from intracellular organelles. Oneform of Ca2+ release involves the activation of ryanodine-and methylxanthine-sensitive stores (28). The actions ofryanodine and caffeine on dorsal ectoderm were thus exam-ined. In stage 8a ectoderm, caffeine (20 mM) had a pro-nounced effect on internal Ca2+ release, triggering a 30%increase in [Ca2W]i (mean ± SEM, 32 ± 11.2%, n = 8explants, 19 cells measured) (Fig. 3A). [Ca2W+i peaked in 1min and then declined gradually during 20 min. Ectodermcells exposed to caffeine for 30-180 min differentiated intoimmunocytochemically identifiable neuron and glial cellsafter 5-6 days of culture, confirming the neuralizing effect of

1.4

0

1.2

1.1

1.0

0 20 40 60 80

Time, min

FIG. 3. Evidence of Ca2+ stores in cells of the competentectoderm. Ectoderm cells were loaded with 5 uM fluo-3 AM. (A)Action ofcaffeine (20 mM) on [Ca2+]i measured with fluo-3. Caffeinetreatment triggered a strong increase of [Ca2+]i. Note that caffeinewas inductive (Fig. 1). (B) Action of 0.2 uM ryanodine on ectodermcells. An increase in [Ca2+] was observed. Ryanodine was alsoinductive.

rising [Ca2+]i (Fig. 1 E and F). Ryanodine added to theexternal medium triggered a continuous [Ca2+]i increase (Fig.3B), although its amplitude was less than that provoked bycaffeine (mean ± SEM, 7.5 ± 2.6%; n = 5). Like caffeine,ryanodine triggered neural induction in the in vitro systemdespite the weak amplitude of Ca2+ release. Thus a rise in[Ca2+]i, above threshold value, irrespective of the mecha-nism by which [Ca2+]j is increased, is sufficient to neuralizecompetent ectoderm cells.

Activation ofPKC by phorbol esters triggers neural induc-tion (9-12). Protein phosphorylation by PKC is one of themechanisms of Ca2+ channel regulation. Indeed, PKC canexert inhibitory or stimulatory effects on L-type Ca2+ chan-nels (29, 30). This led us to hypothesize that the inductiveeffect of phorbol esters might occur via activation of Ca2+channels, either directly or indirectly (30). [Ca2+]i was mon-itored during application of PMA. Continuous bath applica-tion of 50-500 nM PMA transiently increased [Ca2+]i in adose-dependent manner (mean ± SEM, 6.4 ± 1.7%; n = 8).The fluorescence usually increased gradually, reaching amaximum in 5-10 min (Fig. 4A), and returned to the restinglevel in 10-20 min. Control experiments designed to validatethe direct action ofPMA on PKC activation involved the useof staurosporine, a known PKC inhibitor, and 4a-PMA, aphorbol ester that is inactive with respect to PKC. Theaverage change in fluorescence level triggered by 500 nMPMA was decreased to under 2% when dorsal ectoderm waspreincubated for 40 min with 500 nM staurosporine (Fig. 4B).In addition, 4a-PMA (500 nM) produced no significant effect

1.2

1.1

1.0

1.2

1.1

01.0

1.2

1.1

1.0

1.2

1.1 [1.0

O 10 2C

-C4a-PMA

O 10 2C

0 5 10 15

Time, min

FIG. 4. Modulation of L-type Ca2+ channel activity by PMA.Competent ectoderm (stage 8a) was loaded with fluo-3 AM. (A)Effect ofPMA (500 nM) showing an increase in [Ca2+]i. (B) Inhibitionof PMA effect by staurosporine (500 nM). (C) Competent explanttreated with 4a-PMA (500 nM), a phorbol ester analog that is inactivewith respect to PKC. (D) Effect of nimodipine (10 AM) on the [Ca2+]iincrease provoked by PMA (500 nM). The explant was incubated 10min in nimodipine before PMA addition. Nimodipine was continu-ously present during stimulation by PMA.

PMA

) 10 20

-B+ staurosporine

PMA

I1~~~~~~~~~~~~~~~~~~~~~~~~~~~

0 10 20 30 40

B

ryanodine

'1l

- D + nimodipine

PMAI

Developmental Biology: Moreau et al.

(

12642 Developmental Biology: Moreau et al.

1.2

1.1

0

r%.2

1.0

1.2

1.1

1.0

0 10 20

Time, min

FIG. 5. PKC mediates the activation of L-type Ca2+ channels,but not Ca2+ release from internal stores. (A) Effect of staurosporine(500nM) on competent ectoderm stimulated by Con A (300 Ag/ml).Explants were preincubated for 30 min with staurosporine beforeCon A addition. (Inset) Control showing a fluorescence increase ofthe competent ectoderm loaded with fluo-3 AM triggered by Con A(300 tig/ml) (embryos of the same batch as in A); arrow, addition ofCon A. (B) Effect of caffeine (10 mM) on competent ectodermpreincubated for 30 min with staurosporine (500 nM). (Inset) Controlshowing the effect of caffeine (10 mM) on competent ectoderm;arrow, addition of caffeine.

on the fluorescence level recorded in dorsal explants (<1%increase) (Fig. 4C). Nimodipine (10,uM), an antagonist of LCa2+ channels, blocked the increase in [Ca2+]i induced by 500nM PMA (n = 3), indicating that the [Ca2+]i increase, in thiscase, was mostly due to L Ca2+ channels (Fig. 4D). Theseresults raised the question of whether or not the effect of ConA on [Ca2+]i was also dependent on the activation of L Ca2+channels by PKC. In fact, the [Ca2+]i increase triggered byCon A (300 pg/ml) was abolished when dorsal explants were

incubated for 30 min with 500 nM staurosporine (Fig. 5 A andInset), thus suggesting that theCa2+ entry triggered by ConA is indeed controlled by PKC-dependent phosphorylationprocesses. Furthermore, neural induction was totally inhib-ited when Con A-stimulated ectoderms were preincubatedfor 30 min with 500 nM staurosporine, since in all cases (n =10) only epidermal differentiation was observed. To furtherconfirm that PKC acts on plasma membrane Ca2+ channels,dorsal explants incubated for 40 min with 500 nM staurospo-rine were still able to release Ca2+ upon treatment with 10mM caffeine (Fig. SB and Inset). No significant differencewas observed in fluorescence intensity between caffeine-stimulated dorsal explants incubated with or without stauro-sporine, although the duration of the transient was slightlyreduced (30%) in the presence of staurosporine. Hence, PKCplays an essential role on the plasma membrane Ca2+ chan-nels and not on cytosolic Ca2+ stores.

DISCUSSION

The role played by the ionic environment in neural inductionhas been pointed out (16), but the direct effect of ionicmovements in this process was not documented. The initialexperiments reported here demonstrate that competent cellsof dorsal ectoderm of P. waltl undergo an increase in [Ca2+]in response to Con A. Similarly, Sater et al. (15) have shownthat neural induction in Xenopus is accompanied by an

increase in internal pH. Therefore, intracellular alkaliniza-tion and internal Ca2+ increase are among the earliest knownresponses to neural induction. The idea that a [Ca2+]i increaseis a direct cause in triggering neural induction is supported bythe following lines of evidence: (i) in our experimentalparadigms, neural induction was always accompanied by an

increase in [Ca2]1i; (ii) in the absence of neural induction(Con A-treated stage 6 ectoderm), [Ca2+]i increase never

developed; and (iii) when [Ca2+]i increase was inhibited byBAPTA, neural induction was abolished. BAPTA-treatedectoderm cells developed into epidermis. Neuralization trig-gered in ectoderm associated with mesoderm was totallyblocked by BAPTA. Therefore, this direct relationship be-tween[Ca2W]i increase and neural induction is apparently notrestricted to artificial neuralizing factors. Our results suggestthat the increase in [Ca2+]i triggered by ConA is mainly dueto the activation of L-type Ca2+ channels. Our data, togetherwith the fact that Con A binds to the a2 subunit of the L-typechannel on PC12 rat pheochromocytoma cells (31), imply thatthis channel is indeed the primary molecular target of thelectin in neural induction. However, the relationship betweena putative receptor of the neuralizing signal and L Ca2+channels remains unknown. Con A treatment did not elicit a

[Ca2+]i increase in stage 6 ectoderm, suggesting that L-typechannels at this stage are absent or not functional. Thissuggests that incorporation into the plasma membrane ofectoderm and/or acquisition of functionality of L-type Ca2+channels may play an important role in the appearance ofcompetence.The sustained increase in [Ca2+]i triggered by Con A was

somewhat intriguing, in light of the rapid inactivation that ischaracteristic of L-type Ca2+ channels. Three hypothesescan be considered. (i) The presence of Con A in the externalmedium for 3 hr is required to trigger neural induction (7, 8,25). This rather long time is justified by the fact that Con Ais internalized in few minutes by the ectoderm cells (25),although the internalization is not a prerequisite for neuralinduction. The continuous presence of Con A may thus exertan iterative effect on Ca2+ channels, resulting in the long-lasting [Ca2+]i increase observed. (ii) This lectin could exertits effect on Ca2+ influx by increasing channel numbers, ashas been suggested for PC12 cells (31). (iii) Finally, theobserved time course of increase is rather suggestive of theintervention of an internal relay, which might be due to a

process referred to as a Ca2+-induced Ca2+ release (CICR).In our system, internal Ca2+ stores can be activated byryanodine or caffeine. Ryanodine is less efficient but also lesspermeant than caffeine in Pleurodeles ectoderm cells, whichmight explain the difference in amplitude of Ca2+ releasebetween the two drugs. CICR may also involve inositoltrisphosphate-sensitive stores (32). However, we have no

evidence for the existence of such stores in ectoderm cells,since no permeant agonist of inositol trisphosphate receptorsis currently available. With the two first hypotheses, whichdo not take into account the inactivation of L-type Ca2+channels, the sole effect of Con A on Ca2+ channels shouldresult in a more transitory [Ca2+]i increase than that observedin our experiments. The third hypothesis is more likely, sincewe have demonstrated the presence of ryanodine-sensitiveCa2+ stores. These releasable Ca2+ stores can constitute an

internal relay for Ca2+ influx, and the long-lasting kineticsobserved in the presence of Con A (Fig. 2A) could be due toa continuous refilling of internal stores from the externalmedium. However, no definitive conclusions can be drawnregarding an eventual CICR acting as an internal relay of theCa2+ influx upon neural induction triggered by Con A. Thelong-lasting effect of caffeine could suggest that caffeine actson its other target-i.e., inhibition of phosphodiesterase-totrigger an increase in cAMP. In addition, direct control ofL-type Ca2+ channels by PKA has been described (33),suggesting that the observed effect of caffeine could be dueto the activation of L channels. However, this mechanismseems unlikely, since the Ca2+ release triggered by caffeinewas not modified in the presence of staurosporine, which, as

demonstrated in this study, totally inhibits the activation ofL channels by PMA. The activation ofPKC by phorbol esters

.A 1.2 + staurosporine

conA 1.00 5 30

0 1 0 20 30

+ staurosporineB 1.2-

caffein 0

0 0 201~~~-J

Proc. Natl. Acad. Sci. USA 91 (1994)

Proc. Natl. Acad. Sci. USA 91 (1994) 12643

has been described to be involved in neural induction (10-12).Here we suggest that one role of PKC may be to regulateL-channel activity. Such a role has been described in numer-ous systems (29, 30). In our system, the effect of PMA mayalso be explained by an upmodulation of Ca2+ channels. Infact, our results showing that the [Ca2+]i increase triggered byCon A is abolished by the PKC inhibitor staurosporine andthat the [Ca2+], increase triggered by PMA is inhibited bynimodipine confirm the control of this Ca2+ permeability byPKC. Consequently, neural induction mediated by phorbolesters (10-12) may be partially interpreted as a direct actionof PKC on Ca2+ channels. Phosphorylation by PKC couldincrease channel activity by various mechanisms: PKC couldact by direct phosphorylation of a component of L channels(30). It has been suggested that stimulation ofPKC in Aplysianeurons recruits a covert type of Ca2+ channel (34). On theother hand, it has been shown in many cell types that PKCdownmodulates the activity of other voltage-activated chan-nels, particularly K+ and Na+ channels. Therefore the in-duced depolarization may indirectly activate voltage-dependent Ca2+ channels (35). The total inhibition by stau-rosporine of Con A-triggered neural induction can beexplained by at least two mechanisms: an inactivation of Lchannels or a direct effect of Con A on PKC. The latterpossibility is unlikely, since PKC is a regulatory componentwhich is located downstream (30).Neural induction takes place as a result of the activation of

specific sets of genes. How then might an increase in Ca2+activate genes? Protein products of immediate early genesappear to play a critical role in long-term changes in cellfunction (36). During differentiation of PC12 cells stimulatedby depolarizing agents, an increase in [Ca2+]i resulting fromthe activation of L-type Ca2+ channels leads to the activationof immediate early genes such as c-fos orjun-B, in <20 minafter the opening of the channels. The mechanism by whichCa2+ influx activates transcription of c-fos orjun-B has beencharacterized (37). It involves the phosphorylation of tran-scription factor CREB (cAMP response element-bindingprotein) which is directly mediated by a calmodulin-dependent protein kinase. The function ofCREB then couldbe to integrate various signaling pathways resulting in geneexpression. Among these, one can include follistatin, whichis responsible for neural induction by blocking the activinreceptor (5, 6). In our system, a similar mechanism ofCREBphosphorylation following a [Ca2+]i increase can be consid-ered. In this respect, it is interesting that a transient variationof [Ca2+], (whatever its provenance, either from externalmedium or from internal stores) of <30 min (i.e., caffeine orBay K 8644 effects) is sufficient to drive the competentectoderm toward neural determination.

We are grateful to Anne Warner, Julian Smith, Pierre Guerrier, andPhilippe Cochard for reading the manuscript and for criticisms,suggestions, and correcting the English version. We thank FrancoiseFoulquier for technical assistance. This work was supported byCentre National de la Recherche Scientifique, by a research grantfrom Association Frangaise contre les Myopathies (no. 4350432), andby Centre National des Etudes Spatiales.

1. Spemann, H. & Mangold, M. (1924) Arch. Mikrosk. Anat.

Entwicklungmesh. 100, 599-638.

2. Saxen, L. (1989) Int. J. Dev. Biol. 33, 21-48.3. Tiedeman, H. & Born, J. (1978) Roux's Arch. Dev. Biol. 184,

285-299.4. Lamnb, T., Knecht, A., Smith, W., Stachel, S., Economides,

A., Stahl, N., Yancapopoulos, G. & Harland, R. M. (1993)Science 262, 713-718.

5. Hemmati-Brivaulou, A. & Melton, D. A. (1994) Cell 77, 273-281.

6. Hemmati-Brivaulou, A., Kelly, 0. G. & Melton, D. A. (1994)Cell 77, 283-295.

7. Takata, K., Yamamoto, K. Y. & Ozawa, R. (1981) Roux'sArch. Dev. Biol. 190, 92-96.

8. Duprat, A. M., Gualandris, L. & Rouge, P. (1983) J. Embryol.Exp. Morphol. 10, 171-187.

9. Davids, M., Loppnow, B., Tiedemann, H. & Tiedemann, H.(1987) Roux's Arch. Dev. Biol. 196, 137-140.

10. Otte, A. P., Koster, C. M., Snoek, G. T. & Durston, A. J.(1988) Nature (London) 324, 618-620.

11. Otte, A. P., Van Run, P., Heldelveld, M., Rian-Driel, R. &Durston, A. J. (1989) Cell 58, 641-648.

12. Otte, A. P. & Moon, R. T. (1992) Cell 68, 1021-1029.13. Collett, J. W. & Steele, R. E. (1992) Dev. Biol. 152, 194-198.14. Collett, J. W. & Steele, R. E. (1993) Dev. Biol. 158, 487-495.15. Sater, A. K., Alderton, J. M. & Steinhardt, R. A. (1994) De-

velopment (Cambridge, U.K.) 120, 433-442.16. Barth, L. G. & Barth, L. J. (1964) Biol. Bull. (Woods Hole,

MA) 127, 413-427.17. Saint Jeannet, J. P., Huang, S. & Duprat, A. M. (1990) Dev.

Biol. 141, 93-103.18. Gualandris, L., Rouge, P. & Duprat, A.-M. (1985) J. Embryol.

Exp. Morphol. 86, 39-51.19. Gallien, L. & Durocher, M. (1957) Biol. Bull. Fr. Belg. 91,

97-114.20. Cornell-Bell, A. H., Finkbeiner, S. M., Cooper, M. S. &

Smith, S. J. (1990) Science 247, 470-473.21. Duprat, A. M., Saint Jeannet, J. P., Pituello, F., Huang, S.,

Boudannaoui, S., Kan, P. & Gualandris, L. (1990) Int. J. Dev.Biol. 34, 149-156.

22. Soula, C., Sagot, Y., Cochard, P. & Duprat, A.-M. (1990) Int.J. Dev. Biol. 34, 351-364.

23. Leikola, A. (1963) Ann. Soc. Zool. 25, 2-50.24. Kao, J., Harootunian, A. & Tsien, R. W. (1989) J. Biol. Chem.

264, 8179-8184.25. Grunz, H. (1985) J. Exp. Embryol. Morphol. 89, Suppl. 349-

364.26. Curtis, B. M. & Catterall, W. A. (1984) Biochemistry 23, 2113-

2118.27. Borsotto, M., Barhanin, J., Norman, R. I. & Lazdunski, M.

(1984) Biochem. Biophys. Res. Commun. 122, 1357-1366.28. McPherson, P. S. & Campbell, K. P. (1993) J. Biol. Chem. 268,

13765-13768.29. Shearman, M. S., Sekiguchi, K. & Nishizuka, Y. (1989) Phar-

macol. Rev. 41, 211-237.30. Yang, J. & Tsien, R. W. (1993) Neuron 10, 127-136.31. Greenberg, D., Carpenter, C. L. & Messing, R. 0. (1987) J.

Neurochem. 48, 888-894.32. Berridge, M. J. (1993) Nature (London) 361, 315-325.33. Armstrong, D. & Eckert, R. (1987) Proc. Natl. Acad. Sci. USA

84, 2518-2522.34. Strong, J. A., Fox, A. P., Tsien, R. W. & Kaczmarek, L. K.

(1987) Nature (London) 325, 714-717.35. Ahlijianian, M. K., Striessnig, J. & Catterall, W. A. (1991) J.

Biol. Chem. 266, 20192-20197.36. Sheng, M. & Greenberg, M. E. (1990) Neuron 4, 477-485.37. Sheng, M., McFadden, G. & Greenberg, M. E. (1990) Neuron

4, 571-582.

Developmental Biology: Moreau et al.