Development 112, 21-31 (1991)Printed in Great Britain (£) The Company of Biologists Limited 1991

21

Inductive differentiation of two neural lineages reconstituted in a

microculture system from Xenopus early gastrula cells

SHOHEI MTTANI and HARUMASA OKAMOTO*

Department of Newobwlogy, Institute of Brain Research, Faculty of Medicine, University of Tokyo, 7-3-1-Hongo, Bunkyo-ku, Tokyo, 113,Japan

*To whom correspondence should be addressed

Summary

Neural induction of ectoderm cells has been reconsti-tuted and examined in a microculture system derivedfrom dissociated early gastrula cells of Xenopus laevis.We have used monoclonal antibodies as specific markersto monitor cellular differentiation from three distinctectoderm lineages in culture (Nl for CNS neurons fromneural tube, Mel for melanophores from neural crestand E3 for skin epidermal cells from epidermallineages).

CNS neurons and melanophores differentiate whendeep layer cells of the ventral ectoderm (VE, prospectiveepidermis region; 150 cells/culture) and an appropriateregion of the marginal zone (MZ, prospective mesodermregion; 5-150 cells/culture) are co-cultured, but not incultures of either cell type on their own; VE cellscultured alone yield epidermal cells as we havepreviously reported. The extent of inductive neuraldifferentiation in the co-culture system strongly dependson the origin and number of MZ cells initially added toculture wells. The potency to induce CNS neurons ishighest for dorsal MZ cells and sharply decreases asmore ventrally located cells are used. The same

dorsoventral distribution of potency is seen in the abilityof MZ cells to inhibit epidermal differentiation. Incontrast, the ability of MZ cells to induce melanophoresshows the reverse polarity, ventral to dorsal. These dataindicate that separate developmental mechanisms areused for the induction of neural tube and neural crestlineages.

Co-differentiation of CNS neurons or melanophoreswith epidermal cells can be obtained in a single well ofco-cultures of VE cells (150) and a wide range ofnumbers of MZ cells (5 to 100). Further, reproducibledifferentiation of both neural lineages requires intimateassociation between cells from the two gastrula regions;virtually no differentiation is obtained when cells fromthe VE and MZ are separated in a culture well. Theseresults indicate that the inducing signals from MZ cellsfor both neural tube and neural crest lineages affect onlynearby ectoderm cells.

Key words: Xenopus embryo, neural induction, monoclonalantibodies, microculture.

Introduction

Ectoderm gives rise to three distinct lineages (neuraltube, neural crest and epidermis) during vertebrateneurulation (Slack, 1983). In amphibians, the separ-ation of the two neural lineages from the epidermal oneappears to arise as consequence of determinativecellular interactions during gastrulation, which pre-cedes neurulation. The ectoderm overlying the invagi-nated dorsal mesoderm forms the neural plate, which isdelineated by the neural fold (Schroeder, 1970; Warner,1985). The neural plate and a part of the neural foldproduce the neural tube, which eventually differen-tiates into the central nervous system (CNS). The restof the neural fold, which fails to be incorporated intothe neural tube, gives rises to the neural crest. Its

descendents eventually differentiate into the peripheralnervous system, melanophores and a variety of otherderivatives. The ectoderm outside the neural fold formsepidermis and placodes for lens or sensory epithelia(Slack, 1983).

Experimental manipulations such as grafting orexplanting of embryonic tissue have elucidated thedevelopmental mechanisms underlying the lineageseparation of gastrula ectoderm in amphibians (Holt-freter, 1938; Spemann, 1938). Ectoderm pursues theepidermal lineage autonomously, but inductive actioninvolving the close association of the dorsal mesodermis required to produce the neural lineages. Theseclassical observations have recently been confirmedmolecularly with the aid of a variety of specific markerssuch as monoclonal antibodies (mAbs) or monospecific

22 5. Mitani and H. Okamoto

antiserum that can distinguish the three ectodermlineages (Akers et al. 1986; Jacobson and Rutishauser,1986; Mitani, 1989; Jones and Woodland, 1989).

In spite of intensive research after the discovery ofthe neural induction (Spemann and Mangold, 1924), itis still not known how the inducing signal(s) istransmitted from mesoderm to ectoderm cells, nor whatthe nature of this signal is (Warner, 1985). Much lessattention has been given to the question of theseparation of the neural crest and neural tube lineages(Nieuwkoop, 1985; Mitani, 1989; Moury and Jacobson,1989). One approach to these problems is to use aculture system of early gastrula cells in which theinductive differentiation of neural lineages can beobtained reproducibly from undetermined ectodermcells. We recently developed a microculture system ofearly gastrula cells from Xenopus laevis (Mitani andOkamoto, 1989a) in which the autonomous differen-tiation of epidermal and muscular lineages proceedsfrom cells of the ventral ectoderm (VE) and themarginal zone (MZ), respectively. Here we report theapplication of this microculture method to the co-culturing of cells from these two regions of the gastrulato analyze the inductive differentiation of the neurallineages in a quantitative way.

Cells from the dorsal MZ (DMZ), which give rise tothe middle part of the dorsal mesoderm, have thehighest activity to induce neurons of neural tube origin(CNS neurons), whereas cells from the remaining MZs,which contribute the marginal part of the dorsalmesoderm in Xenopus (Keller, 1976), show a higheractivity to induce melanophores than those from DMZ.Gose association between VE and MZ cells is requiredfor both types of inductive cellular differentiation andthe inducing signals appear to be locally effective.These results suggest that neural tube and neural crestlineages are induced in normal development separatelyby a localized action of the different parts of the dorsalmesoderm that underlie the prospective neural plateand neural fold regions of ectoderm, respectively.Preliminary accounts of the present study appeared inan abstract form (Okamoto and Mitani, 1987; Mitaniand Okamoto, 19896).

Materials and methods

Animal careMethods for keeping frogs and for obtaining embryos havebeen described previously (Mitani and Okamoto, 1988).

Microculture system for early gastrula cellsEarly gastrula embryos of Xenopus laevis (stage 10} or 10iaccording to Nieuwkoop and Faber, 1967) were used.Methods for cultunng early gastrula cells were essentially asdescribed before (Mitani and Okamoto, 1989a) and aresummarized below. Up to 40 dejellied embryos were dissectedin a modified Barth solution (MBS; Gurdon, 1977) to isolatethe ventral ectoderm (VE) and portions of the marginal zone(MZ). These fragments were disaggregated by incubating inCa2+, Mg2+-defiaent MBS containing 1% bovine serum

albumin (BSA, Fraction V, Wako) and 20/igml 1 gentamy-cin, both being mcluded m all subsequent media. After theouter pigmented layer of each fragment was removed and theremaining cells were thoroughly dissociated and resuspendedin standard MBS (pH7.25), the desired number of cells(5-150 in 10-20 jA) from each gastrula region were distributedseparately or in combination (co-culture) into each microcul-ture well of Terasaki plates (total 200-400 wells for a series ofexperiments). Unless otherwise indicated, the plates werecentrifuged at a low speed (max. 600 revs mm"1) to facilitatereaggregation of cells in each microculture well. The cellswere washed with 20 fA of fresh MBS (pH7.75) over a periodof 3 to 4h. After the final centrifugation, each culture wellcontained 12 /il of culture medium (standard culture con-dition). For culture periods longer than 24 h, half the culturemedium was substituted by 67 % RPMI 1640 supplementedwith 10 mM NaHCO3 after around 20 h of culture andsubsequently a 1:1 mixture of this medium and standard MBSwas used for substitution of half the culture medium everyother day.

Staining of cultured cellsCultures were terminated by the addition of 20 fA of fixative(0.5 % paraformaldehyde in BSA-free MBS containing 0.05to O.Wmgml"1 poly-L-lysine) After approximately lhincubation on ice, the same volume was removed from eachwell and fresh 20/A fixative was added. After fixation forfurther 9 to 17 h on ice, culture plates werewashed by gentledipping into PBS, PBS containing 50 mM glycine and Tns-buffered saline (TBS) successively over at least 2 days in arefrigerator.

For indirect lmmunofluorescence, each culture wellreceived 10/i of a first layer mAb containing 1% (for E3,specific for epidermis; Mitani and Okamoto, 1989a), 0.25%(for Mul, specific for myotomal muscle; Mitani and Oka-moto, 1989a), 0.1 % (for Mel, specific for melanophores andretinal pigment cells; Mitam, 1989) or none (for Nl, specificfor CNS neurons; Mitani and Okamoto, 1988) of NP-40 Afterincubation for 2h at room temperature, the plates werewashed by gentle dipping into TBS solution. The washingsolution was changed several times over 4 h (for E3, Mel andMul) or overnight (for Nl) at 4°C. The plates were finallyremoved from the TBS solution and excess TBS in the plateswas aspirated. Each well then received 10[A of FTTC-conjugated second layer antibody (MBL) at an appropriatedilution: affinity-purified rabbit anti-mouse IgG (specific forH+L-chains) containing 0 25% (for E3 and Mul) or 0.1%(for Mel) NP-40, or rabbit anti-mouse IgM (specific for fi-chain) for Nl. After lh incubation at room temperature andwashing as before, DAPI-staining and photographic record-ing were done as previously described (Mitani and Okamoto,1989a).

When cultured cells were doubly stained with mAbs(Nl-Mel, N1-E3, Mel-E3 or Mul-E3 combinations), theywere first stained with Nl, Mel or Mul and the stainingrecorded. The cultures were then stained with the secondmAb (Mel or E3). This reiterative double staining wascarried out because all mAbs belong to the same unmuno-globulin class (IgG) except for Nl which belongs to IgM class.FTTC-conjugated rabbit anti-mouse IgG was used as secondlayer antibody in both the first and second stainings forMel-E3 and Mul-E3 combinations, whereas FTTC-conju-gated rabbit anti-mouse IgM for the first Nl and rhodamine-conjugated goat anti-mouse IgG (Cappel) for the second Melor E3 staining were used in Nl-Mel and N1-E3 combi-nations, respectively.

Neural induction in cell culture from Xenopus 23

Results

Reconstitution of neural induction in a microculturesystemIn the previous study, we used a microculture system todemonstrate the autonomous differentiation of ventralectoderm (VE) cells and marginal zone (MZ) cells fromearly gastrula (Mitani and Okamoto, 1989a). In thisreport, we show that the co-culture of these cells resultsin the induction of neural cells; during co-culture theundifferentiated ectoderm cells give rise to neural cellsinstead of epidermal cells.

The MZ (prospective mesoderm region) of earlygastrula (stage 10£) was divided operationally into 4

portions along the dorsoventral axis of embryo asillustrated in Fig. 1A. Fate mapping study has shownthat all these 4 portions contribute more or lesssubregions of the dorsal mesoderm (Keller, 1976). Thepotency of cells from each region of MZ to support thedifferentiation of CNS neurons in the co-culture withVE cells from stage 10i was examined using an mAb Nlas a probe. The spatiotemporal specificity of thisantibody has been extensively studied in Xenopuslarvae (Mitani and Okamoto, 1988); it shows a highaffinity for neurons of neural tube origin both in situ andin vitro. The results are shown in Fig. IB (someexamples of typical cultures and their staining) andFig. 1C (cumulative data). Reproducible neuronal

0 50 100 150Initial number of MZ cells in culture

Fig. 1. Inductive differentiation of neuronal cells in a microculture system. (A) Portions of the marginal zone (MZ) used inour experiments. Vegetative view of an early gastrula (stage 10i). DMZ, dorsal marginal zone; DLMZ, dorsolateralmarginal zone; VLMZ, ventrolateral marginal zone; VMZ, ventral marginal zone. (B) Neuronal cell differentiation in co-cultures. Deep layer cells from the ventral ectoderm (VE, 150 cells) were cultured with those from DMZ (100 cells,al-a3), DLMZ (100 cells, bl-b3) or VLMZ (150 cells, cl-c3) These cultures were fixed after incubation for 138h andexamined for the expression of Nl antigen (a3-c3). (al-cl) Phase-contrast appearance; (a2-c2) nuclear staining withDAPI. Bar=50/an (C) Dorsoventral distribution of MZ potency to induce neuronal cell differentiation. Data werecollected from a single series of experiments. The proportion of positive culture for the Nl antigen is plotted against initialnumber of cells in culture from the four regions of MZ indicated in A; • , DMZ, O, DLMZ; • , VLMZ; V, VMZ. 12-26culture wells were examined for each point.

24 5. Mitani and H. Okamoto

differentiation is obtained only in co-cultures of cellsfrom the VE and dorsal MZ (DMZ), being dependenton the initial cell number of the latter added in a culturewell, as judged from the Nl antigen expression(Fig. 1C). Neither VE nor DMZ cells have detectableNl binding when cultured on their own (data notshown). Cells from dorsolateral MZ (DLMZ) alsoinduce Nl staining in co-cultures with VE cells, but thefrequency (Fig. 1C) and extent of the antigen ex-pression are considerably less than found in DMZ co-cultures (compare Fig. lBa3 with lBb3). There isevidently an inclined distribution of the activity toinduce neuronal cell differentiation within the MZ ofearly gastrula. It is highest in the dorsal portion of MZand decreases rather sharply towards the ventral side.Fig. 1C gives a quantitative measure of this inclineddistribution.

To trace DMZ cells in co-cultures, they were labelledby tetramethyl rhodamine isothiocyanate (Heasman etal. 1984). When 20 to 80 of the labelled DMZ cells wereco-cultured with 150 unlabelled VE cells, most of DMZcells were seen to sort out as one or two clumps and

extrude from a central mass of VE cells within 12 h ofculture. Since Nl stain always resided in the differen-tiated VE cell mass, a large proportion of neuronal cellswas certainly derived from VE cells. However, in mostcases a small number of labelled DMZ cells remained inthe VE cell mass until the end of culture. The possiblecontribution of these DMZ cells themselves to theneuronal cell population, though little if any, is nottotally excluded at the moment (Okamoto, unpublishedobservation).

We next analyzed the potency of each region of MZto suppress epidermal differentiation in the co-culturesystem with an mAb E3. This antibody is highly reactivewith differentiated epidermal cells both in situ and incultures composed of VE cells alone (Mitani andOkamoto, 1989a and also Fig. 2Aa3). Cells from DMZinhibit E3 antigen expression partially (Fig. 2Ab3) orcompletely (Fig. 2Ac3) when co-cultured with VE cells,whereas cells from other sections of the MZ to do so toa slight extent (exemplified in Fig. 2Ad3) or actuallypromote the expression (data not shown). A quantitat-ive analysis reveals an inclined distribution of activity

0 50 100Initial number of MZ cells in culture

Fig. 2. Differentiation of epidermal cells in a microculture system. (A) Epidermal differentiation in culture. Deep layercells from the VE were cultured alone (150 cells, al-a3) or in combination with those from DMZ (50 cells, bl-b3 andcl-c3) or VMZ (50 cells, dl-d3). Cultures were fixed after incubation for 40h and examined for the expression of E3antigen (a3-d3). (al-dl) Phase-contrast appearance; (a2-d2) nuclear staining with DAPI Bar=50/an. (B) Dorsoventraldistribution of MZ potency to suppress epidermal differentiation Data were collected from a single series of experiments.The proportion of cultures positive for the E3 antigen is plotted against initial number of cells in culture from the four MZregions; • , DMZ;O, DLMZ, • , VLMZ; V, VMZ. 13-21 culture wells were examined for each point.

Neural induction in cell culture from Xenopus 25

within the MZ again (Fig. 2B). The potency to suppressepidermal differentiation is highest for dorsal MZ andsharply decreases in more ventral MZ. The polarity andextent of the inclination seem fairly parallel with thoseobserved for the MZ ability to induce neuronal celldifferentiation.

DMZ is the source of the middle part of dorsalmesoderm (Keller, 1976), which is known to beresponsible for the induction of the neural tissue fromthe overlying ectoderm in normal development. Thispart of ectoderm could differentiate into epidermis,when experimentally freed from the influence of thedorsal mesoderm (Holtfreter, 1938; Kintner and Mel-ton, 1987). The data described so far, thus, stronglysuggest that our co-culture system can reconstitutenormal neural induction in vitro.

Inductive differentiation of melanophores in co-cultureDuring the present study, we frequently observedpigmented cells differentiating in co-cultures, notnecessarily with DMZ cells but with other MZ cells.These pigmented cells were thought to be melano-phores, which are one of the derivatives of neural crest.We then examined whether the neural crest lineage isalso induced by MZ from undifferentiated ectodermand which part of MZ has the highest activity to inducethe lineage. An mAb Mel was used as a marker for thispurpose, which has a high affinity for melanophores ofneural crest origin and retinal pigment cells in Xenopusembryos (Mitani, 1989).

Typical staining patterns of various co-cultures areshown in Fig. 3Aa to d and 3Ba3. Mel-positive cellshave a star-like appearance and dispersed throughoutthe culture well. Most of these cells contain blackpigment (Fig. 3Bal). All these traits are characteristicsof neural crest-derived melanophores in situ. Cells fromneither the VE nor the MZ cultured on their own yieldMel-positive cells, as exemplified in Fig. 3Bb3 and c3.The degree of melanophore differentiation depends onthe number of MZ cells initially added to culture wellsas judged from the proportion of positive wells for theMel-antigen expression (Fig. 4A) or from the meannumber of Mel-positive cells per well directly counted(Fig. 4B). When the initial number of VE cells is 150,about 20 VMZ or VLMZ cells are sufficient to inducemelanophores in every culture well examined, yieldingan average of 25 Mel-positive cells per well. Theseresults strongly suggest that the differentiation ofmelanophores also needs inductive cellular interactionbetween ectoderm and MZ, as in case of that of CNSneurons.

The potency of cells from the MZ to inducemelanophores is highest for ventral tissue and graduallydecreases as more dorsal tissue is used (Fig. 4). Thispolarity is opposite to that observed for the potency toinduce CNS neurons (Fig. 1C) or to inhibit epidermalcell differentiation (Fig. 2B). These results indicate thedifference in developmental mechanisms for the induc-tion of neural crest and neural plate lineages and raisethe interesting possibility that a significant portion ofthe neural crest lineage (probably for its trunk to tail

region) is induced by the action of mesoderm other thanthe chordamesoderm which is derived from DMZ and ismainly responsible for the induction of neural tubelineage (see Discussion).

Increasing the number of DMZ cells in co-culturesometimes results in the production of a large aggregateof Mel-positive cells containing black pigment(Fig. 3Ca, c) instead of dispersed melanophores. Theseaggregates were seen only occasionaly in co-cultureswith increased numbers of DLMZ cells, but never in co-cultures with VMZ or VLMZ cells. It is tempting tospeculate that the aggregates contain retinal pigmentcells that can also be recognized by Mel, a possibilitythat will be tested in future studies.

Co-differentiation of multiple lineages of ectoderm ina single culture wellHaving reconstituted the inductive differentiation ofneural lineages in the co-culture system, we investi-gated the cellular mechanisms underlying the inductionprocesses, in particular, whether the inducing signal(s)from MZ cells readily diffuses throughout the culturemedium, affecting all of the ectodermal cells in the well,or is locally active and affects only some of the cells.

Our data support the hypothesis of a localized actionof the inducing signal(s). Cells derived from neural tubeand epidermal lineages, for instance, can be seen in thesame culture well, though at separate sites within a cellaggregate as demonstrated by double staining(Fig. 5Ac, d). However, often the E3-positive epider-mal cells had mostly disappeared by the time the Nl-positive neuronal cells were detectable, possibly be-cause of the lack of support from connective tissue cells.To avoid this difficulty, we tested for multiple cell typesin duplicate sets of co-cultures. A combination of 25DMZ and 150 VE cells yields Nl antigen expression inall wells examined in one set and E3 antigen expressionin 80% wells in the other set of co-cultures (Fig. 5C);we may expect co-differentiation of CNS neurons andepidermal cells in a single culture well at the rate of80%. In another series of experiments shown inFig. 5B, co-culturing of DMZ and VE cells in the samecell number combination results in the expression ofboth Nl and E3 antigen in about 70 % culture wells induplicate sets of co-cultures; we may expect a rate of co-differentiation of at least 40% in this series. About 5DMZ cells are sufficient to activate Nl antigenexpression in more than 20 % of E3-positive cultures(Fig. 5B), whereas even 100 DMZ cells are not enoughto inhibit E3-antigen expression in all of Nl-positivecultures (Fig. 5C). These results suggest that CNSneurons co-differentiate with epidermal cells in a singleculture well more or less over a wide range of numbersof inducing DMZ cells. Further, changing culturevolume per well (6 fA to 24 /il) in the DMZ cell titrationexperiment (Fig. 5B) does not appear to affect signifi-cantly the titration profile.

We also expected the coexistence of cells derivedfrom neural crest and epidermal lineages from acomparison of titration curves shown in Fig. 2B and 4.Indeed, cultures with VMZ, VLMZ or DLMZ cells,

26 S. Mitani and H. Okamoto

Fig. 3. Inductivedifferentiation ofmelanophores in amicroculture system. (A) 150VE cells were co-cultured withcells from DMZ (a), DLMZ(b), VLMZ (c) or VMZ (d).After incubation for 62 h,cultures were fixed andexamined for the expression ofMel antigen The cultureshown in d was successivelystained with E3 to show thepresence of an epidermal cellaggregate (e). The focal planesin d and e are slightlydifferent. (B) 150 VE cells and20 VMZ cells were co-cultured(al-a3) or cultured on theirown (VE cells alone, bl-b3;VMZ cells alone, cl-c3).Incubation time was 62 h.(al-cl) Phase-contrastappearance; (a2-c2) nuclearstaining with DAPI; (a3-c3)Mel staining. (C) 150 VE cellsand 50 DMZ cells were co-cultured. After incubation for130 h the culture was fixed anddoubly stained with Mel (c)and Nl (d). (a) Phase-contrastappearance; (b) nuclearstaining with DAPI.Bar=50/an in A-C

100

3 50 ->

oOL.

Neural induction in cell culture from Xenopus 27

A a t

10 20Initial number of MZ cells in culture

10 20Initial number of MZ cells in culture

Fig. 4. Ventrodorsal distribution of MZ potency to inducemelanophore differentiation Data were collected from theexperiments in Fig. 3A, B. The proportion of Mel-positivecultures (A) or the mean number of Mel-positivemelanophores per culture well (B) is plotted against initialnumber of MZ cells added to the culture: • , DMZ; O,DLMZ; • . VLMZ, V, VMZ. 12-21 culture wells wereexamined for each point.

which stain with the melanophore-specific antibodyMel (Fig. 3Ab to d), also have E3-positive epidermalcells (Fig. 3Ae). Double staining with Mel and Nlantibodies revealed coexistence of the two neurallineages in a single culture well composed of VE andDMZ cells (data not shown), though its frequency waslow as expected from titration profiles (Fig. 4 andFig. 5B). When the large Mel-positive aggregates areseen (Fig. 3C), some regions are Mel-positive(Fig. 3Cc) and others are Nl-positive (Fig. 3Cd).

The most straightforward explanation of the abovedata is that the inducing signals from MZ cells for thetwo neural lineages are effective only over a shortdistance within a cell aggregate in culture.

Requirement of intimate cell association for inductiveaction of MZ cellsBy aggregating cells by centrifugation, we have alsodemonstrated that close association of VE and MZ cellsis required for the induction of neural lineages. WhenVE and appropriate MZ cells are aggregated, both CNSneurons and melanophores differentiate (exp. 1 inFig. 6A, B). However, when MZ cells are preaggre-gated about 30min before the addition of dissociated(exp. 2) or preaggregated (about 30min, exp. 3) VE

0 10 20Initial number of DMZ cells in culture

o so IOOZ Initial number of DMZ cells in culture

Fig. 5. Co-differentiation of neuronal and epidermal cellsin a single culture well (A) 150 VE cells were co-culturedwith 24 DMZ cells. After 130 h incubation they weredoubly stained with Nl (c) and E3 (d). (a) Phase-contrastappearance; (b) nuclear staining with DAPI. Bar=50//m.(B and C) 150 VE cells were co-cultured with increasingnumbers of DMZ cells (two independent series ofexperiments differing in the number of DMZ cell added).Half of the cultures in each series were fixed after 40 hincubation and examined for the expression of E3 antigen,whereas the rest were fixed after 130 h incubation andexamined for the presence of Nl antigen. The proportionof E3-positive (O, A, D), or Nl-positive ( • , A, • ) wellsis plotted against initial number of DMZ cells in culture.Culture volume was 6 (O, • ) , 12 (A, A) or 24 ( • , • )^lwell"1. 13-20 culture wells were examined for eachpoint.

cells, far fewer CNS neurons (Fig. 6A) and melano-phores (Fig. 6B) arise. Neuronal or melanophoredifferentiation is restored when the two aggregates arebrought into close contact by recentrifugation (exp. 4 inFig. 6A, B).

28 5. Mitani and H. Okamoto

exp. 1 exp. 2 exp. 3 exp. 4

•lOO-i

I 50 -

0-

glOOnon

aO

g 50H

o

0-

B

Fig. 6. Dependence of neuronal cell and melanophoredifferentiation on intimate cell association. 150 VE cellswere co-cultured with 100 DMZ cells for 129 h (A) or with50 lateral marginal zone (DLMZ+VLMZ) cells for 61 h(B) according to the four experimental protocols illustratedat the top of the figure (exp. 1 to 4; O, VE cell; • , MZcell; see text for detail). Columns; proportion of Nl-positive (A), or Mel-positive (B) cultures. 12-20 culturewells were examined for each column.

Discussion

One of the main results of the present study is thesuccessful reconstitution of inductive differentiation ofnot only neural tube but also neural crest lineages fromearly gastrula cells in a microculture system. Togetherwith our previous demonstration of the autonomousdifferentiation of the epidermal lineage in culture(Mitani and Okamoto, 1989a), we can now follow thedifferentiation of all three major lineages arising fromXenopus gastrula ectoderm in a microculture system.

Inductive differentiation of neural tube lineage in amicroculture systemA sharp dorsoventral distribution of potency is seen inthe ability of MZ cells to induce neuronal cells of neuraltube origin (Fig. 1C). This distribution is in agreementwith both classical studies and more recent reports thatindicate that DMZ cells can serve as a strong organizerfor the CNS (Spemann, 1938; Akers et al. 1986;Jacobson and Rutishauser, 1986; Kintner and Melton,1987; Sharpe et al. 1987; Jones and Woodland, 1989).VE cells never give rise to CNS neurons when culturedalone (Fig. 1C) and lose their competence to respondDMZ cells to yield CNS neurons in the course ofisolated culture (data not shown). Furthermore our dyemarking experiments in culture suggest that a largeproportion of CNS neurons arise from VE cells, but not

from DMZ cells. Thus, our culture system of dis-sociated and reaggregated cells appears to supportnormal neural induction.

Deuchar (1970) co-cultured explants and groups ofcells from Xenopus early gastrula and showed that forsuccess of neurahzation a certain number of ectodermcells, but not of dorsal lip cells is critical. Althoughquantification of cell number and neural differentiationwas not necessarily thorough in the study, herobservation appears to be fairly consistent with ours(Fig. 5B and data not shown). On the other hand, someresults of previous reports (Barth and Barth, 1959;Godsave and Slack, 1989) apparently contradict thepresent ones. They observed that neural cells differen-tiated from cultured amphibian ectoderm cells in theabsence of MZ cells. However, Barth and Barth (1963)showed later that substance(s) released from the agarsubstrate at the bottom of operating dish had theinducing activity to Rana pipiens ectoderm cells. Theyalso obtained the neural induction from Rana ectoderm(clumps of about 125 cells) exposed to various saltsolutions (Barth and Barth 1963, 1969), but the highlevel of Ca2+, Mg2"1" or Mn in these solutions seemsrather unphysiological. Godsave and Slack (1989)cultured Xenopus embryonic cells in microculture wellsas in our experiments. There are, however, somedifferences in culture conditions between the twostudies. Godsave and Slack used animal cap cells fromstage 8 (age 5 h) instead of VE cells from stage 10J (age10 h) as in our experiments, culture wells coated withlaminin and fibronectin instead of uncoated wells andy-globuhn (fraction n) from bovine plasma in culturemedium instead of bovine serum albumin. Amongthese, the difference of age of embryos used seems mostimportant; the state of commitment of the animal capcells is likely to be unstable, whereas that of the VEcells is already biased toward the epidermal differen-tiation. Isolated culture of the animal cap cells in thepresence of y-globulin and some extracellular materialsmay favor their commitment to the neural direction.

Since the number of cells and extent of neuraldifferentiation can be quantified and also the manner ofcellular interactions can be easily modified, we believethat the culture system described in the present studyopens up the possibilities of analysing the normalcourse of neural induction quantitatively under avariety of experimental conditions, including that fortesting inducing substances.

Inductive differentiation of neural crest lineage in amicroculture systemWe have shown that cells from a broad MZ sectorsupport melanophore differentiation of neural crestorigin in co-cultures with VE cells (Fig. 4). Cells fromeither the VE or MZ alone yield no melanophores mculture. Further, preculturing of VE cells on their own,which leads to the autonomous epidermal differen-tiation, suppresses their competence to respond to MZcells to produce melanophores as in case of neuronalcell induction (data not shown). These results suggestthat separation between epidermal and neural crest

Neural induction in cell culture from Xenopus 29

lineages is brought about by the inductive action of MZcells on undetermined ectoderm cells. However, it isalso shown that ventral and not dorsal MZ cells bestsupport melanophore differentiation. This difference inthe induction of melanophores and CNS neurons mayreflect differences in the inducing mechanisms of neuralcrest and neural tube lineages, as previously suggested(Mitani, 1989). Mitani showed that melanophoresdifferentiated abundantly even in the larvae whose CNSwas severely damaged by the injection of heparin ordextran sulfate into the blastocoel cavity of Xenopusearly embryos.

Developmental mechanism for the formation of neuralcrest.Neural crest cells are known to arise from a part of theneural fold that delineates the neural plate in vertebratedevelopment (Slack, 1983). Both the neural plate andfold structures are induced from the part of ectodermthat is underlaid by the dorsal mesoderm duringgastrulation. Since we have shown that close associationbetween cells from VE and MZ (prospective mesodermregion) is required for the induction of both neural crestand tube lineages, it is highly likely that the marginalpart of dorsal mesoderm that underlies the prospectiveneural fold region of ectoderm at late gastrula producesinducing signal(s) for the neural crest lineage; the signalis locally active and, thus, limits the spatial extent of theneural plate which is separately induced by the localizedaction of the remaining central part of dorsal mesoderm(Jones and Woodland, 1989).

We have obtained an unexpectedly high inducingactivity of VMZ cells. However, understanding of theintricate morphogenetic movement that forms thedorsal mesoderm from MZs during Xenopus gastru-lation, can explain our present results; in Xenopus,precursor cells for the dorsal mesoderm are present in abroad MZ sector (dorsal to ventralmost) of the earlygastrula (Keller, 1976; Gerhart and Keller, 1986). Cellsof the involuting MZ of early gastrula converge stronglytoward the dorsal midhne with concomitant extensionin the animal-vegetal direction that corresponds to theanteroposterior axis of the late gastrula, after turninginside the gastrula (Gerhart and Keller, 1986). Thisconvergent extension of MZ cells begins first and is mostvigorous at the prospective dorsal midline of embryoand spreads laterally. As a result of these complicatedcellular movements, the dorsal mesoderm of lategastrula is not necessarily derived solely from DMZ butalso from other MZs of the early gastrula; the wholecephalic region and the central part of the trunk-to-tailregion of the dorsal mesoderm arising from DMZ,whereas the remaining marginal part of the trunk-to-tailregion of the dorsal mesoderm derives from other MZsincluding VMZ (Keller, 1976). Thus, if the marginalpart of the dorsal mesoderm, which underlies theprospective neural fold region of ectoderm, induces theneural crest, as we have suggested, it is not surprisingthat cells from a broad MZ sector of early gastrulaincluding VMZ are more or less capable of supportingmelanophore differentiation. It is well known that head

and trunk neural crest have a different set of derivatives(Hopwood et al. 1989). The ventrodorsal distribution inthe potency of MZ cells that we have observed formelanophore differentiation may be due to this regionaldifference in the developmental fate of neural crestcells; more caudal neural crest cells would give rise tomore melanophores. If DMZ cells exhibited theirability to induce neuronal cells in preference tomelanophores, this would also contribute to theventrodorsal distribution in the MZ cell potency.

There could, of course, be other explanations for theinducing potency of VMZ cells observed in vitro. VMZcells for instance gain the inducing potency as a result ofthe in vitro operative procedures. Dale and Slack (1987)have presented evidence that ventrovegetal blasto-meres from Xenopus embryos can have a dorsalinductive potency partially activated by microsurgery.However, they have also shown that by stage 7 theseventrovegetal blastomeres are less sensitive to theoperative procedures. Furthermore, we explored theartificial effect of our in vitro procedures on MZ cellsfrom stage 10£ gastrula by examining the potency ofthese MZ cells to differentiate into muscle cells on theirown. Cells (50/well) from each of the four MZ portionswere cultured under the same conditions as described inthis article and differentiated muscle cells were quanti-fied using an mAb Mul as a specific probe (Mitani andOkamoto, 1989a). Mean numbers of Mul-positivemuscle cells per well were 7.1±1.4 for VMZ cells,7.6±0.9 for VLMZ cells, 19.6±2.0 for DLMZ cells and0.36±0.19 for DMZ cells (mean±s.E.). The relativeratio of these figures agrees with that expected from thefate map (Keller, 1976). Thus, although the possibilityis not totally excluded that we overestimate the activityof VMZ cells to some extent due to our operativeprocedures, its degree is likely to be small. It is alsopossible that the ectoderm is sensitized to differentiatemore readily into melanophores by the operativeprocedures. However, this mechanism of sensitizationdoes not necessarily lead to overestimation of theactivity of VMZ cells, as compared with those of cellsfrom other MZs. When MZ cell activities werecompared, cells from each of the four MZ portions wereco-cultured with ectoderm cells that were derived fromthe same large pooled cell suspension.

There have been so far, at least, two different viewson the neural crest formation from ours (Nieuwkoop,1985; Moury and Jacobson, 1989). Nieuwkoop hassuggested that the neural crest is formed in theperipheral region of the neural anlage when theneuroectoderm begins to lose its competence forthe inducer that is propagating laterally from thechordamesoderm of the dorsal midline. According tothis view, we could expect that the potency of MZ tosupport melanophore differentiation should be sharplybiased toward a DMZ sector, because the chordameso-derm is mainly derived from the DMZ. However, theresults we have obtained are totally inconsistent withthis expectation. Since the view assumes that a ratherweak interaction between the inducer and respondingcells results in the neural crest formation, it also

30 S. Mitani and H. Okamoto

predicts that a smaller number of DMZ cells would beneeded for melanophore differentiation than for neur-onal cell differentiation and that the preculturing of VEcells before addition of MZ cells, which has led to a lossof their competence to an appropriate extent, wouldfacilitate melanophore differentiation. However, theseare again not supported in our experiments.

Moury and Jacobson (1989) claim that the localinteraction of neural and epidermal tissue is responsiblefor the formation of neural crest; when a lateral piece ofneural plate was transplanted into the ventral epider-mis, it gave rise to melanophores. However, otherexplanations could account for the observation. Theneural plate transplants that they used may havealready received the inducing signal locally from theunderlying dorsal mesoderm. It would be, then,possible that these transplants, especially their lateral-most portions, autonomously produced the neuralcrest, when placed in the ventral epidermis andremoved far from the influence of the chordamesodermwhich strongly favors the induction of neural tubelineage. It is also noteworthy that according to thissecond view, we should obtain the greatest degree ofdifferentiation of melanophores in co-cultures of VEcells with DMZ cells, because in this combinationepidermal and neuronal cells co-differentiated mostefficiently. However, the results we have obtained areinconsistent with this prediction.

Signal transmission in neural lineage inductionThe second major result of this study is that closeassociation between cells of the VE and MZ is essentialfor the induction of both neural tube and crest lineages(Fig. 6). Moreover, preliminary experiments to showan effect of conditioned culture medium from MZ cellson induction of neural lineages failed to do so(Okamoto, unpublished observation), in contrast to theprevious result (Niu and Twitty, 1953). Together withour demonstration of localized action of the inducingsignals within a cell aggregate, readily diffusiblesubstances from MZ cells do not seem to play a majorrole in the induction processes.

The requirement of close cell contact for the 'neuralinduction' has also been indicated using embryonicpieces from Xenopus laevis (Tacke and Gruntz, 1988),or cleavage-arrested blastomeres from Halocynthiaroretzi, a sea squirt (Okado and Takahashi, 1988).These and our data indicate that inducing signals aretransmitted via either direct cell-to-cell interactions orsubstances that are diffusible but effective only over ashort distance like neurotransmitters at synaptic junc-tions in the nervous system or some of growth factorsthat are sequestered to extracellular material. In theformer case, molecules associated with outer surface ofMZ cells, for instance, could bind to surface receptorson VE cells and initiate the induction. The transfilterexperiments of Toivonen et al. (1975) seem to be inaccord with the latter idea, although the extent ofneural differentiation is pointed out to be poor(Warner, 1985). Examination of the inducing activities

of killed MZ cells or membrane fractions from them willbe helpful to decide between the two possibilities.

We wish to thank Professor K. Takahashi for his helpfulsuggestions and discussions throughout this work. We are alsograteful to Dr M. Chalfie for his critical reading of themanuscript. This work was supported by grants in aid forscientific research from the Ministry of Education, Scienceand Culture, Japan.

References

AKERS, R M , PHILLIPS, C R. AND WESSELLS, N K (1986).Expression of an epidermal antigen used to study tissueinduction in the early Xenopus laevis embryo Science 231,613-616.

BARTH, L G AND BARTH, L J (1959) Differentiation of cells ofRana pipiens gastrula in unconditioned medium / Embryolexp Morph 7, 210-222

BARTH, L G AND BABTH, L. J (1963) The relation betweenintensity of inductor and type of cellular differentiation of Ranapipiens presumptive epidermis Bio! Bull mar biol Lab ,Woods Hole 124, 125-140

BAR™, L. G AND BARTH, L J (1969). The sodium dependence ofembryonic induction Devi Biol 20, 236-262

DALE, L AND SLACK, J M W (1987) Regional specificationwithin the mesoderm of early embryos of Xenopus laevisDevelopment 100, 279-295.

DEUCHAR, E M (1970) Neural induction and differentiation withminimal numbers of cells. Devi Biol 22, 185-199.

GERHART, J AND KELLER, R (1986). Region-specific cell activitiesin amphibian gastrulation A Rev. Cell Biol 2, 201-229.

GODSAVE, S. F AND SLACK, J M W. (1989) Clonal analysis ofmesoderm induction in Xenopus laevis Devi Biol 134, 486^490

GURDON, J B (1977). Methods for nuclear transplantation inAmphibia.' Meth Cell Biol 16,125-139.

HEASMAN, J , WYLIE, C C , HAUSEN, P AND SMITH, J C (1984)Fates and states of determination of single vegetal poleblastomeres of Xenopus laevis Cell 37, 185-194

HOLTFRETER, J (1938). Differenzirungspotenzen isolierter Teile derUrodelengastrula Wilhelm Roux' Arch EntwMech Org 138,522-656

HOPWOOD, N D , PLUCK, A. AND GURDON, J B (1989). AXenopus mRNA related to Drosophila twist is expressed inresponse to induction in the mesoderm and the neural crestCell 59, 893-903

JACOBSON, M AND RUTISHAUSER, U. (1986) Induction of neuralcell adhesion molecule (NCAM) in Xenopus embryos. DeviBiol 116, 524-531.

JONES, E A AND WOODLAND, H R (1989) Spatial aspects ofneural induction in Xenopus laevis Development 107, 785-791

KELLER, R. E (1976) Vital dye mapping of the gastrula andneurula of Xenopus laevis II. Prospective areas andmorphogenetic movements of the deep layer Devi Biol 51,118-137

KINTNER, C R AND MELTON, D. A (1987) Expression ofXenopus N-CAM RNA in ectoderm as an early response toneural induction. Development 99, 311-320.

MITANI, S. (1989) Retarded gastrulation and altered subsequentdevelopment of neural tissues in hepann-injected Xenopusembryos Development 107, 423-435

MITANI, S AND OKAMOTO, H (1988) Monoclonal antibodiesagainst larval nervous system of Xenopus laevis, theirspecificities and application to analysis of neural developmentNeuroscience 25, 291-305.

MITANI, S. AND OKAMOTO, H. (1989a). Embryonic development ofXenopus studied in a cell culture system with tissue-specificmonoclonal antibodies Development 105, 53-59.

MITANI, S AND OKAMOTO, H (19896). The differentiation ofmelanophore in vitro from embryonic cells of Xenopus laevisJpn J Physiol 39 suppl., S27

MOURY, J. D AND JACOBSON, A. G. (1989) Neural fold formation

Neural induction in cell culture from Xenopus 31

at newly created boundaries between neural plate and epidermisin the Axolotl. Devi Biol. 133, 44-57.

NIEUWKOOP, P. D. (1985). Inductive inyteractions in earlyamphibian development and their general nature. /. Embryol.exp. Morphol. 89 suppl., 333-347.

NIEUWKOOP, P. D. AND FABER, J. (1967). Normal Table ofXenopus laevis (Daudin). Amsterdam: North-HollandPublishing Company.

Niu, M. C. AND TWTTTY, V. C. (1953). The differentiation ofgastrula ectoderm in medium conditioned by axial mesoderm.Proc. nam. Acad. Sci. U. S. A. 39, 985-989.

OKADO, H. AND TAKAHASHI, K. (1988). A simple 'neural induction'model with two interacting cleavage-arrested ascidianblastomeres. Proc. nam. Acad. Sci. U.S.A. 85, 6197-6201.

OKAMOTO, H. AND MITANI, S. (1987). Reconstitution of the neuralinduction process in vitro. Neurosci. Res. suppl. 5, S61.

SCHROEDER, T. E. (1970). Neurulation in Xenopus laevis. Ananalysis and model based upon light and electron microscopy. /.Embryol. exp. Morph. 23, 427-462.

SHARPE, C. R., FRITZ, A., DEROBERTIS, E. M. AND GURDON, J. B.(1987). A homeobox-containing marker of posterior neuraldifferentiation shows the importance of predetermination inneural induction. Cell 50, 749-758.

SLACK, J. M. W. (1983). From Egg to Embryo. DeterminativeEvents in Early Development. Cambridge: Cambridge UniversityPress.

SPEMANN, H. (1938). Embryonic Development and Induction. NewHeaven: Yale University Press.

SPEMANN, H. AND MANGOLD, H. (1924). Ober Induction vonEmbryonalanlagen durch Implantation artfremder Organisation.Arch, mikrosk. Anat. EntwMech. 100, 599-638.

TACKE, L. AND GRUNZ, H. (1988). Close juxtaposition betweeninducing chordamesoderm and reacting neuroectoderm is aprerequisite for neural induction in Xenopus laevis. Cell Differ.24, 33-44.

TOIVONEN, S., TARIN, D., SAXEN, L., TARIN, P. J. ANDWARTIOVAARA, J. (1975). Transfilter studies on neural inductionin the newt. Differentiation 4, 1-7.

WARNER, A. E. (1985). Factors controlling the early developmentof the nervous system. In Molecular Bases of Development (ed.G. M. Edelman, W. E. Gall and W. M. Cowan), pp. 11-34.New York: John Wiley and Sons.

(Accepted 28 January 1991)