Effects of HNK-1 monoclonal antibody on the substratum attachment
and survival of neural crest and sclerotome cells in culture
E. J. SANDERS and E. CHEUNG
Department of Wivstologv, University of Alberta, Edmonton, Alberta T6G 2117, Canada
Summary
The sclerotome portion of the differentiating em-bryonic chick somite becomes infiltrated byneural crest cells prior to its dispersal towards theembryonic axis. This means that sclerotome cellsexplanted into culture for the purpose of examin-ing their interactions in vitro are contaminatedwith a proportion of neural crest cells. The pur-pose of this study was to explore the neural crestcell adhesion epitope recognized by the HNK-1monoclonal antibody and ways in which thisantibody can be used to eliminate neural crestcells from mixed culture by selective cytotoxicity.Immunofluorescence technique, under the con-ditions used here, indicated that the antibodyappeared to stain all the mesenchymal (i.e. neuralcrest) cells emigrating from pieces of embryonicneural tube in culture. Examination of the effectsof HNK-1 suggests that the antibody interacts withsubstratum-binding sites on the neural crest cellsurface. On fibronectin-coated substrata the anti-body tended to inhibit neurite outgrowth but leftthe cells relatively well spread, while on laminin
substrata the effect was to discourage both neuriteextension and cell spreading, causing cell retrac-tion. These results suggest that the cell surfaceepitope recognized by HNK-1 influences neuriteoutgrowth, neurite adhesiveness or both. Failureof cell spreading on laminin suggests interactionwith the laminin binding sites on the cell body.Elimination of the crest cells from mixed culturewith sclerotome was achieved by culturing thecells in the presence of HNK-1 antibody andcomplement during the period required for com-plete cell outgrowth from the sclerotome explant.This treatment resulted in cultures that, onfurther incubation, were entirely HNK-1 negativeand that failed to show either the presence ofneurite-bearing cells or of cells reacting positivelyto anti-neurofilament monoclonal antibody. Wethus appear to have succeeded in removing thecrest cells from the mixed population.
Sclerotome cells differentiate from the early avianembryonic somites and break out ventromedially fromthe somites to migrate toward the neural tube andnotochord, where they differentiate further into chon-drocytes. In this location they then lay down cartilageand thus contribute to the formation of the axialskeleton (Hall, 1977; Lash & Vasan, 1978). Theseevents represent a series of well-defined interactionsthat, although complex, appear to involve only alimited set of tissues. For this reason problems ofsclerotome differentiation have engendered muchinterest and have been studied by a variety of tech-niques, many of which have relied on the examination
of sclerotome cells that have been explanted intoculture (Fisher & Solursh, 1979; Bellairs et cil. 1980;Sanders, 1986; Sanders & Prasad, 1983; Newgreen,1984; Newgreen et al. 1986).
Such cultures of dissected sclerotome were earlierconsidered to be uncontaminated by other cell types,especially since pigment cells do not develop underthese conditions. However, more recent observationsshow that neural crest cells infiltrate part of thesclerotome and use it as a major route for their dispersal(Rickmann et cil. 1985), so that this must be taken intoaccount in interpreting the results of in vitro exper-iments involving sclerotome. The development of amonoclonal antibody that specifically interacts with'most or all' early migrating neural crest cells (Vincent
115
& Thiery, 1984), and its identity with the HNK-1antibody, which also recognizes a leucocyte subpopu-lation (Tucker et al. 1984), provides a tool for making adistinction between sclerotome and neural crest cells,both in situ (Bronner-Fraser, 1986) and in vitro (Sternet al. 1986).
The epitope recognized by HNK-1 on neural crestcells is unknown. We have examined the effects of theantibody on cell adhesiveness to different substrata. Wesuggest that its effects are related to its interferencewith cell spreading and neurite extension on lamininand, to a lesser extent, fibronectin substrata.
There remains, however, the difficulty of conductingstudies on pure cultures of sclerotome cells uncontami-nated by cells of an entirely different population.Results obtained in previous studies of 'sclerotome'cultures may indeed by compromised by the un-detected presence of the crest cells. We describe amethod for eliminating crest cells from these mixedpopulations, based on selective cytotoxicity of HNK-1for neural crest cells. We also show that HNK-1apparently binds to all the crest cells emigrating fromneural tube cultures and that no identifiable neural cellsremain in the sclerotome culture after treatment.
Materials and methods
Chick embryos were incubated to bring them approximatelyto stage 17, as described by Hamburger & Hamilton (1951),at which time they showed 25-30 somites. They were freedfrom their yolk, rinsed in Tyrode's saline and incubated for30 min in 0' 1 % trypsin solution in order to loosen the tissuesprior to dissection. The cells were not dissociated from oneanother by this procedure.
Sclerotome tissue from somites number 10-20 was dis-sected out using tungsten needles and transferred in a drop ofsaline onto a cleaned coverslip contained in a 35 mm diameterPetri dish. The dermo-myotome portion of the somite wasremoved and excluded from the explant. The exclusion ofdermo—myotome was verified by reference to the morphologyof cells in the resulting cultures (Bellairs et al. 1980). Thesaline was replaced with several drops of medium 199 plus10 % foetal bovine serum and gentamycin. The Petri dish wascovered and the culture placed in a humid incubator at 37°Covernight.
Neural crest cultures were prepared using pieces of neuraltube from the trunk level opposite the four or five mostrecently formed somites. The pieces were carefully cleanedfree of ectoderm, notochord and sclerotome and incubated asabove. Contaminating tissue could easily be seen adhering tothe tube, so that removal of such tissue could be performedwith confidence. These cultures were incubated for up to 4days before fixing and staining.
After incubation, the coverslips were washed with jets ofwarm Tyrode's saline and fixed with formol saline for 2h.This and all subsequent steps were performed at roomtemperature. The cultures were then washed with phosphate-buffered saline (PBS) and treated with 1 % bovine serum
albumin (BSA) in PBS for 30 min in order to block non-specific staining. After a further wash with PBS the coverslipswere incubated in HNK-1 monoclonal antibody (mouse anti-human leu-7, Becton Dickinson Canada Inc.), diluted 1:100(v/v) in PBS, for 1 h. They were then rinsed again andtreated with polyclonal (Calbiochem) or monoclonal (BectonDickinson Canada Inc.) fluorescein-conjugated anti-mouseIgM, diluted 1:5, for 1 h. After a final wash the coverslipswere mounted onto glass slides using Fluormount medium(Fisher Scientific). Control sections, in which the primaryantibody was replaced by PBS, showed only a low back-ground fluorescence, or none at all.
Living cultures were treated with HNK-1 alone at di-lutions of 1:10, 1:50 and 1:100, or at 1:10 in the presence ofcomplement (pooled guinea-pig complement, 1: 100, Gibco)for up to 20 h. Some cultures were then fixed and stained withHNK-1 by indirect immunofluorescence to identify neuralcrest cells, while others were washed with PBS and reincu-bated with normal culture medium for a further 1-4 daysbefore staining. Control cultures were incubated with a non-specific monoclonal IgM (Anti-Ly.2.2). This antibody wasalso tested for cell surface binding by immunofluorescenceusing the same fluorescent anti-mouse IgM second antibody.
Coverslips were coated with fibronectin and laminin, asdescribed previously (Sanders, 1980; Newgreen, 1984), bytreating them with these glycoproteins at a concentration ofSOjUgml"1 in serum-free medium for 45 min at room tem-perature. After rinsing with serum-free medium the tissuewas explained onto the coverslips in the usual way.
Neurofilaments were detected using a monoclonal antibodyto neurofilament polypeptide 160 kD (Boehringer-Mann-heim). Cultures on coverslips were fixed with acetone at 4°Cfor 10 min, washed with PBS and incubated with 1 % BSAfor 30 min at room temperature. Drops of antibody solution(1:3, dilution) were placed on the cultures for l h at 37°C.After rinsing with PBS three times the specimens wereincubated with fluorescein-conjugated anti-mouse IgG (Cal-biochem) for 1 h at 37°C, washed and mounted. Controls,performed by replacing the neurofilament antiserum withPBS, failed to stain.
Fluorescence micrographs were made on Tri-X film(Kodak) using a Zeiss Photomicroscope equipped withepifluorescence and a 40X Planapo immersion objective,(NA=l-0) .
Results
Neural tube cultures
As shown previously (Vincent & Thiery, 1984), piecesof neural tube gave rise to an outgrowth of individuallymigrating neural crest cells that, after 1 day in culture,stained positively for the HNK-1 epitope (Fig. 1). Asin the earlier description, we found some variation inthe intensity of staining among these cells in thepopulation, but we also found that all of these outgrow-ing cells showed some degree of staining. On someoccasions an epithelial sheet derived from the neuraltube itself was present and stained with a low intensitythat was close to background staining. These sheets
116 E.J. Sanders and E. Cheung
Fig. 1. IINK-1-labelled neural crest cells at the periphery of a neural tube explant after 20h in culture. All cells presentshow some degree of staining. The magnification of this and all micrographs is X480.Fig. 2. HNK-1-labelled neural crest cells at the periphery of a sclerotome explant after 20h in culture. Unlabelledsclerotome cells are dispersed between them.Fig. 3. After 4 days in culture neurite outgrowth from HNK-1 positive cells is extensive. These neurites overlie thesclerotome cells.Fig. 4. A bipolar neurone with neurites in a sclerotome culture after 4 days. Phase-contrast.
were always close to the site of origin and clearlydistinguishable from the individual fibroblast-like crestcells at the margin of the explant (Newgreen et al.1982).
Continued incubation of these explants on untreatedglass coverslips indicated that the HNK-1 epitope wasretained for at least 4 days in culture. By this time therehad occurred a considerable outgrowth of HNK-1positive neurites (see Figs 3,5).
Sclerotome cultures
After 24 h in culture, neural crest cells comprisedapproximately 20% of the sclerotome culture. At theperiphery of the culture they were distributed singly orin small groups on the glass substratum (Fig. 2), buttowards the centre of the explant they frequently
formed an extensive network of cells on the surface ofthe continuous underlying sclerotome cell layer. Theappearance of the crest cells always on the surface of thesclerotome sheet is somewhat unusual, since the dorsalsurface of many cultured embryonic cells is inhospi-table to the adhesion of other cells (Sanders, 1982).
By phase-contrast microscopy at this time, the twotypes of cells were indistinguishable. At 2 clays, thecrest cells had begun to form neurites, again on thesurface of the culture, which became increasinglyextensive over the two succeeding days (Fig. 3). Thecell bodies were relatively small at this time in compari-son with the underlying sclerotome cells and for thisreason it appeared as if the crest cell number wasreduced, although it was not. By this time the crestcells were distinguishable by their neurites with carefulobservation by phase-contrast microscopy (Fig. 4).
Neural crest and sclerotome cells in culture 117
They were, however, difficult to identify against thebackground of densely packed sclerotome cells, and itwas impossible to determine to which cell body eachncurite belonged. The neurites themselves were easilyconfused with retraction fibres from sclerotome cells,and those crest cells not bearing prominent neuriteswere not distinguishable. Pigment cells did not appearin any of the cultures under these conditions.
Influence of the substratum on HNK-1 bindingOn fibronectin and laniinin substrata for 20 h theappearance of the neural crest cells, but not thesclerotome cells, was markedly altered (Table 1). Thecrest cells had rapidly developed extensive networks offine neurites (Fig. 5) and even ganglion-like structures,more characteristic of explants cultured in excess of 4
days on untreated glass. This enhanced differentiation,described by previous authors, was apparent with bothglycoproteins, but was significantly more prominentwith fibronectin than with laminin. Neurite outgrowthwas scored by measuring neurite length from cell bodyto growth cone. Cultures containing cells with neuritelengths greater than 100 fim were thus distinguishedfrom those with lesser outgrowth.
The experiments in this section were performed onboth neural crest cultures and on neural crest/sclero-tome mixed cultures. The response of the crest cellswas the same in both cases. Use of the mixed culture,however, provided a control for the specificity of theHNK-1 activity, since the unstained sclerotome cellspresent simultaneously were seen to be unaffected by
Fig. 5. Crest cells cultured for 1 day on a fibronectin substratum produce extensive outgrowths of neurites. Lamininsubstrata produce similar, though less-prominent, outgrowths.Fig. 6. Crest cells cultured for 20 h in the continuous presence of HNK-1 antibody (1 :S0) and then detected by thatantibody, showing the retraction and blebbing. Compare with Fig. 2. The sclerotome cells are unaffected.Fig. 7. Crest cells explanted on a fibronectin substratum and treated with HNK-1 antibody (1:50) for 20h. Neuriteoutgrowth is inhibited, but the cells are still spread. Unstained cells are sclerotome.Fig. 8. Crest cells explanted on a laniinin substratum and treated with HNK-1 antibody for 20h. Neurite outgrowth ininhibited and the cells are retracted. Unstained cells are sclerotome.
118 li.J. Sanders and E. Cheung
Table 1. The effects of HNK-1 antibody treatment for 20 h on cell spreading and nenrite outgrotvth
Substratum
GlassFibronectinLaminin
No
Spreading
+ +
+
HNK-1
Neurites
+ HNK-1
Spreading
(1:50)
Neurites
+ Non-speci
Spreading
+ +
+
fie IgM
Neurites
Cell spreading: + + , broad lamellipodia present; +, cells spread, but fusiform or fibroblastic; - , cells rounded. Neurite outgrowth:+ + , neurites present in excess of 100(im length; +, all neurites less than 100^m length; —, no neurites.
Numbers in parentheses indicate the figure that illustrates the condition.
the treatments. Each experimental condition was re-peated a minimum of five times. Each run consisted ofcultures in triplicate with simultaneous controls.
Living cultures on untreated glass incubated withHNK-1 antibody (1:10, 1:50 or 1:100) for 20 h andthen stained with that antibody indicated that there wasno cytotoxic effect under these conditions, althoughcrest cell spreading (but not that of sclerotome cells)was reduced and accompanied by apparent blebbing(Fig. 6 and Table 1). This response was dose-depen-dent, with the 1:100 dilution being ineffective. Thatthese treatments did not kill the cells was shown bycontinued incubation for a further day in the absence ofthe antibody. After this time the crest cells hadresumed their normal appearance.
On fibronectin and laminin substrata in the presenceof HNK-1 antibody for 20 h (Table 1) the crest cellsshowed much reduced neurite outgrowth comparedwith that demonstrated on these substrata in theabsence of antibody for an equal length of time(compare Figs 7, 8 with Fig. 5). On fibronectin thesecells remained spread under these conditions, failing toshow the blebbing appearance of controls on plain glassin the presence of antibody (compare Fig. 7 withFig. 6). In many cases these cells were more highlyspread than untreated controls on an untreated sub-stratum (compare Fig. 7 with Fig. 1). The fibronectinapparently protected the cells from antibody-inducedretraction, but not from antibody-induced inhibition ofdifferentiation. Laminin differed from fibronectin inthat cells on this substratum in the presence of HNK-1lost both their neurites and their well-spread appear-ance (Fig. 8).
These effects were not produced when HNK-1 wasreplaced with the non-specific monoclonal IgM anti-Ly.2.2, despite the fact that this control antibody alsobound to the crest cells as judged by the sameimmunofluorescence technique.
HNK-1 cytotoxicity
In the presence of complement (1:100) and HNK-1antibody (1:10) for 20 h the crest cells, although still
apparent (Fig. 9), were rounded and highly con-densed. The lethality of the cytotoxic effect was shownby culturing for a further day in medium withoutantibody or complement, after which HNK-1 positivecells were absent (Fig. 10). Complement alone had nosuch effect. The sclerotome cultures on untreated glassbegan spreading by 4-5 h after explantation and com-pleted their outgrowth by about 15 h. Termination oftreatment with the antibody plus complement, prior totheir complete emigration from the tissue mass,resulted in incomplete killing of the positive neuralcrest cells, suggesting that spreading might be a pre-requisite for cytotoxicity.
When the HNK-l/complement treated cultureswere incubated for up to three more days in normalmedium, no HNK-1 positive cells appeared, nor didneurites appear or cells resembling those shown inFig. 4. Staining for neurofilaments using an anti-neurofilament antibody gave a negative result(Fig. 11), whereas 4-day neural crest cell culturesshowed intensely staining neurites with this antibody(Fig. 12).
Discussion
The neural crest cells contaminating the sclerotomecultures come not only from the tissue spaces sur-rounding the trunk-level somites (Newgreen & Erick-son, 1986), but also from within the somites themselveswhere they are migrating prior to the formation ofsympathetic ganglia (Le Douarin, 1982; Rickmann etal. 1985; Guillory & Bronner-Fraser, 1986). Indeed,although not widely appreciated, the appearance ofneural crest cells in such cultures has been described,using techniques for the detection of neurotransmitters(Fauquet et al. 1981). However, without staining byspecific markers, the presence of the crest cells isimpossible to demonstrate in cultures prior to 3 daysafter explantation, and even after this time crest cellsnot bearing neurites will remain unidentified. Thepresence of deposited extracellular materials (presum-ably primarily fibronectin) favours a switch in thedifferentiation of the crest cells from a melanocyte line
Neural crest and sclerotome cells in culture 119
Fig. 9. Crest cells after 20 h in the presence of HNK-1 antibody and complement. The cells are highly condensed and notable to recover. The sclerotome cells are unaffected.Fig. 10. Cells treated as those in Fig. 9, but cultured for a further day in the absence of HNK-1 antibody. Crest cells arecompletely eliminated from the culture and do not appear over the following 3 days.Fig. 11. Cells from a mixed sclerotome-neural crest culture 4 days after treatment with HNK-1 antibody and complement.The culture has been stained for neurofilaments and reacts negatively.Fig. 12. Neural crest cells after 4 days in culture stained for neurofilaments. Several neural crest cell bodies stain intensely.The neurite of one cell is in focus. Other neurites are present, but in a different focal plane.
to a neuronal one (Sieber-Blum et al. 1981; Loring etal. 1982) with the resulting neurite outgrowth that weshow after 3 days on the untreated substratum. Thefailure of pigment cells to appear in these cultures maypossibly be attributable to the exclusion of the dermo-myotome from the initial explant.
Fibronectin and laminin substrata exert two effectson neural crest cells, which may or may not beconnected. First, it is known that both substratapromote accelerated neurite formation by a variety ofneural cells in culture (Rogers et al. 1983; Lander et al.1985; Akeson & Warren, 1986; Gundersen, 1987; Hallet al. 1987). We show here that this is true also for thecrest cells in mixed culture, although fibronectin wasthe more efficacious of the two glycoproteins. Theneurite-promoting effect may be a non-specific one,
relying merely on increased adhesion, since it is dupli-cated by polylysine coating of the substratum (Stern etal. 1986). Second, newly explanted crest cells respondwithout preference to both fibronectin and lamininsubstrata by undergoing rapid adhesion and spreading(Newgreen, 1984).
Addition of HNK-1 antibody to cultures on thesetwo extracellular matrices blocked the precociousneurite development on both, and on laminin causedextensive (but non-lethal) cell retraction, similar to thatobserved on untreated glass, suggesting an interactionwith the laminin-binding site of the cell. By contrast,the failure to effect retraction from fibronectin appearsto indicate that the antibody does not block the bindingsites for fibronectin on the cell body. Non-specific
120 E. J. Sanders and E. Cheung
monoclonal IgM failed to inhibit the neurite out-growth, leading to the conclusion that the cell surfaceepitope recognized by HNK-1 is associated either withthe mechanism of neurite outgrowth or neurite attach-ment. This is consistent with observations made usingspinal cord and other neurones (McGarry et al. 1985;Riopelle et al. 1986). The neurite-inhibiting activitymay not necessarily be related to the blocking oflaminin sites, since it also occurs on glass untreatedexcept by the presence of the sclerotome cells. Indeed,since it appears that the antibody may also recognizethe neurone-specific N-CAM and LI antigens (Kruseet al. 1984), HNK-1 may well have multiple effects.
Clearly, it should be possible to eliminate neuralcrest cells from sclerotome cultures by explanting onlythe posterior half of the 3-day somite, since at this stagethe crest cells avoid this half (Rickmann et al. 1985;Stern et al. 1986). However, from a technical point ofview, although possible, this would be unreliable andcertainly an obstacle to the collection of large amountsof tissue. Selective elimination using the cytotoxicity ofHNK-1 in the presence of complement is clearly apractical alternative, with the proviso that treatment becontinued at least until cell emigration from the explantis complete (10—15 h). Whether cell attachment andspreading on the substratum is the crucial factor indetermining the lethality of the treatment, or whetherit is simply a matter of the accessibility of the antibodyto the cells has not been determined.
The question arises as to whether all of the neuralcrest cells are eliminated by this treatment. Vincent &Thiery (1984) suggested from their work that theoverwhelming majority of the early crest cells areHNK-1 positive and that the few observed negativecells in cephalic neural crest cultures were probablycontaminating mesenchyme cells. In the present workwe used only trunk levels that we are confident wereuncontaminated. Furthermore, continued incubationof cultures sufficiently treated with HNK-1 and comp-lement failed to show the presence of either HNK-1positive cells or cells reacting with antibodies to neuro-filaments.
We thank the Medical Research Council of Canada for anoperating grant in support of this work.
References
AKESON, R. & WARREN, S. L. (1986). PC12 adhesion andneurite formation on selected substrates are inhibited bysome glycosaminoglycans and a fibronectin-derivedtetrapeptide. Expl Cell Res. 162, 347-362.
BELLAIRS, R., SANDERS, E. J. & PORTCH, P. A. (1980).
Behavioural properties of chick somitic mesoderm andlateral plate when explanted in vitro. J. Embryol. exp.Morph. 56, 41-58.
BRONNER-FRASER, M. (1986). Analysis of the early stagesof trunk neural crest migration in avian embryos usingmonoclonal antibody HNK-1. Devi Biol. 115, 44-55.
FAUQUET, M., SMITH, J., ZILLER, C. & LE DOUARIN, N.
M. (1981). Differentiation of autonomic neuronprecursors in vitm: cholinergic and adrenergic traits incultured neural crest cells. J. Neurosci. 1, 478-492.
FISHER, M. & SOLURSH, M. (1979). The influence of thesubstratum on mesenchyme spreading /;/ vitm. Expl CellRes. 123, 1-14.
GUILLORY, G. & BRONNER-FRASER, M. (1986). An in vitm
assay for neural crest cell migration through the somites.J. Embryol. exp. Morph. 98, 85-97.
GUNDERSEN, R. W. (1987). Response of sensory neuritcsand growth cones to patterned substrata of laminin andfibronectin in vitro. Devi Biol. 121, 423-431.
HALL, B. K. (1977). Chondrogenesis of the somiticmesoderm. Adv. Anat. 53(4), 1-47.
HALL, D. E., NEUGEBAUER, K. M. & REICHARDT, L. F.
(1987). Embryonic neural retina cell response toextracellular matrix proteins: Developmental changesand effects of the cell substratum attachment antibody(CSAT). J. Cell Biol. 104, 623-634.
HAMBURGER, V. & HAMILTON, H. L. (1951). A series ofnormal stages in the development of the chick embryo. J.Morph. 88, 49-92.
KRUSE, J., MAILHAMMER, R., WERNECKE, II., FAISSNER,
A., SOMMER, I., GORIDIS, C. & SCHACHNER, M. (1984).Neural cell adhesion molecules and myelin-associatedglycoprotein share a common carbohydrate moietyrecognized by monoclonal antibodies L2 and HNK-1.Nature, bond. 311, 153-155.
LANDER, A. D., FUJI, D. K. & REICHARDT, L. F. (1985).
Laminin is associated with the neurite outgrowth-promoting factors found in conditioned media. lJmc.natn. Acad. Sci. U.SA. 82, 2183-2187.
LASH, J. W. & VASAN, N. S. (1978). Somitechondrogenesis in vitro. Stimulation by exogenousextracellular matrix components. Devi Biol. 66, 151-171.
LE DOUARIN, N. M. (1982). The Neural Crest. CambridgeUniversity Press.
MCGARRY, R. C , RIOPELLE, R. J. & RODER, J. C. (1985).
Accelerated regenerative neurite formation by a ncuronalsurface epitope reactive with the monoclonal antibody,Leu7. Neurosci. Lett. 56, 95-100.
NEWGREEN, D. F. (1984). Spreading of explains ofembryonic chick mesenchymes and epithelia onfibronectin and laminin. Cell Tiss. Res. 236, 265-277.
NEWGREEN, D. F. & ERICKSON, C. A. (1986). The
migration of neural crest cells. /;;/. Rev. Cvtol. 103,89-145.
NEWGREEN, D. F., GIBBINS, I. L., SAUTER, J.,
WALLENFELS, B. & WUTZ, R. (1982). Ultrastructuraland tissue-culture studies on the role of fibronectin,collagen and glycosaminoglycans in the migration ofneural crest cells in the fowl embryo. Cell Tiss. Res. 221,521-549.
Neural crest and sclerotome cells in culture 121
NEWGREEN, D. F., SCHEEL, M. & KASTNER, V. (1986).
Morphogenesis of sclerotome and neural crest in avianembryos. In vivo and in vitro studies on the role ofnotochordal extracellular material. Cell Tiss. Res. 244,299-313.
RICKMANN, M., FAWCETT, J. W. & KEYNES, R. J. (1985).
The migration of neural crest cells and the growth ofmotor axons through the rostral half of the chick somite.J. Embiyol. exp. Morph. 90, 437-455.
RiOPELLE, R. J., MCGARRY, R. C. & RODER, J. C. (1986).Adhesion properties of a neuronal epitope recognized bythe monoclonal antibody HNK-1. Brain Res. 367,20-25.
ROGERS, S. L., LETOURNEAU, P. C , PALM, S. L.,
MCCARTHY, J. & FURCHT, L. T. (1983). Neurite
extension by peripheral and central nervous systemneurons in response to substratum-bound fibronectin andlaminin. Devi Biol. 98, 212-220.
SANDERS, E. J. (1980). The effect of fibronectin andsubstratum-attached material on the spreading of chickembryo mesoderm cells in vitro. J. Cell Sci. 44, 225-242.
SANDERS, E. J. (1982). Inability of mesoderm cells tolocomote on the modified free surface of epithelial cellsheets in vitro. In Vitro 18, 71-78.
SANDERS, E. J. (1986). Mesoderm migration in the earlychick embryo. In Developmental Biology. AComprehensive Synthesis, vol. 2 (ed. L. Brovvder), pp.449-480. New York: Plenum.
SANDERS, E. J. & PRASAD, S. (1983). The culture of chickembryo mesoderm cells in hydrated collagen gels. .7. exp.Zool. 226, 81-92.
SIEBER-BLUM, M., SIEBER, F. & YAMADA, K. M. (1981).
STERN, C. D., SISODIYA, S. M. & KEYNES, R. J. (1986).
Interactions between neurites and somite cells: inhibitionand stimulation of nerve growth in the chick embryo. J.Embryol. exp. Morph. 91, 209-226.
TUCKER, G. C , AOYAMA, H., LIPINSKI, M., TURSZ, T. &
THIERY, J. P. (1984). Identical reactivity of monoclonalantibodies HNK-1 and NC-1: Conservation invertebrates on cells derived from the neural primordiumand on some leukocytes. Cell Differ. 14, 223-230.
VINCENT, M. & THIERY, J. P. (1984). A cell surface markerfor neural crest and placodal cells. Further evolution inperipheral and central nervous system. Devi Biol. 103,468-481.
{Received 4 November 1987 - Accepted 21 January 1988)
![DATASET DESCRIPTION PCA RESULTS Dataset #1 RNA-Seq of neural cells (MiSeq) [2] 65 cells Ground truth clusters: Group I (Neural Progenitors), Group.](https://static.documents.pub/doc/80x56/56649e415503460f94b33d19/dataset-description-pca-results-dataset-1-rna-seq-of-neural-cells-miseq.jpg)
