Developmental Brain Research, 6 (1983) 259-268 259 Elsevier Biomedical Press

The Effects of Calcium on the Surface Characteristics of Dissociated Fetal

Mouse Cerebral Cells: Morphological Aspects

A. WONG and D. G. JONES*

Department of Anatomy and Human Biology, University of Western Australia, Nedlands, W.A. 6009 (Australia)

(Accepted June 22nd, 1982)

Key words: cerebral cells - - mouse - - surface features - - calcium - - tissue culture - - embryogenesis

Trypsin-EDTA dissociated fetal mouse cerebral cells were cultured over a 48 h period in media containing physiological and low levels of Ca 2÷, and also in media lacking Ca 2+. Quantitative light microscopy studies revealed that 86 ~ of cells grown in the physiological Ca 2+ medium had recognizable cellular processes 1 h after plating and 96 ~ at 48 h. Of the cells grown in media without Ca 2+, 70-80~/o failed to establish cellular processes. In the low Ca 2÷ medium, 7 ~ of the plated cells had no cellular processes at 1 h.

Scanning electron microscopy revealed that cells grown in the physiological Ca 2+ and low Ca 2+ media demonstrated cell types of varying features. One type had rounded cell bodies, with at least one well-defined cell process. A second type demonstrated consider- able ruffling on the cell surface, and a third type had flattened cell bodies with numerous cell processes. Cells grown in the low Ca 2÷ medium showed considerable proliferation of processes up to 15 h after plating. In contrast, cells grown in the Ca2+-free medium had flattened and relatively smooth cell bodies, with few distinct processes. Those cells with processes were unipolar.

These results suggest that dissociation with trypsin-EDTA renders the cell surface active, exposing it to the direct effects of the presence or absence of external Ca 2+. The lack of cell surface features and cellular outgrowth in the Ca 2+ ti'ee cells suggests that Ca 2+ plays an impoitant role in the initial dynamic processes of cell membrane activity during early neural cell adhesion and development.

INTRODUCTION

Dur ing embryogenesis , there is a morphogene t i c

r ea r rangement o f tissues and cells into organized

cel lular pat terns . Ear ly studies by Hol t f re te r and

colleagues 12,27 demons t r a t ed tha t single cells f rom

disaggregated amph ib i an embryos can undergo

readherence and sort ing out, fo rming organized

tissue-specific pa t te rns s imilar to those achieved in

vivo. Since then, numerous studies uti l izing in vi t ro

dissocia t ion and reaggregat ion techniques have been

under t aken to e lucidate the mechanisms o f inter-

cel lular adhes ion and coupled m e m b r a n e surface act ivi ty 14.

M a n y workers have suggested tha t calc ium ions p lay a role in in tercel lular adhesion15, 22-28. In par -

t icular, it has been shown tha t the adhesive p roper -

ties o f embryo chick neural ret inal cells differ,

depending upon the presence or absence o f Ca z+

dur ing preparation15,24-26, 2s. A similar phe nom-

enon has been repor ted in H e L a cells 1°, Chinese

hamster V79 lung cells3,23,~4, zs, Chang conjunc-

t iva cells 2, and in no rma l and t r ans fo rmed B H K

cells 2a,30. I t has also been p r o p o s e d tha t physio logi -

cal changes in the ion concent ra t ions o f Ca z+ and /o r

proteases may br ing abou t a mechanism for the

regula t ion o f adhesive affinities and surface activi-

ties, for the guidance o f cells and their processes,

and for the specification o f their order ly associa- tions25.

The morpho log ica l aspects o f cell surface char-

acteristics dur ing adhes ion and in response to

var ia t ions in the external Ca 2+ level are poo r ly

explored. Takeichi z3 has shown morpho log ica l ly

that Chinese hamste r V79 lung cells, d issocia ted

ei ther with E D T A or t rypsin and Ca 2+, exhibi t an

* To whom reprint requests should be sent.

0165-3806/83/0000-0000/$03.00 © 1983 Elsevier Biomedical Press

260

increased contact area by spreading onto adjacent cells during aggregation in the presence of Ca:',. Similar spreading also occurred with trypsin-EDTA dissociated cells.

The present study was designed to follow changes in the surface features of trypsin-EDTA dissociated cells from fetal mouse cerebra, in the presence and absence of Ca 2~, in order to examine the primary effects of Ca 2~ . Cells were examined with scanning electron microscopy. Three media were used: phys- iological levels of Ca e , low Ca 2E concentrations, and a lack of Ca 2+. The dynamics of neural process outgrowth, reflecting the effects of the presence or absence of Ca" ~-, were determined using quantitative light microscopic techniques.

MATERIALS AND METHODS

Tissue culture preparation Cerebral hemispheres from 50 mouse embryos

(Canberra strain, 14 days in utero) were dissected free of meninges and cut into 0.5-1 mm fragments. Tissue pieces were then rinsed in Ca-Mg-free trypsin diluent (Commonwealth Serum Laboratories (CSL), Australia), followed by enzymatic dissociation in 0.01% trypsin-EDTA (Grand Island Biological Co., NY) in Ca-Mg-free trypsin diluent. Both proce- dures were performed for 20 rain at 37 °C under a mixture of 5 % COz-95 o/, air. The tissue was rinsed with Ca-Mg-free trypsin diluent containing trypsin inhibitor. The tissue fragments were further dis- sociated mechanically by filtering through two layers of miracloth filter (22-25/zm pore size, Cal- biochem).0.4 ml of the dissociated cell suspension, consisting of 3.7 ~ 107 cells/ml, were inoculated into Corning polystyrene flasks (25 cm '~) each con- taining 10 rnl of treatment media: control, calcium- free and low calcium. This gave a final cell sus- pension of 1.4 ~ 106 cells/ml per treatment flask. Each flask also contained 10 poly-L-lysine (10 #g/ml, type l-B, Sigma Chemicals)-coated glass coverslips.

The control medium consisted of Eagle's basal medium (CSL) with 10 °/,~ fetal calf serum (CSL), 1 "~ii L-glutamine (2 mM, Flow Laboratories, Australia), 1 ~o penicillin-streptomycin (50 IU/ml, 50 mcg/ml, Flow Laboratories), and 0.1 ~,, NaHCO3 (CSL). The Ca 2+ level of the control medium was 1.8 raM. The constituents of the Cae~-free medium were identical

to those of the controi medium, except that CaCI2 and fetal calf serum were omitted, The low-Ca ~ medium was the same as the Ca:' +-free medium, plus 10 ',~i fetal call" serum. All flasks were incubated with caps loosely fastened at 37 ~'C, under 5 o,~ CO~-95 ~:~ air circulation~

Cell viability was determined by a Trypan blue exclusion test (0.04,~i in Eagles Minimum Essential Medium salt solution) before the dissociated cells were inoculated into the treatment flasks, and at 24 h after incubation.

Phase-contrast and scanning electron microscopy At intervals of I, 3, 6, 12, 15, 24, 30 and 48 h alter

initial cell plating, one coverslip was removed from each treatment flask. Coverslip~ from the same treatment flasks were used for the various time inter- vals, in order to maintain consistency throughout the experiment. The removed coverslips were trans- ferred into 2.5 I~~ glutaraldehyde (TAAB) in 0.1 M cacodylate buffer and fixed for 1 h at room tempera- ture. A selection of phase-contrast photographs was taken at a magnification of :~:400, using a phase- colltrast microscope. For each treatment, 36 areas were randomly selected for measurement at each time interval.

In order to measure the dynamics of process out- growth of the cells as a function of the external Ca "2 level, a ratio was established. This ratio expresses the number of cells lacking cellular processes (round cells) to the total number of cells counted in a randomly selected area of the phase-contrast pre- parations. The ratios for the 3 treatment groups were thereafter compared and statistically tested using the Student's t-test. In the areas selected for analysis both phase-dark and phase-bright cells were counted at 6 h after plating (obvious cell debris was not counted). At longer time intervals only phase- dark cells were counted, as these cell populations consisted almost entirely of phase-dark cells 16,19.

Following glutaraldehyde fixation, cells were postfixed in I °//,, OsO4 in 0.15 M cacodyIate buffer for 30 rain at 4 °C. Dehydration was carried out by using graded concentrations of ethanol and acetone, The specimens were dried by the critical point drying method, were sputter-coated with gold, and examined under a Philips 500 scanning microscope

at 25 kV.

For analysis of the scanning electron micro- graphs, all cells observed were subdivided into 3 types according to features outlined in the Results section. The only profiles omitted from this analysis were cell debris and obviously degenerating cells during the first 15 h of preparation. Fifteen micro- graphs were taken of each time interval for each group, making 120 micrographs for each of the control, low calcium and calcium free groups over the period studied. The micrographs ranged in mag- nification from × 1700 to 4250.

RESUL T S

After the dissociation procedures, the dissociated cell suspension consisted of 83 ~ viable cells, deter- mined with the Trypan blue exclusion test and counting of cells in a haemocytometer under a light microscope. There was a decrease in viability in all cell cultures, ranging from 10 to 20 ~, after 24 h of iJacubation. The decreases in cell viability did not vary significantly between the 3 treatment groups, although the decrease was greater for the low Ca 2+ and Ca2+-free groups compared to the control. The initial cell plating 20 min after incubation was approximately 2 × 105 cells/cm 2 per flask for each treatment. The number of cells plated did not differ among the 3 groups.

0.85-

c 6 0.75

0-65 o.25-

u

"-~c~ 0.15_

Ca free T

--~",, I.T . . . . . . . . . ~ . . . . . -' 7_

",,,T,,' !

lOW Co

,i- ....... i i ! ~ - ' r ' ~ i ~ 1o m °o f

O" 05 ........... i " ' " ~ ' ~ t _ . . . . . . ~ z . ~ contrOl t

' ~ ~, ,~ 1~ 2'4 sb 4b hours after plating

Fig. 1. Ra t io of the n u m b e r of round cells (cells lacking processes) to the total n u m b e r of cells a t various t ime intervals after plat ing in 3 different media: Ca2+-free, low-Ca~+ and control .

261

Fig. 1 depicts the values and time course of the mean ratio of the number of round cells to the total number of cells in the phase-contrast photographs of the 3 experimental groups. In the CaZ+-free medium, 70-80~ of the plated cells remained rounded, i.e. only 20-30 ~o established cellular pro- cesses. The cells with processes were predominantly unipolar in nature. This percentage remained re- latively consistent throughout the entire experiment, a significant difference (Student's t-test, P < 0.001) only being found when the ratios at 1 h and 30 h were compared. In the control medium at 1 h after plating, 869/0 of cells have recognizable cellular processes. This percentage increases to approxi- mately 96 ~ at 48 h. At time intervals of 12 h and 30 h, the mean ratio values when compared to 1 h after plating showed significant differences (P < 0.001) in both cases. The mean ratios also differed signifi- cantly (P < 0.01) between 12 and 30 h. Of the cells cultured in a low-Ca 2+ medium, only 7 ~ of the initially plated cells remained rounded after 1 h. The mean ratio of the rounded cells to the total number of cells remained lower than in the control group for the first 15 h, following which there was a significant increase in value to about 20 ~o. This higher value was maintained until 48 h. When the mean ratios at time intervals of 12 h and 30 h are compared to 1 h, the differences are significant at the 0.001 level.

Comparisons of the mean ratio of number of round cells to total cell number among the 3 treat- ment groups were made at intervals of I h, 12 h and 30 h. The mean ratios of these groups differ signifi- cantly (P < 0.001) at these intervals, with the ex- ception of the comparison between the control and low-Ca z+ groups at 12 h.

When viewed with the scanning electron micro- scope, details of the surface morphology are evident. One hour after plating, cells grown in the control and low-Ca 2+ media displayed cells of varied morpho- logical properties (Figs. 2-5). The cells have been divided into 3 types, on the basis of their different morphological properties. Of these, one type (A) has a rounded cell body and at least one well-defined cell process (Fig. 2). On some of these cell processes, well-defined growth cones can be observed at the end of the process, and these may establish cellular con- tacts with other cells (Fig. 4). A second cell type (B) has many fine processes extending from the cell

262

Fig. 2. Scanning electron micrograph of cells grown in control med ium ; 1 h after plating. Cell A has a long cellular process with a g rowth cone (gc) at its te rminat ion. Cell C has n u m e r o u s processes and small blebs on the cell surface. - 2320.

Fig. 3. Ano the r cell type, B, is highlighted in this SEM of cells grown in control m e d i u m ; 1 h after plating. Considerable ruffling (rl is seen on the surface of this cell type. ;~ 3910.

263

Fig. 4. Cells grown in control medium; 1 h after plating. Cell types, A and C, are clearly delineated. Type C has small blebs (b) on its surface, and a thin sheet-like lamellipodium (Ip) at one side. Cellular contacts are depicted by arrows, x 4250.

Fig. 5. Cells grown in a low Ca 2+ medium; 1 h after plating. The 3 cell types, A-C, are represented, all having extensive cellular proliferation, d, cell debris, x 2150.

264

Fig. 6. These cells were grown in a CaZ~-free medium (1 h after plating), and have a rounded and flattened appearance. There i~ little cellular proliferation, dc, degenerating ceil. × 2400.

Fig. 7. Control cells after 6 h plating, Numerous cellular processes are seen, with contacts (arrows) betwceia taeighbourillg cell~, Little ruffling of the cell surfaces is evident. /,, 1720.

265

Fig. 8. Cells g rown in low Ca 2+ m e d i u m ; 6 h after plating. Small aggregates of cells are seen. x 1890.

Fig. 9. Cells grown in Ca2+-free m e d i u m ; 6 h after plating. The cells are rounded , with min imal cellular proliferation. The cell bodies are flattened. Cells with processes are unipolar in appearance, x 1890.

266

body, which itself appears to have considerable ruffling on its surface (Fig. 3). A third type (C) has few surface ruffles; instead, small blebs are present on the cell surface (Figs. 2 and 4). Its cell body has a flattened appearance with many processes extending from it. Some of these cells are characterized by a thin sheet-like lamellipodium extending from the edge of the cell body (Fig. 4).

However, cells grown in the Ca2+-free medium do not display such varied cell surface properties. Instead, the cell bodies remain relatively flattened with few distinct processes and few surface features (Fig. 6). Even the cells with cell processes fail to show

characteristic growth cones. After 6 h of plating, the dissociated cells from the

control and low Ca 2÷ group have extended many processes, establishing contacts with neighbouring cells (Fig. 7). Small cell aggregates are also being formed (Fig. 8). There is a decrease in the ruffling displayed on the cell surface. Although the Ca2+-free cells establish contacts, their cell bodies remain rounded and flattened. Cells with processes remain principally unipolar in appearance, with few ob- servable growth cones (Fig. 9).

After 24 h of culturing the low Ca 2,~ and Ca 2 ~-free cells start showing signs of degeneration. This is confirmed by the 10-20~ decrease in cell viability seen under the Trypan blue exclusion test. The decrease in cell viability may account for the in- crease in the mean ratio of round cells to total cell number of the low Ca z+ group after 15 h plating (Fig. 1). By contrast, the cells of the control group retain the extensive cellular proliferation of the earlier stages of the experiment (Figs. 2-4, 7).

DISCUSSION

In the present study, cells were dissociated with trypsin and EDTA, in order to examine morpholo- gically the role of external Ca z+ in cell behavior. The rationale for this stems from the assumption that the composition and activities of cell surfaces play a direct role in the mechanisms of intercellular ad- hesion 22.

Dissociation of cells with trypsin and EDTA, in conjunction with miracloth filtering, resulted in a more complete dissociation than that brought about by mechanical dissociation alone. The latter results

in the production of small clusters of cells (un- published observations). Cell viability however, does not appear to be seriously affected by this dissociation procedure.

The initial number of cells plated was uniform/'or the 3 treatment groups. Hence. the presence or absence of serum and/or Ca 2~ in the media exerts little effect on the initial plating of the cells. In these cultures most of the cells attached to the poly-t- lysine substrate 20 rain after plating. The poly-t- lysine substrate appeared to provide better attach- ment for the trypsin-EDTA dissociated cells than a collagen substrate, with which there is a tendency to form small aggregates of cells (unpublished results I.

The lack of production of cellular processes and the lack of cell surface activity in cells grown in thc absence of Ca 2 ~ (Figs. 1, 6) suggest that Ca 2+ plays a dynamic role in the establishment of cell-to-cell adhesion. This is made possible by the fact that both trypsin and EDTA modify aspects of the cell sur- face. Trypsin. for instance, alters the chemical com- position of cell surfaces by removing plasma mem- brane sialomucopeptides9,1L by altering the distri- bution of receptor sites for lectins 8.tv.ls,21 and by reducing electrophoretic mobility4,~3,zL Treatmem with EDTA also causes a release of cell surface materials1,1~. However. it is as yet not know,~ whether Ca z+ plays a role in cellular repair in response to trypsin and EDTA dissociation.

Of the control cells. 86°/,, had recognizable cel- lular processes (Fig. 1), unlike the findings of Rothman and Cowan 2°. in which only about 50 '}o of cells showed cellular processes. The reason for this difference may be attributed to differences in the media used and also in the brain regions studied 6,7 so. The alteration in the chemical composition of the cell surface by trypsin-EDTA dissociation however, may render the cell membrane active, thereby en- abling Ca 2+ to permeate into the cytoplasm.

In preliminary studies (unpublished results) on the addition of 1 mM EDTA to the Ca2~-free medium, the cells have morphological features similar to those of cells grown m a Ca2+-free me- dium alone. In contrast, cells grown in serum-free medium in the presence of Ca z+. do not demonstrate the characteristic features of Ca 2--free or Ca" ~-free plus EDTA cells "~1. The flattening of the cell bodies and the unipolar characteristics ,~f lhe CaZ -free

267

cells in the present s tudy therefore, do no t appea r to

result f rom the absence o f serum in the media .

Nevertheless , the presence o f serum may compen-

sate ini t ial ly for the effects o f the lack o f Ca 2 ~ in the

media . This is because low Ca 2+ cells, grown in

Ca2+-free med ia in the presence of serum, demon-

s trate cel lular act ivi ty s imilar to tha t o f cont ro l cells

for the first 15 h o f incubat ion.

The var ious morpho log ica l features o f the cont ro l

and low Ca 2+ cells suggest dynamic activi ty o f the

cell m e m b r a n e dur ing the ear ly stages o f develop-

ment. The surface ruffling o f cell type B and the

fo rma t ion o f l amel l ipodia in type C may be the

result o f membrane fluidity 5. The extensive prol ifer-

a t ion o f these cel lular features dur ing the initial

hours o f p la t ing and the decrease in surface ruffling

after 6 h o f p la t ing (Fig. 7) fur ther suggest dynamic

activi ty o f the cell membrane . M e m b r a n e activity,

however, is lacking in cells grown in the Ca2+-free

medium, suggesting that the presence o f external

Ca z+ may p lay an impor t an t role in the dynamic

processes o f surface membrane activi ty and cel lular

ou tg rowth dur ing early neural cell adhesion.

Scanning electron microscopy is a useful tool in

observing membrane surface act ivi ty as c o m p a r e d to

phase-cont ras t l ight microscopy. Unfor tuna te ly , the

morpho log ica l charac te r iza t ion o f the different cell

types is difficult, as the cells are still at a relat ively

pr imit ive stage o f development . Elec t rophysio logi -

cal and immunohis tochemica l procedures may help

to resolve some o f these difficulties.

REFERENCES

1 Allen, A. and Snow, C., The effect of trypsin and eth- ylenediamene-tetraacetate on the surface of cells in tissue culture, Biochem. J., 117 (1970) 32P.

2 Aoyama, H. and Okada, T. S., Mechanisms of cell- to-cell and cell-to-substrate adhesion of human epi- thelial, Chang conjunctiva cells, Cell Struct. Funct., 2 (1977) 281-288.

3 Atsumi, T. and Takeichi, M., Cell association pattern in aggregates controlled by multiple cell-cell adhesion mechanisms, Develop. Growth Differentiation, 22 (1980) 133-142.

4 Barnard, P. J., Weiss, L. and Ratcliff, T., Changes in the surface properties of embryonic chick neural retina ceils after dissociation, Exp. Cell Res., 54 (1969) 293- 301.

5 Ben-Shaul, Y., Ophir, I., Cohen, E. and Moscona, A. A., SEM study of dissociated embryonic cell surface activity, Scan. Electron Microsc., 2 (1977) 29-36.

6 Bray, D., Surface movements during the growth of single explanted neurons, Proc. nat. Acad. Sci. U.S.A., 65 (1970) 905-910.

7 Bray, D., Branching patterns of individual sympathetic neurons in culture, J. Cell BioL, 56 (1973) 702-712.

8 Burger, M. M., A difference in the architecture of the surface membrane of normal and virally transformed cells, Proc. nat. Acad. Sci. U.S.A., 62 (1969) 994--1001.

9 Cook, G. M. N., Heard, D. H. and Seaman, G. V. F., A sialomuco-peptide liberated by trypsin from the human erythrocyte, Nature (Lond.), 188 (1960) 1011- 1012.

10 Deman, J. J., Bruyneel, E. A. and Mareel, M. M., A study on the mechanism of intercellular adhesion, J. Cell Biol., 60 (1974) 641-652.

11 Edidin, M., The release of soluble H-2 alloantigens during disaggregation of mouse embryo tissue by a chelating agent, J. EmbryoL exp. MorphoL, 16 (1966) 519-530.

12 Holtfreter, J. V., Properties and function of the surface

coat in amphibian embryos, J. exp. Zool., 93 (1943) 251-323.

13 Kemp, R. B., Studies on the effect of dissociating agents on the electrophoretic mobility and aggregative com- petence of embryonic chick muscle cells, Cytobios., 1 (1969) 187-196.

14 Lilien, J., Hermolin, J. and Lipke, P., Molecular inter- actions in specific cell adhesion. In D. R. Garrod (Ed.), Specificity of Embryological Interactions, Chapman and Hall, London, 1978, pp. 131-155.

15 Magnani, J. L., Thomas, W. A. and Steinberg, M. S., Two distinct adhesion mechanisms in embryonic neural retina cells. I. A kinetic analysis, Develop. Biol., 81 (1981) 96-105.

16 McCarthy, K. D. and Partlow, L. M., Preparation of pure neuronal and non-neuronal cultures from embry- onic chick sympathetic ganglia: a new method based on both differential cell adhesiveness and the formation of homotypic neuronal aggregates, Brain Res., 114 (1976) 391-414.

17 Moscona, A. A., Embryonic and neoplastic cell sur- faces: availability of receptor for concanavalin A and wheat germ agglutinin, Science, 171 (1971) 905-907.

18 Nicolson, G. L., Topography of membrane concana- valin A sites modified by proteolysis, Nature New Biol., 239 (1972) 193-197.

19 Nishi, R. and Berg, D. K., Dissociated ciliary ganglion neurons in vitro: survival and synapse formation, Proc. nat. Acad. Sci. U.S.A., 74 (1977) 5171-5175.

20 Rothman, S. and Cowan, W. M., A scanning electron microscope study of the in vitro development of dis- sociated hippocampal cells, J. comp. NeuroL, 195 (1981) 141-155.

21 Sela, B.-A., Lis, H., Sharon, N. and Sachs, L., Differ- ent locations of carbohydrate-containing sites in the surface membrane of normal and transformed mam- malian cells. J. Membrane BioL, 3 (1970) 267-279.

22 Steinberg, M. S., Armstrong, P. B. and Granger, R. E., On the recovery of adhesiveness by trypsin-dis- sociated cells, J. Membrane Biol., 13 (1973) 97-128.

268

23 Takeichi, M., Functional correlation between cell adhesive properties and some cell surface proteins, J. Cell BioL, 75 (1977) 464-474.

24 Takeichi, M., Ozaki, H. S., Tokunaga, K. and Okada, T. S., Experimental manipulation of cell surfaces to affect cellular recognition mechanisms, Develop. Biol., 70 (1979) 195-205.

25 Thomas, W. A. and Steinberg, M. S., Two distinct adhesion mechanisms in embryonic neural retina cells. II. An immunological analysis, Develop. Biol., 81 (19811 106-114.

26 Thomas, W. A., Thomson, J., Magnani, J. L. and Stein- berg, M. S., Two distinct adhesion mechanisms in embryonic neural retina cells. III. Functional specificity, Develop. BioL, 81 (1981) 379-385.

27 Townes, P. L. and Holtfreter, J., Directed movements

and selective adhesion of embryonic amphibian cells, J. exp. ZooL, 128 (1955) 53-120.

28 Urishihara, H., Ozaki, H. S. and Takeichi, M., Immuno- logical detection of cell surface components related with aggregation of Chinese hamster and chick embry- onic cells, Develop. Biol., 70 (1979) 206'216.

29 Urishihara, H., Takeichi, M., Hakura, A. and Okada, T.-S., Different cation requirements for aggregating BHK cells and their transformed derivatives, J. Cell Sci., 22 (1976) 685-695.

30 Urishihara, H., Ueda, M.-J., Okada, 1.-S. and Takeichi, M., Calcium -dependent and -independent adhesion of normal and transformed BHK cells, Cell Struct. Funct., 2 (1977) 289-296.

31 Wong, A. and Jones, D. G., SEM study of cell surface features of developing mouse cerebral cells, Aust. Net~- rosci. Proc., 1 (1981) l12P.