“In vitro” effect of GM1 ganglioside on adrenal chromaffin cells

Life Sciences, Vol. 54, No. 12, pp. 823-830, 1994 Copyright © 1994 Elsevier Science Lid Printed in the USA. All rights reserved 0024-3205/94 $6.00 + .00

Pergamon

"IN VITRO" EFFECT OF GM1 GANGLIOSIDE ON ADRENAL CHROMAFFIN CELLS J.A. Colombo, M. Napp and G. Dran* Programa Unidad de Neurobiologla Aplicada (PRUNA)(CEMIC-CONICET), Avda. Galvfin 4102, (1431) Buenos Aires, Argentina. (Received in final form January 3, 1994) Summary Exposure of "in vitro" grown immature and adult adrenal chromaffin cells to concentrations of 10.3 or 10.5 M but not 10.7 M GM1 ganglioside, resulted in significant increase in cell diameter, coupled with reduction of adhesion to substrate within 48 hrs of exposure. None of the GM1 concentrations, with or without serum supplementation, did significantly increase neuritogenesis in chromaffin cells. Immature chromaffm cells underwent neuritogenesis when grown in co-cultures with actively growing astroglia from striatum or cerebral cortex, an effect that was potentiated by NGF admini~ration and blocked by anti-NGF. In neither of the former conditions did 10.6 M GM1 prove to increase the number of neuxite emitting cells nor their mean neuritic length further. It is speculated that GM1 does not perform the neuritogenic role described for central neurons in chromaffin cells, nor does it potentiate NGF effect on neuritogenesis observed in other peripheral neurons. Chromaffm cells are neuroectodermal cell derivatives, characterized by an endocrine phenotype. Yet they will undergo neuritogenesis under certain experimental conditions (7,15,18,20). Gangliosides are natural constituents of cell membranes and, GM1 particularly, have been implicated in several central and peripheral neuronal phenomena including cell recognition, neuritic growth and neuronal regeneration (3, 4, 6, 12, 16). Such effects suggested the analysis of its possible effects on chromaffin cells phenotype, most specifically on the induction of neuritic growth. Such analysis could provide further data regarding both, mechanisms involved in membrane changes leading to neuritogenesis, and defining further optimal conditions for chromaffin cell transplantation in addition to NGF (10,14). Since the above mentioned effects of GM1 have been linked to an interaction with other neurotrophic factors, we submitted chromaffin cells to GM1 in culture conditions of varying trophic activities judging by the incidence and growth of neuritic processes. Some of these results have been akeady published in abstract form (2, 11). Methods Adult male and female Sprague Dawley rats were fed standard laboratory diet and kept under controlled illumination (10/14 hrs., light/dark cycles per day). Adrenal chromaffin cells were * Present address: Academia Nacional de Medicina, Buenos Aires.

824

GM1 and ChromalTm Cells

Vol. 54, No. 12, 1994

obtained from adult males and neonatal (10-13 days old) rats. Animals were rapidly decapitated, their adrenals excised and placed in Hanks or PBS solution without any Ca ++ or Mg ++, and their medullae dissected out, cut into small pieces and incubated in 0.1% collagenese (Wothtington) at 3 7 o c during 45 minutes, followed by mechanical dissociation. After centrifugation, cells were plated on 1-polylysine coated coverslips, at the approximate equivalent o f half neonatal or one quarter medulla per coverslip (6-103 cells/cruZ). 1. GM1 in basal medium: in order to allow cell adhesion to substrate before GM1 treatment, coverslips containing cell dissociates were preincubated during 40 rain in a 10ul drop o f DMEMF12, and then transferred to capsules containing one o f the following concentrations o f GM1 (FIDIA): 10-3 M, 10-5 M, 10-6 M or 10-7 M, in DMEM/F12 plus antibiotics (GIBCO). In some groups 10% foetal calf serum was added to this culture medium. Culture dishes were incubated at 37oc, in CO2 (5%) air (95%) atmosphere. Medium was not renewed for the following 5-7 days. Whenever necessary, t~esh GM1 was added on a 5-7 day basis, at the time when culture media were replaced. Additional cultures were fed with one o f the following media: medium 199 or MEM, with or without foetal calf serum (10%) or 3% LPSR (SIGMA) supplement. All culture conditions were run in triplicates. 2. GM1 in astroglial conditioned media: additionally, samples o f cell dissociates were cultured in glial-conditioned medium (GCM), as previously described (Colombo et aL, submitted). Briefly, striatal foetal cells (El7) were mechanically dissociated in Hanks balanced salt solution without any Ca ++ or Mg ++ and seeded in 1-polylysine coated dishes containing DMEM/F12 medium supplemented with 10% foetal calf serum and antibiotics as above. After attaining cell confluence, the medium was shifted to DMEM/F12 without serum for 72 hrs (GCM). 3. GM1 in chromat]fin-astroglial co-cultures: these were developed on glass coverslips treated with polylysine on one half o f the coverslip where chromaftin cells were to be seeded; astroglial cells obtained by trypsiniTation o f striatal or c.cortex subcultures were seeded at 5x103 cells/cm2 on the remaining half o f the coversfip, and this was placed in a dish containing DMEM/F12 medium. NGF 2.5S (Sigma) was added to some o f these dishes at a concentration o f 50 or 100 ng/ml. GM1 at 10-6 M was added to co-cultures either after 24 or 48 hrs o f initiated NGF treatment. Anti-NGF (Sigma) was added to another set o f capsules containing immature chromaffan cells in GCM or co-cultures, at a concentration enough to block 25 or 50 ng/ml o f 2.5 S NGF. Daily analysis and photographic records o f cultures were obtained with a Zeiss inverted microscope equipped with phase contrast optics. In order to identify chromaftin cells, cultures were fixed in K-dicromate and processed for neutral red stain or Giemsa (8). Astroglia was identified by GFAP (+) and fibronectin (-) immunocytochemistry. Statistical analysis was performed using two tailed Student's t test or one way A N O V A whenever required. Results In general, one week old serum (or LPSR) supplemented cultures were characterized by a prevailing growth offibroblasts that eventually overgrew other cell types. This condition did not allow a reliable counting o f cells after 7 DIV. Immature, and to a lesser extent, adult adrenal cell dissociates cultured in GCM developed neurites as early as 18-20 hrs after seeding. The most consistent effects o f 10-3 M and 10-5 M GM1 concentrations in basal media, with or without supplementation, were reduction o f cell adhesion to substrate (Fig. 1) and an increase in cell size, both o f which were dose dependent . Cells grown in basal medium showed a mean diameter o f 8.5 um while it increased up to 18.5 and 16.2 um in 10-3 and 10-5 M GM1

Vol. 54, No. 12, 1994

GM1 and Chromaffm Cells

825

respectively (Student's L test p<0.05 ), and had a foamy appearence. Both effects appeared to overlap in time, and were most apparent after 24-48 hrs. GM1 at 10-6 M and 10-7 M concentrations did not produce such effects being cell diameters and cell numbers ~milar to those observed in control eunditions (basal culture medium with or without serum supplement). Adrenal medulla cells obtained from immature and adult donors behaved in a similar way, although the former appeared to be more affected by the treatment.

25C -(3 t4--

20C

U3 (D (D

15C

t4 _

O

100

T

ki_

(1) n E D c-

50

none

D Q none 10-3 10-~ 10-7 GM1 TREATMENT (molarity) Fig. 1

Number of cells adherent to a polylysine substrate after seeding (empty bar) or 48 hrs in culture (dashed bars), in DMEM/F12 medium with or without GM1. The number of cells was obtained by counting in five symmetric microscope fields per coverslip, starting from the center (relative density) of cell cultures. Mean +/- SEM *p<0.05 with respect to non-treated group (Student's t test). In none of the conditions tested did GM1 si~,nificantly enhance neuritic growth from identified chromaffan cells beyond the pattern observed in control groups. Since the most efficient culture condition to elicit persistent neuritic growth from chromatfin cells involved co-culturing immature adrenal medulla cells with recently seeded astroglia, the effect of GM1 was also analyzed in such conditions, with and without concomitant NGF treatment. For this purpose, striatal astroglia was seeded as previously described. Nenrites were observable 24 hrs after seeding and kept growing with time (Figs. 3, 4) (df=2; F=5.839; p<0.01). Most reactive cells developed a characteristic monopolar structure, and gave a positive stain with K dicromate and neutral red. Co-cultures receiving NGF showed increased neuritic growth in chromaffm cells (Figs. 2, 3), characterized in several cases by long neurites (Fig. 3B) and increased number of un~olar cells (Chi square, p<0.0001). In these conditions, addition of 10-6 M GM1 either 24 or 48 hrs after NGF treatment was initiated, did not modify such features (Figs. 2A, 3C), and tended to reduce the number of emitting cells (Chi square, p<0.01) (Fig. 2B). For overall group comparisons, ANOVA test was applied (df-=4; F=2.534; p<0.05). Post hoe analysis was performed between groups using a modified Tukey's test due to the unequal number of observations between samples (1). Results indicate a si~ificant (p<0.05) difference between

826

GM1 and Chromaff'mCells

Vol. 54, No. 12, 1994

control and NGF treated cultures at 72 hrs, and between the former and NGF+GM1 group. It may be worth noticing that in all cases mostly primary or, sometimes, secondary neurites were observed. No significant differences were observed between control and GM1 treated groups after 72 hrs in vitro. Addition of anti-NGF, at either concentration tested, blocked neuritogenesis in co-cultures of chrornaffin and astroglial ceils. In none of these cases there was morphological evidence of astroglial changes.

A 250

B lOOt

E 2_ "~' 2 0 0 £_ 4J [D 150 C

T

5O

lO0

T

4A L J

50

C 0

CONTROL

I

25 i

'CONTROLI

01

NGF

NGF GMI

GMI 24 h

GMI 48 h

NGF

NGF GMI

GMI 24 h

GMI 48 h

Fig.2 A) Average neuritic length (urn) of tmipolar chromafl]n cells alter 72 hrs in coculture with striatal astroglia, under different treatments. Mean +/- SEM . B) Percentage ofunipolar cells in similar culture conditions as in A). Discussion

No increase in neuritogenesis was observed "in vitro" in rat chromatfin ceils following GM1 treatment, when cultured under various conditions, some of which fostered neuritic growth by themselves. Additionally, GM1 concentrations higher than 10-6 M resulted most frequently in some form of interference with their adhesion to the polylysine substrate. 10.6 M GM1 concentration was used to examine its effects on immature chromafl~ cells co-cultured with recently seeded striatal astroglia, with and without NGF addition. While NGF potentiated trophic effects of astroglial co-cultures, GM1 did not significantly modify neuritogenesis in any of these conditions. Intervening mechanisms involved in ganglioside promotion of cell survival and neuritic growth yet remain to be established. Gangliosides have been described as being associated with the outer leaflet of the plasma membrane in various possible ways, depending on the physical form (monomer, oilgomer, micellae) of the starting ganglioside (17). Their variety and location has led to envision them as active agents in the process of cell to cell recognition (23) and perhaps in the modulation of receptors for neurotrophic factors (22). Evidence obtained from the present studies

Vol. 54, No. 12, 1994

GM1 and Chromaffin Cells

827

Fig. 3 Immature adrenal medulla cell disociates cocultured in vitro for 72 hrs with striatal astroglia in basal mediunl A) control. B) with NGF 50 ng/mi. C) with NGF plus GM1 10-6 M . Phase contrast microscopy. Bar = 40unl

828

GM1 and Chromaffin Cells

Vol. 54, No. 12, 1994

B f-, 450 E 400

',450 :14oo

350 [42 300 (D C 250

~350 r)300

~

C 250

(9 --

200

200

--

OJ 150 4J U i00

4~ 1 5 0 L

i00

5O

3 (9 C

5o

C

o

0 '

cells

cells

500 /

I

4oot 3OOl250[

C~250

2oot .~_ 15°t

2OO

I

15Ol L

3

i °o[ Trl I I I I I I I I I I I I

C

lOOI

5Ol C' ceils

cells

Fig. 4 Neuritic length of individual unipolar chromaffm cells ~om immature rat donors, under different treatments: 24 hrs (A) and 72 hrs (B) control co-cultures of chromaffiu and striatal astroglial cells in basal medium (C) 72 hrs treatment with NGF 50 ng/ml. (D) as in (C) with addition of 10 .6 M GM1. Number of cells and neurite length were determined in 10-15 microscope fields per condition. suggests that GM1 does not constitute a significant factor in the process by which chromaffi, cells may be instructed to undergo phenotypic transformation to a neuronal cell type. Its effects at high concentrations on cell adhesion to substrate and on cell volume suggest displacement or interference with the synthesis or insertion of cell adhesion molecules into the phqma membrane, coupled with vacuolar degeneration, judging by their morphological aspect under phase contrast microscopy. Increase in cell size partially resembles a cellular change also observed in ganglioside storage diseases, for which disfimilar characteristics have been described between GM1 and GM2 (24). In fact, observed differences among gangliosides in terms of their presence in various cell types (21) and in the expression of cell pathology in gangliosidoses, support the idea of a dissimilar potency to affect cellular processes in any particular cell type. In this respect, it is interesting to note that Unsicker and Wiegandt (21) have descn%ed a positive effect of GDlb and GTlb, but not GM1, on peripheral neuron survival. In 8 ED chick dorsal root ganglion neurons,

Vol. 54, No. 12, 1994

GM1 and Chromaffin Cells

829

GM1 efficiency was 20% while the other two gangliosides showed a 60-80% efficiency. These results were obtained at molar concentrations within the range of those used in the present study (10-4 to 10-5 M). According to the same authors, ciliary ganglion neurons survival was virtually not supported, suggesting specificity regarding neuronal targets. Moreover, the apparent absence of GM1 in PC12 cells (24) and its lack of effect in the present studies casts doubt on its physiological role in these cell types. In regard to this it is interesting to note, that a recent report describes an inhibitory effect of GM1 on PC12 cell growth (25). Other gangliosides, such as disialogangliosides and GM3 may be functionally more significant in such cells. Finally, the observed potentiating effects of NGF in co-cultures and interference by anti-NGF, supports the idea that NGF is an integer molecule for neuritic growth mechanisms in chromaffin cells (18,19), and suggests that astroglial effects on chromaffin cells may depend on an NGF-like factor released by astroglia. Synthesis of NGF by "in vitro" and "in vivo" growing astroglia (5,7), known dependence of chromaffin cells on NGF for neuritic grox~th, and its inhibition by anti-NGF would tend to support the above mentioned possibility. At any rate, it seems worth stressing that in these experiments such trophic effect was exerted by astroglia lacking contact inhibition such as in the case of confluent cultures, t~om which conditioned media were actually obtained (Colombo et al. submitted). It is possible that astroglial synthesized protease inlfibitors (13) could participate in the observed neuritogenic effects. Acknowledgements. The authors wish to thank Dr. J. Paz (CEMIC) for statystical analysis, and Vir~nia Puissant, P.Lotersztain, D. Kestelman and D. Gicovate for technical help. This work was supported by the CONICET (Argentina), Petrolera Argentina San Jorge, a grant from the EEC, Fundaci6n Conectar and the A.von Humboldt Foundation (Germany). References 1. F.J. ANSCOMBE and T.W. TUKEY, Teclmometrics 5 141 (1963). 2. J.A. COLOMBO, G. PEREZ, IL CACCURI and G. DRAN, Ann. Meet. Soc. Neurosci. 15 1366 (1989). 3. C. CUELLO, L. GAROFALO, ILL. KENIGSBERG and D. MAYSINGER, Proc. Natl. Acad. Sci. USA 86 2056-2060 (1989). 4. P.H. FISHMAN, Chem Phys. Lipids 42 137-151 (1986). 5. S.H. FURUKAWA, Y. FURUKAWA, E. SATAYOSH] and I ( HAYASHI, Biochen~ Biophys. Res. Comm~m. 142 395-402 (1987). 6. A. LEON, D. BENVEGNU, IL DAL TOSO, D. PRESTI, L. FACCI, O. GIOGI and G. TOFFANO, J. Neurosci. Res. 12 277-287 (1984). 7. B. LU, M. YOKOYAMA, C.F. DREYFUS and I.B. BLACK, J. Neurosci. 11 318-326 (1991). 8. M.A. MORO, M.G. LOPEZ, L. GANDIA, P. MICHELENA and A.G. GARCIA Anal.Biochem 185 243-248 (1990). 9. K_W. NAUJOKS, S. KORSHING, H. ROHRER and H. THOENEN, Dev. Biol. 9__2_2 365-379 (1982). 10. L. OLSON, E.O. BACKLUND, T. EBENDAL, IL FREEDMAN, B. HAMBERGER, a. HANSSON, B. HOFFER, U. LINDBLOM, B. MEYERSON, I. STROMBERG, O. SYDO and A. SEIGEIL Arch.Neurol. 48 373-381 (1991).

830

GM1 and Chromaffin Cells

Vol. 54, No. 12, 1994

11. G. PEREZ, A. POSSE, M.V. ROSATO SIRI and J.A. COLOMBO, Ann. Meet. Soc. Neurosci. 15 279 (1989). 12. B.A. SABEL, IL DEL MASTRO, G.L. DUNBAR and D. STEIN, Neurosci. Lett. 77 360366 (1987). 13. T.B SHEA, M.L. BEERMAN and ILA. NIXON, J. Neurosci. Res. 31 309-317 (1992). 14. I. STROMBERG, M. HERRERA-MARSCHITZ, U. UNGERSTEDT, T. EBENDAL and L. OLSON Exp.Brain Res. 60 335-349 (1985). 15. A.S. TISCHLER, ILL. PEARLMAN, G. NUNNEMACHE1L G.M. MORSE, ILA. DE LELLIS, H.J. WOLFE and B.E. SHEAD, Cell Tissue Res. 255 525-542 (1982). 16. G. TOFFANO, G. SAVOINI, F. MORONI, G. LOMBARDI, L. CALZA and L.F. AGNATI, Brain Res. 261 163-166 (1983). 17. G. TOFFANO, D. BENVEGNU, A.C. BONETTI, L. FACCI, A. LEON, P. ORLANDO, IL GHIDONI and G. TETTAMANTI, J.Neurochem 35 861-864 (1980). 18. IC UNSICKER, B. KIRSCI-I, U. OTTEN and H. THOENEN, Proc. Natl. Acad. Sci. USA 75 3498-3502 (1978). 19. IC UNSlCKER, T. MILLER and H. HOFFMAN, Dev. Neurosci. 5 412-417 (1982). 20. K_ UNSICKER, J. VEY, J.D. HOFFMAN, T.H. MULLER and A.J. WILSON, Proc. Natl. Acad. Sci. USA 81 2242-2246 (1984). 21. IC UNSlCKER and H. WlEGANDT, Exp. Cell Res. 178 377-389 (1988). 22. S.U. WALKLEY, H.J. BAKER and M.C. RATAZZI, Dev. Brain Res. 51 167-178 (1990). 23. T. YAMAKAWA and Y. NAGAI, TIBS 3 128-131 (1976). 24. H. YAMAMOTO, S. TSUJI and Y. NAGAI, J. Neurochem. 54 513-517 (1990). 25. A.J. YATES and J. CONLEY, Restor. Neurol. Neurosci. 4 51-54 (1992).