Differential effects of bFGF on development of the rat retina

BRAIN RESEARCH ELSEVIER

Brain Research 723 (1996) 169-176

Research report

Differential effects of bFGF on development of the rat retina Shulei Zhao, Colin J. Barnstable

*

Department of Ophthalmology and Visual Science, Yale University School of Medicine, 330 Cedar Street, P.O. Box 208061, New Haven, CT 06520-8061, USA Accepted 7 February 1996

Abstract

A variety of growth factors can influence the expression of differentiated properties by cell types of the developing retina. One unresolved question has been whether these factors can direct the differentiation pathway of uncommitted precursors or whether they act to help the expression of properties by already committed cells. To address this question we have studied the effects of basic fibroblast growth factor (bFGF) on the differentiation of ganglion cells and rod photoreceptors in explant cultures of embryonic rat retinas. Incubation of retinas in the presence of bFGF accelerated the appearance of differentiated ganglion cells and incubation in the presence of anti-bFGF antibodies delayed the appearance, bFGF had no effect on the appearance of differentiated rod photoreceptors as judged by expression of opsin, although all-trans-retinoic acid did increase the number of cells expressing opsin, bFGF inhibited the formation of rod photoreceptor rosettes suggesting that it does influence some properties of rods or the adjacent Mtiller glial cells. The results suggest that bFGF can alter the timing of differentiation of retinal ganglion cells but not direct their production from retinal precursors. Keywords: Development; Neural retina; Ganglion cell; Rod photoreceptor; bFGF; Retinoic acid; Opsin; Tissue culture

1. Introduction The neural retina is an excellent model system for studying differentiation in the CNS because it has a welldefined laminar structure, its five major types of neurons and two types of glia have been well characterized and a wide array of cell-type-specific molecular markers are available to study retinal development both in vivo and in vitro [5-7,22]. Lineage tracing studies have shown that all the cell types in the retina can arise from a single precursor cell [13,27,28,31]. Although there are minor differences among species, birthdating studies have shown that the overall patterns of the sequence of cell birth are remarkably similar across all vertebrates [24,25]. For example, retinal ganglion cells and rod photoreceptors are always the first and among the last cell types to be generated, respectively. There is still considerable debate as to whether development proceeds by the action of a changing microenvironment on pluripotent precursor cells, or whether the potential of precursors changes over time such that

* Corresponding author. Fax: (1) (203) 785-7401. I)006-8993/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S0006-8993 ( 9 6 ) 0 0 2 3 7 - 5

common signals can elicit different responses at different times in development [4,18,29,30]. There is good evidence that diffusible factors, including fibroblast growth factors (FGFs) [t2,17], ciliary neurotrophic factor (CNTF) [9], and retinoic acid [15], can influence retinal cell differentiation in dispersed cultures. Because these studies have shown an increased number or proportion of a particular cell type within a constant total cell number, these results have been interpreted to suggest that the factors are influencing commitment and differentiation from uncommitted precursors. A number of studies have suggested that low density monolayer cultures of retinal cells do not differentiate fully [2,3,26]. For example, while there are many reports of rod photoreceptors expressing the early marker opsin, these cells do not express later markers such as transducin or the rod-specific phosphodiestrase [2,3,23]. In reaggregate cultures or high density clusters in monolayer cultures, rod photoreceptors undergo more complete morphological and molecular differentiation [23,3]. Similarly, Miiller glial cells also undergo more extensive differentiation in reaggregate cultures where extensive cell contacts occur [3]. Comparison of monolayer cultures with explants have shown even

S. Zhao, C.J. Barnstable / Brain Research 723 (1996) 169 176

170

[]

s. Zhao, C.J. Barnstable/ Brain Research 723 (1996) 169-176

more striking differences in that low density monolayers of embryonic cells did not form rod photoreceptors, as judged by opsin expression, whereas explants from the same age of embryos formed rods at an equivalent time to that seen in vivo [26]. To more closely approximate the in vivo condition, we have used retinal explants to study factors that influence retinal cell differentiation at various stages of development. Our results show a clear effect of bFGF on ganglion cell differentiation but not rod photoreceptor differentiation. bFGF did, however, have effects on explant morphology that suggest it influences either Miiller glia or the interactions of these cells with the developing rods.

2. Materials and methods 2.1. Tissue cultures

Pregnant L o n g - E v a n s rats were sacrificed at 12 and 16 days of gestation (El2 and El6, where E0 is defined as the day of conception) and the eyes were removed from embryos as previously described [26]. The RPE and attached tissues were separated from the neural retina by excising around the peripheral margin of the RPE with fine needles under a dissecting microscope. Each separated retina with the developing lens was cultured in individual wells of plastic 24-well plates (Coming Glass Works, Coming, NY) without coating. The culture medium was either UltraCulture (UC), a serum-free medium (BioWhitaker, Walkersville, MD), with 2 mM L-glutamine added, or UC supplemented with 2 mM L-glutamine and 30 / z g / m l human recombinant bFGF (Upstate Biotechnology, Lake Placid, NY), 30 / x g / m l rabbit anti-bFGF antiserum (Sigma, St. Louis, MO), or 1 /xM all-trans-retinoic acid (Sigma, St. Louis, MO). The anti-bFGF antibodies have been shown to block the effect of bFGF. One of the two retinas from each E l 2 embryo was always cultured in UC as a control and the other cultured in the supplemented UC. The two retinas were always harvested and analyzed at the same time to observe the effect of the factors. The tissue cultures were maintained in a humidified atmosphere of 5% CO 2, balance air at 37°C and the medium was changed every 3 days. 2.2. Immunocytochemistry

The cultured retinal tissues were harvested after various periods of time, fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) at 4°C for 2 h, and then

171

infiltrated with 15% sucrose in PBS at 4°C overnight. These tissues were frozen in O.C.T. compound (Miles, Elkhart, IN) and 12 /xm sections were cut on a cryostat and mounted on slides treated with 2% 3-aminopropyltriethoxysilane (Sigma, St. Louis, MO) in acetone. All further procedures were carried out at room temperature. Cryostat sections were preincubated with 5% normal goat serum (NGS) in PBS for 15 min followed by incubation with primary antibodies for at least 1 h. Three mouse monoclonal antibodies used in this study were anti-/3tubulin isotype III (Sigma, St. Louis, MO), anti-neurofilament (70 kDa) antibody 8A1 [33], and anti-rhodopsin antibody RET-P1 [5,33]. After rinsing with a large excess of PBS and a final wash with 5% NGS, sections were incubated for 1 h with fluorescein isothiocyanate (FITC)conjugated goat anti-mouse IgG (Boehringer Mannheim, Indianapolis, IN). After extensive washing in PBS, sections were coverslipped with 50% glycerol in PBS. Sections were viewed and photographed under epi-fluorescence illumination.

3. Results 3.1. bFGF promotes retinal ganglion cell differentiation

Explants of E12 rat retina (where E0 is defined as the day of conception) were dissected free of RPE and other tissues and cultured as previously described [26]. At E l 2 no differentiated cells were seen by immunocytochemistry with antibody to /3-tubulin isotype III(/3-tub), an early marker for differentiating ganglion cells that is expressed while the cells still have a neuroepithelial morphology and span the entire neuroblastic layer [8]. In the E l 2 retinal explants, /3-tub + cells were first observed in medium containing 30 / x g / m l bFGF ( U C / b F G F ) after 36 h in culture, but not in explants cultured in UC or UC containing 3 0 / x g / m l anti-bFGF antibodies. After 40 h in culture, many more /3-tub + cells appeared in bFGF-treated explants and a few in explants cultured in UC, but no /3-tub + cells were detected in explants treated with anti-bFGF antibodies (Fig. 1). /3-tub + cells eventually appeared in anti-bFGF antibody-treated explants after 46 h in culture. This pattern of appearance of/3-tub + cells was obtained in 12 independent experiments. Immunocytochemistry with anti-70 kDa-neurofilament antibody 8A1 confirmed the accelerating effect of bFGF on ganglion cell differentiation (Fig. 2). 8A1 antibody labeled only ganglion cells in the embryonic and newborn rat retina but also recognized horizontal cells a few days

Fig. 1. Expression of /3-tubulin isotype III in El2 retina cultured for 40 h. A, C, E, and G are fluorescence images and B, D, F, and H are phase contrast images. Many more /3-tub+ cells appeared in the retinal explant treated with bFGF (A) compared to that cultured in the serum-free medium UC (C and E). A and C are sections of the left and right retinas, respectively, from the same embryo. No /3-tub+ cells were seen in the explant treated with anti-bFGF antibdies (G). E and G are sections of the left and right retinas, respectively, from the same embryo. NR, neural retina; L, lens.

[]

[]

[]

I

s. Zhao, CJ. Barnstable / Brain Research 723 (1996) 169-176

after birth and in adulthood (data not shown). Expression of neurofilament lagged behind that of/3-tub in the differentiating retinal ganglion cells by about 4 - 8 h. In culture, weak labeling of 8A1 was first observed in bFGF-treated E l 2 retinal explants after 40 h in culture, but not in explants cultured in UC until 48 h and 54 h for explants cultured in UC containing anti-bFGF antiserum. Together, these observations indicate that ganglion cell differentiation was accelerated by bFGF, while the onset was delayed by anti-bFGF antibodies. 3.2. Retinoic acid but not b F G F promotes rod photoreceptor differentiation

At E l 6 rod photoreceptor differentiation has not yet begun, as judged by expression of the visual pigment protein opsin [10]. Isolated opsin + cells were first observed in E l 6 retinal explants after 5 days in UC containing 1 /xM retinoic acid ( U C / R A ) . No opsin + cells were seen in tissues cultured in UC or U C / b F G F . On day 6 in culture, many more opsin + cells were observed in explants cultured in U C / R A while some isolated opsin + cells appeared in explants cultured in UC and U C / b F G F (Fig. 3). Each of these observations was made in at least 16 independent experiments. Thus, treatment with retinoic acid, but not bFGF, speeded up the onset of rod photoreceptor differentiation as compared to controls. After about 6 days in culture, some E16 retinal explants cultured in U C / R A and to a lesser extent those cultured in UC, developed small lumps and pits on the tissue surface. No change was observed in explants cultured in U C / b F G F on day 6. After 8 days in culture, the change in tissue appearance had spread to the entire surface of explants cultured in U C / R A while the same degree of change on explants in UC was observed 2 days later. The surface of explants in U C / b F G F still remained smooth on day 10 in culture. This morphological change was due to invagination of rod photoreceptor cells at outer layer of the explant to form semi-rosette structures as detected by their strong labeling with RET-P1 antibody. The cells were then internalized to form full rosettes (Fig. 4). This process was accelerated by retinoic acid whereas it was inhibited by bFGF.

4. Discussion Previous work has demonstrated that bFGF may be an important factor in the determination of the fate of cell

173

layers in the optic vesicle [17,20,21,33]. The present study has demonstrated that bFGF increases the rate of differentiation of retinal ganglion cells during early developmental stages. Ganglion cell differentiation was slowed but not prevented by antibodies to bFGF, an effect we assume to be due to the blocking of endogenous bFGF. We propose that this growth factor enhances differentiation but probably does not specify ganglion cells from undifferentiated precursors. This conclusion is supported by a recent report in which ganglion cells were formed normally from precursors expressing a dominant negative FGF receptor [19]. bFGF had no apparent effect on rod photoreceptor differentiation from E l 6 explants. Other work has shown that in cultures of dissociated neonatal rat retinal cells treated with bFGF, the number of cells expressing opsin increased as compared with the untreated control [12]. The discrepancy between the two studies is probably due to the age of the tissue used, although dissociated retinal cells may respond to bFGF differently from those in explants. Older cells may be committed to become rod photoreceptors and bFGF may trigger their differentiation, a finding in agreement with studies that show an increase in nuclear protein binding at a key regulatory site in the opsin promoter and an apparent increase in the amount of opsin per cell [11,32]. The different effects of bFGF suggest that as development proceeds, the internal state of retinal cells changes, and therefore, they respond to an extracellular signal differently. Our observation that retinoic acid promotes photoreceptot differentiation is in agreement with previous findings of others [15]. The mechanism of action of retinoic acid is not yet understood. Which of the retinoic acid receptors is responsible for the effect has not been determined, although double knockouts of R X R a and R A R T cause significant abnormalities of eye development [14]. Examination of the upstream flanking regions of genes known to be expressed in differentiated rod photoreceptors has not yet detected any functional retinoic acid response elements, so that the site of action of retinoic acid may well be at an earlier phase of differentiation. It is also possible that the effect of retinoic acid is on other cell types that in turn influence rod photoreceptor differentiation, although because some of the effects of retinoic acid can be observed on isolated rod photoreceptors, it is more likely to be a direct effect [15]. The precocious differentiation induced by retinoic acid was accompanied by the coalescence of the rods into rosettes. These rosettes are reminiscent of structures seen in retinoblastoma as well as the retinal disorganization

Fig. 2. Expression of 70 kDa neurofilaments in El2 retina cultured for 48 h. A, C, E, and G are fluorescence images and B, D, F, and H are phase contrast images, Neurofilaments were strongly expressed in many cells in the inner layer of the retinal explant treated with bFGF (A). A few neurofilament+ cells were seen in the explant cultured in UC (B and E). A and B are sections of the left and right retinas, respectively, from the same embryo. Neurofilament was not expressed in the explant treated with anti-bFGF antibodies (G). E and G are sections of the left and right retinas, respectively, from the same embryo. Some lens cells were also labeled by the antibody (A and G) but the expression did not appear to correlate with bFGF treatment. NR, neural retina; L, lens.

174

S. Zhao, C.J. Barnstable / Brain Research 723 (1996) 169-176

Fig. 3. Sections of El6 retina explants cultured for 6 days labeled with anti-rhodopsin antibody RET-PI. A, C, and E are fluorescence images and B, D, and F are phase contrast images. Many opsin + cells appeared in the explant treated with retinoic acid (A) while only a few isolated opsin + cells were seen in explants cultured in UC (C) or UC with bFGF (E), and the numbers of opsin + cells in these two explants did not differ significantly.

S. Zhao, C.J. Barnstable / Brain Research 723 (1996) 169-176

caused by RPE diseases in animals [1]. The cause of rosette formation is unknown but could be due to aincrease in adhesion between rods or a lessening of the strength of the interactions between the developing rods and other retinal cells. It is possible that retinoic acid has effects on the expression of some of the molecules mediating such interactions as well as on the expression of opsin. Treatment of isolated retinas in culture with b F G F prevented rosette formation, although we do not know whether the effects were exerted on the rod photoreceptors themselves or on adjacent cells such as Mbller glia. The RPE has been shown to be a natural source of FGF, and one of its trophic effects on retina may be to help maintain its normal structure. This effect has also been demonstrated

175

by in vitro studies showing that the presence of RPE cells significantly influenced the structure of aggregates of retinal cells [16]. In summary, our results show that b F G F and retinoic acid car, accelerate the differentiation of ganglion cells and rod photoreceptors, respectively, in explant cultures of embryonic rat retina. This suggests that these factors have an important role in triggering the expression of differentiated properties, but that b F G F alone may not be involved in directing the differentiation of retinal precursors along particular differentiation pathways. Because b F G F had different effects at different times in development, our results further suggest that either the responsiveness (or potential) of the precursor cells is changing with time or that differ-

/

pil ~

Fig. 4. Tissue morphologies of El6 retinal explants cultured for 8 days. The tissue sections were labeled with RET-P1 antibody. A, C, and E are fluorescence images and B, D, and F are phase contrast images. In the explant treated with retinoic acid, invaginationof rod photoreceptors was completed and full rosettes had been formed. In the explant cultured in UC, most photoreceptor cells were invaginatingto form semi-rosettes(C). The tissue surface of the explant treated with bFGF remained smooth and photoreceptors maintainedtheir radial orientation(E).

176

S. Zhao, C.J. Barnstable / Brain Research 723 (1996) 169-176

ent combinations of positive and negative factors are necessary to induce a specific differentiation event.

[14]

Acknowledgements We thank Steven Viviano for excellent technical assistance. This study was supported by NIH grant NS 204083, the Foundation Fighting Blindness and a gift from Mr. James M. Kemper, Jr. SZ was the recipient of a Dr. Charles A. Perera Fellowship of the Fight for Sight research division of Prevent Blindness America and CJB is a Jules and Doris Stein Research to Prevent Blindness Professor.

References [1] Aguirre, G. and Rubin, L., Diseases of the retinal pigment epithelium in animals. In K.M. Zinn and M.F. Marmor (Eds.), The Retinal Pigment Epithelium, Harvard University Press, Cambridge, MA, 1979, pp. 334-356. [2] Akagawa, K. and Barnstable, C.J., Identification and characterization of cell types in monolayer cultures of rat retina using monoclonal antibodies, Brain Res., 383 (1986) 1101-1120. [3] Akagawa, K., Hicks, D. and Barnstable, C.J., Histiotypic organization and cellular differentiation of rat retinal neurons maintained in reaggregate culture, Brain Res., 437 (1987) 298-308. [4] Altshuler, D.M., Turner, D.L. and Cepko, C.L., Specification of cell type in the vertebrate retina. In C.J. Shatz and D.M.K. Lam (Eds.), Development of the Visual System, MIT Press, Cambridge, MA, 1991, pp. 155-165. [5] Barnstable, C.J., Monoclonal antibodies which recognise different cell types in the rat retina, Nature, 286 (1980) 231-235. [6] Barnstable, C.J., Immunological studies of the diversity and development of the mammalian visual system, lmmunol. Rev., 100 (1987) 47-78. [7] Barnstable, C.J., Molecular aspects of development of mammalian optic cup and formation of retinal cell types, Prog. Retina Res., 10 (1991) 69-88. [8] Brittis, P.A. and Silver, J., Exogenous glycosaminoglycans induce complete inversion of retinal ganglion cell bodies and their axons within the retinal epithelium, Proc. Natl. Acad. Sci. USA, 91 (1994) 7539-7542. [9] Fuhrmann, S., Kirsch, M. and Hofmann, H.-D., Ciliary neurotrophic factor promotes chick photoreceptor development in vitro, Development, 121 (1995) 2695-2706. [10] Hicks, D. and Barnstable, C.J., Different rhodopsin monoclonal antibodies reveal different binding patterns on developing and adult retina, J. Histochem. Cytochem., 35 (1987) 1317-1328. [11] Hicks, D. and Courtois, Y., Acidic fibroblast growth factor stimulates opsin levels in retinal photoreceptor cells in vitro, FEBS Lett., 234 (1988) 475-479. [12] Hicks, D. and Courtois, Y,, Fibroblast growth factor stimulates photoreceptor differentiation in vitro, J. Neurosci., 12 (1992) 20222033. [13] Holt, C.E., Bertsch, T.W., Ellis, H.M. and Harris, W.A., Cellular

[15]

[16]

[17]

[18] [19]

[20] [21]

[22] [23]

[24]

[25]

[26]

[27]

[28]

[29]

[30] [31] [32]

[33]

determination in the Xenopus retina is independent of lineage and birth date, Neuron, 1 (1988) 15-26. Kastner, P., Grondona, J.M., Mark, M., Gansmuller, A., LeNeur, M., Decimo, D., Vonesch, J.-L., Golle, P. and Chambon, P., Genetic analysis of RXRa development function: convergence of RXR and RAR signaling pathways in heart and eye morphogenesis, Cell, 78 (1994) 987-1003. Kelley, M.W., Turner, J.K. and Reh, T.A., Retinoic acid promotes differentiation of photoreceptors in vitro, Development, 120 (1994) 2091-2102. Layer, P.G. and Willbold, E., Histogenesis of the avian retina in reaggregation culture: from dissociated cells to laminar neural networks, Int. Rev. Cytol., 146 (1993) 1-47. Lillien, L. and Cepko, C., Control of proliferation in the retina: temporal changes in responsiveness to FGF and TGFc~, Development, 115 (1992) 253-266. Lillien, L., Changes in retinal cell fate induced by overexpression of EGF receptor, Nature, 377 (1995) 158-162. McFarlane, S., Cornel, E., Amaya, E. and Holt, C.E., Expression of a dominant negative FGF receptor in the developing retina, Soc. Neurosci. Abstr., 21 (1995) 1293. Park, C.M. and Hollenberg, M.J., Basic fibroblast growth factor induces retinal regeneration in vivo, Dev. Biol., 134 (1989) 201-205. Pittack, C., Jones, M. and Reh, T.A., Basic fibroblast growth factor induces retinal pigment epithelium to generate neural retina in vitro, Development, 113 (1991) 577-588. Reh, T.A., Cellular interactions determine neuronal phenotypes in rodent retinal cultures, J. Neurobiol., 23 (1992) 1067-1083. Reh, T.A. and Kljavin, I.L, Age of differentiation determines rat retinal germinal cell phenotype: induction of differentiation by dissociation, J. Neurosci., 9 (1989) 4179-4189. Robinson, S.R., Development of the mammalian retina. In S.R. Robinson and B. Dreher (Eds.), Neuroanatomy of the Visual Pathways and their Development, Macmillan, London, 1991, pp. 69-128. Sidman, R.L., Histogenesis of the mouse retina studied with thymidine-3H. In G.K. Smelser (Ed.), The Structure of the Eye, Academic Press, New York, 1961, pp. 487-506. Sparrow, J.R., Hicks, D. and Barnstable, C.J., Cell commitment and differentiation in explants of embryonic rat retina. Comparison to the developmental potential of dissociated retina, Dev. Brain Res., 51 (1990) 69-84. Turner, D.L. and Cepko, C.L., A common progenitor for neurons and glia persists in rat retina late in development, Nature, 328 (1987) 131-136. Turner, D.L., Snyder, E.Y. and Cepko, C.L., Lineage-independent determination of cell type in the embryonic mouse retina, Neuron, 4 (1990) 833-845. Watanabe, T. and Raft, M.C., Rod photoreceptor development in vitro: intrinsic properties of proliferating neuroepithelial cells change as development proceeds in the rat retina, Neuron, 4 (1990) 461-467. Watanabe, T. and Raft, M.C., Diffusible rod-promoting signals in the developing rat retina, Development, 114 (1992) 899-906. Wetts, R. and Fraser, S.E., Multipotent precursors can give rise to all major cell types of the frog retina, Science, 239 (1988) 1142-1145. Yu, X. and Barnstable, C.J., Characterization and regulation of the protein binding to a cis acting element, RET 1, in the rat opsin promoter, J. Mol. Neurosci., 5 (1994) 259-271. Zhao, S., Thornquist, S.C. and Barnstable, C.J., in vitro transdifferentiation of embryonic rat retinal pigment epithelium to neuronal cells, Brain Res., 677 (1995) 300-310.