Influence of a retinal pigment epithelial cell factor(s) on rat retinal progenitor cells

DEVELOPMENTAL BRAIN RESEARCH

ELSEVIER

Developmental Brain Research 93 (1996) 88-99

Research report

Influence of a retinal pigment epithelial cell factor(s) on rat retinal progenitor cells Harold J. Sheedlo *, James E. Turner Department of Anatomy and Cell Biology, North Texas Eye Research Institute, University of North Texas Health Science Center, 3500 Camp Bowie Boulevard, Fort Worth, TX 76107, USA Accepted 20 December 1995

Abstract Retinal development was studied by explant culture of retinas from embryonic and neonatal rats in response to medium conditioned (CM) by a transformed neonatal rat retinal pigment epithelial (tnrRPE) cell line. Retinal explants from embryonic days 26 and 18 and postnatal day 2 Long-Evans rats were cultured for over 3 weeks on a poly-D,L-ornithine-coated surface in RPE-CM only, 10% serum or a serum-free defined medium. By 2 days in vitro, round cells were seen emerging from both embryonic and neonatal retinal explants grown in tnrRPE-CM. With extended time in culture, these round cells had increased in number and were seen in large confluent clusters adjacent to the explants. After 2 weeks in culture, some of these cells had undergone a morphological differentiation as shown by process formation. Insignificant numbers of these same cells were seen in explant cultures grown in 10% serum or serum-free defined medium. When isolated and subcultured, approx. 80% of the round cells from embryonic and neonatal rat retinal explants were densely immunolabeled for opsin and arrestin, both photoreceptor cell markers and neuron-specific enolase, a marker for mature neurons. Cellular retinaldehyde-binding protein, a Miiller cell marker, immunolabeled approx. 30% of the cells from embryonic and neonatal rat retinal explants. In addition, nestin, an intermediate filament protein found only in neuroepithelial cells, was present in approx. 70% of the embryonic cells, but in only less than 1% of the neonatal cells. Based on this immunocytochemical characterization, these round cells are termed retinal progenitor cells and because of their mitogenic capacity under these in vitro conditions, these cells appear to possess stem cell characteristics. Moreover, in a 3-day bioassay, tnrRPE-CM caused a twofold and greater increase in harvested progenitor cells from both neonatal and embryonic explants, while cell numbers in control and growth factor-supplemented cultures showed no increase above the initial plating density. In these studies, CM from cultures of transformed neonatal rat RPE cells promoted the production, survival, proliferation and maturation of retinal progenitor cells from neonatal and embryonic rat retinal explants. Keywords: Retinal pigment epithelium; Trophic factor; Retinal explant; Progenitor cell; Photoreceptor

I. Introduction In the adult, RPE cells are found adjacent to the neural retina and function primarily in phagocytosis of shed outer segments of photoreceptor cells [14]. In addition, these cells form, in part, the blood/retinal barrier, control transport between the neural retina and choriocapillaris and take part in the phototransduction process [6]. Although a large body of information exists concerning the structural and functional characteristics of RPE cells, little is known about trophic functions via paracrine mechanisms between RPE cells and the underlying photoreceptors or other retinal cells. Trophic factors are believed to play important roles in retinal health and disease and have recently been * Corresponding author. Fax: (1) (817) 735 2610. 0165-3806/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PI1 S 0 1 6 5 - 3 8 0 6 ( 9 6 ) 0 0 0 0 8 - 9

considered for their therapeutic value in a number of sight compromising conditions [5]. The RPE has been shown to express in vitro the message, receptor or synthesize a n d / o r secrete a number of proteins, growth factors and cytokines [10]. Since it has been shown recently that trophic requirements appear early in neuronal lineages, at the time of neurogenesis, it is possible that the RPE play a role in these events by releasing a factor(s) that influences the developmental phenomena [23]. RPE may play a very direct and prominent role in the development and function of the retina as suggested by the fact that the RPE develops prior to other retinal cells and is positioned adjacent to that part of the retina destined to produce photoreceptor cells [8]. It would, therefore, seem possible that RPE may directly influence rod a n d / o r cone differentiation. A number of studies have demonstrated

H.J. Sheedlo,J.E. Turner~DevelopmentalBrain Research 93 (1996) 88-99

that RPE or its conditioned medium is responsible for photoreceptor cell maturation [18], including outer segment development [36,38]. RPE cells also appear to play a role in the proliferation of retinal progenitor cells and the eventual organization of the retinal layers possibly through secretion of diffusible factors [24,41]. In fact, RPE cells have been shown to produce a number of trophic factors including basic fibroblast growth factor (bFGF), insulinlike growth factor (IGF) and platelet-derived growth factor (PDGF) [7,11,25,30,391. In addition, the cytokines interleukin-6 and -8, macrophage colony stimulating factor (MCSF) and transforming growth factor-/3 (TFG-/3) are secreted in vitro by RPE cells [10,19,28]. In previous studies, medium conditioned by neonatal rat RPE cells promoted the survival of photoreceptor cells, a majority of which expressed opsin, in dissociated neonatal retinal cell cultures. If allowed to attach to Miiller cells, these presumptive photoreceptor cells began to express signs of morphological differentiation (i.e., process formation). In contrast, photoreceptor cells in defined medium did not survive under these conditions [16,31,32]. In addition, CM from passaged transformed neonatal rat (tnr) RPE cell cultures affected the upregulation of mRNA for c-fos and NGFI-A, two immediate early genes, in neonatal rat RPE cells [20]. The present study has demonstrated that medium conditioned by cultures of tnrRPE cells caused round cells to emerge from explants of retinas from embryonic and neonatal rats and populate the culture dish surface in large numbers. When isolated, subcultured and characterized, these progenitor cells continued to express predominately differentiated photoreceptor cell markers and survived and proliferated in low density cultures in the presence of tnrRPE-CM, but not in serum, defined medium or growth factors. The harvesting of a large, self-perpetuating population of progenitor cells, with stem cell characteristics, in response to an RPE secreted factor(s), presents many opportunities to study the role of the RPE in retinal development. These progenitor cells may also serve as excellent candidates in replacement therapy through transplantation techniques.

2. Materials and methods 2.1. Animals

The embryonic day 16 and 18 (El6 and El8) and postnatal day 2 (PN2) rats used in this study were progeny of Long-Evans rats purchased from Harlan Sprague Dawley (Indianapolis, IN). The PN2 rats were sacrificed by an overdose of the anesthetic isoflurane (AErrane, Anaquest, Liberty Corner, N J) prior to enucleation, while pregnant rats were sacrificed by CO 2 overdose and the fetuses were decapitated before eye removal. All animals were handled in accordance with NIH and AAALAC guidelines.

89

2.2. RPE cell isolation and culture and conditioned medium production

A dispase method [13,46] was used to isolate RPE cells from eyes of 6-8-day-old Long-Evans rats. Eyes were incubated in 10 units dispase/ml (Collaborative Res., Bedford, MA, 40352) for 30 min, then RPE sheets were removed from the neural retina and dissociated into a cell suspension. The cells were grown to confluence in a medium consisting of Dulbecco's minimum essential medium (DMEM) supplemented with F12 nutrient (Gibco, Grand Island, NY), 10% fetal bovine serum (FBS, Sigma, St. Louis, MO) and kanamycin (10 /xg/ml) (Sigma) and gentamicin (5 /xg/ml) (Sigma). One particular RPE cell culture at second passage became spontaneously transformed, but continued to express RPE cell characteristics, specifically cytokeratins [20,44]. Cultures of this transformed RPE cell line, from passages 2-52, were used in this study. For conditioned medium production, confluent RPE cell cultures were thoroughly rinsed with DMEM/F12, then incubated in defined medium consisting of DMEM/F12, antibiotics, with or without 1% ITS + (40352, Collaborative Res., Bedford, MA) [31,32]. The medium was allowed to condition for 2-3 days. Conditioned medium was collected at the end of this time period, sterile filtered (0.2 /xm) and stored at - 7 0 ° C prior to use. 2.3. Retinal explant preparation and culture

Retinas from El6, El8 and PN2 rats were cut into small pieces (1-2 mm), then seeded onto a polyornithine-coated surface (50 /zg/ml, Sigma, poly-D,L-ornithine, P8638) on glass coverslips or plastic (Falcon). The explants were grown in CM diluted 1:1 with defined medium, defined medium only or defined medium supplemented with 10% FBS for over 3 weeks. The explant cultures were examined and photographed on a daily basis with an Olympus phase contrast microscope [34,35]. 2.4. Retinal cell isolation and culture

Round cells that emerged from explants grown in tnrRPE-CM were isolated following either gentle triturafion or a brief (5-10 min) incubation in 0.25% trypsin. After collection, the cells were separated from explant contamination by allowing the heavier material to settle, then collecting the round cells. Prior to culture, the cells were examined by phase contrast microscopy to determine purity, then counted. Approx. 2000 cells were seeded onto polyornithine-coated coverslips (12 mm diameter) in defined medium and fixed 1-2 h later with 4% paraformaldehyde in 0.01 M phosphate-buffered saline (PBS), pH 7.4. Other cells were grown in the presence of 50% tnrRPE-CM, 10% serum, defined medium or growth factors - epidermal (EGF), macrophage colony stimulating

90

H.J. Sheedlo, J.E. Turner/Developmental Brain Research 93 (1996) 88-99

( M C S F ) , n e r v e ( N G F ) , p l a t e l e t - d e r i v e d ( P D G F ) , b a s i c fib r o b l a s t ( b F G F ) , t r a n s f o r m i n g (TGF-ce or T G F - / 3 1 ) at 1 - 2 5 0 n g in a 4 0 0 /xl total v o l u m e a n d f i x e d 7 d a y s later.

Cells in the 7 - d a y c u l t u r e g r o u p s w e r e c o u n t e d w i t h a p h a s e c o n t r a s t m i c r o s c o p e u s i n g a 10 X o b j e c t i v e f o l l o w ing a p r e v i o u s l y d e s c r i b e d p r o c e d u r e [32].

Fig. 1. Progenitor cells from embryonic days 16 and 18 (El6 and El8) and postnatal day 2 (PN2) rat retinal explants. A: after 2 days, a few round progenitor cells (arrowheads) were located in close proximity to the explants (Exp) from El6 rats. B: after 6 days, an increased number of El6 progenitor cells was observed, when compared to the earlier culture periods (A). These cells appeared to form large clumps (arrowheads) outside the explant (Exp). C: explants (Exp) from El6 rats grown in defined medium for 6 days exhibited no progenitor cell production or visible signs of cellular outgrowth. The extremely small cells (arrowheads) were broken free from the explant at the time of seeding. D: after 13 days in vitro, a large number of El6 progenitor cells (arrowheads) were observed, especially adjacent to these explants (Exp) where they were still densely packed. E: a large number of progenitor cells were observed in El8 retinal explant (Exp) cultures after 7 days in vitro. Note also the processes that extended from the explant periphery (arrows). F: by 9 days, progenitor cells were also found in large clusters (arrowheads) in explant cultures (Exp) from PN2 rat retinas. Bar = 100 /xm.

H.J. Sheedlo, J.E. Turner~Developmental Brain Research 93 (1996) 88-99

2.5. Retinal progenitor cell proliferation bioassay and data analysis Progenitor cells isolated to purity after migrating from PN2 explants were diluted to 100 or 200 c e l l s / # l in ITS + defined medium. The wells of a 96-well microtiter plate were precoated with polyornithine and cells were plated at 2000 cells/well. CM from tnrRPE cells was added to constitute 10%, 25%, 50% or 75% of the total 100 /~1 volume, with the remainder of the volume consisting of ITS ÷ defined medium. The defined medium for RPE cell conditioning in this study was DMEM + antibiotics (DM). The proliferation controls were 10-75% DM added to 2000 cells, while the assay reagent controls were 10-75% DM without cells. Each of the controls consisted of a 100 /~1 total volume. Triplicate wells were used for each individual experimental and control condition. MCSF, NGF or EGF in at least two different concentrations ranging from 1 to 200 ng were also added to triplicate wells in a 100 /zl total volume. The cells were allowed to incubate for 3 days in 5% CO 2 at 37°C. Progenitor cells from El8 explants were plated in 96-well plates (2000/well) and grown in tnrRPE-CM or DM, then assayed 3 days later. These cells were also plated on coverslips, grown in tnrRPE-CM and counted 7 days later. A non-radioactive, colorimetric, cell proliferation assay (G5421, Promega, Madison, WI) was used to determine the total number of surviving cells following the manufacturer's protocol in 96-well plates. This assay is based on the conversion of a tetrazolium salt into formazan by enzymes of mitochondria in living cells. For this assay, 50 /zl of phenazine methosnlfate (PMS) was mixed with 1000 /xl of a tetrazolium salt (MTS) and 20 /xl of this solution was added to each well. After a 1 h incubation, the color development was read at 490 nm with a BioRad model 450 microtiter plate reader. The absorbance is directly proportional to the number of living cells. To determine cell number, a standard curve was generated at the day of plating using 500, 1000, 2000, 3000, 4000, 5000 and 6000 cells/well in triplicate in a 100 /xl volume of ITS + defined medium. After a 1-2 h incubation, 20 /zl of the PMS + MTS mixture was added and the absorbance was read at 490 nm as described above. The data were plotted using the Cricket Graph 1.3 software on a Macintosh computer. The total number of surviving cells was determined using the equation generated by the simple curve fit. The following equation was generated from one standard curve in this study: y = 484.69 + 2.7079x4; R 2 = 0.990; y = number of surviving cells; x = CM a b s o r b a n c e - control absorbance. The StatView SE + Graphics program was used to analyze all data, specifically using the two-way analysis of variance (ANOVA). Differences within each experimental group and when these individual groups were compared to control groups were assessed by Fisher PLSD multiple-range tests. The level of statistical signficance was defined as

91

P < 0.05. The colorimetric assay results were confirmed by visual counts from coverslips grown in tnrRPE-CM under similar conditions.

Fig. 2. Progenitor cells from PN2 rat retinal explants grown in tnrRPE-CM for 12-25 days. A: those cells near the retinal explant (Exp), after 12 days in vitro, were primarily round, although a few process-bearing cells (arrowheads) were detected among the round cells. These explants were grown in seventh-passaged tnrRPE-CM. B: after 25 days in vitro in third-passaged tnrRPE-CM, the progenitor cells emerging from explants exhibited multiple morphological conformations. Some of the cells remained round (arrowheads), while others had a single process (arrows). C: bipolar (arrowhead) and multi-polar (arrows) cells from retinal explants grown in third-passaged tnrRPE-CM were clearly evident after 25 days in vitro. Bar = 100 p~m.

92

H.J. Sheedlo, J.E. Turner/Developmental Brain Research 93 (1996) 88-99

2.6. Immunocytochemistry For immunocytochemistry retinal explant and progenitor cell cultures were fixed with 4% paraformaldehyde for 10 min. The explants and isolated cells were examined for photoreceptor-specific markers - opsin [26] and arrestin (L.A. Donoso, Philadelphia, PA), a mature neuron marker - neuron-specific enolase (NSE, Sigma), an early neuroepithelial cell marker - nestin [17], and a glia cell marker CRALBP (J. Saari, Seattle, WA) following a previously described procedure [31-33]. Briefly, after treatment with 10% normal rabbit or goat serum (Sigma), explant and progenitor cell cultures were incubated with antibodies against opsin, arrestin, NSE, nestin or CRALBP diluted 1:30 to 1:200 in 1% normal serum overnight at 4°C. Following rinses with PBS, the cultures were incubated with goat antimouse IgG-, goat antirabbit IgG- or rabbit antisheep IgG-fluorescein isothiocyanate (FITC) conjugate (Vector Laboratories, Burlingame, CA) diluted 1:200 for 1

h at room temperature. For controls, cultures were incubated in 1% normal serum, but otherwise identically treated. After mounting in VectaShield mounting medium (Vector Laboratories), the cultures were examined with a Nikon microphot-FXA light microscope using epifluorescent optics. The number of progenitor cells on coverslips positively labeled for each of the retinal cell markers listed above was counted in 20-30 random areas of the coverslip.

3. Results

Retinal explants from El6 rats cultured in tnrRPE-CM exhibited round cells at their periphery after 2 days in vitro (Fig. 1A). The number of these cells in El6 explant cultures increased after 6 days in vitro (Fig. 1B), while those explants grown for 6 days in defined medium appeared to be degenerating, with no evidence of round cell

Fig. 3. Microscopic examination of subcultured retinal progenitor cells isolated from PN2 rat retinal explants. A: the retinal progenitor cells appeared to form clusters (arrowheads) when cultured at high density in tnrRPE-CM for 6 days. This phenomenon would be indicative of cell proliferation. B: cells plated at the same density as in A, but grown in 10% serum for 5 days, showed only a few cell clusters and a reduced total number of cells. C: serum-free defined medium also did not affect noticeable proliferation or clustering of the progenitor cells. Note, however, many of the ceils at this time period appeared to form processes (arrowheads). D: PN2 progenitor cells were plated at low density in 12th-passage tnrRPE-CM for 11 days. During this time period these progenitor cells became aligned in parallel rows. Note that many of the cells were bipolar (arrowheads) at this stage and the processes appeared to position these cells in rows. Bar = 100 /xm.

H.J. Sheedlo, J.E. Turner~Developmental Brain Research 93 (1996) 88-99

production (Fig. 1C). At approx. 2 weeks in vitro, these cells appeared in a semi-confluent state at the periphery of El6 explants (Fig. 1D). Retinal explants from El8 rats in tnrRPE-CM had numerous round cells at their periphery after 7 days in vitro (Fig. 1E). By 9 days, in retinal explant cultures from PN2 rats, the round cells had also reached a semi-confluent state at the periphery of the explants in a manner similar to that seen for embryonic explant cultures (Fig. 1F). After several weeks in culture, the round cells which emerged from the PN2 rat retinal explants had undergone some morphological differentiation. For example, after 12 (Fig. 2A) and 25 (Fig. 2B) days in vitro, many processbearing cells were seen among the round cells. After over 3 weeks in culture, some of the cells became bipolar, while other cells had three or more processes (Fig. 2C). The round cells were successfully isolated to purity from PN2 rat retinal explant cultures, subcultured, then grown for 5 days at high density in 50% tnrRPE-CM, 10%

93

serum or serum-free defined medium. The cells grown in tnrRPE-CM under these conditions appeared in greater numbers and clustered into small groups (Fig. 3A), a phenomenon which was observed to a lesser degree when these cells were cultured in serum (Fig. 3B) or not at all in a serum-flee defined medium (Fig. 3C). However, progenitor cells cultured in defined medium for 5 days, after a previous exposure to tnrRPE-CM, expressed a greater process formation than CM- or serum-treated cells (compare Fig. 3C to Fig. 3A,B). The cell clustering in tnrRPECM could be indicative of cell division, which was subsequently verified in a cell proliferation assays shown in Fig. 8 or a facilitation of cell aggregation. PN2 retinal cells cultured in tnrRPE-CM for 11 days were observed to exhibit a bipolar morphology and some were aligned in parallel rows (Fig. 3D). The round cells derived from the PN2 rat retinal explants grown in tnrRPE-CM were isolated, then briefly subcultured, densely immunostained for two photorecep-

Fig. 4. Immunocytochemical characterization of isolated and subcultured progenitor cells from PN2 retinal explants. A: most of the subcultured progenitor cells (arrowheads) from PN2 rat retinal explants immunostained for the photopigment opsin. Note that one of these cells had a single-immunostained process (arrowhead). B: arrestin also immunolabeled most of these progenitor cells. Note that most of the immunostained cells were round, although an elongated cell showed light immunoreactivity (arrowhead). C: neuron-specific enolase, a cytoplasmic enzyme, lightly immunostained many of the progenitor cells (arrowheads). D: a few of the progenitor cells were also immunostained with antibodies against cellular retinaldehyde binding protein (arrowheads), a typically used Miiller cell marker. Note that most of the cells in this field exhibited only background staining (arrows). Bar = 50 ~ m (A,C); 100 /zm (B,D).

H.J. Sheedlo, J.E. Turner/Developmental Brain Research 93 (1996) 88-99

94

tor-specific markers opsin (Fig. 4A) and arrestin (Fig. 4B). In addition, these cells were also immunoreactive with antibodies against neuron-specific enolase, a mature neuron marker (Fig. 4C). Furthermore, some of the subcultured retinal cells immunolabeled for a Miiller cell marker - CRALBP (Fig. 4D). Therefore, these round cells will subsequently be referred to as retinal progenitor cells. Approx. 80% of the progenitor cells isolated from PN2 retinal explants were immunostained for opsin, arrestin and neuron-specific enolase. In addition, 20-30% of these cells showed immunoreactivity for the Miiller cell marker, CRALBP. However, less than 1% of the cells labeled for the early neuroepithelial marker nestin (Fig. 5). Thus, these progenitor cells expressed primarily a mature photoreceptor phenotype. Cells that exhibited a round morphology at the periphery and within retinal explants of El8 rats were also densely immunolabeled with the photopigment opsin (Fig. 6A). In addition, cells that had undergone a morphological differentiation distant from these retinal explants exhibited opsin immunoreactivity (Fig. 6B), while other cells were immunostained with arrestin antibodies (Fig. 6C). Retinal explants from El6 rats showed cells and processes which were densely immunoreactive for the intermediate filament protein nestin (Fig. 6D). Nestin-immunolabeled processes were also observed at the periphery of El6 retinal explants (Fig. 6E), while cells which had migrated from these explants were also stained with nestin (Fig. 6F). Progenitor cells were also isolated from El6 and El8 retinal explants grown in tnrRPE-CM, dissociated and allowed to attach to coverslips prior to analysis. These cells were then examined immunocytochemically for retinal cell markers. Many of the El6 progenitor cells were

1D • 100 J= _m

_E o

densely immunoreactive for the photopigment opsin (Fig. 7A). El8 progenitor cells were immunolabeled with NSE, a mature neuron marker (Fig. 7B), while others were immunostained with CRALBP, a Miiller cell marker (Fig. 7C). When expressed as a percentage of total progenitor cells isolated from El6 explants, the approximate retinal cell marker distribution was as follows: opsin (81%); arrestin (76%); neuron-specific enolase (79%); CRALBP (18%); and nestin (67%). A sensitive bioassay method was used to determine the survival and mitogenic activity of a tnrRPE-CM factor(s) or known growth factors directed toward progenitor cells isolated from PN2 and El8 rat retinal explants. The bioassay results revealed that 10-75% concentrations of the tnrRPE-CM caused a twofold or greater increase in PN2 progenitor cells after only 3 days. However, two-way ANOVA showed no significant difference in cell number when contrasting each concentration of the tnrRPE-CM. In contrast, growth factors EGF, NGF and MCSF did not promote cell proliferation of PN2 progenitor cells after 3 days in vitro. Only limited cell survival at no greater than 50% of the original plating density was observed in these growth factor-supplemented cultures in contrast to a 200300% increase over this plating density in tnrRPE-CMtreated cultures (Fig. 8). Furthermore, progenitor cells from El8 retinal explants increased three- and sevenfold in response to 50% tnrRPE-CM after 3 and 7 days, respectively, while cells in the control conditions numbered less than 50% of the original plating density (Fig. 9). Factor(s) secreted by tnrRPE-CM also caused significant proliferation of El6 progenitor cells after 7 days in vitro. However, known growth factors, such as EGF, PDGF, MCSF, bFGF, NGF, TGF-/31 and TGF-a, were only able to provide minimal cell survival at no greater than 20% of the original plating density. These 7-day in vitro results were obtained by actual cell counts performed on fixed cells on coverslips and thus confirm the 96-well enzyme bioassay. In addition, El6 progenitor cells, which increased 160% above plating density in response to tnrRPE-CM, maintained a neuron-to-glial distribution of 90% to 10%, respectively, similar to what was shown for PN2 progenitor cells described in Fig. 5 (Fig. 10).

L

4. Discussion

°

Opsin

Arrestin

NSE

CRALBP

Nestin

Fig. 5. PN2 rat progenitor cells expressing neuronal or glial markers. The progenitor cells were isolated after emerging from PN2 rat retinal explant cultures grown in tnrRPE-CM. The cells were plated onto polyornithine for 1-2 h, fixed, then immunostained for photoreceptor or Miiller cell markers. Over 80% of these progenitor cells immunolabeled for photoreceptor markers opsin and arrestin and neuron-specific enolase, a mature neuron enzyme. However, only about 30% of these cells immunostained for the Miiller cell marker, CRALBP, and less than 1% for nestin.

Within the nervous system, neurotrophic factors participate in proliferation of neuroepithelial stem cells, cell lineage sorting and neuron maturation [2,4,12,15,29]. These factors may stimulate some cells to undergo mitosis or promote survival a n d / o r differentiation of those cells not capable of division [27]. Growth factors or cytokines may play a role in controlling the pathways of development as they do in regulating cell division and activities in adult tissues. Several factors have been shown to stimulate cellular events in vitro, but only at late developmental

H.J. Sheedlo, J.E. Turner/Developmental Brain Research 93 (1996) 88-99

periods [23]. To date, very few retinal mitogenic- or fate-inducing factors have been identified, although a developmentally regulated, diffusible factor was proposed to cause retinal progenitor cells to differentiate into photore-

95

ceptor cells in culture [1,43]. For example, bFGF, taurine, retinoic acid and S-laminin were shown to influence photoreceptor cell differentiation in low density cultures, which suggests that these factors may perform such functions in

Fig. 6. Immunocytochemical characterization of retinal explants from E l 6 and El8 rats cultured in tnrRPE-CM. A: opsin immunolabeled cells (arrowheads) were observed as components and at the periphery of explants (E) from El8 rats. B: some of the cells that migrated from the El8 rat retinal explants became bipolar and immunostained for the photopigment opsin. C: other cells from El8 rat retinal explants were densely immunoreactive for a photoreceptor~specific protein, arrestin. D: cells (arrowheads) and processes (arrows) which projected from El6 rat retinal explants (E) were densely immunostained for the intermediate filament protein nestin, which is an early neuroepithelial cell marker. E: nestin-immunostained process (arrowheads) rimmed the periphery of E l 6 rat retinal explants (E). F: nestin immunolabeled cells were also observed distant from the El6 rat retinal explant. Note that many of the immunolabeled cells were process bearing (arrowheads). Bar = 100 /zm (A,E,F); 50 /zm (B-D).

96

H.J. SheedIo, .I.E. Turner / Deuelopmental Brain Research 93 (1996) 88-99 8000 7000 8ooo

50OO .Q

E = z

4000

•

300o

o

2000 1000 o o

10 CM

25

50

75

EGF

Concentration/Growth

MCSF

NGF

Factor

Fig. 8. The proliferation of PN2 rat progenitor cells in response to a factor(s) secreted by tnrRPE cells and growth factors. Progenitor cells were isolated from PN2 rat retinal explants, plated at 2000 cells/well in triplicate and cell numbers were measured using a cell proliferation assay. Cells were grown in tnrRPE-CM or EGF, MCSF and NGF at 1-200 ng for 3 days and the total number of cells was counted to determine the mitogenic response for each condition. These results were taken from seven separate experiments performed in triplicate with conditioned medium from tnrRPE cell cultures at passages 2, 3, 4 and 12. The bars represent the mean value + standard error (S.E.M.). A statistically significant increase in cell number was determined when comparing each of the tnrRPE-CM values with the control (0% CM) conditions ( P < 0.05). In addition, cell numbers in RPE-CM cultures were significantly greater than those cultures grown in EGF, MCSF and NGF ( P < 0.05).

normal rats. In a previous report, these retinal progenitor cells were not found in neonatal rat retinal explant cultures supplemented separately with seven known growth factors, 20000

•

18000

[]

[]

DM (3 day enzyme bioassay) tnrRPE-CM (3 day enzyme bioassay) tnrRPE-CM (7 day cell counts)

--f-

16000 14000

~

12000

z

10000

w

Fig. 7. Immunocytochemical characterization of subcultured El6 or El8 rat retinal progenitor ceils. A: isolated and subcultured progenitor cells from the El6 rat retinal explants expressed the photopigment opsin. Note that one cell had a short process (arrow). B: most of the El8 progenitor cells (arrowheads) immunolabeled for the mature neuron marker, neuron-specific enolase, as shown here at low magnification to reveal the total number of plated cells. C: a few of the progenitor ceils from El8 rat retinal explants were also immunoreactive for the Mtiller cell marker, CRALBP. Note that one cell had remained round (arrow), while the other cell formed two short flattened processes (arrowheads). Bar = 50 /.~m (A,C); 100 /zm (B).

vivo [1,3]. Our results reported here suggest that neonatal rat RPE cells secrete a factor(s) in vitro that also promotes the survival, proliferation and maturation of progenitor cells from retinal explants of neonatal and embryonic

8000

f r i l l

6000

[//ZZ/ ,.<<<<~

4000

V//'.©

2000"

t ! ! ! ! ~ I J l f { I 1 1 1 1

E18

Retinal

Cells

Fig. 9. The proliferation of El8 rat progenitor cells in response to a factor(s) secreted by mrRPE cells. Retinal progenitor ceils from El8 rat explants were plated at 2000 cells and grown in defined medium or 50% tnrRPE-CM for 3 days in a 96-well enzyme bioassay system or on coverslips and grown in 50% tnrRPE-CM for 7 days. The bars represent the mean values _+standard error (S.E.M.). This experiment was performed in triplicate with three different conditioned media. A statistically significant increase in cell number was detected in cultures supplemented with tnrRPE-CM at 3 days when compared to cultures in defined medium at the same time period (P < 0.05).

H.J. Sheedlo, J.E. Turner / DeL,elopmental Brain Research 93 (1996) 88-99 2500

2000 93.2% Opsin ~) .Q E

1500

m CRALBP

10.0"/.

Positive Positive

z

0

1000

500-

0

~ I--

~

LU

Z

O

iFig. 10. The proliferation of El6 rat progenitor cells in response to mrRPE-CM and growth factors. Retinal progenitor cells were isolated from El6 rat explants, then 1250 cells were cultured onto coverslips for 7 days in second-passaged tnrRPE-CM at a 50% concentration or growth factors at 20-250 ng in a total 4 0 0 / z l volume. The total number of cells in each condition was determined by actual counts and the bars represent the mean values _+standard error (S.E.M.). Note that only tnrRPE-CM promoted proliferation of these progenitor cells, while the growth factors only affected less than 20% cell survival. A statistically significant difference was determined when comparing tnrRPE-CM values with control and growth factor cell numbers ( P < 0.05). In addition, over 90% of the progenitor cells grown in the tnrRPE-CM were immunoreactive for opsin, while only 10% of the cells were immunolabeled for CRALBP.

including bFGF and EGF, or in explant cultures exposed to trypsin- or heat-treated tnrRPE-CM [34]. These latter results suggest that the component(s) of tnrRPE-CM that promoted the explant and cellular responses may not be a known growth factor, but still has a protein moiety. Retinal cell fates are most likely not determined until the last mitotic cycle which was revealed by lineage tracing experiments through the use of trophic factors. In vivo labeling of single progenitor cells showed that these cells can generate diverse cell types. Thus, late stage progenitor cells were able to give rise to a combination of either photoreceptors, bipolar, amacrine a n d / o r Miiller cells at or near their final division cycle [1]. In our study, a factor(s) secreted by neonatal RPE cells was shown to stimulate retinal progenitor cells to express primarily a photoreceptor cell phenotype. Specifically, 70-80% of these progenitor cells, harvested after emerging from either embryonic or PN2 retinal explants, demonstrated opsin and arrestin photoreceptor cell markers. This is approximately the same percentage of photoreceptor cells estimated to be present in the adult retina [45]. Approx. 30% of these cells labeled for the Miiller cell marker CRALBP, but these cells may also coexpress alternate late developing retinal cell markers as has been demonstrated for multipotential cells in other areas of the central nervous system [22]. A coexpression of cell markers may also occur within the

97

harvested embryonic progenitor cells when considering that almost 70% of these cells also labeled for nestin, an immature neuroepithelial cell marker, although 70-80% of these cells also expressed mature neuron markers. In turn, few of the neonatal progenitor cells expressed nestin, which indicated a higher level of maturation. Thus, a factor(s) in tnrRPE-CM possibly directs progenitor cells toward their full maturation potential as photoreceptors, as well as other retinal cell types. These results suggest that retinal cell fate is mainly independent of lineage but possibly determined by environmental factors [1]. Factors of this description may be crucial to control retinal differentiation and establishment of synaptic contacts. Recent studies demonstrated that El5 rat retinal progenitor cells were more capable of cell division than those at PN1 and that EGF, TGF-a and FGF were active mitogens for these cells [4,23]. Such mitotic signals could be produced by the RPE which directly apposes the ventricular or mitotic surface of the retina [8]. Indeed, RPE cells have been reported to produce a variety of growth factors [10] and could, therefore, provide such mitotic and differentiation signals or stimulate retinal production of such factors. In confirmation of this hypothesis, RPE-CM has been shown to stimulate cell division of RPE cells, astrocytes and fibroblasts [9]. These investigations inferring possible RPE mitotic factors were corroborated by studies showing a more direct RPE effect on retinal mitogenesis and organization. For example, the use of chick reaggregate and strip retinal cultures showed the importance of RPE in establishing the polarity, stratification and cell division within the chick retina. Specifically, retinas cultured in the presence of RPE demonstrated a significant increase in the number of dividing cells particularly in the area of the presumptive photoreceptor cell layer adjacent to the RPE. Under these conditions, retinas with associated RPE cells were signficantly thicker and better stratified than those cultured without RPE [24,41,42]. Furthermore, removal of the retina from the rat eye and its RPE caused a twofold decrease in mitotic activity in explant culture [23]. This evidence is in agreement with data from this study that a mitogenic factor is produced by RPE cells that sustains such activity in a progenitor cell population within the neuroepithelial layer of the developing retina. Although several known growth factors have been found to stimulate retinal progenitor cell proliferation [4,21,23], a factor(s) in tnrRPE-CM was found to be a much more potent mitogen. Specifically, EGF, NGF, MCSF, PDGF, TGF-a, TGF-/31 and bFGF could only partially sustain a portion of the original cell number by 10-50% compared to 150-300% for tnrRPE-CM. The fact that progenitor cells were continually self-perpetuating in the presence of tnrRPE-CM may mean that an RPE factor(s) is capable of indefinitely maintaining these cells in a stem cell condition or has transformed them. The above studies suggest that RPE cell trophic factors may be essential for normal photoreceptor cell mitosis and

98

H.J. Sheedlo, ,I.E. Turner/Developmental Brain Research 93 (1996) 88-99

differentiation. However, the definitive role that RPE cells play through trophic elements in retinal cell development has yet to be defined. Significantly, a pigment epithelialderived factor has been recently isolated from fetal human RPE conditioned medium which caused cytodifferentiation of retinoblastoma cells [37,40]. These and future investigations may provide a category of novel molecules that affect mitogenic and fate determinations of many retinal cell types.

Acknowledgements The authors wish to thank Ann-Marie Brun-Zinkernagel for her assistance with microscopic preparation and analysis and Rich Dear for his help in tissue culture and cell analysis. This investigation was supported by NIH grant EY04337. Portions of this work were presented at the Association for Research in Vision and Ophthalmology meeting, Fort Lauderdale, FL, May 1995 and the Ocular and Molecular Biology Symposium II, San Diego, CA, August 1995.

References [1] Altshuler, D. and Cepko, C., A temporally regulated, diffusible activity is required for rod photoreceptor development in vitro, Development, 114 (1992) 947-952. [2] Altshuler, D.M., Turner, D.L. and Cepko, C.L., Specification of cell type in the vertebrate retina, In D.M.K. Lam and C.J. Shatz (Eds.), Development of the Visual System, MIT Press, Cambridge, MA, 1991, pp. 37-58. [3] Altshuler, D., LoTurco, J.J., Rush, J. and Cepko, C., Taurine promotes the differentiation of a vertebrate retinal cell type in vitro, Development, 119 (1993) 1317-1328. [4] Anchan, R.M., Reh, T.A., Angello, J., Balliet, A. and Walker, M., EGF and TFG-ct stimulate retinal neuroepithelial cell proliferation in vitro, Neuron, 6 (1991) 923-936. [5] Barinaga, M., Neurotrophic factors enter the clinic, Science, 264 (1994) 772-774. [6] Bok, D., Retinal photoreceptor-pigment epithelium interactions: Friedenwald Lecture, Invest. Ophthalmol. Vis. Sci., 12 (1995) 1659-1694. [7] Bost, L.M., Oataki-Keen, A.E. and Hjelmeland, L.M., Coexpression of FGF-5 and bFGF by the pigment epithelium in vitro, Exp. Eye Res., 55 (1992) 727-734. [8] Braekevelt, C.R. and Hollenberg, M.J., The development of the retina of the albino rat, Am. J. Anat., 127 (1970) 281-302. [9] Bryan, J. and Campochiaro, P.A., A retinal pigment epithelial cell derived growth factor(s), Arch. Ophthalmol., 104 (1986) 422-425. [10] Campochiaro, P.A., Cytokine production by retinal pigmented epithelial cells, Int. Rev. Cytol., 146 (1993) 75-82. [11] Campochiaro, P.A., Sugg, R., Grotendorst, G. and Hjelmeland, L.M., Retinal pigment epithelial ceils produced PDGF-like proteins and secrete them into their media, Exp. Eye Res., 49 (1988) 217-227. [12] Cattaneo, E. and McKay, R., Proliferation and differentiation of neuronal stem cells regulated by nerve growth factor, Nature, 347 (1990) 762-765. [13] Chang, C.W., Roque, R.S., DeFoe, D.M. and Caldwell, R.B., An improved method for isolation and culture of pigment epithelial cells from rat retina, Curr. Eye Res., 10 (1991) 1081-1086.

[14] Clark, V.M., The cell biology of the retinal pigment epithelium. In R. Adler and D. Farber (Eds.), The Retina, Academic Press, Orlando, FL, 1986, pp. 129-169. [15] Fredericksen, K. and McKay, R.D.G., Proliferation and differentiation of rat neuroepithelial precursor cells in vivo, J. Neurosci., 8 (1988) 1144-1151. [16] Gaur, V., Liu, H. and Turner, J.E., RPE conditioned medium stimulates photoreceptor cell survival, neurite outgrowth and differentiation in vitro, Exp. Eye Res., 54 (1992) 645-659. [17] Hockfield, S. and McKay, R.D.G., Identification of major cell classes in the developing mammalian nervous system, J. Neurosci., 5 (1985) 3310-3328. [18] Hollyfield, J.G. and Witkovsky, P., Pigmented retinal epithelium involvement in photoreceptor development and function, J. Exp. Zool., 189 (1974) 357-378. [19] Jaffe, G.J., Peters, W.P., Roberts, W., Kurtzberg, J., Stuart, A., Wany, A.M. and Stoudemire, J.B., Modulation of macrophage colony stimulating factor in cultured human retinal pigment epithelial cells, Exp. Eye Res., 54 (1992) 595-603. [20] Jaynes, C.D., Sheedlo, H.J., Agarwal, N., O'Rourke, K. and Turner, J.E., Miiller and retinal pigment epithelial (RPE) cell expression of NGFI-A and c-fos mRNA in response to medium conditioned by the RPE, Mol. Brain Res., 32 (1995) 329-337, [21] Kelley, M.W., Turner, J.K. and Reh, T.A., Regulation of proliferation and photoreceptor differentiation in fetal human retinal cell cultures, Invest. Ophthalmol. Vis. Sci., 36 (1995) 1280-1289. [22] Kentroti, S. and Vernadakis, A., Early neuroblasts are pluripotential: colocalization of neurotransmitters and neuropeptides, J. Neurosci. Res., 41 (1995) 696-707. [23] Lillian, L.E. and Cepko, C.L., Control of proliferation in the retina: temporal changes in responsiveness to FGF and TGF-a, Development, 115 (1992) 253-266. [24] Liu, L., Cheng, S.H., Jiang, L.Z., Hansmann, G. and Layer, P.G., The pigmented epithelium sustains cell growth and tissue differentiation of chicken retinal explants in vitro, Exp. Eye Res., 46 (1988) 801-812. [25] Martin, D.M., Yee, D. and Feldman, E.L., Gene expression of the insulin-like growth factor and their receptors in cultured retinal pigment epithelial cells, Mol. Brain Res., 12 (1992) 181-186. [26] Papermaster, D.S., Preparation of antibodies to rhodopsin and the large protein of rod outer segments, Methods Enzymol., 8l (1982) 240-246. [27] Pincus, D.W., Dicicco-Bloom, E. and Black, I., Trophic mechanisms regulate mitotic neuronal precursors: role of vasoactive intestinal peptide (VIP), Brain Res., 663 (1994) 51-60. [28] Planck, S.R., Dang, T.T., Graves, D., Tara, D., Anset, J.C. and Rosenbaum, J.T., Retinal pigment epithelial cells secreted interleukin-6 in response to interleukin-l, Invest. Ophthalmol. Vis. Sci., 33 (1992) 78-82. [29] Reynolds, B.A., Tetzlaff, W. and Weiss, S., A multipotent EGF-responsive striatal embryonic progenitor cell produces neurons and astrocytes, J. Neurosci., 12 (1992) 4565-4574. [30] Schweigerer, L., Malerstein, B., Neufeld, G. and Gospodarowicz, D., Basic fibroblast growth factor is synthesized in cultured retinal pigment epithelial cells, Biochem. Biophys. Res. Commun., 143 (1987) 934-940. [31] Sheedlo, H.J., Li, L. and Turner, J.E., Effects of RPE cell factors secreted from permselective fibers on photoreceptor cell survival in vitro, Brain Res., 587 (1992) 327-337. [32] Sheedlo, H.J., Li, L., Fan, W. and Turner, J.E., Retinal pigment epithelial cell support of photoreceptor survival in vitro, In Vitro Cell. Dev. Biol., 31 (1995) 330-333. [33] Sheedlo, H.J., Jaynes, D., Bolan, A.L. and Turner, J.E., Miillerian glia in dystrophic rodent retinas: an immunocytochemical analysis, Dev. Brain Res., 85 (1995) 171-180. [34] Sheedlo, H.J. and Turner, J.E., Retinal progenitor cell proliferation stimulated from rat retinal explants by media conditioned by rat RPE. Invest. Ophthalmol. Vis. Sci., 36 (1995) $58.

H.J. Sheedlo, J.E. Turner~Developmental Brain Research 93 (1996) 88-99

[35] Sheedlo, H.J. and Turner, J.E., Photoreceptor survival and development in culture. In N.N. Osborne and G.J. Chader (Eds.), Progress in Retinal and Eye Research, Vol. 15, Ch. 16, Elsevier, Amsterdam, 1996, pp. 127-137. [36] Spoerri, P.E., Ulshafer, R.J., Ludwig, H.C., Allen, C.B. and Kelley, K.C., Photoreceptor cell development in vitro: influence of pigment epithelium conditioned medium on outer segment differentiation, Eur. Z Cell. Biol., 46 (1988) 362-367. [37] Steele, F.R., Chader, G.J., Johnson, L.V, and Tombran-Tink, J., Pigment epithelium-derived factor: neurotrophic activity and identification as a member of the serine protease inhibitor gene family, Proc. Natl. Acad. Sci. USA, 90 (1992) 1526-1530. [38] Stiemke, M.M., Landers, R.A., A1-Ubaidi, M.R. and Hollyfield, J.G., Photoreceptor outer segment development in Xenopus laevis: influence of the pigment epithelium, Dev. Biol., 162 (1994) 169-180. [39] Takagi, H., Yoshimura, N., Tanihara, H. and Honda, Y., Insulin-like growth factor-related genes, receptors and binding proteins in cultured human retinal pigment epithelial cells, Invest. Ophthalmol. Vis. Sci., 35 (1994) 916-923. [40] Tombran-Tink, J. and Johnson, L.V., Neuronal differentiation of

[41]

[42]

[43] [44]

[45] [46]

99

retinoblastoma cells induced by medium conditioned by human RPE cells, Invest. Ophthalmol. Vis. Sci., 30 (1989) 1700-1707. Vollmer, G. and Layer, P.G., An in vitro model of proliferation and differentiation of the chick retina: coaggregates of retinal and pigment epithelial cells, J. Neurosci., 6 (1986) 1885-1896. Vollmer, G., Layer, P.G. and Gierer, A., Reaggregation of embryonic chick retina cells: pigment epithelial cells induce a high order of stratification, Neurosci. Lett., 48 (1984) 191-196. Watanabe, T. and Raff, M.C., Diffusible rod-promoting signals in the developing rat retina, Development, 114 (1992) 899-906. Wordinger, R.J., Sheedlo, H.J., Fan, W., Agarwal, N. and Turner, J.E., Fibroblast growth factor (FGF) expression is upregulated by immortalized neonatal rat RPE cells, Invest. Ophthalmol. Vis. Sci., 36 (1995) $97. Young, R.W., Cell differentiation in the retina of the mouse, Anat. Rec., 212 (1985) 199-205. Young, R.W., Sheedlo, H.J., Li, L. and Turner, J.E., 5-Bromo-2-deoxyuridine labelling of rabbit retinal pigment epithelial cells grown on polycarbonate membrane substrate, Vet. Comp. Ophthalmol., 4 (1994) 71-76.