Molecular Vision 2001; 7:27-35 <http://www.molvis.org/molvis/v7/a5/> Received 26 September 2000 | Accepted 1 February 2001 | Published 13 February 2001

Download Reprint

Lactose supports Müller cell protein expression patterns in the absence of the retinal pigment epithelium

Monica M. Jablonski, Alessandro Iannaccone
 
 

Retinal Degeneration Research Center, Department of Ophthalmology, The University of Tennessee Health Science Center at Memphis, Memphis, TN

Correspondence to: Monica M. Jablonski, Ph.D., The University of Tennessee Health Science Center at Memphis, Retinal Degeneration Research Center, Department of Ophthalmology, 956 Court Avenue, Memphis, TN, 38163; Phone: (901) 448-7572; FAX: (901) 448-7193; email: mjablonski@mail.eye.utmem.edu

Abstract

Purpose: We have previously shown that lactose, but not mannose, promotes the assembly of nascent photoreceptor outer segments in the absence of the retinal pigment epithelium (RPE). The purpose of the present study was to determine if, in addition to the improved outer segment assembly observed in the presence of lactose, biosynthetic changes in Müller cells could also be detected.

Methods: The RPE was removed from intact isolated Xenopus embryonic eyes that were allowed to complete differentiation in Niu-Twitty medium, Niu-Twitty with mannose, or Niu-Twitty with lactose. Control retinas matured in vitro with an adherent RPE. Retinal morphology was evaluated for organized folding of outer segment membranes and cell loss. In addition, the expression of three Müller cell proteins, glial fibrillary acidic protein (GFAP), cellular retinaldehyde-binding protein (CRALBP), and glutamine synthetase, was examined.

Results: In control retinas, GFAP is undetectable, CRALBP heavily immunolabels Müller cells, and radial patterns of glutamine synthetase immunoreactivity are present. In the absence of the RPE, Müller cells upregulate GFAP expression, CRALBP labeling is present at a slightly reduced level, and glutamine synthetase immunolabeling is negligible. Neither mannose nor lactose modify significantly the expression of CRALBP. Similarly, both compounds completely prevent the upregulation of GFAP. However, normal glutamine synthetase expression was observed only in the presence of lactose, but not in the presence of mannose. Statistical analyses of slot blot-based protein quantification confirmed our immunochemical results.

Conclusions: In RPE-deprived retinas supplemented with lactose, Müller cells were morphologically normal. The proper photoreceptor outer segment morphogenesis observed under these conditions was uniquely associated with normal levels of glutamine synthetase expression. The exact significance of this finding with respect to photoreceptor outer segment morphogenesis is unknown. We suggest that glutamine synthetase may be a marker of Müller cell metabolic or structural integrity that may reflect the enhanced ability of these cells, in the presence of lactose, to support photoreceptor outer segment morphogenesis.

Introduction

The retina is a complex tissue composed of several different neuronal cell types, along with radial glia, the Müller cells. Müller cells and photoreceptors are embryologically and physically coupled. During development, a single retinal progenitor cell gives rise to a columnar arrangement of Müller cells, photoreceptors, and a subset of inner retinal neurons [1]. Within the column of cells, Müller cells surround photoreceptors from the synaptic terminals to the inner segments [2]. Several laboratories also speculate that Müller cells offer metabolic and trophic support to promote photoreceptor survival [1,3,4]. When the retinal pigment epithelium (RPE) is removed physically or carries a genetic mutation, both Müller cells and photoreceptors undergo degenerative or reactive changes. Retinal detachment induces rapid photoreceptor degeneration [5] and an increase in Müller cell expression of various cytoskeletal proteins [6,7], including GFAP [8,9]. In the Royal College of Surgeons (RCS) rat, a mutation in an RPE-expressed gene [10] results in photoreceptor degeneration unless neurotrophic substances are injected into the subretinal space [11] or an RPE transplant is performed [12]. Müller cells also are affected in this retinal degeneration model, with an elevated expression of GFAP and apical process hypertrophy into the subretinal space [13]. Under areas of transplanted RPE, however, less GFAP immunoreactivity is detected and apical process sprouting is inhibited [14], suggesting that Müller cells also are stabilized by transplanted RPE.

We have demonstrated previously that lactose, but not mannose, protects against the photoreceptor dysmorphogenesis induced by RPE removal by supporting the organized assembly of nascent outer segment membranes and normal steady-state expression levels of photoreceptor proteins [15-17]. In addition, we have shown recently that lactose-protected retinas are characterized ultrastructurally by the presence of photoreceptor calycal processes, an organized filamentous actin cytoskeleton within photoreceptors, by the presence of apical processes on Müller cells, and by the presence of a structurally intact adherens junction between the two cell types at the outer limiting membrane. All of these subcellular characteristics are absent in dysmorphic retinas, implicating that these structures may be required for the proper assembly and stability of organized outer segments [18]. We also have documented that targeted disruption of Müller cell metabolism with a-aminoadipic acid or fluorocitrate results in disorganization of photoreceptor outer segments despite normal levels of opsin expression [19]. Cummulatively, these results suggest that that Müller cells interact with photoreceptors through mechanisms that may regulate, at least in part, the assembly of photoreceptor outer segment membranes.

The first purpose of the present study was to determine whether cultured intact eyes from Xenopus embryos have normal steady-state expression patterns of Müller cell proteins when compared to eyes that matured in vivo. Subsequently, we sought to investigate whether selective biosynthetic changes within Müller cells can provide a partial explanation for lactose-mediated protection of photoreceptors in RPE-deprived retinas [15-17]. To address this issue, three Müller cell markers were evaluated: glial fibrillary acidic protein (GFAP), cellular retinaldehyde binding protein (CRALBP), and glutamine synthetase. Each of these Müller cell proteins has been shown to be up- or downregulated in response to retinal injury (i.e., upregulation of GFAP [8,9] and downregulation of glutamine synthetase [20]), the appropriate cell-cell interactions (i.e., glutamine synthetase [21-24]), or to provide a specific metabolic function within Müller cells (i.e, glutamine synthetase [1] and CRALBP [25,26]). Herein we demonstrate that during three days in vitro, Müller cells developed normally in the presence of an adherent RPE, as assessed by morphological, immunocytochemical, and protein criteria. We also show that RPE removal alters profoundly the immunolabeling and steady-state protein expression patterns of Müller cells. Mannose, which does not allow for the development of well organized outer segments [15-17], prevents GFAP upregulation, allows for a slight but significant decrease in the steady-state protein level of CRALBP, and allows for a drastic reduction in the expression of glutamine synthetase. Lactose supports the expression of both GFAP and glutamine synthetase to levels that are comparable to control conditions in which the RPE was attached. It did not, however, prevent the slight reduction in the expression level of CRALBP. These findings suggest a link between Müller cell expression of glutamine synthetase and the ability of Müller cells to support the organized assembly of photoreceptorouter segments.

Methods

Culture of developing retinas

The experimental culture preparation used in these studies has been previously described in detail [15-19,27,28]. The handling of animals was in accordance with the Declaration of Helsinki and The Guiding Principles in the Care and Use of Animals (DHEW Publication NIH 80-23). Adult Xenopus laevis were induced to breed with injections of human chorionic gonadotropin (Sigma Chemical Co., St. Louis, MO). Embryos were staged by external morphological criteria according to Nieuwkoop and Faber [29]. Eyes were removed from embryos at stage 33/34, just as photoreceptor outer segments are beginning to form. At this stage, the retina is neither fully stratified nor are retinal cells fully differentiated [29,30]. Therefore, the final steps of retinal morphogenesis, including outer segment elaboration and assembly, occur in vitro under controlled experimental conditions.

When removed from the embryo, the eyes are not yet surrounded by the sclera, leaving the posterior segment covered only by the RPE layer. Taking advantage of this characteristic, the overlying RPE was gently peeled away from the neuroepithelium, leaving the underlying retina exposed to the culture medium. Intact eye rudiments without an adherent RPE were cultured in Niu-Twitty medium alone [31], Niu-Twitty containing 5 mM mannose (Sigma) or Niu-Twitty containing 5 mM lactose (Eastman Kodak Company, Rochester, NY). Control eyes were allowed to mature in vitro in the presence of an adherent RPE in Niu-Twitty medium alone (i.e., no sugars added). Groups of 20 eyes for each experimental condition were maintained cornea-side down in 35 mm sterile Petri dishes maintained in a moist chamber for three days at 23 °C, after which they were fixed or frozen, as appropriate for the subsequent analysis. Using this culture protocol, retinas from fully intact eye rudiments have reached approximately stage 42 of the in vivo developmental scale, characterized by complete stratification of the retina and mature photoreceptors with well developed outer segments that express opsin and rds/peripherin in the proper location and quantity [17,30,32]. All experiments were repeated in triplicate.

Morphological assessment and immunocytochemistry

After three days of in vitro development, eyes were grossly examined under a dissecting microscope for integrity and smoothness of the neuroepithelial surface to ensure that all rudiments were intact. Any rudiment that exhibited an uneven surface or had many loose cells associated with it was discarded. Four eyes from three separate experimental sessions were evaluated in each series of experiments. For structural analysis, eyes were fixed in Tucker fixative and processed as previously described [16-19,30,33,34]. To ensure that photoreceptors of equivalent stages of maturation were compared, structural analyses were performed on tissue sections taken exclusively from the posterior pole region of the retina. For immunocytochemical localization analysis, eyes were fixed in Davidson fixative (32% ethanol, 2% formalin, 11% acetic acid), dehydrated, and embedded in Unicryl (Electron Microscopy Sciences, Fort Washington, PA). One-mm thick sections taken through the posterior pole of the eye were cut and collected on microscope slides. Immunocytochemistry was performed as previously described [17,19,34]. The following antibodies were used: A2D4 anti-GFAP [35], 1:1000 dilution; anti-CRALBP [25], 1:5000 dilution; and anti-glutamine synthetase, 1:1000 dilution (Chemicon International Inc., Temecula, CA). The appropriate gold-conjugated goat anti-mouse or anti-rabbit secondary antibodies were applied to the tissue sections (1:50 dilution for 2 h, ultrasmall gold particle size) followed by silver enhancement (Electron Microscopy Services). Additional retinal sections were processed identically except for exclusion of primary antibody. Retinal sections were viewed on a Nikon Eclipse E400 microscope equipped with Sensys Color Camera (Photometrics, Tucson AZ) and images were collected using MetaMorph Imaging System software (Universal Imaging Corporation, West Chester, PA). Two images of each retinal sectionwere collected, a brightfield image that showed the morphology of the tissue, and another image taken with epipolarized light that showed only the immunolabeling pattern. The epipolarized image was pseudo-colored and merged with the brightfield image so that the specific immunolabeling patterns could be distinguished easily. In Figure 1, Figure 2, Figure 3, and Figure 4, the immunopositive labeling appears red. Because the retinas were not osmicated, the structure of photoreceptor outer segments is not readily seen. For comparison of outer segment membranous organization and location, the reader is referred to the tissue sections presented in the "A" panels of those figures.

Quantification of steady-state Müller cell protein levels

To quantify Müller cell proteins, eyes were harvested from the various culture conditions after three days, ground, and solubilized with sodium cholate detergent (Sigma). Retinal extracts of equal volume were applied in duplicate to Hybond P membrane (Amersham Pharmacia Biotech, Piscataway, NJ) using a Bio-Dot SF microfiltration apparatus (BioRad, Hercules, CA). Twenty eyes per condition from each of the three individual experimental sessions were analyzed using this procedure. Pooling of retinas was necessary due to the small size of each intact eye (i.e., approximately 150 mm in diameter and 25 mg of total protein per eye). Quantification of proteins was performed using the ECF Western blotting kit (Amersham Pharmacia Biotech, Piscataway, NJ) was used as described previously [17,19,34]. The following primary antibodies were used: anti-GFAP [35] at 1:5,000 dilution; anti-CRALBP [25] at 1:10,000 dilution; and anti-glutamine synthetase (Chemicon International Inc., Temecula, CA) at 1:5,000 dilution. Blots were scanned on a Storm 860 Imaging system and data were quantified using ImageQuant software version 1.1 (Molecular Dynamics, Sunnyvale, CA). For each experimental session, data were normalized to the protein levels of eyes obtained from the same batch of embryos that were allowed to develop in vivo to stage 42; i.e., the equivalent stage of development reached by our retinas in vitro at the end of the three-day culture period [30]. This normalization effectively circumvented the inherent variability in absolute protein expression levels occurring from session to session, mainly due to the use of different parental pairs of Xenopus laevis frogs, each with distinct genetic backgrounds. In so doing, the in vitro eyes were always compared to in vivo eyes from sibling embryos, thereby limiting the residual variance to the specific experimental conditions.

Statistical analyses

Statistical analyses were performed with SAS statistical software (SAS Institute, Inc. Cary, NC). First, steady-state protein levels of stage 42 embryo eyes developed in vivo were compared with those of intact eyes that developed for three days in vitro by means of an unpaired t-test procedure. Subsequently, differences between the four in vitro conditions (i.e., with RPE, without RPE only, and without RPE with either mannose or lactose added) were statistically analyzed by one-way analysis of variance (ANOVA). Average steady-state protein levels for three groups of 10 eyes under the four experimental conditions were obtained as described above (n=12). P values less than 0.05 were considered statistically significant.

Results

Morphology and Müller cell protein expression patterns in retinas with an adherent RPE

Figure 1 illustrates representative morphology and immunocytochemical localization patterns of control retinas that were allowed to complete morphogenesis in the presence of an adherent RPE. Under these conditions, the retina was composed of three distinct cell layers with even borders separated by plexiform layers. There was no apparent cell loss within any of the cell layers. The outer segments of photoreceptors were highly organized into cylindrical profiles separated by interphotoreceptor matrix. Occasional diamond-shaped profiles, likely Müller cell nuclei, were noted in the inner nuclear layer (Figure 1A).

In the presence of the RPE, GFAP immunolabeling was undetectable, as would be expected in a healthy retina (Figure 1B). CRALBP labeling was heavy within both Müller cells and the RPE (Figure 1C). The radial Müller cell processes and diamond-shaped profiles within the inner nuclear layer were well demarcated with immunopositive-reaction product. This radial pattern, present throughout the entire thickness of the retina, is identical to the physical location of Müller cells within the retina [2], and corroborates the immunolocalization pattern demonstrated by other groups [36,37]. Some branching of radial processes was visible in the inner plexiform layer. In addition to being localized to the Müller cell radial processes, heavy CRALBP immunolabeling was present vitreally at the inner limiting membrane and sclerally in the area of the outer limiting membrane. A lower level of immunolabeling was also appreciable in the interphotoreceptor space immediately beyond the outer limiting membrane, coincident with the localization of Müller cell apical processes [36]. Glutamine synthetase immunolabeling also followed a radial pattern similar to that of CRALBP; however, the Müller cell endfeet were not as heavily labeled (Figure 1D). The level of glutamine synthetase immunoreactivity was heaviest over the inner and outer plexiform layers. At these locations, synaptic communication occurs between neurons, and Müller cells remove extracellular glutamate, converting it to glutamine via catalysis with glutamine synthetase [3]. It therefore follows that a high concentration of glutamine synthetase would be localized in these regions where it would protect retinal neurons by removing potentially harmful glutamate from the extracellular spaces, as suggested by Gorovits [38]. A lower level of glutamine synthetase labeling was also apparent at the outer limiting membrane Figure 1D. No CRALBP or glutamine synthetase labeling was detectable beyond the inner segment region of the photoreceptors Figure 1C,D.

Morphology and Müller cell expression patterns of dysmorphic retinas

The morphology and immunolocalization patterns of Müller cell markers of dysmorphic and mannose-exposed eyes are presented in Figure 2 and Figure 3, respectively. In retinas that completed morphogenesis in the absence of the RPE, the cytoarchitecture of the retina was markedly altered. Photoreceptor outer segment membrane structure was disorganized, with little evidence of normal disc stacking. Many photoreceptors were also lacking an inner segment (Figure 2A). Gaps within the nuclear layers, created by either cell loss or retinal edema, were also present. The same array of outer segment abnormalities was observed when retinas were maintained in vitro in the presence of mannose (Figure 3A), a non-permissive sugar [15,16,18].

Concurrent with these structural changes, the immunolabeling patterns of Müller cell markers were substantially altered in retinas undergoing dysmorphogenesis induced by RPE removal. In these eyes, GFAP immunoreactivity was detectable across the entire retinal thickness, corresponding to the localization of Müller cell radial processes (Figure 2B). The labeling extended from the inner to the outer limiting membranes. No GFAP labeling was present sclerad to the outer limiting membrane, indicating that hypertrophy of Müller cell processes did not occur. Müller cell expression of CRALBP (Figure 2C) was slightly reduced compared to healthy retinas. In addition, fewer nuclei were outlined in the inner nuclear layer (compare to Figure 1C). The CRALBP label traversed the entire retinal thickness and ended abruptly at the outer limiting membrane. No glutamine synthetase immunolabeling was detected in these retinas (Figure 2D).

In the absence of the RPE, the Müller cell immunolabeling patterns of retinas exposed to mannose was very similar to those of eyes cultured in Niu-Twitty medium alone, with the exception of GFAP. Mannose suppressed the upregulation of GFAP expression that was induced by removal of the RPE (Figure 3B; compare to Figure 2B). CRALBP continued to be expressed (Figure 3C; compare to Figure 2C), while glutamine synthetase immunopositive labeling remained negligible (Figure 3D; compare to Figure 2D).

Morphology and Müller cell expression patterns of retinas cultured in lactose-containing medium

Retinal morphology and Müller cell protein expression patterns of retinas incubated in lactose-containing medium are illustrated in Figure 4. In the absence of the RPE, retinas exposed to lactose maintained normal stratification with minimal cell loss and/or swelling in the inner and outer nuclear layers (Figure 4A). Lactose-exposed retinas also showed well-organized outer segments. The structure of the outer segment in lactose-exposed retinas resembled very closely that of outer segments elaborated under control conditions in which the RPE was present (compare to Figure 1A).

Lactose also was associated with near normal expression of all Müller cell-specific proteins. In lactose-exposed retinas, the upregulation of GFAP expression that should have been induced by RPE removal was prevented (Figure 4B; compare to Figure 2B). The retinal pattern of CRALBP expression (Figure 4C) was similar to control conditions with immunolabeling at the level of the outer and inner limiting membranes and in the radial processes. In addition, immunopositive labeling was present in the area of photoreceptor inner and proximal outer segments, beyond that indicated in control eyes. The level of glutamine synthetase labeling was maintained in lactose-exposed eyes (Figure 4D). Similar to the immunolabeling pattern noted with CRALBP, glutamine synthetase immunoreaction product was present sclerad to the outer limiting membrane, in the proximal outer segment region.

Quantification of steady-state protein levels of Müller cell markers: In vivo versus in vitro

Eyes that developed for three days in vitro expressed steady-state levels of the three investigated Müller cell markers that were not significantly different compared to stage 42 in vivo levels. In Figure 5, the protein levels are not shown for eyes from stage 42 embryos; however, the data from the in vitro control have been normalized to that measured in the in vivo control. Specifically, GFAP levels were very low, expressed at 99.5±2.5% of the in vivo levels (p=0.7486), CRALBP was widely detectable and expressed at 97.5±1.1% of the in vivo levels (p=0.0605), and glutamine synthetase too was expressed at levels undistinguishable from the in vivo state (100.5±2.9%; p=0.7803).

Quantification of steady-state protein levels of Müller cell markers: Comparisons between in vitro experimental conditions

The relative steady-state protein levels of the Müller cell markers from the in vitro experiments are presented graphically in Figure 5. Protein levels are shown normalized to that of embryos at the equivalent in vivo developmental stage (stage 42). Figure 5A illustrates findings for GFAP. The overall test for differences among the four groups was highly significant (F=29.56, p=0.0001). Removal of the RPE induced a 30-fold increase in GFAP expression (3,115.3±960.4%, p<0.0001 vs control). Both mannose and lactose prevented the upregulation of GFAP (102.9±7.4%, and 99.7±3.0%, respectively). These levels were not different from the control state (p=0.9933 and 0.9996, respectively), and were significantly lower than the RPE-deprived state (p<0.0001). Figure 5B illustrates findings for CRALBP. The overall test for differences in CRALBP expression levels among the four conditions was statistically significant (F=6.47; p=0.0156). There was a modest but significant decrease in the steady-state protein level of CRALBP in retinas with dysmorphic photoreceptors (86.3±5.9%, p=0.0037 and 91.0±2.1%, p=0.0457) for those eyes without the RPE and with the addition of mannose, respectively. The addition of lactose did not mitigate the downregulation of this protein (87.7±2.5%, p=0.0078 vs control). Figure 5C presents findings for glutamine synthetase. The overall test for differences in glutamine synthetase expression levels among the four conditions was highly significant (F=1792.86; p<0.0001). The expression of glutamine synthetase showed a differential regulation compared to both GFAP and CRALBP. In the dysmorphic retinas, including those exposed to mannose, the expression of glutamine synthetase was negligible (on average, less than 3% of control retinas, p<0.0001). However, the addition of lactose supported the expression of this protein to virtually normal levels (97.0±2.6%, p=0.0910 vs control and p<0.0001 vs the two other experimental conditions).

Discussion

When intact eyes are placed into culture in our experimental paradigm, Müller cells are not fully mature. Holt et al. previously demonstrated that at stage 33/34, only 45% of Müller cells in Xenopus retinas are post-mitotic [39]. However, our present results demonstrate that, after three days under our culture conditions, intact Xenopus laevis embryo eyes maintained with an adherent RPE express normal patterns and steady-state expression levels of Müller markers compared to eyes from stage 42 embryos. In addition, using a culture protocol identical to that used herein, we previously showed that, after three days of in vitro development, the subcellular cytoarchitecture of Müller cells reflects complete and normal cytomorphogenesis [18]. These results indicate that in the presence of the RPE, intact Xenopus eyes undergo normal development under our culture conditions, thus adding strength to the value of this model for evaluating Müller cell and photoreceptor maturation, for evaluating the efficacy of glio- and neuroprotective agents, and importantly, for evaluating interactions between the two cell types.

Evidence in the literature supports the stabilizing effect of the RPE upon both photoreceptors and Müller cells. Retinal detachment and RPE dystrophy trigger degenerative changes in both Müller cells and photoreceptors [5-8,11-14,30]. In the present study, removal of the RPE induced marked alterations in the expression patterns of the Müller cell markers similar to what has been described in the literature. GFAP upregulation, a widely accepted indicator of non-specific Müller cell stress [9], was only detectable in RPE-deprived retinas. These results are similar to what is documented in various forms of retinal injury induced by retinal detachment [6-8], light damage [9,40,41], a genetic defect in the RPE (i.e., the RCS rat) [14], and age-related macular degeneration [42]. However, we did not detect GFAP immunoreactivity sclerad to the outer limiting membrane, as has been described in other models of degeneration in which a glial scar is formed in the subretinal space [13,43]. This indicates that, in our model of retinal degeneration, no massive glial hypertrophy occurred.

Surprisingly, the presence of mannose in the incubation medium prevented the upregulation of GFAP, although it did not promote proper outer segment assembly. This suggests (a) that low levels of GFAP expression were not correlated with organized outer segment assembly, and (b) that mannose exerted this effect on Müller cells independently of any detectable effect upon photoreceptors. In these same retinas, CRALBP was expressed across the entire retinal thickness, although quantification indicated a slight but statistically significant reduction. The measured reduction in the level of CRALBP corroborates data from other groups who have described a downregulation of CRALBP expression during photoreceptor or Müller cell injury [20,44,45]. In the Müller cells of dysmorphic retinas, including those exposed to mannose, glutamine synthetase was not detectable by immunocytochemical assays, and was measurable only at negligible levels by immunoblot analyses (less than 3% of control levels).

We previously documented that lactose has a permissive effect upon the synthesis and assembly of photoreceptor outer segments in RPE-deprived retinas [15-17]. Herein we demonstrate that the expression of glutamine synthetase within Müller cells is preserved in retinas incubated in lactose-containing medium. Given that the upregulation of GFAP that is induced by RPE removal is abscent in cultures incubated in mannose-containing medium, it is unlikely that GFAP and its related functions play a role in the proper folding of photoreceptor outer segment membranes. Because CRALBP expression was slightly reduced three days after RPE removal, and neither mannose nor lactose affected the steady-state expression of this protein, our results also suggest that, while this enzyme may play a critical role in visual pigment regeneration [25], its expression is not necessary for organized photoreceptor outer segment assembly. Among the Müller cell markers evaluated in this study, only glutamine synthetase was correlated with properly formed outer segments. Glutamine synthetase is localized exclusively in Müller cells, where it is a potent neuroprotectant [38] and a key enzyme in glial-neuronal neurotransmitter recycling [46]. Evidence in the literature also indicates that the expression of glutamine synthetase may be indicative of Müller cell differentiation [47]. Results from the present study suggest that its expression is also associated with proper folding of nascent photoreceptor outer segment membranes.

Evidence in the literature demonstrates that the expression of glutamine synthetase is dependent upon cell-cell contact between Müller cells and retinal neurons [21-24,48]. A recent study by Prada et al. also indicates that precise contact with neurons at the level of the outer plexiform or outer nuclear layer is essential for Müller cells to express the glutamine synthetase gene [24]. We have recently provided corroborating evidence by demonstrating that properly structured adherens junctions that compose the outer limiting membrane are present exclusively in retinas with an adherent RPE or in lactose-protected retinas [18]. The present study reveals that glutamine synthetase is expressed only when adherens junctions are present. Both adherens junctions [18] and glutamine synthetase expression are lacking in RPE-deprived and mannose-exposed retinas. A link between glutamine synthetase expression and integrity of the cytoskeletal network of the retina has been recently provided by Oren et al. [49], who demonstrated that, subsequent to depolymerization of the actin or tubulin cytoskeleton, glutamine synthetase expression is suppressed concomitant with an induction in c-Jun within Müller cells. Oren et al. proposed that information regarding cell contact between Müller cells and retinal neurons may be relayed to the nucleus by the cytoskeletal network [49]. Interestingly, several laboratories, including our own, have demonstrated that disruption of the retinal cytoskeleton is associated with, or responsible for, aberrant outer segment assembly [18,50-52].

In summary, in conjunction with our previous work [18,19], the results of the present study show that preservation of Müller cell protein expression levels and ultrastructure are associated with structurally intact retinas with properly assembled photoreceptor outer segments. This observation does not allow us to establish either a direct cause-effect relationship between glutamine synthetase expression in Müller cells and photoreceptor outer segment assembly, or a specific temporal sequence. In the chain of events linking these two phenomena, one may be responsible for the other, or vice-versa, or either one may be the marker for a third, related but yet uncharacterized, event. A simple explanation for preserved glutamine synthetase expression patterns may be that the lactose-mediated support of photoreceptor integrity causes a greater photoreceptor synaptic activity, with increased release of glutamate. This, in turn, may lead Müller cells to upregulate glutamine synthetase to enzymatically convert the released glutamate into glutamine. However, this would not explain the absence of GFAP upregulation in eyes treated with mannose, a non-permissive sugar with which no appreciable beneficial effect on outer segment organization is observed [18]. In fact, in preliminary experiments with tritiated galactose (another sugar that supports outer segment assembly [15,16]), we have observed selective accumulation of galactose in Müller cells, and not in photoreceptors (unpublished observation). This suggests that the effect of permissive sugars towards photoreceptors may be exerted via Müller cell mediation. If this hypothesis is true, we suggest that, in the absence of the RPE, lactose may be supporting Müller cell morphogenesis and allow these cells to differentiate normally. In turn, this effect of lactose upon Müller cells may support photoreceptor outer segment assembly via an as of yet uncharacterized mechanism (of which glutamine synthetase may well be only a marker, and not an effector), possibly related to the formation of normal adherens junctions with adjacent photoreceptors [18] and an intact Müller cell cytoskeleton. This hypothesis is currently being evaluated in our laboratory.

Acknowledgements

The authors gratefully acknowledge Andre DeBellevué and Mike Davis from Advanced Scientific Inc., Meraux, LA for their assistance with collecting the images presented in this manuscript. The following have generously provided antibodies for our use: Dr. Amico Bignami, Harvard Medical School, Boston, MA and Dr. John Saari, University of Washington, Seattle, WA. This study was supported by grants from the National Eye Institute grant EY10853 (MMJ), the International Retinal Research Foundation (Birmingham, AL) and an unrestricted departmental award from Research to Prevent Blindness Inc., New York. MMJ is a Research to Prevent Blindness William and Mary Greve Special Scholar.

References

1. Reichenbach A, Stolzenburg JU, Eberhardt W, Chao TI, Dettmer D, Hertz L. What do retinal muller (glial) cells do for their neuronal 'small siblings'? J Chem Neuroanat 1993; 6:201-13.

2. Robinson SR, Dreher Z. Muller cells in adult rabbit retinae: morphology, distribution and implications for function and development. J Comp Neurol 1990; 292:178-92.

3. Newman E, Reichenbach A. The Muller cell: a functional element of the retina. Trends Neurosci 1996; 19:307-12.

4. Cao W, Wen R, Li F, Cheng T, Steinberg RH. Induction of basic fibroblast growth factor mRNA by basic fibroblast growth factor in Muller cells. Invest Ophthalmol Vis Sci 1997; 38:1358-66.

5. Guerin CJ, Lewis GP, Fisher SK, Anderson DH. Recovery of photoreceptor outer segment length and analysis of membrane assembly rates in regenerating primate photoreceptor outer segments. Invest Ophthalmol Vis Sci 1993; 34:175-83.

6. Lewis GP, Matsumoto B, Fisher SK. Changes in the organization and expression of cytoskeletal proteins during retinal degeneration induced by retinal detachment. Invest Ophthalmol Vis Sci 1995; 36:2404-16.

7. Okada M, Matsumura M, Ogino N, Honda Y. Muller cells in detached human retina express glial fibrillary acidic protein and vimentin. Graefes Arch Clin Exp Ophthalmol 1990; 228:467-74.

8. Erickson PA, Fisher SK, Guerin CJ, Anderson DH, Kaska DD. Glial fibrillary acidic protein increases in Muller cells after retinal detachment. Exp Eye Res 1987; 44:37-48.

9. Eisenfeld AJ, Bunt-Milam AH, Sarthy PV. Muller cell expression of glial fibrillary acidic protein after genetic and experimental photoreceptor degeneration in the rat retina. Invest Ophthalmol Vis Sci 1984; 25:1321-8.

10. D'Cruz PM, Yasumura D, Weir J, Matthes MT, Abderrahim H, LaVail MM, Vollrath D. Mutation of the receptor tyrosine kinase gene Mertk in the retinal dystrophic RCS rat. Hum Mol Genet 2000; 9:645-51.

11. Faktorovich EG, Steinberg RH, Yasumura D, Matthes MT, LaVail MM. Photoreceptor degeneration in inherited retinal dystrophy delayed by basic fibroblast growth factor. Nature 1990; 347:83-6.

12. Li L, Turner JE. Optimal conditions for long-term photoreceptor cell rescue in RCS rats: the necessity for healthy RPE transplants. Exp Eye Res 1991; 52:669-79.

13. Roque RS, Caldwell RB. Muller cell changes precede vascularization of the pigment epithelium in the dystrophic rat retina. Glia 1990; 3:464-75.

14. Li L, Sheedlo HJ, Turner JE. Muller cell expression of glial fibrillary acidic protein (GFAP) in RPE-cell transplanted retinas of RCS dystrophic rats. Curr Eye Res 1993; 12:841-9.

15. Stiemke MM, Hollyfield JG. Outer segment disc membrane assembly in the absence of the pigment epithelium: the effect of exogenous sugars. Brain Res Dev Brain Res 1994; 80:285-9.

16. Stiemke MM, Hollyfield JG. Effect of sugars on photoreceptor outer segment assembly. In: Anderson RE, LaVail MM, Hollyfield JG, editors. Degenerative Diseases of the Retina. New York: Plenum Publishing; 1995. p. 129-37.

17. Jablonski MM, Wohabrebbi A, Ervin CS. Lactose promotes organized photoreceptor outer segment assembly and preserves expression of photoreceptor proteins in retinal degeneration. Mol Vis 1999; 5:16 <http://www.molvis.org/molvis/v5/a16/>.

18. Jablonski MM, Ervin CS. Closer look at lactose-mediated support of retinal morphogenesis. Anat Rec 2000; 259:205-14.

19. Jablonski MM, Iannaccone A. Targeted disruption of Muller cell metabolism induces photoreceptor dysmorphogenesis. Glia 2000; 32:192-204.

20. Lewis GP, Guerin CJ, Anderson DH, Matsumoto B, Fisher SK. Rapid changes in the expression of glial cell proteins caused by experimental retinal detachment. Am J Ophthalmol 1994; 118:368-76.

21. Linser P, Moscona AA. Induction of glutamine synthetase in embryonic neural retina: localization in Muller fibers and dependence on cell interactions. Proc Natl Acad Sci U S A 1979; 76:6476-80.

22. Linser P, Saad AD, Soh BM, Moscona AA. Cell contact-dependent regulation of hormonal induction of glutamine synthetase in embryonic neural retina. Prog Clin Biol Res 1982; 85:445-58.

23. Vardimon L, Fox LE, Cohen-Kupiec R, Degenstein L, Moscona AA. Expression of v-src in embryonic neural retina alters cell adhesion, inhibits histogenesis, and prevents induction of glutamine synthetase. Mol Cell Biol 1991; 11:5275-84.

24. Prada FA, Quesada A, Dorado ME, Chmielewski C, Prada C. Glutamine synthetase (GS) activity and spatial and temporal patterns of GS expression in the developing chick retina: relationship with synaptogenesis in the outer plexiform layer. Glia 1998; 22:221-36.

25. Saari JC, Huang J, Possin DE, Fariss RN, Leonard J, Garwin GG, Crabb JW, Milam AH. Cellular retinaldehyde-binding protein is expressed by oligodendrocytes in optic nerve and brain. Glia 1997; 21:259-68.

26. Saari JC. Biochemistry of visual pigment regeneration: the Friedenwald lecture. Invest Ophthalmol Vis Sci 2000; 41:337-48.

27. Lolley RN, Farber DB, Rayborn ME, Hollyfield JG. Cyclic GMP accumulation causes degeneration of photoreceptor cells: simulation of an inherited disease. Science 1977; 196:664-6.

28. Hollyfield JG, Witkovsky P. Pigmented retinal epithelium involvement in photoreceptor development and function. J Exp Zool 1974; 189:357-78.

29. Nieuwkoop PD, Faber J, editors. Normal table of Xenopus laevis (Daudin). Amsterdam: North Holland Publishing Co.; 1956.

30. Stiemke MM, Landers RA, al-Ubaidi MR, Rayborn ME, Hollyfield JG. Photoreceptor outer segment development in Xenopus laevis: influence of the pigment epithelium. Dev Biol 1994; 162:169-80.

31. Jacobson AG. Amphibian cell culture, organ culture, and tissue dissociation. In: Wilt FH, Wessells NK, editors. Methods in developmental biology. New York: Thomas Y. Crowell; 1967. p. 531-42.

32. Hollyfield JG, Rayborn ME. Photoreceptor outer segment development: light and dark regulate the rate of membrane addition and loss. Invest Ophthalmol Vis Sci 1979; 18:117-32.

33. Kancherla V, Kedzierski W, Travis GH, Jablonski MM. Abnormal formation of outer segments in Xenopus laevis eye rudiments cultured with rds antisense oligonucleotides. In: Hollyfield JG, Anderson RE, LaVail MM, editors. Retinal degenerative disease and experimental therapy. New York: Plenum; 1999. p. 419-29.

34. Jablonski MM, Tombran-Tink J, Mrazek DA, Iannaccone A. Pigment epithelium-derived factor supports normal development of photoreceptor neurons and opsin expression after retinal pigment epithelium removal. J Neurosci 2000; 20:7149-57.

35. Dahl D, Crosby CJ, Sethi JS, Bignami A. Glial fibrillary acidic (GFA) protein in vertebrates: immunofluorescence and immunoblotting study with monoclonal and polyclonal antibodies. J Comp Neurol 1985; 239:75-88.

36. Bunt-Milam AH, Saari JC. Immunocytochemical localization of two retinoid-binding proteins in vertebrate retina. J Cell Biol 1983; 97:703-12.

37. Eisenfeld AJ, Bunt-Milam AH, Saari JC. Localization of retinoid-binding proteins in developing rat retina. Exp Eye Res 1985; 41:299-304.

38. Gorovits R, Avidan N, Avisar N, Shaked I, Vardimon L. Glutamine synthetase protects against neuronal degeneration in injured retinal tissue. Proc Natl Acad Sci U S A 1997; 94:7024-9.

39. Holt CE, Bertsch TW, Ellis HM, Harris WA. Cellular determination in the Xenopus retina is independent of lineage and birth date. Neuron 1988; 1:15-26.

40. Burns MS, Robles M. Muller cell GFAP expression exhibits gradient from focus of photoreceptor light damage. Curr Eye Res 1990; 9:479-86.

41. de Raad S, Szczesny PJ, Munz K, Reme CE. Light damage in the rat retina: glial fibrillary acidic protein accumulates in Muller cells in correlation with photoreceptor damage. Ophthalmic Res 1996; 28:99-107.

42. Madigan MC, Penfold PL, Provis JM, Balind TK, Billson FA. Intermediate filament expression in human retinal macroglia. Histopathologic changes associated with age-related macular degeneration. Retina 1994; 14:65-74.

43. Fan W, Lin N, Sheedlo HJ, Turner JE. Muller and RPE cell response to photoreceptor cell degeneration in aging Fischer rats. Exp Eye Res 1996; 63:9-18.

44. Sheedlo HJ, Jaynes D, Bolan AL, Turner JE. Mullerian glia in dystrophic rodent retinas: an immunocytochemical analysis. Brain Res Dev Brain Res 1995; 85:171-80.

45. Rich KA, Figueroa SL, Zhan Y, Blanks JC. Effects of Muller cell disruption on mouse photoreceptor cell development. Exp Eye Res 1995; 61:235-48.

46. Germer A, Jahnke C, Mack A, Enzmann V, Reichenbach A. Modification of glutamine synthetase expression by mammalian Muller (glial) cells in retinal organ cultures. Neuroreport 1997; 8:3067-72.

47. Vardimon L, Ben-Dror I, Havazelet N, Fox LE. Molecular control of glutamine synthetase expression in the developing retina tissue. Dev Dyn 1993; 196:276-82.

48. Vardimon L, Fox LL, Degenstein L, Moscona AA. Cell contacts are required for induction by cortisol of glutamine synthetase gene transcription in the retina. Proc Natl Acad Sci U S A 1988; 85:5981-5.

49. Oren A, Herschkovitz A, Ben-Dror I, Holdengreber V, Ben-Shaul Y, Seger R, Vardimon L. The cytoskeletal network controls c-Jun expression and glucocorticoid receptor transcriptional activity in an antagonistic and cell-type-specific manner. Mol Cell Biol 1999; 19:1742-50.

50. Williams DS, Linberg KA, Vaughan DK, Fariss RN, Fisher SK. Disruption of microfilament organization and deregulation of disk membrane morphogenesis by cytochalasin D in rod and cone photoreceptors. J Comp Neurol 1988; 272:161-76.

51. Arikawa K, Williams DS. Acetylated alpha-tubulin in the connecting cilium of developing rat photoreceptors. Invest Ophthalmol Vis Sci 1993; 34:2145-9.

52. Hale IL, Fisher SK, Matsumoto B. The actin network in the ciliary stalk of photoreceptors functions in the generation of new outer segment discs. J Comp Neurol 1996; 376:128-42.