Muller Cell Expression of Gliol Fibrillory Acidic Protein offer Genetic and Experimental Photoreceptor Degeneration in the Rat

Retina
Amy J. Eisenfeld, Ann H. Bunr-Milam, and P. Vijoy Sarrhy
Glial fibrillary acidic protein (GFAP) is normally found in astrocytes. In the normal rat retina at all ages, only astrocytes stain for GFAP. This staining pattern is also found in RCS rats with inherited retinal dystrophy younger than 38 days. Beginning on day 38, when about 61% of the photoreceptors have degenerated, a few GFAP-positive fibers span the retina from the inner limiting membrane to the external limiting membrane. By day 41 and at all later ages examined, the radial fibers of Muller cells are labeled throughout the retina. To determine if the expression of GFAP in Muller cells is a response to photoreceptor necrosis or might be a direct effect of the mutant gene, we induced photoreceptor degeneration in normal, adult Sprague-Dawley rats by exposing them to constant light for variable periods of time. After 3 days in constant light, there is a 20% reduction in the number of photoreceptors and many Muller cells are positive for GFAP. Immunoblot studies confirmed that the anti-GFAP reacted with a single protein from retina that corresponded in molecular weight and Triton-insolubility to GFAP. The immunoblots also corroborated the results from anti-GFAP immunostaining of control and experimental retinas. These results indicate that Muller cells express GFAP immunoreactivity in response to experimentally as well as genetically induced photoreceptor degeneration. Invest Ophthalmol Vis Sci 25:1321-1328, 1984

Mtiller cells are the major type of non-neuronal cells in the vertebrate retina. Although they are morphologically similar to radial glia and Bergmann glial cells, unlike these cells, the Muller cells normally do not express the glial cell-specific protein, glial fibrillary acidic protein (GFAP).1"4 However, it has been reported that Muller cells accumulate GFAP in response to neuronal injury1'2 and degeneration.3-4 We have used antibodies to GFAP and indirect immunofluorescence to examine the accumulation of GFAP in Muller cells in response to genetically induced photoreceptor degeneration in the Royal College of Surgeons (RCS) rat. The RCS rat has an inherited retinal dystrophy resulting in the loss of photoreceptors between the ages of 18 and 60 days.5 Since any alterations in Mtiller cells might be a direct effect of the mutant gene, we also examined Muller cells in Sprague-Dawley rats maintained in constant

light, a condition that causes loss of photoreceptors in albino rats.6 These models of photoreceptor degeneration result in loss of photoreceptors with little damage to the remainder of the retina.6'7 It was of interest to determine if the different forms of photoreceptor degeneration might result in any differences in the time course of appearance of GFAP in the Muller cells.

Materials and Methods

Animals Pink-eyed dystrophic (RCS) rats (from breeding pairs provided by Dr. Matthew La Vail, University of California, San Francisco), maintained in a 12-hr light/ 12-hr dark environment, were used as a model for inherited photoreceptor cell degeneration. Rats of the same ages from a congenic strain without inherited retinal dystrophy (RCS-rdy+) were used as controls. All procedures involving animals were performed in adherence to the ARVO Resolution on the Use of Animals in Research. Photoreceptor degeneration was experimentally induced in 45-day-old Sprague-Dawley rats (Bellevue, WA) by placing them in constant light (CL) for periods of 1 day to 8 weeks. These rats were kept in transparent cages with stainless steel wire bar covers. In addition to 24-hr overhead room light, two lamps,

From the Department of Ophthalmology, University of Washington School of Medicine, Seattle, Washington. Supported in part by NIH Research grant nos. EYO7O13, EY01311, EY03523, EY03664, and EYO173O and in part by an unrestricted grant from Research to Prevent Blindness, Inc. Dr. Bunt-Milam is the recipient of a William and Mary Greve International Scholar Award from Research to Prevent Blindness, Inc. Submitted for publication: April 13, 1984. Reprint requests: Amy J. Eisenfeld, PhD, Department of Ophthalmology RJ-10, University of Washington, Seattle WA 98195.

1321

1322

INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / November 1984

Vol. 25

each containing two 15-watt fluorescent bulbs were positioned 18 cm above the bottom of the cage. These conditions resulted in an incident luminance of approximately 200-ft candles at the floor of the cage. The temperature in the cage was 24 1 C. Age-matched Sprague-Dawley rats maintained in a 12-hr overhead room light/ 12-hr dark environment, were used as controls. All rats were enucleated under ether anesthesia between 1:00 and 2:30 PM. After a slit was made in the cornea, the lens and vitreous were removed and the globe was immersed in 4% formalin in 0.13 M phosphate buffer (pH 7.4). Immunofluorescence After 6 hr in fixative at room temperature, the eyes were bisected, transferred to 30% sucrose in 0.13 M phosphate buffer and stored at 4C overnight. Sections were cut at a thickness of 20 nm using a cryostat at 20C. The sections were mounted on chrome alum-gelatin coated slides and air-dried overnight at room temperature. Plastic rings (0.75-cm diameter) were mounted with fingernail polish around the sections to form incubation wells. Each well contained an experimental section of retina (RCS or CL-damaged) and a control section. The sections were treated for 10 min at room temperature with 1% goat serum and 4% bovine serum albumin (BSA) in phosphate buffered saline (PBS) followed by overnight incubation at 4C in GFAP antiserum diluted 1:100 in PBS containing 0.3% Triton X-100. The GFAP antiserum, provided by Dr. Larry Eng (Veterans Medical Center; Palo Alto, CA), was raised in rabbits against GFAP obtained from multiple sclerosis plaques.8 Control sections were treated identically with an IgG fraction from preimmune rabbit serum. Sections were washed twice (15 min each) with PBS at room temperature and incubated for 30 min at room temperature in the dark in sheep anti-rabbit IgG-fluorescein isothiocyanate (Cappel Laboratories), diluted 1:50 in PBS with 0.3% Triton X-100. After two 10-min washes in phosphate buffer, the plastic rings were removed and the sections were coverslipped with 80% glycerol in 0.13 M phosphate buffer containing 5% n-propyl gallate.9 The sections were examined with a Zeiss microscope equipped for epifluorescence. Measurement of Outer Nuclear Layer After enucleation, eyes were stored in fixative at 4C overnight. They were bisected vertically just temporal to the optic nerve head. The bisected eyes were washed for several hours in phosphate buffer and then dehydrated through a graded series of

ethanol. The half of the eye including the optic nerve head was embedded in plastic (Sorvall Embedding Medium), with the cut edge of the eyecup placed flat to allow for sectioning along the inferior-superior plane, including the optic nerve head. In some cases the eyes were bisected after 6 hr in fixative, and the half without the optic nerve head was processed for immunofluorescence, while the remaining half was stored in fixative overnight. Sections were cut at a thickness of 2.5 ^m on a Sorvall JB-4 microtome and stained for 30 sec with 10% Richardson's stain. A section through the optic nerve head from each eye was chosen and the thickness of the outer nuclear layer was measured 250 ixm, 500 fim, and 750 /xm from the optic nerve head in the superior and inferior hemispheres. The mean and standard error of the mean of these six measurements were calculated. Preparation of Triton X-100 Insoluble Proteins Triton-insoluble proteins were obtained from retina according to Pruss et al.10 Four to six retinas were homogenized in ice-cold PBS containing the protease inhibitors p-chloromercuribenzoate, phenyl methane sulfonyl fluoride and o-phenanthraline, each at 1 raM concentration. The homogenate was centrifuged at 8000 g for 10 min at 4C. The pellet was extracted with PBS containing 0.6 M KC1, 0.5% Triton X-100 and the protease inhibitors, and centrifuged at 8000 g for 10 min at 4C. After a second extraction with Triton X-100, the pellet was washed four times in ice-cold PBS and solubilized by boiling in SDSpolyacrylamide sample buffer." In the light damage experiments, Sprague-Dawley rats had been exposed to constant light for 3 or 7 days. Control animals had been kept in 12-hr light/12-hr dark cycle. RCS rats were 45 and 70 days old. Congenic, 45-day-old ^y"1" rats were used as controls. Polyacrylamide Gel Electrophoresis and Electroblotting to Nitrocellulose One dimensional SDS-polyacrylamide gel electrophoresis (PAGE) was performed according to the procedure of Fairbanks et al" using protein standards ranging in molecular weight from 15-94,000. Proteins were transferred from PAGE gels to nitrocellulose membranes (BIORAD) in a Hoefer Transphor electrophoresis apparatus at 0.8 mV for 1 hr at room temperature. 12 After 1 hr blocking in Tris-buffered saline (TBS) containing 3% BSA, the blots were treated overnight in anti-GFAP diluted in TBS and 1% BSA. After several washes, the blots were incubated with goat anti-rabbit IgG (Cappel) for 1 hr (1:1000 dilution in TBS and 1% BSA). Following a 30-min exposure to the peroxidase-antiperoxidase complex

No. 11

GFAP IN MULLER CELLS / Eisenfeld er al.

1023

Fig, 1. Light micrographs of normal, constant light-damaged and RCS rat retinas. A, Normal Sprague-Dawley rat retina. O, outer segments; ON, outer nuclear layer; IN, inner nuclear layer; 1, inner plexiform layer; G, ganglion cell layer. B, Sprague-Dawley rat retina after 3 days in constant light. C, 38-day-old RCS rat. Note the decreased thickness of the ON. The inner retina appears normal.

(1:500 in PBS and 1% BSA), the blots were stained in Tris-saline containing 4-chloronaphthol (0.5 mg/ ml) and hydrogen peroxide (0.025%).13 Results Thickness of Outer Nuclear Layer Examples of normal, CL-damaged and RCS retinas (Figs. 1A-C) illustrate the extent of photoreceptor degeneration. The amount of damage caused by a 3-day exposure of CL was somewhat variable from animal to animal. The thickness of the outer nuclear layer of a control Sprague-Dawley rat was 46.0 1.0 /xm (mean SEM, n = 3) while for rats kept in CL for 3 days it ranged from 22-47 nm with a mean of 37.7 2.7 nm (n = 11), representing a 20% loss of photoreceptors (Fig. IB). In the 38-day-old RCS-rdy+ rats without inherited retinal degeneration, the thickness of the outer nuclear layer was 45.6 1.0 jum (n = 6). In 38-day-old RCS retinas with inherited retinal degeneration, the outer nuclear layer thickness was only 17.8 0.6 ^m (n = 7), representing an average reduction by 61% (Fig. 1C). No abnormalities were apparent in the inner retina in either condition of photoreceptor degeneration (Fig. IB, C).

Immunofluorescence

Sections treated with preimmune serum showed only autofluorescence that was pale green for the neural retina and yellow for erythrocytes (Figs. 2A> 3A). In control Sprague-Dawley rats, GFAP staining was confined to filamentous structures in the innermost retina, including the nerve fiber and ganglion cell layers and encircling blood vessels (Fig. 2B). GFAP positive cells were also abundant in the optic nerve head. From their location and morphology, these GFAP positive cells were interpreted as astrocytes. In some cases, there was a light, finely particulate staining in the outer segment layer. After 1 day in CL, the GFAP staining did not differ from that in control retinas. After 3 days in CL, variable numbers of radially oriented processes were stained, some more strongly than others. These processes extended from the inner limiting membrane through the inner plexiform and inner nuclear layers, with occasional positive fibers in the outer nuclear layer (Fig. 2C). This pattern of staining closely matched the distribution of Muller cell processes and appeared to represent the appearance of GFAP reactivity in Muller cells. Staining was most intense

1324

INVESTIGATIVE OPHTHALMOLOGY & VI5UAL SCIENCE / November 1984

Vol. 25

Fig. 2, Fluorescence micrographs of control and constant light-damaged retinas treated with antibodies to GFAP. A, Control section, exposed to constant light for 3 days and treated with preimmune serum. B, Normal retina. Only astrocytes () stain for GFAP. C, Retina exposed to constant light for 3 days. Miiller cell processes (') express GFAP immunoreactivity. D, Retina exposed to constant light for 2 weeks. Miiller cells (>) stain for GFAP. ON, outer nuclear layer (X576).

against the inner limiting membrane, and in the ganglion cell and innermost inner plexiform layer. Although the amount of photoreceptor cell loss varied

from animal to animal after 3 days in CL, every retina examined showed Miiller cell staining. The staining pattern remained the same at 2 weeks in CL

No. 11

GFAP IN MULLEP, CELLS / Eisenfeld er d .

1325

tig. 3. Huorescence micrographs of RCS retinas treated with antibodies to GFAP. A, Control section. A 38-day-old retina treated with preimmune serum. B, A 25-day-old RCS retina. Only astrocytes () express GFAP immunoreactivity. C, A 38-day-old RCS retina. Muller cell processes () first stain for GFAP at this age. D, A 6-month-old RCS retina. At this later stage of degeneration, Muller cells () stain intensely for GFAP. Accumulations of lipofuscin () are seen in the pigment epithelium. ON, outer nuclear layer (X576).

(Fig. ID) and at 8 wk when very few photoreceptors could be found. In the RCS rat, the GFAP staining was restricted to the astrocytes until day 32, as seen in control retinas (Figs. 2B, 3B). Beginning on day 32, an occasional Muller fiber stained lightly for GFAP. Muller cell processes throughout the retina were stained consistently with anti-GFAP only after day 38 (Fig. 3C). At this time the Muller end feet and innermost Muller radial processes stained most intensely. As the photoreceptor degeneration progressed, the number of GFAP positive fibers increased, and

they spanned the retina from the inner to the external limiting membranes. At advanced stages of degeneration (6 months; Fig. 3D), the Muller processes were thickened and very heavily stained.
Characterization of anti-GFAP

In order to ascertain that the protein stained by anti-GFAP was indeed GFAP, electroblot analysis was performed on proteins from normal and degenerated retinas. Results from the anti-GFAP electroblot experiments are presented in Figure 4. Although

1
67k

43k-

67k

43k-

No. 11

GFAP IN MULLER CELLS / Eisenfeld er ol.

1327

Fig. 4. Characterization of GFAP antibody in normal, constant light-damaged (A) and RCS (B) retinas. Lanes 1-3 are SDS-polyacrylamide gels stained with Coomassie Blue. Lanes 4-6 are PAP stained immunoblots. Molecular weights calculated from protein standards are shown on the left. A, Constant light-damaged retinas. Lanes (1, 4), normal retina; (2, 5) retina exposed to constant light for 7 days; (3, 6) retina exposed to constant light for 3 days. B, RCS retinas; Lanes (1,4) normal retina; (2, 5) 40-day-old RCS retina; (3, 6) 70-day-old RCS retina.

several protein bands were seen in the Coomassie Blue-stained acrylamide gels, anti-GFAP stained only one or at most two adjacent bands. Using molecular weight markers, the size of the anti-GFAP reacting protein was estimated at 50,000 for CL-damaged retinas and 47,000 for RCS retinas. The Tritoninsoluble nature of the protein, as well as its apparent molecular weight, indicate that the protein stained in the immunofluorescence studies is GFAP. The nitrocellulose blots showed that a small amount of GFAP was present in both RCS-rdy+ and the CL control retinas. This was in accord with the immunocytochemical staining of astrocytes in sections of these retinas. Further, it appeared that the amount of GFAP increased with progressive photoreceptor loss in both the RCS and light damaged retinas, corroborating the immunocytochemical observation of increased GFAP in Miiller cells in both conditions.

Discussion
The initial appearance of GFAP immunoreactivity was seen in Miiller cells from CL and RCS retinas only after substantial loss of photoreceptors. This loss, as determined by measurements of the outer nuclear layer, reflected a 20% decrease in CL damaged retinas and a 61% decrease in RCS rats. There are several possible explanations for this apparent difference in the degree of photoreceptor loss before GFAP reactivity was detected. (1) The outer nuclear layer thickness measurements indicated the degree of death and dropping out of photoreceptor cells. Since the metabolic status of the remaining photoreceptors was not monitored, even a morphologically normal photoreceptor might already be altered functionally. Therefore, the outer nuclear layer thickness might not be an accurate measure of the state of degeneration.14 (2) Although both conditions resulted ultimately in the loss of photoreceptors, the etiologies of the two forms of degeneration are thought to differ. In the RCS rat, the genetic defect has been localized to the pigment epithelial cell, which is unable to phagocytose shed outer segments. 1516 This leads to accumulation of outer segment debris and subsequent photoreceptor degeneration. The cause of photoreceptor death in CL exposure is unknown, but several mechanisms have been considered, including lipid peroxidation,17

photo-oxidation and retinol-induced membranolysis.18 These unrelated mechanisms leading to photoreceptor death might result in a different time course of GFAP accumulation in Muller cells. (3) Finally, there might be a direct effect on other cell types in the retina, leading to GFAP accumulation in Muller cells. This would seem more likely in the CL condition, where exposure to CL might have a primary effect on Muller cells, resulting in a more rapid accumulation of GFAP. This, of course, remains hypothetical at the present time. Our results provide further evidence that Muller cells express GFAP immunoreactivity following degeneration of apparently a single cell type, the photoreceptor. The time course of GFAP expression here in Muller cells after CL damage is quite similar to the increased anti-GFAP stainability in astrocytes at 48 hr following a stab wound of the brain 19 and in Muller cells after optic nerve section or penetrating wounds of the eye.1 The different time course of Muller cell gliosis in the RCS rat, as well as the actual significance of increased GFAP expression in astrocytes and Muller cells in pathologic conditions 820 are topics for future study. This study has shown that retinas with environmentally and genetically caused photoreceptor degeneration may provide useful models for the elucidation of Muller cell functions, including reaction to injury. The ability to induce accumulation of GFAP in a cell type not normally expressing this protein should facilitate the study of the mechanisms of reactive gliosis in other parts of the central nervous system. Key words: Muller cells, glial fibrillary acidic protein, photoreceptor degeneration, RCS rat, light damage

Acknowledgments
The authors wish to thank Dr. Larry Eng for the antiserum to GFAP; Dr. Matthew La Vail for the RCS rats; Dr. J. C. Saari for critical review of the manuscript; Mr. G. Garwin and Ms. I. Klock for technical assistance; Mr. B. Clifton and Ms. D. Cannon for photographic help; and Ms. J. Seng for secretarial assistance.

References
1. Bignami A and Dahl D: The radial glia of Muller in the rat retina and their response to injury. An immunofluorescence study with antibodies to the glial fibrillary acid (GFA) protein. ExpEye Res 28:63, 1979.

1328

INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / November 1984

Vol. 25

2. O'Dowd DK and Eng LF: Immunocytochemical localization of the glial fibrillary acid (GFA) protein in the Mueller cell of the human retina. Soc Neurosci 5:431, 1979. 3. Drager UC and Edwards DL: Antibodies to intermediate filaments reveal abnormalities in retinas of mice with photoreceptor degeneration. ARVO Abstracts. Invest Ophthalmol VisSci24(Suppl):115, 1983. 4. Shaw G and Weber K: The structure and development of the rat retina: an immunofluorescence microscopical study using antibodies specific for intermediate filament proteins. Eur J Cell Biol 30:219, 1983. 5. LaVail MM: Analysis of neurological mutants with inherited retinal degeneration. Invest Ophthalmol Vis Sci 31:638, 1981. 6. Noell WK, Walker VS, Kang BS, and Berman S: Retinal damage by light in rats. Invest Ophthalmol 5:450, 1966. 7. Eisenfeld AJ, LaVail MM, and LaVail JH: Assessment of possible transneuronal changes in the retina of rats with inherited retinal dystrophy: Cell size, number, synapses and axonal transport by retinal ganglion cells. J Comp Neurol 223:22, 1984. 8. Eng LF and DeArmand SJ: Immunocytochemical studies of astrocytes in normal development and disease. Adv Cell Neurobiol 3:145, 1982. 9. Giloh H and Sedat JW: Fluorescence microscopy: reduced photobleaching of rhodamine and fluorescein protein conjugate by n-propyl gallate. Science 217:1252, 1982. 10. Pruss RM, Mirsky R, Raff MC, Thorpe R, Dowding AJ, and Anderton BH: All classes of intermediate filaments show a common antigenic determinant defined by a monoclonal antibody. Cell 27:419, 1981. 11. Fairbanks G, Steck TL, and Wallach DFH: Electrophoretic

analysis of the major polypeptides of the human erythrocyte membrane. Biochemistry 10:2606, 1971. 12. Towbin H, Staehelin T, and Gordon J: Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc Natl Acad Sci USA 76:4350, 1979. 13. Bunt-Milam AH and Saari JC: Immunocytochemical localization of two retinoid-binding proteins in vertebrate retina. J Cell Biol 97:703, 1983. 14. O'Steen WK and Donnelly JE: Chronologic analysis of variations in retinal damage in two strains of rats after short-term illumination. Invest Ophthalmol Vis Sci 22:252, 1982. 15. Bok D and Hall MD: The role of the pigment epithelium in the etiology of inherited retinal dystrophy in the rat. J Cell Biol 14:73, 1962. 16. Herron WL, Riegel BW, Myer OE, and Rubin ML: Retinal dystrophy in the rata pigment epithelial disease. Invest Ophthalmol 8:595, 1969. 17. Wiegand RD, Giusto NM, Rapp LM, and Anderson RE: Evidence for rod outer segment lipid peroxidation following constant illumination of the rat retina. Invest Ophthalmol Vis Sci 24:1433, 1983. 18. Noell WK: Possible mechanisms of photoreceptor damage by light in mammalian eyes. Vision Res 20:1163, 1980. 19. Dixon RG and Eng LF: Glial fibrillary acidic protein in the retina of the developing albino rat: An immunoperoxidase study of paraffin embedded tissue. J Comp Neurol 195:305, 1981. 20. Eng LF and DeArmand SJ: Immunochemistry of the glial fibrillary acidic protein. Prog Neuropath 5:19, 1983.