Journal of Neuroscience Research 29:549-559 (1 991)
Rapid Communication Epidermal Growth Factor and Transforming Growth Factor a Induce c-fos Gene Expression in Retinal Muller Cells In Vivo S.M. Sagar, R.H. Edwards, and F.R. Sharp Neurology Service, Veterans Affairs Medical Center (S.M.S., F.R.S.) and Departments of Neurology (S.M.S., R.H.E., F.R.S.) and Physiology (F.R.S.). and Hormone Research Institute (R.H.E.), University of California, San Francisco, California
Epidermal growth factor (EGF) and transforming growth factor a (TGFa) are peptides that act at a common receptor and are mitogenic for immature astrocytes and trophic for developing brain neurons in vitro. However, a role for these growth factors in the mature nervous system has not been established. To investigate the actions of EGF and TGFa in the adult central nervous system (CNS) in vivo, the growth factors were injected into the vitreous cavity of adult male rabbits. After varying intervals, the retinas were examined for c-jos mRNA by Northern blot hybridization or Fos (and Fos-related antigen) protein by immunocytochemistry EGF induction of c-fos mRNA occurs within 30 min and persists more than 4 hr. Fos nuclear immunostaining is induced selectively in nuclei of Muller cells by both EGF and TGFa. Fos-like immunoreactivity appears within 1 hr and persists more than 9 hr after EGF injection. These observations demonstrate that mature retinal Muller cells respond to exogenously applied EGF and TGFa in vivo, although the effect of the growth factors is not necessarily direct. The expression of c-fos and other immediate early genes provides a shortterm marker that can be used to investigate the role of growth factors in normal retinal physiology and responses to injury.
.
Key words: growth factor, proto-oncogene, retina, glia
INTRODUCTION Epidermal growth factor (EGF) and transforming growth factor a (TGFa) are peptides that were originally identified as mitogens. They are widely distributed in mammalian tissues (Carpenter et al., 1986; Derynck, 1988) and share a common receptor, which is a transPublished 1991 by Wiley-Liss, Inc.
membrane protein with tyrosine kinase activity. EGFlike immunoreactive material is detectable in mammalian brain, although the chemical identity and function of this material are unknown (Fallon et al., 1984; Schaudies et al., 1989). Material hybridizing with a probe for EGF mRNA on dot blots is present in rat brain (Rall et al., 1985), although at low levels, and in bovine retina (Fassio et al., 1989). TGFa immunoreactive material and mRNA are also present in mammalian brain (Code et al., 1987; Wilcox and Derynck, 1988) and retina (Fassio et al., 1989). In addition, a third peptide that binds to the EGF receptor has been identified in a macrophage-like cell line, although it is not known whether it is present in brain (Higashiyama et al., 1991). Finally, EGF receptor immunoreactivity is present in rat brain (Gomez-Pinilla et al., 1988); and EGF binding sites are present in all layers of bovine neural retina except for the photoreceptor outer segments (Fassio et al., 1989). The biological function of EGF and TGFa in the central nervous system (CNS) is unknown. EGF is mitogenic for brain astrocytes derived from neonatal rodent brain and maintained in cell culture (Leutz and Schachner, 1981; Simpson et al., 1982; Guentert-Lauber and Honegger, 1985; Westphal et al., 1988; Huff et al., 1990); EGF receptor immunoreactivity is present on developing, but not mature, brain astrocytes in vivo (Gomez-Pinilla et al., 1988; Werner et al., 1988). Therefore, EGF or TGFa may regulate glial cell division during development. Although the role for EGF and TGFa in
Received April 24, 1991; revised June 5 , 1991; accepted June 6 , 1991. Address reprint requests to Dr. Stephen M . Sagar, Neurology Service (127), VA Medical Center, 4150 Clement Street, San Francisco, CA 94121. Dr. Edwards is now at the Department of Neurology, University of California, Los Angeles, CA 90024.
550
Sagar et al.
the mature CNS is speculative, they may participate in the reaction to CNS injury (Nieto-Sampedro et al., 1988). In order to seek evidence for a physiological role for EGF and TGFa in mammalian CNS, the peptides were tested for their ability to induce c-fos gene expression in vivo. The c-fos gene, a cellular immediate early gene, is among the first genes expressed in response to a mitogenic signal (Curran and Morgan, 1987). Many immediate early genes code for transcription factors that bind to specific DNA sequences in regulatory regions of target genes and influence their expression. Hence the immediate early gene products are part of the mechanism by which growth factor signals alter gene expression. Moreover, immediate early genes are expressed in response to a variety of extracellular signals in addition to growth factors. These include differentiation factors, hormones, changes in the ionic environment, and, in the case of neurons, neurotransmitters (Sheng and Greenberg, 1990). The best studied immediate early gene is c-fos. In the mature CNS in vivo, c-fos is expressed at a low level in unstimulated animals. However, the gene is rapidly expressed in neurons in response to sudden increases in neuronal activity (Morgan et al., 1987; Hunt et al., 1987; Dragunow and Robertson, 1987; Sagar et al., 1988) and in both neurons and glia in response to injury (Dragunow and Robertson, 1988; Sharp et al., 1989). In vitro, neurons and glia express c-fis in response to growth factors and other stimuli (Arenander et al., 1989; Condorelli et al., 1989; Hisanaga et al., 1990). Induction of c-fos gene expression in response to EGF or TGFa in vivo would strengthen the evidence that c-fos participates in growth factor signaling and would provide evidence of a biological effect of the growth factors. Moreover, the localization of c-fos gene expression might identify target cells of growth factors. Therefore, the effect of intravitreal injections of EGF and TGFa on c-fos gene expression in the retina was examined. These studies were performed in retina to take advantage of its anatomic simplicity and accessibility. In addition, the retina may be exposed to test agents in vivo by intravitreal injection without directly injuring neural tissue.
mals are sedated with pentobarbital, 40-50 mg i.v., and the eyes are topically anesthetized with 1% proparicaine. Injections of growth factors dissolved in phosphate-buffered saline (PBS) (i.e., EGF) or 0.15 M NaCl (i.e., TGFa) are made in a 50-p.1 volume into multiple sites of the vitreal cavity through a 30-gauge needle. Control injections consist of 50 pl of the appropriate vehicle. After varying intervals, rabbits are sacrificed with a lethal dose of pentobarbital, 200-250 mg i.v., followed by bilateral thoracotomy . Posterior eye cups (retina plus attached choroid, sclera and a stump of optic nerve) are immersion fixed for immunocytochemistry . Alternatively, retinas are immediately dissected and frozen on dry ice for extraction of RNA.
Northern Blotting Frozen retinas are each homogenized with a polytron in l-ml 6 M guanidinium isothiocyanate, 50 mM Tris HCl, pH 7.4, 10 mM EDTA, 8% 2-mercaptoethanol; 7 ml 4 M LiCl is added to the homogenate, and nucleic acid is precipitated at 4°C for 48-60 hr. Following centrifugation at 15,OOOg for 30 min, the precipitate is washed in 4 M urea, 2 M LiC1; incubated in 10 mM Tris-HC1, pH 7.4, 1 mM EDTA, 0.2% sodium dodecyl sulfate (SDS), 0.2 mg/ml RNase-free proteinase K at 37°C for 1 hr; frozen and thawed twice to dissolve the RNA; and extracted with phenol-chloroform. Total cellular RNA is precipitated in ethanol and redissolved in 10 mM Tris HCl, pH 7.5, 1 mM EDTA; 5 p g RNA is denatured in 50% formamide, electrophoresed through a 1.5% agaose gel in the presence of 2.2 M formaldehyde, and transferred to a Nylon membrane (Nytran, Schleicher and Schuell, Keene, NH) by capillary blotting. Gels are stained with ethidium bromide prior to transfer, to ensure uniform gel loading. The membrane is hybridized with a full-length [32P]riboprobe synthesized using SP6 polymerase and a plasmid (pSP65 -fos 1A) containing full-length rat c-fos cDNA in reverse orientation downstream of an SP6 promotor. This plasmid was generously provided by Dr. Tom Curran (Roche Institute of Molecular Biology, Nutley, NJ). Hybridization buffer contains 50% formamide, 1.5 X SSPE, 1% SDS, 0.5% nonfat dry milk, 0.2 mg/ml sheared denatured salmon testis DNA, and 0.5 mg/ml yeast RNA. Hybridization is performed at 65"C, and membranes are washed at a maximum stringency of MATERIALS AND METHODS 0.2X SSC, 0.1% SDS at 65°C and exposed to Kodak Animals SB5 X-ray film with an intensifying sceen at -70°C Adult male New Zealand white rabbits are main- overnight. tained on a 12-hr light-dark cycle with free access to food and water. To suppress light-induced c-fos expres- Immunocytochemistry Except as noted, immunocytochemistry was persion in retinal neurons (Sagar and Sharp, 1990), rabbits are maintained in a lighted room overnight prior to formed with an affinity-purified polyclonal rabbit antisegrowth factor injection. For intravitreal injections, ani- rum (RlB6) raised to a synthetic peptide representing
c-fos Expression in Muller Cells
551
E S E S E S E S E S E S
28s
+
18s Fig. 1. Northern blot of 5 pg per lane total cellular RNA extracted from retinas of eyes injected with 0.5 p.g EGF (E lanes) or with phosphate-buffered saline (S lanes). The membrane was probed for c-fos. Animals were sacrificed at the indicated times (min, hr).
residues 132-154 of Fos, a sequence conserved among Fos and Fos-related antigens. This antiserum has been shown to recognize Fos and at least two lower-molecular-weight antigens on Western blots of HeLa cells (Aronin et al., 1990). The actual retinal antigens recognized in the experiments described here are unknown but can be assumed to be a mixture of authentic Fos and antigenically and presumably functionally related nuclear proteins termed Fos-related antigens (Sonnenberg et al., 1989). For immunostaining with RlB6, posterior eye cups are immersed in freshly prepared PLP (0.05 M sodium phosphate buffer, pH 7.4, 0.1 M L-lysine, 0.01 M sodium rn-periodate, 0.5% paraformaldehyde), prepared according to McLean and Nakane (1974), for 4 hr at 4°C. The eye cups are washed in 0.1 M sodium phosphate buffer, pH 7.4 (PB), and the neural retinas are dissected from the pigment epithelium. The whole retinas are incubated at room temperature in PB containing 0.1% bovine serum albumin (BSA), 2% normal goat serum, and 0.2% Triton XlOO (PB-G) for at least 30 min at room temperature. Affinity-purified antiserum R1B6 is added at a final dilution of 1/50 or 1/100, and the retinas are incubated at 4°C on a rotating table at 40 rpm for 72-96 hr. Following washing with PB, sites of antibody binding are visualized using the avidin-biotin-peroxidase procedure (Vectastain Kit, Vector Laboratories, Burlingame, CA) according to the manufacturer’s instructions. Secondary antiserum and ABC reagent are diluted in PB-G, incubation times are 2-2% hr, and diaminobenzidine is used as the chromagen. Where stated, immunocytochemistry was performed with a monoclonal antibody (LA041, Microbiological Associates, Bethesda, MD) raised to a synthetic peptide representing residues 4-17 (with serine at position 15 omitted) of Fos. This sequence is not highly conserved between Fos and known Fos-related antigens,
and immunoprecipitation experiments verify that this antibody is specific for authentic Fos (De Togni et al., 1988). The immunocytochemical protocol is identical to that used for the polyclonal antiserum with the following modifications: posterior eye cups are immersion fixed in 4% paraformaldehyde in PB rather than PLP, and normal horse serum replaces normal goat serum in incubation buffer (PB-H). Ammonium sulfate-fractionated mouse ascites fluid containing the antibody is used at dilutions of 1/3,000-1/50,000, depending on the lot and length of storage. For double-label irnmunocytochemistry of Fos and vimentin, retinas are first reacted with the rabbit polyclonal antibody RIB6 as described above. After washing in PB, the retinas are reacted with antivimentin mouse monoclonal antibody (BioGenex Laboratories, Dublin, CA) diluted 1/500 in PB-H for 16 hr at 4°C. The retinas are then reacted with FITC-conjugated horse antimouse IgG (Sigma Chemical Co., St. Louis, MO) diluted in PB-H, washed in PB , and embedded in either 4% agarose in PI3 (for Vibratome sectioning) or polyacrylamide (Johnson and Blanks, 1984) for cryostat sectioning. Sections 50 pm thick are cut with a Vibratome or 10- or 20-pm thick with a cryostat. Sections are mounted in buffered glycerol, pH 9.0, and examined by both brightfield and epifluorescence illumination. Immunocytochemical controls include inclusion of 1 pg/ml synthetic peptide antigen during incubation of tissue with primary anti-Fos antibodies. This abolishes nuclear staining. For the antivimentin antibody, omission of primary antiserum abolishes Muller cell cytoplasmic staining.
RESULTS Northern blots demonstrate low basal level of c-fos mRNA in light-adapted, saline-injected retinas (Fig. 1).
Fig. 2 . Time course of Fos immunostaining with polyclonal antiserum R l B 6 . The inferior mid-periphery of retinas exposed by intravitreal injection to (A) PBS 1 hr prior to sacrifice or to EGF, 0.5 Fg; (B) 1 hr; (C) 3 hr; and (D) 9 hr prior to sacrifice are pictured. The plane of focus is the inner nuclear layer. Scale bar = 20 h m .
c-fos Expression in Muller Cells
553
Fig. 3. Fos immunocytochemistry with monoclonal antibody LA041. Retinas were examined 3 hr after an intravitreal injection of (A) 0.5 c1.g EGF or (B) PBS. The plane of focus is the inner nuclear layer of the flat-mounted retinas. Scale bar = 20 Fm.
Within 30 min of an intravitreal injection of 0.5 pg EGF, c-fos mRNA levels increase markedly and remain high for at least 2 hr. The EGF-treated retinas have detectable
levels of c-fos mRNA for as long as 8 hr following growth factor injections. The appearance of Fos protein following EGF in-
Fig. 4.
c-fos Expression in Muller Cells
jections is demonstrated immunocytochemically in Figures 2 and 3. Retinas of control, vehicle-injected eyes demonstrate little Fos immunostaining with either antibody (Figs. 2A, 3B). Within 1 hr of an injection of 0.5 pg EGF, Fos nuclear immunostaining is present in a large number of nuclei in the inner nuclear layer (INL) (Fig. 2B). At 3 hr, a similar pattern of immunostaining is seen with both antibodies (Figs. 2C, 3A), although the intensity of staining with the polyclonal antibody is generally greater. With antiserum RlB6, which recognizes both authentic Fos and Fos-related antigens, substantial immunostaining consistently persists for 9 hr (Fig. 2D). Although Fos immunostaining more than 12 hr after EGF injection is inconsistent, in some retinas a small amount of nuclear immunostaining is seen for up to 25 hr (not shown). With the monoclonal antibody LA041 thought to be specific for authentic Fos, Muller cell nuclear immunostaining disappears within 6 hr of the growth factor injections (not shown). The mosaic of immunostained nuclei is seen over much, but not all, of the area of retinal whole mounts. In general, the intensity of staining is greatest in the central retina, but the precise distribution of staining varies between animals and is presumed to represent variability of injection sites and diffusion of growth factor in the vitreal cavity. The Fos-immunostained nuclei in EGF-treated retinas belong to Muller cells, the principal glial cell type of retina. The size and polygonal shape of the immunostained nuclei in flat mounts are consistent with that conclusion. The vast majority of immunostained nuclei are in the center of the INL (Fig. 4), the location of most Muller cell nuclei. Finally, double-label immunocytochemistry demonstrates vimentin-immunoreactive cytoplasm surrounding Fos immunoreactive nuclei (Fig. 4B,C).
Fig. 4. Retinal cross sections. A: 50-k.m thick cross section of a retina immunostained with antibody RIB6 3 hr after the injection of 0.5 p,g EGF. The photoreceptor layer is at the top, and the ganglion cell layer is at the bottom. The immunostained nuclei are all in the inner nuclear layer (INL), the width of which is indicated by the bar. Most immunostained nuclei are in the middle sublayer of the INL. The staining of photoreceptor outer segments is nonspecific. B: Cross section through the periphery of a retina immunostained for Fos 3 hr after the injection of EGF. Four immunostained nuclei in the INL are indicated by the arrows. C: The same field as in B , photographed with epifluorescence illumination. The section was double labeled for Fos and vimentin as described under Materials and Methods. The four Fos-immunoreactive nuclei are surrounded by vimentin-immunoreactive cytoplasm. Scale bars = 20 p,m.
555
Fos-immunostained Muller cell nuclei were seen following the intravitreal injection of EGF at doses as low as 0.02 pg (Fig. 5). With low doses, the intensity of nuclear immunostaining and the retinal area occupied by immunostained nuclei was less than with higher doses. As expected, TGFa, at doses of 5-50 ng, also induces Fos nuclear immunostaining in Muller cell nuclei (Fig. 6).
DISCUSSION EGF is a mitogen for astrocytes derived from embryonic or newborn rat brain and maintained in cell culture (Leutz and Schachner, 1981; Simpson et a]., 1982; Guentert-Leuber and Honegger, 1985; Westphal et al., 1988; Huff et al., 1990). EGF also stimulates cell division in cultured malignant astrocytes (Westphal and Herrmann, 1989), and EGF receptor gene amplification is a frequent feature of glioma cells (Liebermann et al., 1985). Moreover, EGF induces the expression of the c-fus gene in astrocytes in cell culture (Arenander et al., 1989; Condorelli et al., 1989; Hisanaga et al., 1990). Therefore, EGF or a related growth factor may have a role in astrocyte function, at least during development. The present observations extend these findings by demonstrating that exogenously applied EGF and TGFa induce c-fos expression in vivo in adult Muller cells, the major glial cell type of rabbit retina. This expression is reflected in increased levels of both c-fos mRNA, as demonstrated by Northern hybridization, and of Fos protein, as demonstrated by immunocytochemistry . The identity of the Fos-like nuclear antigens expressed in retina in response to EGF and TGFa in not entirely known. At least a portion of the antigen is authentic Fos, as demonstrated by immunocytochemistry with the amino terminally directed monoclonal antibody LA041. Moreover, Northern blots demonstrate specific c-fos mRNA. Therefore, the c-fos gene is expressed in response to EGF administration. Related genes, coding for Fos-related antigens, are probably expressed as well. The persistence of immunostaining with the antiserum RlB6, which recognizes both Fos and Fos-related antigens, for a longer time than staining with the monoclonal antibody LA041, which is specific for authentic Fos, suggests that both Fos and Fos-related antigens are expressed following EGF stimulation (Sonnenberg et al., 1989; Sharp et al., 1991). The identification of the Fos-expressing cells as Muller cells rests on a number of morphologic criteria, including co-localization with vimentin-immunoreactive cytoplasm (see Fig. 4). In rabbit retina, vimentin is a specific Muller cell marker outside the region of the medullary rays (the horizontal band of myelinated fibers leading to the optic disc) (Schnitzner, 1985). In addition,
Fig. 5 .
c-fos Expression in Muller Cells
557
Fig. 6. TGFa-induced Fos nuclear immunostaining. Flat mount of a retina immunostained with antibody RIB6 3 hr after the injection of 0.05 pg TGFa. The pattern of immunostained nuclei in the INL is identical to that seen with EGF. Scale bar = 20 pm.
the cells containing Fos-immunoreactive nuclei have other characteristics of Muller cells, including nuclear size, shape, and location. The ability to identify immunostained cells by these criteria is a major advantage of the retina in these studies. We cannot rigorously exclude the possibility that other retinal cell types, including astrocytes and microglia, may also respond to EGF and TGFa . Fossio et al. (1989) demonstrated the presence of TGFa-immunoreactive material, TGFa mRNA, and EGF binding in bovine retina. The present observations demonstrate a cellular response to exogenous TGFa and EGF in adult rabbit retina. Remaining to be discovered is the biological function of endogenous EGF and TGFa. It has been suggested that an EGF- (or TGFa)-like molecule is involved in the response to brain injury (NietoSampedro et al., 1985; Nieto-Sampedro, 1988). Brain injury is associated with the appearance of EGF receptor immunoreactivity on the suface of astrocytes and the
Fig. 5. EGF dose response. Retinas were immunostained with antibody R1B6 3 hr after intravitreal injections of (A) 0.5 p g EGF, (B) 0.1 pg EGF, (C) 0.02 p g EGF, or (D) PBS. In each case, the inferior central retina is pictured. The retinas pictured in A and B had immunostained nuclei over almost their entire area, whereas the retina pictured in C had immunostained nuclei in only a small area of central retina. The maximum intensity of staining in that retina is shown. Scale bar = 20 pm.
disappearance of a factor that inhibits EGF receptor function (Nieto-Sampedro, 1988). These observations are consistent with the involvement of an EGF-like factor in the astrocytic response to brain injury. EGF, TGFa, or a similar factor may have an analogous role in the retina. Although EGF is mitogenic for immature astrocytes in tissue culture, there is no direct demonstration that EGF or TGFa are mitogenic for adult astrocytes or Muller cells in vivo (Yong et al., 1988). Nevertheless, EGF or TGFa may mediate glial reaction and proliferation following injury to the adult CNS (Nieto-Sampedro et al., 1985). It may well be possible to use c-fos expression as a marker of biological response to test this hypothesis. These experiments in vivo have limitations. First, they do not in themselves demonstrate that EGF and TGFa act directly on Muller cells. As yet, the evidence for a direct action of the growth factors on glia is only circumstantial. The anatomic distribution of EGF binding in bovine retina is consistent with a Muller cell localization, but a localization on multiple neuronai cell types cannot be excluded (Fossio et al., 1989). In dissociated cells from rodent brain, EGF binding is much higher on astrocytes than on neurons (Wang et al., 1989); in rodent brain, EGF receptor-like immunoreactivity is prominant on developing, but not adult, astrocytes (Gomez-Pinilla et al., 1988). Second, the failure to demonstrate Fos nuclear immunostaining in retinal neurons following EGF injections does not exclude a functional role
558
Sagar et al.
for EGF or TGFa in neurons. In fact, there is evidence that EGF is a trophic factor for immature brain neurons (Morrison et al., 1987). Third, the effects of large doses of exogenously applied growth factors may not be identical to the effects of endogenous agents. In this regard, the actual concentrations of growth factors that reach the cellular receptors in these experiments is unknown. Finally, since the function of Fos in the mature nervous system-that is, the identity of its target genes and its effect on the transcription of those genes-is unknown, the role of Fos in Muller cell physiology remains unknown. In particular, since c-fos expression is a general consequence of activation of many intracellular signaling pathways, it cannot be equated with mitosis. In spite of these limitations, c-fos expression by Muller cells in response to EGF and TGFa provides a convenient marker for a biological effect of these growth factors in retina. Although many immediate early genes are expressed in response to growth factor stimulation, the low basal level and rapid time course of c-fos expression make it an experimentally convenient marker of cellular activation. The expression of c-fos and other immediate early genes may therefore be used to elucidate the role of growth factors in the cellular responses to retinal injury. It will be of interest to determine whether EGF and TGFa are mitogenic for mature Muller cells in vitro or in vivo and whether various forms of retinal injury induce c-fos expression in Muller cells.
ACKNOWLEDGMENTS The authors thank Karen Baner and Kathleen J. Hicks for expert technical assistance and Dr. James Morgan and Dr. Tom Curran, Roche Institute of Molecular Biology, Nutley , NJ, for helpful discussions concerning this work and for providing reagents, including the cDNA probe for c-fos. This work was supported by the National Institute of Neurologic Diseases and Stroke (NS 27448 and NS28167) and by the Department of Veterans Affairs Merit Review Program (S.M.S. and F.R.S.). R.H.E. is the recipient of a Clinician-Investigator Development Award from the National Institute of Neurological Diseases and Stroke.
REFERENCES Arenander AT, Lim RW, Varnum BC, Cole R, de Vellis J , Herschman HR (1989):TlS gene expression in cultured rat astrocytes: Induction by mitogens and stellation agents. J Neurosci Res 23:247-256. Aronin N, Sagar SM, Sharp FR, Schwartz WJ (1990): Light induces Fos in the suprachiasmatic nuclei. Proc Natl Acad Sci USA 875959-5962. Carpenter G, Goodman L, Shaver L (1986): The physiology of epi-
dermal growth factor. In Kahn P, Graf T (eds): “Oncogenes and Growth Control.” Berlin: Springer-Verlag, pp 65-69. Chiu R, Boyle WJ, Meek J, Smeal T, Hunter T, Karin M (1988): The c-Fos protein interacts with c-JuniAP-1 to stimulate transcription of AP-1 responsive genes. Cell 54:541-552. Code RA, Seroogy KB, Fallon JH (1987): Some transforming growth factor-alpha connections and their colocalization with enkephalin in the rat central nervous system. Brain Res 421:401-405. Cohen DR, Curran T (1988): fra-1: A serum-inducible, cellular immediate-early gene that encodes a Fos-related antigen. Mol Cell Biol 8:2063-2069. Condorelli DF, Kaczmarek L, Nicoletti F, Arcidiacomo A, Dell’Albani P, lngrao F, Magri G, Malaguarnera L, Avola R, Messina A, Giuffrida Stella AM (1989): Induction of protooncogene fos by extracellular signals in primary glial cell cultures. J Neurosci Res 23:234-239. Curran T (1988): Thefos oncogene. In Reddy EP (ed): “The Oncogene Handbook.” New York: Elsevier, pp 307-326. Curran T, Franza BR Jr (1988): Fos and Jun: The AP-I connection. Cell 55:395-397. Curran T, Morgan JI (1987): Memories offos. BioEssays 7:255-258. De Togni P, Niman H, Raymond V, Sawchenko P, Verma IM (1988): Detection of fos protein during osteogenesis by monoclonal antibodies. Mol Cell Biol 8:2251-2256. Derynck R (1988): Transforming growth factor a. Cell 54593-595. Dragunow M, Robertson HA (1987): Kindling stimulation induces c-fos protein(s) in granule cells of rat dentate gyrus. Nature 329:441-442. Dragunow M, Robertson HA (1988): Brain injury induces c-fos protein(s) in nerve and glial-like cells in adult mammalian brain. Brain Res 455:295-299. Fallon JH, Seroogy KB, Loughlin SE, Morrison RS, Bradshaw RA, Knauer DJ, Cunningham DD (1984): Epidermal growth factor immunoreactive material in the central nervous system: Location and development. Science 224:1107-1109. Fassio JB, Brackman EB, Jumblatt M, Greaton C, Henry JL, Geoghegan TE, Barr C, Schultz GS (1989): Transforming growth factor alpha and its receptor in neural retina. Invest Ophthalmol Vis Sci 30:1916-1922. Gomez-Pinilla F, Knauer DJ, Nieto-Sampedro M (1988): Epidermal growth factor receptor immunoreactivity in rat brain. Development and cellular localization. Brain Res 438:385-390. Guentert-Lauber B, Honegger P (1 985): Responsiveness of astrocytes in serum-free aggregate cultures to epidermal growth factor: Dependence on the cell cycle and the epidermal growth factor concentration. Dev Neurosci 7:286-295. Higashiyama S , Abraham JA, Miller J, Fiddes JC, Klagsbrun M (1991): A heparin-binding growth factor secreted by macrophage-like cells that is related to EGF. Science 251:936-939. Hisanaga K, Sagar SM, Hicks KJ, Swanson RA, Sharp FR (1990): c-fos proto-oncogene expression in astrocytes associated with differentiation or proliferation but not depolarization. Mol Brain Res 8:69-75. Huff KR, Schreier W, Ibric L (1990): Proliferation-related responses in rat astrocytes to epidermal growth factor. Int J Dev Neurosci 8:255-266. Hunt SP, Pini A, Evan G (1987): Induction of c-fos-like protein in spinal cord neurons following sensory stimulation. Nature 328: 632-634. Johnson LV, Blanks JC (1984): Application of acrylamide as an embedding medium in studies of lectin and antibody binding in the vertebrate retina. Curr Eye Res 3:969-974. Leutz A, Schachner M (1981): Epidermal growth factor stimulates
c-fos Expression in Muller Cells DNA-synthesis of astrocytes in primary cerebellar cultures. Cell Tissue Res 220:393-404. Libermann TA, Nusbaum HR, Razon N, Kris R, Lax I, Soreq H, Whittle N, Waterfield MD, Ullrich A, Schlessinger J (1985): Amplification, enhanced expression and possible rearrangement of EGF receptor gene in primary human brain tumors of glial origin. Nature 313:144-147. McLean IW, Nakane PK (1974): Periodate-lysine-paraformaldehyde fixative. A new fixative for immunoelectron microscopy. J Histochem Cytochem 22: 1077-1083. Morgan J1, Cohen DR, Hempstead JL, Curran T (1987): Mapping patterns of c-fos expression in the central nervous system after seizure. Science 237:192-197. Momson RS, Kornblum HI, Leslie FM, Bradshaw RA (1987): Trophic stimulation of cultured neurons from neonatal rat brain by epidermal growth factor. Science 238:72-75. Nieto-Sampedro M (1988): Astrocyte mitogen inhibitor related to epidermal growth factor receptor. Science 240: 1784-1786. Nieto-Sampedro M, Saneto RP, de Vellis Cotman CW (1985): The control of glial populations in brain: Changes in astrocyte mitogenic and morphogenic factors in response to injury. Brain Res 343:320-328. Rail LB, Penschow JD, Niall HD, Coghlan JP (1985): Mouse preproepidermal growth factor synthesis by the kidney and other tissues. Nature 313:228-231. Sagar SM, Sharp FR (1990): Light induces a Fos-like nuclear antigen in retinal neurons. Molec Brain Res 7: 17-21, Sagar SM, Sharp FR, Curran T (1988): Expression of c-fos protein in brain: Metabolic mapping at the cellular level. Science 240: 1328-1331. Schaudies RP, Christian EL, Savage CR Jr (1989): Epidermal growth factor immunoreactive material in the rat brain. Localization and identification of multiple species. J Biol Chem 264: 1044710450. Schnitzer J (1985): Distribution and immunoreactivity of glia in the retina of the rabbit. J Comp Neurol 240:128-142.
559
Sharp FR, Gonzalez MF, Hisanaga K, Mobley WC, Sagar SM (1989): Induction of the c-fos gene product in rat forebrain following cortical lesions and NGF injections. Neurosci Lett 100:117122. Sharp FR, Sagar SM, Hicks K, Hisanaga K (1991): c-fos mRNA, Fos and Fos-related antigen induction by hypertonic saline and stress. J Neurosci (in press). Sheng M, Greenberg ME (1990): The regulation and function of c-fos and other immediate early genes in the nervous system. Neuron 4:477-485. Simpson DL, Momson R, de Vellis J, Henchman HR (1982): Epidermal growth factor binding and mitogenic activity on purified populations of cells from the central nervous system. J Neurosci Res 8:453-462. Sonnenberg JL, Macgregor-Leon PF, Curran T, Morgan JI (1989): Dynamic alterations occur in the levels and composition of transcription factor AP-1 complexes after seizure. Neuron 3: 359-365. Wang S-L, Shiverick KT, Oglivie S, Dunn WA, Raizada MK (1989): Characterization of epidermal growth factor receptors in astrocytic glial and neuronal cells in primary culture. Endocrinology 124:240-247. Werner MH, Nanney LB, Stoscheck CM, King LE (1988): Localization of immunoreactive epidermal growth factor receptors in human nervous system. J Histochem Cytochem 36:8 1-86. Westphal M, Hemnann H-D (1989): Growth factor biology and oncogene activation in human gliomas and their implications for specific therapeutic concepts. Neurosurgery 25:68 1-694. Westphal M, Brunken M, Rohde E, Hermann D-D (1988): Growth factors in cultured human glioma cells: Differential effects of FGF, EGF and PDGF. Cancer Lett 38:283-296. Wilcox JH, Derynck R (1988): Localization of cells synthesizing transforming growth factor-alpha mRNA in the mouse brain. J Neurosci 8: 1901-1904. Yong VW, Kim SU, Pleasure DE (1988): Growth factors for fetal and adult human astrocytes in culture. Brain Res 444:59-66.
Rapid Communication Epidermal Growth Factor and Transforming Growth Factor a Induce c-fos Gene Expression in Retinal Muller Cells In Vivo S.M. Sagar, R.H. Edwards, and F.R. Sharp Neurology Service, Veterans Affairs Medical Center (S.M.S., F.R.S.) and Departments of Neurology (S.M.S., R.H.E., F.R.S.) and Physiology (F.R.S.). and Hormone Research Institute (R.H.E.), University of California, San Francisco, California
Epidermal growth factor (EGF) and transforming growth factor a (TGFa) are peptides that act at a common receptor and are mitogenic for immature astrocytes and trophic for developing brain neurons in vitro. However, a role for these growth factors in the mature nervous system has not been established. To investigate the actions of EGF and TGFa in the adult central nervous system (CNS) in vivo, the growth factors were injected into the vitreous cavity of adult male rabbits. After varying intervals, the retinas were examined for c-jos mRNA by Northern blot hybridization or Fos (and Fos-related antigen) protein by immunocytochemistry EGF induction of c-fos mRNA occurs within 30 min and persists more than 4 hr. Fos nuclear immunostaining is induced selectively in nuclei of Muller cells by both EGF and TGFa. Fos-like immunoreactivity appears within 1 hr and persists more than 9 hr after EGF injection. These observations demonstrate that mature retinal Muller cells respond to exogenously applied EGF and TGFa in vivo, although the effect of the growth factors is not necessarily direct. The expression of c-fos and other immediate early genes provides a shortterm marker that can be used to investigate the role of growth factors in normal retinal physiology and responses to injury.
.
Key words: growth factor, proto-oncogene, retina, glia
INTRODUCTION Epidermal growth factor (EGF) and transforming growth factor a (TGFa) are peptides that were originally identified as mitogens. They are widely distributed in mammalian tissues (Carpenter et al., 1986; Derynck, 1988) and share a common receptor, which is a transPublished 1991 by Wiley-Liss, Inc.
membrane protein with tyrosine kinase activity. EGFlike immunoreactive material is detectable in mammalian brain, although the chemical identity and function of this material are unknown (Fallon et al., 1984; Schaudies et al., 1989). Material hybridizing with a probe for EGF mRNA on dot blots is present in rat brain (Rall et al., 1985), although at low levels, and in bovine retina (Fassio et al., 1989). TGFa immunoreactive material and mRNA are also present in mammalian brain (Code et al., 1987; Wilcox and Derynck, 1988) and retina (Fassio et al., 1989). In addition, a third peptide that binds to the EGF receptor has been identified in a macrophage-like cell line, although it is not known whether it is present in brain (Higashiyama et al., 1991). Finally, EGF receptor immunoreactivity is present in rat brain (Gomez-Pinilla et al., 1988); and EGF binding sites are present in all layers of bovine neural retina except for the photoreceptor outer segments (Fassio et al., 1989). The biological function of EGF and TGFa in the central nervous system (CNS) is unknown. EGF is mitogenic for brain astrocytes derived from neonatal rodent brain and maintained in cell culture (Leutz and Schachner, 1981; Simpson et al., 1982; Guentert-Lauber and Honegger, 1985; Westphal et al., 1988; Huff et al., 1990); EGF receptor immunoreactivity is present on developing, but not mature, brain astrocytes in vivo (Gomez-Pinilla et al., 1988; Werner et al., 1988). Therefore, EGF or TGFa may regulate glial cell division during development. Although the role for EGF and TGFa in
Received April 24, 1991; revised June 5 , 1991; accepted June 6 , 1991. Address reprint requests to Dr. Stephen M . Sagar, Neurology Service (127), VA Medical Center, 4150 Clement Street, San Francisco, CA 94121. Dr. Edwards is now at the Department of Neurology, University of California, Los Angeles, CA 90024.
550
Sagar et al.
the mature CNS is speculative, they may participate in the reaction to CNS injury (Nieto-Sampedro et al., 1988). In order to seek evidence for a physiological role for EGF and TGFa in mammalian CNS, the peptides were tested for their ability to induce c-fos gene expression in vivo. The c-fos gene, a cellular immediate early gene, is among the first genes expressed in response to a mitogenic signal (Curran and Morgan, 1987). Many immediate early genes code for transcription factors that bind to specific DNA sequences in regulatory regions of target genes and influence their expression. Hence the immediate early gene products are part of the mechanism by which growth factor signals alter gene expression. Moreover, immediate early genes are expressed in response to a variety of extracellular signals in addition to growth factors. These include differentiation factors, hormones, changes in the ionic environment, and, in the case of neurons, neurotransmitters (Sheng and Greenberg, 1990). The best studied immediate early gene is c-fos. In the mature CNS in vivo, c-fos is expressed at a low level in unstimulated animals. However, the gene is rapidly expressed in neurons in response to sudden increases in neuronal activity (Morgan et al., 1987; Hunt et al., 1987; Dragunow and Robertson, 1987; Sagar et al., 1988) and in both neurons and glia in response to injury (Dragunow and Robertson, 1988; Sharp et al., 1989). In vitro, neurons and glia express c-fis in response to growth factors and other stimuli (Arenander et al., 1989; Condorelli et al., 1989; Hisanaga et al., 1990). Induction of c-fos gene expression in response to EGF or TGFa in vivo would strengthen the evidence that c-fos participates in growth factor signaling and would provide evidence of a biological effect of the growth factors. Moreover, the localization of c-fos gene expression might identify target cells of growth factors. Therefore, the effect of intravitreal injections of EGF and TGFa on c-fos gene expression in the retina was examined. These studies were performed in retina to take advantage of its anatomic simplicity and accessibility. In addition, the retina may be exposed to test agents in vivo by intravitreal injection without directly injuring neural tissue.
mals are sedated with pentobarbital, 40-50 mg i.v., and the eyes are topically anesthetized with 1% proparicaine. Injections of growth factors dissolved in phosphate-buffered saline (PBS) (i.e., EGF) or 0.15 M NaCl (i.e., TGFa) are made in a 50-p.1 volume into multiple sites of the vitreal cavity through a 30-gauge needle. Control injections consist of 50 pl of the appropriate vehicle. After varying intervals, rabbits are sacrificed with a lethal dose of pentobarbital, 200-250 mg i.v., followed by bilateral thoracotomy . Posterior eye cups (retina plus attached choroid, sclera and a stump of optic nerve) are immersion fixed for immunocytochemistry . Alternatively, retinas are immediately dissected and frozen on dry ice for extraction of RNA.
Northern Blotting Frozen retinas are each homogenized with a polytron in l-ml 6 M guanidinium isothiocyanate, 50 mM Tris HCl, pH 7.4, 10 mM EDTA, 8% 2-mercaptoethanol; 7 ml 4 M LiCl is added to the homogenate, and nucleic acid is precipitated at 4°C for 48-60 hr. Following centrifugation at 15,OOOg for 30 min, the precipitate is washed in 4 M urea, 2 M LiC1; incubated in 10 mM Tris-HC1, pH 7.4, 1 mM EDTA, 0.2% sodium dodecyl sulfate (SDS), 0.2 mg/ml RNase-free proteinase K at 37°C for 1 hr; frozen and thawed twice to dissolve the RNA; and extracted with phenol-chloroform. Total cellular RNA is precipitated in ethanol and redissolved in 10 mM Tris HCl, pH 7.5, 1 mM EDTA; 5 p g RNA is denatured in 50% formamide, electrophoresed through a 1.5% agaose gel in the presence of 2.2 M formaldehyde, and transferred to a Nylon membrane (Nytran, Schleicher and Schuell, Keene, NH) by capillary blotting. Gels are stained with ethidium bromide prior to transfer, to ensure uniform gel loading. The membrane is hybridized with a full-length [32P]riboprobe synthesized using SP6 polymerase and a plasmid (pSP65 -fos 1A) containing full-length rat c-fos cDNA in reverse orientation downstream of an SP6 promotor. This plasmid was generously provided by Dr. Tom Curran (Roche Institute of Molecular Biology, Nutley, NJ). Hybridization buffer contains 50% formamide, 1.5 X SSPE, 1% SDS, 0.5% nonfat dry milk, 0.2 mg/ml sheared denatured salmon testis DNA, and 0.5 mg/ml yeast RNA. Hybridization is performed at 65"C, and membranes are washed at a maximum stringency of MATERIALS AND METHODS 0.2X SSC, 0.1% SDS at 65°C and exposed to Kodak Animals SB5 X-ray film with an intensifying sceen at -70°C Adult male New Zealand white rabbits are main- overnight. tained on a 12-hr light-dark cycle with free access to food and water. To suppress light-induced c-fos expres- Immunocytochemistry Except as noted, immunocytochemistry was persion in retinal neurons (Sagar and Sharp, 1990), rabbits are maintained in a lighted room overnight prior to formed with an affinity-purified polyclonal rabbit antisegrowth factor injection. For intravitreal injections, ani- rum (RlB6) raised to a synthetic peptide representing
c-fos Expression in Muller Cells
551
E S E S E S E S E S E S
28s
+
18s Fig. 1. Northern blot of 5 pg per lane total cellular RNA extracted from retinas of eyes injected with 0.5 p.g EGF (E lanes) or with phosphate-buffered saline (S lanes). The membrane was probed for c-fos. Animals were sacrificed at the indicated times (min, hr).
residues 132-154 of Fos, a sequence conserved among Fos and Fos-related antigens. This antiserum has been shown to recognize Fos and at least two lower-molecular-weight antigens on Western blots of HeLa cells (Aronin et al., 1990). The actual retinal antigens recognized in the experiments described here are unknown but can be assumed to be a mixture of authentic Fos and antigenically and presumably functionally related nuclear proteins termed Fos-related antigens (Sonnenberg et al., 1989). For immunostaining with RlB6, posterior eye cups are immersed in freshly prepared PLP (0.05 M sodium phosphate buffer, pH 7.4, 0.1 M L-lysine, 0.01 M sodium rn-periodate, 0.5% paraformaldehyde), prepared according to McLean and Nakane (1974), for 4 hr at 4°C. The eye cups are washed in 0.1 M sodium phosphate buffer, pH 7.4 (PB), and the neural retinas are dissected from the pigment epithelium. The whole retinas are incubated at room temperature in PB containing 0.1% bovine serum albumin (BSA), 2% normal goat serum, and 0.2% Triton XlOO (PB-G) for at least 30 min at room temperature. Affinity-purified antiserum R1B6 is added at a final dilution of 1/50 or 1/100, and the retinas are incubated at 4°C on a rotating table at 40 rpm for 72-96 hr. Following washing with PB, sites of antibody binding are visualized using the avidin-biotin-peroxidase procedure (Vectastain Kit, Vector Laboratories, Burlingame, CA) according to the manufacturer’s instructions. Secondary antiserum and ABC reagent are diluted in PB-G, incubation times are 2-2% hr, and diaminobenzidine is used as the chromagen. Where stated, immunocytochemistry was performed with a monoclonal antibody (LA041, Microbiological Associates, Bethesda, MD) raised to a synthetic peptide representing residues 4-17 (with serine at position 15 omitted) of Fos. This sequence is not highly conserved between Fos and known Fos-related antigens,
and immunoprecipitation experiments verify that this antibody is specific for authentic Fos (De Togni et al., 1988). The immunocytochemical protocol is identical to that used for the polyclonal antiserum with the following modifications: posterior eye cups are immersion fixed in 4% paraformaldehyde in PB rather than PLP, and normal horse serum replaces normal goat serum in incubation buffer (PB-H). Ammonium sulfate-fractionated mouse ascites fluid containing the antibody is used at dilutions of 1/3,000-1/50,000, depending on the lot and length of storage. For double-label irnmunocytochemistry of Fos and vimentin, retinas are first reacted with the rabbit polyclonal antibody RIB6 as described above. After washing in PB, the retinas are reacted with antivimentin mouse monoclonal antibody (BioGenex Laboratories, Dublin, CA) diluted 1/500 in PB-H for 16 hr at 4°C. The retinas are then reacted with FITC-conjugated horse antimouse IgG (Sigma Chemical Co., St. Louis, MO) diluted in PB-H, washed in PB , and embedded in either 4% agarose in PI3 (for Vibratome sectioning) or polyacrylamide (Johnson and Blanks, 1984) for cryostat sectioning. Sections 50 pm thick are cut with a Vibratome or 10- or 20-pm thick with a cryostat. Sections are mounted in buffered glycerol, pH 9.0, and examined by both brightfield and epifluorescence illumination. Immunocytochemical controls include inclusion of 1 pg/ml synthetic peptide antigen during incubation of tissue with primary anti-Fos antibodies. This abolishes nuclear staining. For the antivimentin antibody, omission of primary antiserum abolishes Muller cell cytoplasmic staining.
RESULTS Northern blots demonstrate low basal level of c-fos mRNA in light-adapted, saline-injected retinas (Fig. 1).
Fig. 2 . Time course of Fos immunostaining with polyclonal antiserum R l B 6 . The inferior mid-periphery of retinas exposed by intravitreal injection to (A) PBS 1 hr prior to sacrifice or to EGF, 0.5 Fg; (B) 1 hr; (C) 3 hr; and (D) 9 hr prior to sacrifice are pictured. The plane of focus is the inner nuclear layer. Scale bar = 20 h m .
c-fos Expression in Muller Cells
553
Fig. 3. Fos immunocytochemistry with monoclonal antibody LA041. Retinas were examined 3 hr after an intravitreal injection of (A) 0.5 c1.g EGF or (B) PBS. The plane of focus is the inner nuclear layer of the flat-mounted retinas. Scale bar = 20 Fm.
Within 30 min of an intravitreal injection of 0.5 pg EGF, c-fos mRNA levels increase markedly and remain high for at least 2 hr. The EGF-treated retinas have detectable
levels of c-fos mRNA for as long as 8 hr following growth factor injections. The appearance of Fos protein following EGF in-
Fig. 4.
c-fos Expression in Muller Cells
jections is demonstrated immunocytochemically in Figures 2 and 3. Retinas of control, vehicle-injected eyes demonstrate little Fos immunostaining with either antibody (Figs. 2A, 3B). Within 1 hr of an injection of 0.5 pg EGF, Fos nuclear immunostaining is present in a large number of nuclei in the inner nuclear layer (INL) (Fig. 2B). At 3 hr, a similar pattern of immunostaining is seen with both antibodies (Figs. 2C, 3A), although the intensity of staining with the polyclonal antibody is generally greater. With antiserum RlB6, which recognizes both authentic Fos and Fos-related antigens, substantial immunostaining consistently persists for 9 hr (Fig. 2D). Although Fos immunostaining more than 12 hr after EGF injection is inconsistent, in some retinas a small amount of nuclear immunostaining is seen for up to 25 hr (not shown). With the monoclonal antibody LA041 thought to be specific for authentic Fos, Muller cell nuclear immunostaining disappears within 6 hr of the growth factor injections (not shown). The mosaic of immunostained nuclei is seen over much, but not all, of the area of retinal whole mounts. In general, the intensity of staining is greatest in the central retina, but the precise distribution of staining varies between animals and is presumed to represent variability of injection sites and diffusion of growth factor in the vitreal cavity. The Fos-immunostained nuclei in EGF-treated retinas belong to Muller cells, the principal glial cell type of retina. The size and polygonal shape of the immunostained nuclei in flat mounts are consistent with that conclusion. The vast majority of immunostained nuclei are in the center of the INL (Fig. 4), the location of most Muller cell nuclei. Finally, double-label immunocytochemistry demonstrates vimentin-immunoreactive cytoplasm surrounding Fos immunoreactive nuclei (Fig. 4B,C).
Fig. 4. Retinal cross sections. A: 50-k.m thick cross section of a retina immunostained with antibody RIB6 3 hr after the injection of 0.5 p,g EGF. The photoreceptor layer is at the top, and the ganglion cell layer is at the bottom. The immunostained nuclei are all in the inner nuclear layer (INL), the width of which is indicated by the bar. Most immunostained nuclei are in the middle sublayer of the INL. The staining of photoreceptor outer segments is nonspecific. B: Cross section through the periphery of a retina immunostained for Fos 3 hr after the injection of EGF. Four immunostained nuclei in the INL are indicated by the arrows. C: The same field as in B , photographed with epifluorescence illumination. The section was double labeled for Fos and vimentin as described under Materials and Methods. The four Fos-immunoreactive nuclei are surrounded by vimentin-immunoreactive cytoplasm. Scale bars = 20 p,m.
555
Fos-immunostained Muller cell nuclei were seen following the intravitreal injection of EGF at doses as low as 0.02 pg (Fig. 5). With low doses, the intensity of nuclear immunostaining and the retinal area occupied by immunostained nuclei was less than with higher doses. As expected, TGFa, at doses of 5-50 ng, also induces Fos nuclear immunostaining in Muller cell nuclei (Fig. 6).
DISCUSSION EGF is a mitogen for astrocytes derived from embryonic or newborn rat brain and maintained in cell culture (Leutz and Schachner, 1981; Simpson et a]., 1982; Guentert-Leuber and Honegger, 1985; Westphal et al., 1988; Huff et al., 1990). EGF also stimulates cell division in cultured malignant astrocytes (Westphal and Herrmann, 1989), and EGF receptor gene amplification is a frequent feature of glioma cells (Liebermann et al., 1985). Moreover, EGF induces the expression of the c-fus gene in astrocytes in cell culture (Arenander et al., 1989; Condorelli et al., 1989; Hisanaga et al., 1990). Therefore, EGF or a related growth factor may have a role in astrocyte function, at least during development. The present observations extend these findings by demonstrating that exogenously applied EGF and TGFa induce c-fos expression in vivo in adult Muller cells, the major glial cell type of rabbit retina. This expression is reflected in increased levels of both c-fos mRNA, as demonstrated by Northern hybridization, and of Fos protein, as demonstrated by immunocytochemistry . The identity of the Fos-like nuclear antigens expressed in retina in response to EGF and TGFa in not entirely known. At least a portion of the antigen is authentic Fos, as demonstrated by immunocytochemistry with the amino terminally directed monoclonal antibody LA041. Moreover, Northern blots demonstrate specific c-fos mRNA. Therefore, the c-fos gene is expressed in response to EGF administration. Related genes, coding for Fos-related antigens, are probably expressed as well. The persistence of immunostaining with the antiserum RlB6, which recognizes both Fos and Fos-related antigens, for a longer time than staining with the monoclonal antibody LA041, which is specific for authentic Fos, suggests that both Fos and Fos-related antigens are expressed following EGF stimulation (Sonnenberg et al., 1989; Sharp et al., 1991). The identification of the Fos-expressing cells as Muller cells rests on a number of morphologic criteria, including co-localization with vimentin-immunoreactive cytoplasm (see Fig. 4). In rabbit retina, vimentin is a specific Muller cell marker outside the region of the medullary rays (the horizontal band of myelinated fibers leading to the optic disc) (Schnitzner, 1985). In addition,
Fig. 5 .
c-fos Expression in Muller Cells
557
Fig. 6. TGFa-induced Fos nuclear immunostaining. Flat mount of a retina immunostained with antibody RIB6 3 hr after the injection of 0.05 pg TGFa. The pattern of immunostained nuclei in the INL is identical to that seen with EGF. Scale bar = 20 pm.
the cells containing Fos-immunoreactive nuclei have other characteristics of Muller cells, including nuclear size, shape, and location. The ability to identify immunostained cells by these criteria is a major advantage of the retina in these studies. We cannot rigorously exclude the possibility that other retinal cell types, including astrocytes and microglia, may also respond to EGF and TGFa . Fossio et al. (1989) demonstrated the presence of TGFa-immunoreactive material, TGFa mRNA, and EGF binding in bovine retina. The present observations demonstrate a cellular response to exogenous TGFa and EGF in adult rabbit retina. Remaining to be discovered is the biological function of endogenous EGF and TGFa. It has been suggested that an EGF- (or TGFa)-like molecule is involved in the response to brain injury (NietoSampedro et al., 1985; Nieto-Sampedro, 1988). Brain injury is associated with the appearance of EGF receptor immunoreactivity on the suface of astrocytes and the
Fig. 5. EGF dose response. Retinas were immunostained with antibody R1B6 3 hr after intravitreal injections of (A) 0.5 p g EGF, (B) 0.1 pg EGF, (C) 0.02 p g EGF, or (D) PBS. In each case, the inferior central retina is pictured. The retinas pictured in A and B had immunostained nuclei over almost their entire area, whereas the retina pictured in C had immunostained nuclei in only a small area of central retina. The maximum intensity of staining in that retina is shown. Scale bar = 20 pm.
disappearance of a factor that inhibits EGF receptor function (Nieto-Sampedro, 1988). These observations are consistent with the involvement of an EGF-like factor in the astrocytic response to brain injury. EGF, TGFa, or a similar factor may have an analogous role in the retina. Although EGF is mitogenic for immature astrocytes in tissue culture, there is no direct demonstration that EGF or TGFa are mitogenic for adult astrocytes or Muller cells in vivo (Yong et al., 1988). Nevertheless, EGF or TGFa may mediate glial reaction and proliferation following injury to the adult CNS (Nieto-Sampedro et al., 1985). It may well be possible to use c-fos expression as a marker of biological response to test this hypothesis. These experiments in vivo have limitations. First, they do not in themselves demonstrate that EGF and TGFa act directly on Muller cells. As yet, the evidence for a direct action of the growth factors on glia is only circumstantial. The anatomic distribution of EGF binding in bovine retina is consistent with a Muller cell localization, but a localization on multiple neuronai cell types cannot be excluded (Fossio et al., 1989). In dissociated cells from rodent brain, EGF binding is much higher on astrocytes than on neurons (Wang et al., 1989); in rodent brain, EGF receptor-like immunoreactivity is prominant on developing, but not adult, astrocytes (Gomez-Pinilla et al., 1988). Second, the failure to demonstrate Fos nuclear immunostaining in retinal neurons following EGF injections does not exclude a functional role
558
Sagar et al.
for EGF or TGFa in neurons. In fact, there is evidence that EGF is a trophic factor for immature brain neurons (Morrison et al., 1987). Third, the effects of large doses of exogenously applied growth factors may not be identical to the effects of endogenous agents. In this regard, the actual concentrations of growth factors that reach the cellular receptors in these experiments is unknown. Finally, since the function of Fos in the mature nervous system-that is, the identity of its target genes and its effect on the transcription of those genes-is unknown, the role of Fos in Muller cell physiology remains unknown. In particular, since c-fos expression is a general consequence of activation of many intracellular signaling pathways, it cannot be equated with mitosis. In spite of these limitations, c-fos expression by Muller cells in response to EGF and TGFa provides a convenient marker for a biological effect of these growth factors in retina. Although many immediate early genes are expressed in response to growth factor stimulation, the low basal level and rapid time course of c-fos expression make it an experimentally convenient marker of cellular activation. The expression of c-fos and other immediate early genes may therefore be used to elucidate the role of growth factors in the cellular responses to retinal injury. It will be of interest to determine whether EGF and TGFa are mitogenic for mature Muller cells in vitro or in vivo and whether various forms of retinal injury induce c-fos expression in Muller cells.
ACKNOWLEDGMENTS The authors thank Karen Baner and Kathleen J. Hicks for expert technical assistance and Dr. James Morgan and Dr. Tom Curran, Roche Institute of Molecular Biology, Nutley , NJ, for helpful discussions concerning this work and for providing reagents, including the cDNA probe for c-fos. This work was supported by the National Institute of Neurologic Diseases and Stroke (NS 27448 and NS28167) and by the Department of Veterans Affairs Merit Review Program (S.M.S. and F.R.S.). R.H.E. is the recipient of a Clinician-Investigator Development Award from the National Institute of Neurological Diseases and Stroke.
REFERENCES Arenander AT, Lim RW, Varnum BC, Cole R, de Vellis J , Herschman HR (1989):TlS gene expression in cultured rat astrocytes: Induction by mitogens and stellation agents. J Neurosci Res 23:247-256. Aronin N, Sagar SM, Sharp FR, Schwartz WJ (1990): Light induces Fos in the suprachiasmatic nuclei. Proc Natl Acad Sci USA 875959-5962. Carpenter G, Goodman L, Shaver L (1986): The physiology of epi-
dermal growth factor. In Kahn P, Graf T (eds): “Oncogenes and Growth Control.” Berlin: Springer-Verlag, pp 65-69. Chiu R, Boyle WJ, Meek J, Smeal T, Hunter T, Karin M (1988): The c-Fos protein interacts with c-JuniAP-1 to stimulate transcription of AP-1 responsive genes. Cell 54:541-552. Code RA, Seroogy KB, Fallon JH (1987): Some transforming growth factor-alpha connections and their colocalization with enkephalin in the rat central nervous system. Brain Res 421:401-405. Cohen DR, Curran T (1988): fra-1: A serum-inducible, cellular immediate-early gene that encodes a Fos-related antigen. Mol Cell Biol 8:2063-2069. Condorelli DF, Kaczmarek L, Nicoletti F, Arcidiacomo A, Dell’Albani P, lngrao F, Magri G, Malaguarnera L, Avola R, Messina A, Giuffrida Stella AM (1989): Induction of protooncogene fos by extracellular signals in primary glial cell cultures. J Neurosci Res 23:234-239. Curran T (1988): Thefos oncogene. In Reddy EP (ed): “The Oncogene Handbook.” New York: Elsevier, pp 307-326. Curran T, Franza BR Jr (1988): Fos and Jun: The AP-I connection. Cell 55:395-397. Curran T, Morgan JI (1987): Memories offos. BioEssays 7:255-258. De Togni P, Niman H, Raymond V, Sawchenko P, Verma IM (1988): Detection of fos protein during osteogenesis by monoclonal antibodies. Mol Cell Biol 8:2251-2256. Derynck R (1988): Transforming growth factor a. Cell 54593-595. Dragunow M, Robertson HA (1987): Kindling stimulation induces c-fos protein(s) in granule cells of rat dentate gyrus. Nature 329:441-442. Dragunow M, Robertson HA (1988): Brain injury induces c-fos protein(s) in nerve and glial-like cells in adult mammalian brain. Brain Res 455:295-299. Fallon JH, Seroogy KB, Loughlin SE, Morrison RS, Bradshaw RA, Knauer DJ, Cunningham DD (1984): Epidermal growth factor immunoreactive material in the central nervous system: Location and development. Science 224:1107-1109. Fassio JB, Brackman EB, Jumblatt M, Greaton C, Henry JL, Geoghegan TE, Barr C, Schultz GS (1989): Transforming growth factor alpha and its receptor in neural retina. Invest Ophthalmol Vis Sci 30:1916-1922. Gomez-Pinilla F, Knauer DJ, Nieto-Sampedro M (1988): Epidermal growth factor receptor immunoreactivity in rat brain. Development and cellular localization. Brain Res 438:385-390. Guentert-Lauber B, Honegger P (1 985): Responsiveness of astrocytes in serum-free aggregate cultures to epidermal growth factor: Dependence on the cell cycle and the epidermal growth factor concentration. Dev Neurosci 7:286-295. Higashiyama S , Abraham JA, Miller J, Fiddes JC, Klagsbrun M (1991): A heparin-binding growth factor secreted by macrophage-like cells that is related to EGF. Science 251:936-939. Hisanaga K, Sagar SM, Hicks KJ, Swanson RA, Sharp FR (1990): c-fos proto-oncogene expression in astrocytes associated with differentiation or proliferation but not depolarization. Mol Brain Res 8:69-75. Huff KR, Schreier W, Ibric L (1990): Proliferation-related responses in rat astrocytes to epidermal growth factor. Int J Dev Neurosci 8:255-266. Hunt SP, Pini A, Evan G (1987): Induction of c-fos-like protein in spinal cord neurons following sensory stimulation. Nature 328: 632-634. Johnson LV, Blanks JC (1984): Application of acrylamide as an embedding medium in studies of lectin and antibody binding in the vertebrate retina. Curr Eye Res 3:969-974. Leutz A, Schachner M (1981): Epidermal growth factor stimulates
c-fos Expression in Muller Cells DNA-synthesis of astrocytes in primary cerebellar cultures. Cell Tissue Res 220:393-404. Libermann TA, Nusbaum HR, Razon N, Kris R, Lax I, Soreq H, Whittle N, Waterfield MD, Ullrich A, Schlessinger J (1985): Amplification, enhanced expression and possible rearrangement of EGF receptor gene in primary human brain tumors of glial origin. Nature 313:144-147. McLean IW, Nakane PK (1974): Periodate-lysine-paraformaldehyde fixative. A new fixative for immunoelectron microscopy. J Histochem Cytochem 22: 1077-1083. Morgan J1, Cohen DR, Hempstead JL, Curran T (1987): Mapping patterns of c-fos expression in the central nervous system after seizure. Science 237:192-197. Momson RS, Kornblum HI, Leslie FM, Bradshaw RA (1987): Trophic stimulation of cultured neurons from neonatal rat brain by epidermal growth factor. Science 238:72-75. Nieto-Sampedro M (1988): Astrocyte mitogen inhibitor related to epidermal growth factor receptor. Science 240: 1784-1786. Nieto-Sampedro M, Saneto RP, de Vellis Cotman CW (1985): The control of glial populations in brain: Changes in astrocyte mitogenic and morphogenic factors in response to injury. Brain Res 343:320-328. Rail LB, Penschow JD, Niall HD, Coghlan JP (1985): Mouse preproepidermal growth factor synthesis by the kidney and other tissues. Nature 313:228-231. Sagar SM, Sharp FR (1990): Light induces a Fos-like nuclear antigen in retinal neurons. Molec Brain Res 7: 17-21, Sagar SM, Sharp FR, Curran T (1988): Expression of c-fos protein in brain: Metabolic mapping at the cellular level. Science 240: 1328-1331. Schaudies RP, Christian EL, Savage CR Jr (1989): Epidermal growth factor immunoreactive material in the rat brain. Localization and identification of multiple species. J Biol Chem 264: 1044710450. Schnitzer J (1985): Distribution and immunoreactivity of glia in the retina of the rabbit. J Comp Neurol 240:128-142.
559
Sharp FR, Gonzalez MF, Hisanaga K, Mobley WC, Sagar SM (1989): Induction of the c-fos gene product in rat forebrain following cortical lesions and NGF injections. Neurosci Lett 100:117122. Sharp FR, Sagar SM, Hicks K, Hisanaga K (1991): c-fos mRNA, Fos and Fos-related antigen induction by hypertonic saline and stress. J Neurosci (in press). Sheng M, Greenberg ME (1990): The regulation and function of c-fos and other immediate early genes in the nervous system. Neuron 4:477-485. Simpson DL, Momson R, de Vellis J, Henchman HR (1982): Epidermal growth factor binding and mitogenic activity on purified populations of cells from the central nervous system. J Neurosci Res 8:453-462. Sonnenberg JL, Macgregor-Leon PF, Curran T, Morgan JI (1989): Dynamic alterations occur in the levels and composition of transcription factor AP-1 complexes after seizure. Neuron 3: 359-365. Wang S-L, Shiverick KT, Oglivie S, Dunn WA, Raizada MK (1989): Characterization of epidermal growth factor receptors in astrocytic glial and neuronal cells in primary culture. Endocrinology 124:240-247. Werner MH, Nanney LB, Stoscheck CM, King LE (1988): Localization of immunoreactive epidermal growth factor receptors in human nervous system. J Histochem Cytochem 36:8 1-86. Westphal M, Hemnann H-D (1989): Growth factor biology and oncogene activation in human gliomas and their implications for specific therapeutic concepts. Neurosurgery 25:68 1-694. Westphal M, Brunken M, Rohde E, Hermann D-D (1988): Growth factors in cultured human glioma cells: Differential effects of FGF, EGF and PDGF. Cancer Lett 38:283-296. Wilcox JH, Derynck R (1988): Localization of cells synthesizing transforming growth factor-alpha mRNA in the mouse brain. J Neurosci 8: 1901-1904. Yong VW, Kim SU, Pleasure DE (1988): Growth factors for fetal and adult human astrocytes in culture. Brain Res 444:59-66.
