Neurotrophic Factor Receptors in Epiretinal Membranes After Human Diabetic Retinopathy

This study was carried out in accordance with the tenets of the Helsinki Declaration. Informed consent was obtained from each patient for the collection of samples. Criteria for inclusion in the study were age <80 years, absence of renal or hematological diseases or uremia, no administration of chemotherapy or life-support measures, and the fewest possible chronic pathologies other than diabetes. The ERMs were surgically removed from consecutive eyes with secondary ERM after PDR (30 eyes) or idiopathic ERM (control subjects, 15 eyes) undergoing pars plana vitrectomy and membrane peeling. Membranes were dissected from the retinal surface with horizontal scissors or membrane pick. Samples derived from 19 of the PDR patients (aged 57 ± 8 years, duration of diabetes 17 ± 8 years) and 15 control subjects (aged 68 ± 7 years) were processed for RT-PCR analysis. The remainder of the PDR samples (from subjects aged 47 ± 18 years, duration of diabetes 13 ± 4 years) were processed for immunohistochemistry. These samples were embedded in optimum cutting temperature compound (Miles Laboratories, Naperile, IL), flash-frozen in liquid nitrogen, and then stored at −80°C.

Total cellular RNA was prepared using Isogen reagent (Nippon Gene, Tokyo) according to the manufacturer’s protocol. Surgical materials were immediately homogenized in 800 μl of isogen, and 200 μl of chloroform was then added. After centrifugation at 4°C, the aqueous phase was collected, and total RNA was precipitated with an equal volume of isopropanol. RNA was then dissolved in 12 μl of water treated with diethyl pyrocarbonate. Next, 0.1 μg of RNA extracted from each sample was reverse-transcribed into first-strand cDNA using the Superscript Preamplification System (Gibco, Paisley, Scotland, U.K.) and oligo-dT primers.

RT-PCR analysis was performed as previously described (25). A volume of 1 μl of each cDNA product from the reverse transcription procedure was used as the template for PCR amplification in a reaction mixture containing PCR buffer (10 mmol/l Tris-HCl, pH 8.3, 50 mmol/l KCl, 1.5 mmol/l MgCl₂), 0.2 mmol/l dNTPs, a 0.2 mmol/l set of oligonucleotide primers, and 2.5 units Taq DNA polymerase in a final volume of 50 μl. Complementary DNA reverse-transcribed from total RNA was amplified by using primers specific for trkA (sense: 5′-GCA TCT GGA GCT CCG TGA TC-3′; antisense: 5′-CTC TGC CCA GCA CGT CAA GT-3′), trkB (sense: 5′-AAG ACC CTG AAG GAT GCC AG-3′; antisense: 5′-AGT AGT CAG TGC TGT ACA CG-3′), trkC (sense: 5′-CTA CAA CCT CAG CCC GAC CA-3′; antisense: 5′-GCT GTA GAC ATC TCT GGA CA-3′), p75^NTR (sense: 5′-TGG ACA GCG TGA CGT TCT CC-3′; antisense: 5′-GAT CTC CTC GCA CTC GGC GT-3′), GFRα1 (sense: 5′-AAG CAC AGC TAC GGG ATG CT-3′; antisense: 5′-GGT CAC ATC TGA GCC ATT GC-3′), GFRα2 (sense: 5′-ACG AGA CCC TCC GCT CTT TG-3′; antisense: 5′-GGG AGG CTT CGT AGA ACT CCT C-3′), Ret (sense: 5′-TGG CAA TTG AAT CCC TTT TT-3′; antisense: 5′-ATG CCA TAG AGT TTG TTT TC-3′), bFGF (sense: 5′-GAG AAG AGC GAC CCT CAC A-3′; antisense: 5′-TAG CTT TCT GCC CAG GTC C-3′), and glyceraldehyde-3-phosphate dehydrogenase (G3PDH) (sense: 5′-ACC ACA GTC CAT GCC ATC AC-3′; antisense: 5′-TCC ACC ACC CTG TTG CTG TA-3′).

PCR was performed after initial denaturation at 94°C for 3 min. Each cycle consisted of a heat-denaturation step at 94°C for 15 s; annealing of primers at either 55°C (bFGF and G3PDH), 56°C (trkB and trkC), 58°C (trkA, p75^NTR, and GFRα2), or 60°C (GFRα1 and Ret) for 2 min; and then polymerization at 72°C for 2 min. The expected sizes of the amplified cDNA fragments of trkA, trkB, trkC, p75^NTR, GFRα1, GFRα2, Ret, bFGF, and G3PDH were 565, 406, 448, 371, 441, 339, 439, 277, and 452 bp, respectively. Negative controls for PCR were performed using “templates” derived from RT reactions lacking either reverse transcriptase or total RNA. After 35 cycles, 15 μl of each reaction mixture were electrophoresed on a 2% Tris borate–EDTA agarose gel and stained with ethidium bromide. In some experiments, human cDNA acquired from skeletal muscle (26) or testis (Takara, Kyoto, Japan) was used as positive control.

Frozen sections (7-μm thick) were cut by a cryostat, mounted on 3-aminopropyltriethoxysilane–coated glass slides, and air-dried at room temperature. For immunohistochemical analysis, the sections were fixed in ice-cold acetone and then washed with PBS. The sections were incubated with normal donkey serum for 30 min to block nonspecific staining. They were stained with a goat polyclonal antibody against GFRα2 (1.0 μg/ml, Santa Cruz Biotechnology, Santa Cruz, CA). For double-labeling immunofluorescence studies, they were then incubated with a mouse monoclonal antibody against vimentin (1×; Zymed, San Francisco, CA) or glutamine synthetase (1.0 μg/ml; Chemicon, Temecula, CA) (27). GFRα2 was visualized with fluorescein isothiocyanate–conjugated donkey anti-goat IgG (200×; Jackson Immunoresearch, West Groove, PA), and vimentin or glutamine synthetase was visualized with Cy3-conjugated donkey anti-mouse IgG (200×; Jackson Immunoresearch). The sections were examined by a confocal laser scanning microscope (Olympus, Tokyo). GFRα2 primary antibody preabsorbed by blocking peptide (Santa Cruz Biotechnology) was used for negative controls. The adjacent section of each specimen was stained with hematoxylin and eosin.

Data are presented as the means ± SEM, except as noted. The χ² test was used to test for significance of the difference between presence of various mRNAs examined in PDR versus control subjects. Statistical significance was accepted at P < 0.05.

Expression of receptors for trophic factors in ERMs was examined by RT-PCR analysis. As shown in Tables 1 and 2, mRNAs encoding high-affinity (trkA, trkB, and trkC) and low-affinity (p75^NTR) NT receptors were expressed at similar levels in both PDR and idiopathic ERM (control) patients. Next, we examined the expression of GDNF receptors (Fig. 1). Surprisingly, GFRα2 mRNA (339 bp) was detected in 17 of 19 (89%) PDR patients, but in only 2 of 15 (13%) control subjects (P < 0.0001). On the other hand, GFRα1 mRNA (441 bp) was detected in 12 of 15 (80%) control subjects, but in only 8 of 19 (42%) PDR patients (P = 0.0258). Ret mRNA (439 bp) was detected in 6 of 19 (32%) PDR patients and 5 of 15 (33%) control subjects (P = 0.9135). We also examined the relationship between GFRα and Ret mRNA expressions in both PDR and control patients. Interestingly, GFRα2-positive and Ret-negative patients were found among 12 of 19 (63%) of those with PDR, but in none (0%) of the control subjects. In addition, GFRα1-positive and Ret-negative subjects were found among 8 of 15 (53%) control subjects, but in only 2 of 19 (11%) PDR patients.

Immunohistochemical analysis was performed to identify the GFRα2 protein expression in ERMs after PDR. GFRα2 protein was detected in all of the 11 samples examined (example in Fig. 2A). The immunoreactivity was completely abolished when primary antibody was preabsorbed with blocking peptide (Fig. 2D). GFRα2 may be involved in the formation of glial components in ERMs (10,28), so we examined whether GFRα2 is coexpressed with the glial cell–specific marker vimentin. ERMs contained a large area composed of glial cells (Fig. 2B), and many cells in that area were double-labeled with GFRα2 (Fig. 2C). Similar results were obtained in experiments using glutamine synthetase, another glial cell marker (data not shown).

Previous studies have shown the high expression level of bFGF in secondary ERMs, coincident with ocular diseases such as PDR and PVR (3,10,16,17,19). We also examined the bFGF mRNAs and detected them in 13 of 19 (68%) PDR patients and 10 of 15 (67%) control subjects (P = 0.9135) (Tables 1 and 2). In 6 of 11 (55%) PDR samples, bFGF immunoreactivity was observed in the large area and partially double-labeled with vimentin or GS (data not shown), as well as GFRα2. Although it was not specific for PDR, we could also determine high expression levels of bFGF mRNA and protein in ERMs.

This study was the first to demonstrate high expression levels of GFRα2 mRNA in ERMs after PDR. Such a high expression level was specific for GFRα2 and was not observed in other NT receptors and Ret. In addition, we showed that the GFRα2 protein is coexpressed with glial cell–specific markers. Together with the recent findings that GDNF functions are not restricted to neurons only, but also implicated in glial differentiation (29), our results suggest the possibility that GFRα2 may be involved in the abnormal glial proliferation that leads to the formation and development of PDR membranes. The mechanism of ERM formation in PDR is still unknown, but its first step is thought to be the neovascularization after the interaction of retinal vessels with the vitreous. In this process, many angiogenic factors (e.g., angiogenin, platelet-derived endothelial cell growth factor, vascular endothelial cell growth factor, and prostaglandins) play an important role (9,30–32). Thus, blocking these angiogenic factors may lead to the inhibition of pathological angiogenesis and ERM formation.

However, ERM is composed of many cell types, such as retinal pigment epithelial cells, glial cells, fibroblasts, myofibroblasts, fibrous astrocytes, macrophages, and endothelial cells (1,5–7). Because these cells surround newly formed vessels, PDR membranes may be stimulated by trophic factors released from vascular components. In turn, because glial cells produce and/or store growth factors, which activate vascular cells (29,33–35), they may further stimulate neovascularization as well as the proliferation of other cell types. Thus, inhibiting the enlargement of the glial component may be one of the promising strategies to prevent the development of ERMs. On the other hand, expression of another GFRα receptor, GFRα1, was found to be significantly higher in idiopathic ERMs. Idiopathic ERMs are thought to be nonangiogenic “fibroglial membrane,” and the glial cells appear to derive from the underlying retina, traversing the inner limiting membrane of the retina through surface breaks (1,5–7). Thus, the retina may be the important endogenous supplier of GDNF and/or NTN in idiopathic ERMs. In fact, though not determined in the human retina, GDNF and NTN are widely localized to the rodent retina (28,36,37). Such a different route and condition of trophic factor supply may upregulate different GFRα receptor subtypes in PDR and idiopathic ERM. Recent studies have shown that GFRα receptors can deliver the intracellular GDNF signaling in the absence of Ret (38). Because GFRα receptors are usually coexpressed with and mediate GDNF binding to Ret (39,40), imbalance in the expressions of these two receptors may represent one of the causes of ERM development.

Finally, although bFGF is reported to play a key role in angiogenesis after PDR and PVR (5,6,8–12), high expression levels of bFGF were detected in both primary and secondary ERMs (Table 1). These results may suggest that bFGF is a general and important player in the formation of preretinal tissues under various conditions (10,11). On the other hand, because high GFRα2 expression levels were specific to PDR membranes (Table 1 and Fig. 1), GFRα2 may be involved in the pathogenesis of ERMs after PDR. If so, inhibiting biological activity of GFRα2 by specific blocker, antisense oligonucleotide, or neutralizing antibodies may be useful in preventing the development of PDR membranes, especially in avoiding membrane recurrence. Further investigations on the precise role of GFRα2 in the process of membrane formation will be needed.

This study was supported by Ministry of Health, Labour and Welfare of Japan; the Ministry of Education, Culture, Sports, Science and Technology of Japan; a Human Frontier Science Program long-term fellowship (LT00170–2001/B to T.H.); and a postdoctoral fellowship from Uehara Memorial Foundation (to C.H.).

We thank Pia Lippincott for critical reading of the manuscript.