Survival of Axotomized Retinal Ganglion Cells in Peripheral Nerve–Grafted Ferrets

Abstract

purpose. Peripheral nerve (PN) grafting to the optic nerve stump stimulates not only axonal regeneration of the axotomized retinal ganglion cells (RGCs) into the grafted PN but also their survival. The purpose of the present study was to determine the number, distribution, and soma diameter of only surviving RGCs without regenerated axons and surviving RGCs with regenerated axons in PN-grafted mammals.

methods. A segment of PN was grafted to the optic nerve stump of adult ferrets. Two months after the PN grafting, surviving RGCs with regenerated axons were retrogradely labeled with granular blue (GB) and stained with RGC-specific antibody C38. Surviving RGCs without regenerated axons were identified as C38-positive cells without GB labeling.

results. Twenty-one percent of RGCs survived axotomy after PN grafting in the area centralis (AC), whereas 47% survived in the peripheral retina. Twenty-six percent of surviving RGCs in the AC exhibited axonal regeneration, which was higher than that in the peripheral retina. Soma diameter histograms revealed that RGCs with regenerated axons showing both GB and C38 positivity were in the large soma diameter ranges. In contrast, the soma diameter distribution of surviving RGCs that did not have regenerated axons showed a peak in the smaller soma diameter ranges.

conclusions. The present data suggest that PN grafting promotes survival of axotomized RGCs more effectively in the peripheral retina than in the AC. Among surviving RGCs, the larger cells exhibited axonal regeneration into the grafted PN, whereas the axons of smaller cells did not to regenerate in either the AC or the peripheral retina.

Injury to neuronal axons in the central nervous system results in degeneration of the somata, and surviving neurons never exhibit axonal regeneration to their targets in mature mammals. ^{1 2} Autologous peripheral nerve (PN) grafting promotes axonal regeneration and survival of central nervous system neurons. ³ The grafted PN supplies central nervous system neurons with some neurotrophic factors such as brain-derived neurotrophic factor, to stimulate axonal regeneration and survival. ⁴  

Retinal ganglion cells (RGCs) provide a good model to study axonal regeneration and functional recovery of central nervous system neurons after PN grafting. Recent studies have demonstrated that RGCs in adult mammals can survive axotomy and exhibit axonal regeneration after PN grafting. ^{3 5 6 7 8} These regenerated axons can make synaptic contacts with target neurons in the visual centers ^{9 10} and even transmit visual information. ^{11 12 13} In these previous reports, only those RGCs with regenerated axons into the grafted PN (R-RGCs) were investigated concerning the recovery of visual functions. 

R-RGCs can be detected easily by retrograde labeling using tracers injected into the grafted PN. The number of R-RGCs represents only a small percentage (1%–5%) in rodents ^{9 14} and cats. ⁸ However, a more RGCs than R-RGCs have been reported to be detectable using stains specific for the retinal tissue after PN grafting. ^{7 15} It could be suggested that there are two types of surviving RGCs in PN-grafted retina; the R-RGCs and surviving RGCs without regenerated axons (S-RGCs). Studies concerning S-RGCs have not been conducted yet, because of the difficulty in detecting the cells. The number of S-RGCs in PN-grafted rats was reported by Villegas–Pérez et al. ¹⁶ They prelabeled RGCs with a fluorescent dye in intact animals and then performed PN grafting to the optic nerve stump. However, the labeling method for intact RGCs occasionally results in mislabeling of nonneuronal cells and displaced amacrine cells because of leakage of dye from degenerated RGCs, ¹⁷ and misidentification of labeled cells as RGCs cannot be completely ruled out. 

To clarify the number and soma distribution pattern of S-RGCs in the retina of PN-grafted mammals, we performed a double-labeling study of RGCs with retrograde labeling by injection of a tracer into the graft and staining with monoclonal antibody C38 that we had previously succeeded in isolating as an RGC-specific marker. ¹⁷ First, we confirmed the specificity of C38 immunoreactivity for intact ferret RGCs. Second, we applied C38 antibody to PN-grafted ferret retina to detect both types of surviving RGCs in combination with a retrograde labeling method for R-RGCs. We show the soma distribution pattern and diameter spectrum of both types of surviving RGCs. Our findings suggest a region-dependent effect of PN grafting on axotomized RGCs in promoting axonal regeneration and/or cell survival. 

Methods

Animals and Transplantation Surgery

Sixteen male sable ferrets (Marshall Farm, North Rose, NY, imported by Charles River Japan, Kanagawa, Japan) weighing 0.8 to 1.5 kg were used in the present study. The treatment of all animals was in strict accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. They were anesthetized with an intramuscular injection of ketamine (0.8 mg/kg body weight) and sodium pentobarbital (0.4 mg/kg body weight). The transplantation surgery was performed according to the methods used for adult cats. ⁸ In brief, the left optic nerve was transected 4 to 6 mm posterior to the eyeball, taking care to avoid injury to the ophthalmic artery. The anterior branch of the left sciatic nerve was excised and sutured autologously to the optic nerve stump with nylon sutures. The other end of the graft was sutured to the temporalis muscle. 

Retrograde Labeling of RGCs with GB

A small piece of gelatin sponge (Spongel, Yamanouchi; Tokyo, Japan) soaked with p-amidinophenyl p-(6-amidino-2-indolyl) phenyl ether (granular blue, GB; Sigma, St. Louis, MO) was implanted in the space behind the optic stump after optic nerve transection in intact animals. For the PN-grafted animals, the Spongel was implanted into the grafted PN at least 10 mm peripheral to the suture to label the R-RGCs (Fig. 1) 2 months after the transplantation surgery. ^{8 15} The ferrets were then killed after 2 days. 

Immunolabeling with C38

C38 antibody is a monoclonal antibody newly isolated in our laboratory as an RGC marker in the flatmounted retinas of rat and cat. ¹⁷ Immunohistochemical methods used in the present study were those used in our previous studies. ^{15 17 18}  

For preparation of retinal flatmounts, the ferrets were perfused transcardially with 4% paraformaldehyde in 0.1 M phosphate buffer. The posterior eyecup was separated from the vitreous body and postfixed with 4% paraformaldehyde solution in phosphate buffer for 1 hour at room temperature. The neural retina was carefully isolated from the retinal pigmented epithelium and incubated overnight with C38 antibody at 4°C. 

For preparation of retinal vertical sections, the enucleated eyeballs were embedded in O.C.T. cryocompound (Tissue-Tek, Miles; Elkhart, IN), and 6- to 8-μm cryosections were cut vertically from the dorsal to the ventral region through the optic disc. After preincubation in 10% normal goat serum, the sections were incubated with C38 antibody for 1 hour at room temperature. C38 immunoreactivities were visualized with fluorescein isothiocyanate–conjugated anti-mouse IgG (Oregon Teknika, Aurora, PA). 

Immunolabeling with C38 antibody for PN-grafted ferret retina was performed using the same staining protocol as that used for the intact retina. 

Determination of Cell Number and Soma Size

Cell counting and measurement of soma diameter were performed in the area centralis (AC) and in an area 2.5-mm dorsal to the AC (the peripheral retina). Labeled cells were counted in an area of 0.29 × 0.19 mm in the AC and 0.58 × 0.19 mm in the peripheral retina by using a stage scanner connected to a computer. The cell density in each area was calculated as labeled cells per square millimeter. Cells that were partly included in the frame were counted on one side and excluded on the other. 

The soma contour of each C38- and/or GB-labeled cell was outlined by camera lucida at ×600 magnification. The soma diameter was expressed as the mean of the maximum diameter and the diameter orthogonal to it and rounded to the nearest whole number. 

Results

Specificity of C38 for the Intact Ferret RGC

When the vertical sections of the intact ferret retina were reacted with C38 antibody, positive immunofluorescence was conspicuous in the ganglion cell layer (GCL) and nerve fiber layer (Fig. 2 A). The labeling pattern was uniform in all retinal regions, from the central to the peripheral retina. Most labeled cells had round or oval-shaped somata. Immunopositive somata that were larger than 20 μm in diameter were observed in the GCL (indicated by the arrows). The labeling was most intense in the cytoplasm, whereas the nucleus showed only weak staining. The morphology, large size, and distribution pattern of the immunopositive somata are characteristic of RGCs. 

When the flatmounted retinas prelabeled retrogradely with GB were reacted with C38 antibody, C38-immunopositive somata were detectable in all retinal regions (Figs. 3 A, 3C). In the region 2.5-mm dorsal to the AC (the peripheral retina; Fig. 3C ), C38-positive somata were distributed more sparsely than in the AC (Fig. 3A) . To confirm that the C38-positive cells were RGCs, we examined their GB positivity in the same frames (Figs. 3B 3D) . The arrows indicate identical somata in both frames. A small number of GB-positive cells in Figure 3B were detectable as C38-negative in Figure 3A . From the microscopic observation, more than 90% of the cells were double labeled with C38 and GB in the intact retina (Fig. 3) . 

C38-positive cells and GB-positive cells were sampled from defined areas of the AC and the peripheral retina of six ferrets (Table 1) . In the AC, 97% of the GB-positive cells were also C38 positive, whereas in the peripheral retina, 105% were C38 positive. The numbers of C38-positive cells were almost equal to those of GB-positive cells in both retinal regions. The density of C38-positive cells was consistent with previous reports of the density of RGCs in the ferret retina. ^{19 20 21} On the basis of double positivity with retrogradely labeling dye, the soma shape of immunopositive cells, and their number, we can clearly identify C38-positive cells as RGCs in the intact ferret retina. 

Detection of C38- and GB+C38-Positive Cells in PN-Grafted Retina

Figure 1 demonstrates our strategy for detecting both types of surviving cells, R-RGCs and S-RGCs, using double-labeling methods. When PN-grafted retinas were reacted with C38 antibody, fewer C38-immunopositive somata were detectable in the GCL than in the intact retina (compare Figs. 4 A, 4C with Figs. 3A 3C , respectively). A micrograph at higher magnification clearly shows that the cytoplasm was intensely stained with C38 antibody similar to that in the intact retina (Fig. 4E) . 

When the specimens were reexamined in the same frame, GB-positive somata were conspicuous in the GCL (Figs. 4B 4D) . The number of GB-positive cells was reduced compared with the number in the intact retina (Figs. 3B 3D) . All the GB-positive somata (indicated by the arrows) were double labeled with C38 antibody. This double-labeling study demonstrated that the number of C38-positive cells was greater than that of the GB-positive cells in the PN-grafted retina both in the AC (Figs. 4A 4B) and the peripheral retina (Figs. 4C 4D) . The arrowheads indicate C38-positive cells without GB labeling in the AC and in the peripheral retina (Figs. 4A 4C) . Some of the GB-positive cells extended GB-positive axons (an example is indicated by the arrow in Figs. 4E 4F ). Some GB-positive axons extended up to the optic disc. From these labeling patterns and the C38 specificity for RGCs, we conclude that GB+C38-positive cells were surviving RGCs with regenerated axons (R-RGCs), and C38-positive cells without GB labeling were surviving RGCs without regenerated axons (S-RGCs). 

The Ratio of S-RGCs and R-RGCs with Intact RGCs

The densities of C38-positive cells and GB+C38-positive cells were determined as cell numbers per millimeter square in both regions (Table 2) . Figure 5 A shows the comparison of the densities of C38-positive cells and GB+C38-positive cells in the AC with those in the peripheral retina in the PN-grafted retina. The densities of C38-positive cells and GB+C38-positive cells were higher in the AC than in the peripheral retina. This indicates that the cell densities of S-RGCs and R-RGCs were higher in the AC than in the peripheral retina. 

Figure 5B shows the percentages of C38-positive cells in the PN-grafted retina against that of intact RGCs. The density of C38-positive cells in both regions as shown in Table 1 was considered as the RGC density in intact retina. Twenty-one percent of RGCs were stained with C38 in the AC, and 47% in the peripheral retina. This indicates that the survival rate of RGCs in the peripheral retina (47%) was higher than in the AC (Table 2) . 

Twenty-six percent of C38-positive cells were double labeled with GB in the AC and 3% in the peripheral retina, respectively. This indicates that 26% of surviving RGCs (S-RGCs + R-RGCs) in the AC exhibited axonal regeneration into the grafted PN and 3% in the peripheral retina. These percentages indicate the axonal regeneration rates among surviving RGCs (S-RGCs + R-RGCs) in each region, which are determined by the proportion of R-RGCs among surviving RGCs. The regeneration rate in the surviving RGCs in the AC was greater than that in the peripheral retina (Table 2) . 

The percentage of GB+C38-positive cells was 5% among intact RGCs in the AC of the PN-grafted retina and 2% in the peripheral retina (Fig. 5B) . The regeneration rates of intact RGCs were similar between the two regions (Table 2) . 

This quantitative study showed that the axonal regeneration rate of surviving RGCs (S-RGCs + R-RGCs) in the AC was higher than that in the peripheral retina. By contrast, in the peripheral retina, the survival rate of RGCs was more than 47% and higher than that in the AC (Fig. 5B , Table 2 ). 

Soma Diameter Distribution of C38-Positive Cells and GB+C38-Positive Cells in PN-Grafted Retina

Figure 6 shows the soma diameter histograms in both retinal regions. The histograms clearly show that the distribution of GB+C38 double-labeled cells was partial to large-sized somata range in both retinal regions. Two peaks are detectable in both regions: One is the peak of C38-positive only cells corresponding to S-RGCs (white bars) ranging from medium- to small-sized somata, and the other is the peak of GB+C38 double-labeled cells corresponding to R-RGCs (black bars) in the large somata range. 

Soma diameter of C38-positive cells without GB labeling (white bars) ranged from 8.3 to 22.5 μm, with a mean diameter (±SD) of 13.0 ± 2.9 μm (n = 162) in the AC (Fig. 6A) , and from 8.8 to 20.1 μm with a mean diameter of 12.5 ± 2.0 μm (n = 148) in the peripheral retina (Fig. 6B) . Soma diameter of GB+C38-positive cells (black bars) ranged from 10.7 to 25.4μ m with a mean diameter of 18.1 ± 2.7 μm (n = 48) in the AC and from 15.2 to 26.2 μm with a mean diameter of 20.7 ± 4.8 μm (n = 5) in the peripheral retina. To define a detailed spectrum of GB+C38-positive somata in the peripheral retina, we sampled from larger areas (1.36 × 0.92 mm) of eight PN-grafted retinas. The GB+C38-positive somata ranged from 13.2 to 26.8 μm with a mean diameter of 20.1 ± 4.0 μm (n = 30) in the larger peripheral retinal areas (Fig. 6C) . 

Both in the AC and the peripheral retina, the soma diameter of GB+C38-positive cells was larger than that of C38-positive cells without GB labeling. We conclude that the soma diameter of R-RGCs is greater than that of S-RGCs in both retinal regions. 

Discussion

Difference in Survival Rates of RGCs between Two Retinal Regions

In the present study, a greater percentage of RGCs in the peripheral retina survived axotomy after PN grafting than in the AC. By contrast, the regeneration rate of surviving RGCs (S-RGCs + R-RGCs) was higher in the AC than in the peripheral retina (Fig. 5B , Table 2 ). This result implies that the PN graft stimulates axonal regeneration more effectively for surviving RGCs in the AC, whereas it supports cell survival of RGCs more effectively in the peripheral region. It is easy to understand that factors secreted from the grafted PN are more effective near the site of operation, the optic disc. The greater number of R-RGCs in the AC than in the peripheral retina can be easily explained by this mechanism. However, this mechanism cannot explain the smaller percentage of S-RGCs in the AC than in the peripheral retina. Other mechanisms probably underlie the enhanced promotion of cell survival in the peripheral retina. 

Berkelaar et al. ²² reported that a greater number of RGCs degenerate when the optic nerve lesion is closer to the eye ball. They speculated that the survival rate of RGCs could be correlated with the length of the optic nerve from the transected site to the soma. Because the present optic nerve transection was performed intraorbitally, RGCs in the peripheral retina had longer optic nerve fibers than those in the AC. The factors secreted from the grafted PN may affect receptors on the intraretinal optic fibers rather than the somata of axotomized cells in promoting cell survival. Histochemical localization of neurotrophin receptor p75 in rat RGCs supports this speculation; it could be suggested that p75 is localized not in the somata but in the optic nerve fibers and/or terminals. ^{23 24} The proportion of S-RGCs may be increased by trophic factor support available from the remainder of the intraretinal optic fiber. 

The expression of C38 antigen may have changed in the axotomized RGCs. There is the possibility that C38-negative S-RGCs reduces the survival rate in the AC. On the other hand, axonal transport may be poorly functioning in regrowing axons of R-RGCs, and consequently GB-negative R-RGCs reduce the regeneration-rate in the peripheral retina. 

In the present study, we investigated survival and axonal regrowth of RGCs in one of the peripheral regions, at 2.5 mm dorsal to the AC. We cannot rule out that the 2.5-mm dorsal region is not a representative region for other peripheral regions. Further detailed analysis must be performed in several peripheral and midperipheral regions. 

Difference in Soma Size between S-RGCs and R-RGCs

The histograms comparing the soma diameters of GB+C38 double-labeled cells with those of only C38-positive cells clearly showed that the soma diameter of R-RGCs was larger than that of S-RGCs in both retinal regions (Fig. 6) . Among all surviving RGCs (S-RGCs + R-RGCs), RGCs with larger soma diameter had regenerated axons, whereas the smaller cells did not exhibit axonal regeneration. Alternatively, somata of the surviving RGCs may have enlarged when their axons regenerated. Soma enlargement was also detectable in RGCs exhibiting axonal regeneration after PN grafting in the adult cat. ⁸ The soma diameter spectrum of R-RGCs shifted to larger size ranges than those of intact RGCs in PN-grafted cat retina. ⁸ This suggests that the somata of RGCs enlarge when the axons of surviving RGCs regenerate after PN grafting. Additionally, the present results suggest that among intact RGCs, large cells are better able to survive axotomy and to regrow their axons. Lucifer yellow or horseradish peroxidase injection into both types of surviving RGCs can reveal the morphologic changes of the somata and dendritic trees and which subtype of RGCs has a better ability to survive axotomy after PN grafting. 

Conclusion

We have reported the region-dependent effect of PN grafting on axonal regeneration and survival of RGCs. Further attempts to enhance axonal regeneration and cell survival, such as injection of neurotrophic factors, should be performed in each of the retinal regions and from the viewpoint of enhancing S-RGC density. ^{25 26 27} The present results suggest that a trial to enhance survival rate in the AC can increase the numbers of R-RGCs in the AC. The contribution to visual transmission by S-RGCs, which is more than that of R-RGCs, could enhance recovery of visual function. Our findings lend encouragement to the challenge of enhancing the number of S-RGCs that can be induced to exhibit axonal regeneration and provide a basis for functional recovery. 

 Supported by grants-in-aid from the Ministry of Science, Education, Culture and Sports of Japan; The Special Coordination Funds for Promoting Science and Technology; the Japan Prevention of Blindness Foundation; and Terumo Life Science Foundation.

 Submitted for publication December 29, 1998; revised March 26, 1999; accepted April 9, 1999.

 Proprietary interest category: N.

 Corresponding author: Jun Kosaka, Department of Anatomy, Okayama University Medical School, 2-5-1 Shikata-cho, Okayama 700-8558, Japan. E-mail: [email protected]

 

Figure 1.

 

Schematic representation of the experimental design to distinguish R-RGCs from S-RGCs. The optic nerve was transected and a segment of autologous left sciatic nerve was sutured to the optic nerve stump. Two months after the surgery, GB was injected into the grafted PN. Two days later, the retinas were labeled with C38 antibody and examined under a fluorescence microscope. Surviving RGCs that did not have regenerated axons (S-RGCs) were only labeled with C38, indicated as (2) C38(+) & GB(−), whereas surviving RGCs with regenerated axons projecting into the PN graft for a long distance up to the GB injection site (R-RGCs) were double-labeled with C38 and GB, indicated as (1) C38(+) & GB(+).

Figure 2.

 

Localization of immunoreactivity with C38 antibody in vertical sections of intact retina. (A) Photomicrograph showing C38 immunoreactivity in the midperipheral region of the intact ferret retina. Immunopositivity was detected in the GCL and the nerve fiber layer. The arrowheads indicate immunopositive somata in the GCL, with diameters larger than 20 μm. (B) Negative control. The vertical section reacted with preimmune mouse serum. The levels of the retinal layers are identical between (A) and (B). GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer. Bar, 50 μm.

Figure 3.

 

Double-labeling study of the intact ferret RGC in flatmounts. (A, C) Fluorescein isothiocyanate staining showing C38-immunoreactive cells in the AC (A) and the region 2.5 mm dorsal to the AC, the peripheral retina (C), focusing on the GCL under a B (Blue)-filter. (B, D) GB labeling showing retrogradely labeled RGCs in the AC (B), and in the peripheral retina (D), focusing on the GCL under a UV-filter. (A) and (B), (C) and (D) are the same frames, respectively. The small arrows indicate identical cells in each frame. The small arrowheads indicate intraretinal nerve fibers of RGCs that are identically detected in (C) and (D). Bar, 50 μm.

Table 1.

 

Number of C38-Positive and GB-Positive Cells in Normal Ferret Retina

	Central	Peripheral
C38-positive cells (cells/mm2)	2360 ± 218	453 ± 25
GB-positive cells (cells/mm2)	2421 ± 185	431 ± 19
C38/GB (%)	97	105

 Cells were labeled with C38 and labeled with GB, and their densities were compared in the area centralis and the peripheral retina. The proportion of all C38-positive cells to GB-positive cells are shown in the bottom row.

Figure 4.

 

Double-labeling study with C38 antibody and GB against PN-grafted axotomized retina. (A, C, E) C38 immunoreactivity. (B, D, F) GB-positive cells. (A, B) In the AC. (C, D) In the peripheral retina. (A) The retina labeled with C38 in the AC focusing on the GCL. The somata in the GCL were strongly stained with C38. (B) The same frame as in (A). The arrows indicate cells double labeled with C38 and GB, whereas the arrowheads indicate C38-positive cells without GB labeling in (A through D). (C) The retina labeled with C38 in the peripheral retina. (D) The same frame as in (C). GB labeling in the peripheral retina. (E, F) Photomicrographs at a higher magnification. A cell with a large soma in the GCL was stained with C38 in (E) and with GB in (F) as indicated by the arrow. The C38 staining showed a mesh-like appearance throughout the cytoplasm with faint staining of the nucleus (E). An intraretinal nerve fiber of the C38-positive cell was also stained with GB in (F). Two cells indicated by the arrowheads in (E) did not show any GB positivity (F). They were S-RGCs. All the GB-positive cells in (B, D, F) were double labeled with C38 antibody (A, C, E). Bar, 50 μm (A through D); 20 μm (E, F).

Table 2.

 

Number of C38-Positive and GB-Positive Cells in PN-Transplanted Ferret Retina

	Central	Peripheral
C38-positive cells (cells/mm2)	489 ± 20	213 ± 21
Survival rate for intact RGCs*	21%	47%
GB-positive cells (cells/mm2), †	129 ± 48	7 ± 3
Regenerating rate for intact RGCs*	5%	2%
GB/C38: regeneration rate for surviving RGCs	26%	3%

*  The proportion for the C38-positive cells in intact retina is shown in Table 1 .

†  All GB-positive cells were double labeled with C38.

Figure 5.

 

Comparison between the AC and the peripheral retina of the densities of C38-positive cells and GB+C38-positive cells in PN-grafted retina. Each cell density is listed in Table 2 . (A) Open bars indicate the normal RGC density shown in Table 1 . The RGC density in the peripheral retina is less than that in the AC. Shaded bars indicate C38-positive cells (surviving RGCs = S-RGCs + R-RGCs) in PN-grafted retina. Black bars indicate GB+C38 double-positive cells (R-RGCs). (B) Proportion of C38-positive cells and GB+C38-positive cells among intact RGCs. Shaded bars indicate C38-positive cells (S-RGCs). Black bars indicate GB+C38-positive cells (R-RGCs). White bars putatively indicate the proportion of degenerated cells derived by subtraction of the S-RGC density from the intact RGC density.

Figure 6.

 

Soma diameter histograms of C38-positive cells and GB+C38-positive cells in PN-grafted retina. (A) In the AC, RGCs were sampled from seven PN-grafted retinas. (B) In the peripheral retina, RGCs were sampled from eight PN-grafted retinas. White bars indicate only C38-positive cells (S-RGCs). Black bars indicate GB+C38-positive cells (R-RGCs). Black bars are distributed in the large soma range in both retinal regions. (C) A larger sampling area for GB+C38-positive cells in the peripheral retina. To define a detailed spectrum showing GB+C38-positive somata in the peripheral retina, we sampled from larger areas (1.36 × 0.92 mm) of eight PN-grafted retinas. The distribution pattern obtained in this larger sampling area also suggests that GB+C38-positive cells (R-RGCs) are distributed in the large soma range in the peripheral region.