c-Jun Expression in Surviving and Regenerating Retinal Ganglion Cells: Effects of Intravitreal Neurotrophic Supply

Abstract

purpose. To investigate c-jun expression in surviving and axon-regenerating retinal ganglion cells (RGCs) and the effect of intravitreal neurotrophic supply on c-jun expression.

methods. All animals underwent optic nerve transection (ONT) 0.5 mm behind the eyeball. Some animals underwent a replacement of the optic nerve with an autologous sciatic nerve graft (SNG) to allow axonal regrowth. To provide a neurotrophic supply, a peripheral nerve (PN) segment or brain-derived neurotrophic factor (BDNF)/ciliary neurotrophic factor (CNTF) was applied intravitreally. The time course of c-jun expression was first examined in both surviving and regenerating RGCs. Then, c-jun expression was examined in surviving and regenerating RGCs 3 weeks after intravitreal BDNF/CNTF treatment. Animals with vehicle eye injection were used as the control. Fluorescent dye was used for retrograde labeling of surviving (applied behind the eyeball) and regenerating (applied at the distal end of the SNG) RGCs. All retinas were immunohistochemically stained for c-jun.

results. c-Jun was not detected in normal RGCs, but weak expression was seen in surviving RGCs after ON injury. The proportion of c-jun–positive (+) RGCs among surviving cell population was 52.6% to 86.5% 2 to 6 weeks after ONT. Among regenerating RGCs, more than 80% expressed c-jun in all treatment groups, a proportion that was significantly higher after CNTF treatment (90.7%). In addition, c-jun expression was much stronger in intensity and the c-jun⁺ nuclei were much larger in regenerating than in surviving RGCs.

conclusions. c-Jun expression in RGCs was upregulated after injury. Most regenerating RGCs were c-jun⁺, and the intensity of c-jun expression was higher in regenerating than in surviving RGCs. CNTF also upregulated c-jun expression in RGCs.

Traumatic injury to the optic nerve (ON) and pathologic conditions such as glaucoma, diabetes, Leber’s congenital ON atrophy, retinitis pigmentosa, and ischemic optic injury are the major causes of blindness. ^{1 2 3 4 5 6 7} In these conditions, many retinal ganglion cells (RGCs) inevitably die by apoptosis and some of the RGCs, although the neurotrauma survive, often fail to regrow their injured axons into their central target. It has been reported that c-jun, one of the immediate early genes, is related to neuronal apoptosis and axonal regeneration. ^{8 9 10 11 12 13} However, when comparing c-jun expression in axon-regenerating RGCs with that in surviving ones, the results are not entirely consistent. Hull and Bähr ⁸ found that there were significantly more c-jun–positive (+) RGCs in the population with axons regenerating into a peripheral nerve (PN) graft 1 month after the surgery (70%) compared with the surviving RGC population 5 days (70%) to 2 weeks (38%) after ON injury. In contrast, Robinson ⁹ has reported very high percentages of c-jun–expressing RGCs in both surviving populations throughout the entire examination period of 1 to 21 days and regenerating population 1 month after surgery. 

Although neurotrophic factors, such as brain-derived neurotrophic factor (BDNF) and ciliary neurotrophic factor (CNTF), have been widely shown to promote survival or axon regeneration of injured RGCs, ^{14 15} no information is available on the effects that these neurotrophic factors exert on c-jun expression in surviving and regenerating RGCs after ON injury. It is well known that axotomized RGCs can regrow their axons into a peripheral nerve (PN) graft that is attached to the transected ON. ^{15 16 17 18} The PN secretes numerous diffusible neurotrophic factors, such as BDNF and CNTF, and thus is capable of enhancing RGC survival and axon regeneration when implanted in the eye. ^{14 17 19} In this study, we first studied the time course of c-jun expression in surviving and regenerating RGCs, and then we investigated the proportion of c-jun⁺ RGCs in surviving and regenerating RGCs with or without intraocular BDNF/CNTF application 3 weeks after surgery. In addition, we examined the intensity of c-jun immunohistochemical staining and the size of c-jun⁺ nuclei under these conditions. These data provide valuable information on the relationship between c-jun expression and axonal regeneration, the effect of neurotrophic factor treatment, and possible specificity in cell morphology of c-jun⁺ RGCs. 

Methods

Young adult (6–8 weeks) golden hamsters were used in the study. Hamsters are often chosen for our studies, because we have systemically investigated the axonal regeneration of RGCs in this rodent species. ^{14 17 19} This study was carried out in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The surgical procedure was also approved by the University of Hong Kong Animal Ethics Committee. 

All animals were anesthetized with pentobarbitone sodium (50 mg/kg body weight, intraperitoneally) for all operations. The left optic nerves in all animals were exposed and transected intraorbitally with a pair of microsurgery scissors. Transection was made approximately 0.5 mm from the posterior pole of the eye, avoiding damage of the blood vessels underneath. In a pilot study, two fluorescent dyes, one gold (Fluorogold; FG; n = 4; Fluorochrome, Englewood, CO) and the other yellow (Diamidino Yellow; DY; n = 4; Sigma-Aldrich, St. Louis, MO), soaked in gelfoam (Pharmacia & Upjohn, Uppsala, Sweden) were applied at the ON stump to label the surviving RGCs immediately after the ON transection. A survival time of 2 days was given, and then the animals were perfused with 4% paraformaldehyde, and the retinas were dissected. To achieve a flat retina on the slide, a cut was made through each quadrant, and the retinas were placed onto gelatin-coated slides. The fluid on the slide was dried with a piece of tissue paper, the retina was flattened by using a fine pen brush and coverslipped with glycerol (Merck, Darmstadt, Germany), and the number of labeled RGCs was counted to compare the labeling efficiency of these two methods. Similar staining efficiency was seen in both methods (Figs. 1A 1B) , thus confirming the efficiency of FG labeling. Note that we have shown that efficiency of DY labeling is high. ²⁰ However, injection difficulties were encountered using DY, as it tended to block the tip of the micropipette. We also wanted to observe the somata morphology of labeled RGCs, but DY only labels nuclei. Thus, FG, which labels cytoplasm of RGCs, was chosen for the remainder of the study. The use of FG for retrograde labeling of RGCs and its efficiency have also been shown in our studies. ^{14 15 17 18}  

All animals were divided into the following paradigms: 

Time-Course of c-Jun Expression

c-Jun expression in RGCs soon (3 hours to 8 days) after ON axotomy has been reported, ^{8 9} and thus the time course of c-jun expression in this study was investigated only at 2, 3, 4, and 6 weeks after ON injury in animals with (1) optic nerve transection (ONT), (2) replacement of the ON with a sciatic nerve graft (SNG), and (3) placement of an SNG+PN segment in eye (SNG+PN). Because no further surgery was performed in the ONT group, this group was used for the c-jun expression study in surviving RGCs. In the other two groups with SNG, a 1-cm segment of autologous sciatic nerve was obtained from the left leg and attached to the ocular stump of the ON to support axon regeneration. The distal end of the PN graft was sutured under the scalp. One of the two groups also received a small segment of peroneal nerve in the eye. After a hole was made in the limbus of the superior part of the left eye, the segment of desheathed PN, 2 mm in length, was inserted into the vitreous with the aid of a micropipette. The hole was then closed with a 10-0 suture. In our previous studies, ^{17 19} no obvious damage was seen in the retina 4 weeks after PN insertion. Occasionally, an inserted PN segment was seen sticking to the peripheral part of the retina at the insertion site. c-Jun expression was studied in regenerating RGCs in these two SNG groups. 

c-Jun Expression 3 Weeks after Intravitreal Neurotrophic Factor Treatment in Animals with ONT Only

For this study, the animals received ONT only and were divided into the following four groups: (1) Animals in the vehicle-injected group (SNG+H₂O; n = 5) received 2 μL of distilled water intravitreally; (2) BDNF-treated animals (n = 6) received intravitreal injections of 5 μg/2 μL BDNF (recombinant human BDNF; Regeneron Pharmaceuticals, Tarrytown, NY); (3) CNTF-treated animals (n = 5) received intravitreal injections of 2 μg/2 μL CNTF (recombinant human CNTF; Regeneron); and (4) BDNF+CNTF-treated animals (n = 6) received intravitreal injections of 5 μg BDNF and 2 μg CNTF in 2 μL. Repeated injections were performed every 5 days thereafter with a total of four injections in each animal. Eye injections were performed at the peripheral part of the retinas using a micropipette. ^{15 17} All animals were allowed to survive for 3 weeks. Two additional intact animals were used for c-jun immunohistochemistry in normal animals. 

c-Jun Expression 3 Weeks after Intravitreal Neurotrophic Factor Treatment in Animals with ONT+SNG

In this part, in addition to ONT, the animals received an attachment of ON with an SNG. The intravitreal BDNF/CNTF interventions and survival time were the same as in the ONT groups. Thus, the animals in the ONT+SNG group were divided into the following four treatment groups: (1) vehicle-injected group (SNG+H₂O; n = 6) and (2) BDNF- (n = 6), (3) CNTF- (n = 7), and (4) BDNF+CNTF-treated (n = 6) animals. 

Fluorescent Tracer FG Application

FG was applied 3 days before the animals were killed. For animals that received ON transection only, FG soaked in gelfoam (Pharmacia Upjohn, Uppsala, Sweden) was applied to the newly cut ocular stump to label the surviving RGCs. For animals in the groups with SNG, 1 μL of 2% FG (dissolved in 0.9% sodium chloride and 0.01% Triton X-100) was injected into the distal end of the graft 2 to 3 mm away from the ligated end through a micropipette, to label the RGCs that had regrown their axons into the graft. Two additional animals received injection of FG into the SNG immediately after surgery, and the animals were perfused 3 days later to check whether the injection method would allow the dye to diffuse to the ON head and label surviving RGCs nonspecifically. Note that RGCs start to regrow their injured axons into an SNG only 4 to 5 days after ON injury. ²¹  

Immunohistochemistry

The eyes with multiple injections showed no obvious damage or infections. All animals were killed with an overdose of pentobarbital sodium. Retinas were dissected and fixed in 4% paraformaldehyde in 0.1 M Tris buffer for 45 minutes. After washes in Tris buffer and blocking in 5% goat serum medium for 30 minutes, the retinas were incubated with primary antibody raised in rabbit against c-jun/activator protein (AP)-1 protein (Oncogene, Manhasset, NY) overnight at 4°C. The retinas were washed three times with PBS (10 minutes each), and further immunohistochemical staining was performed by incubating the retinas with FITC-conjugated goat anti-rabbit secondary antibody (Vector Laboratories, Burlingame, CA) for 60 minutes at room temperature. c-Jun–expressing cells in the retinas were observed under a fluorescence microscope (DMR; Leica, Heidelberg, Germany). 

Data Collection and Analysis

The number of stained cells in the retinas was estimated with a systematic sampling method at 40× magnification. In each quadrant of the retina, labeled cells were counted within a grid (area, 200 × 200 μm) moving from the optic disc at 500-μm intervals toward the peripheral retina. The areas of the retinas were measured using a computerized digital tablet, and the number of labeled cells was obtained by multiplying the mean density in each retina by the size of the retinal area. In each group, the number of FG-labeled RGCs, c-jun⁺ cells, and double-labeled RGCs was counted. Furthermore, the cell size of FG-labeled RGCs and the area of c-jun⁺ RGC nuclei in randomly selected parts (area, 200 × 200 μm each) of the retinas were measured with an image-analysis system (Neurolucida; MicroBrightField, Colchester, VT) at 40× magnification. The nuclei measurement in the ONT group was obtained from surviving RGCs, whereas in the SNG groups, they were obtained from regenerating RGCs. For statistical comparison, Bonferroni’s test was used to compare intragroup statistical differences after one-way analysis of variance (ANOVA). 

Results

The eyes with multiple injections showed no obvious damage or infection. No FG-labeled RGCs were detected in the two retinas of the animals with 3-day survival time, in which the FG was applied into the SNG immediately after transplantation, indicating that there was no diffusion of the dye to the optic disc. Similar staining efficiency and the number of labeled RGCs were obtained in the FG- and DY-labeling methods (Figs. 1A 1B , respectively), indicating the effectiveness of FG staining in this study. 

Time Course of c-Jun Expression

Average Number of Surviving RGCs at Different Time Points after ONT and the Average Proportion of c-Jun⁺ RGCs.

The number of surviving RGCs dramatically decreased after ONT. The average number ± standard deviation of surviving RGCs 2 to 6 weeks after ONT is shown in Figure 2A . Briefly, there was a mean of 11,397 ± 1,187 surviving RGCs per retina 2 weeks after ONT. This number decreased to 7064 ± 2891 per retina at 3 weeks after ONT, 4986 ± 2914 per retina 4 weeks after ONT and 4043 ± 945 per retina 6 weeks after ONT. Note that the average number of RGCs in normal adult hamsters is 65,890 ± 5,684 per retina ²² (Fig. 2A) . Thus, more than 80% of RGCs died by 2 weeks after the ON injury, and by 6 weeks, nearly 94% were lost (Fig. 2A) . 

In normal retinas from intact eyes, no c-jun⁺ RGCs were observed. c-Jun staining was restricted to the nuclei of the cells, and the cytoplasm or dendrites were not labeled for c-jun (Figs. 1D 1F 1H) . c-Jun expression was significantly upregulated 2 weeks after ONT, and counting of FG- and double-labeled RGCs revealed that the average proportion of c-jun⁺ RGCs was 64.2% ± 2.27% (n = 4, Fig. 2B ) among the surviving RGC population. This result indicates that c-jun expression is significantly upregulated by ON injury, thus confirming previous reports. ^{10 23} Three weeks after ONT, the proportion of c-jun⁺ RGCs increased to 86.5% ± 4.87% (n = 4; Figs. 1C 1D 2B ) in the surviving RGC population. This percentage is significantly (P < 0.01) higher than the 2-week value. The proportion decreased slightly to 70.3% ± 12.1% (n = 5) at 4 weeks after ONT, and it is not significantly different from the percentages at 2 or 3 weeks after ONT. The proportion of c-jun⁺ RGCs in surviving RGCs decreased further to 52.6% ± 8.94% (n = 4). This value is not significantly different from that observed at the 2-week time point, but it is significantly lower than the percentages at the 3-week or 4- week time points (P < 0.001 and P < 0.05, respectively; Fig. 2B ). 

The Average Proportion of c-Jun⁺ RGCs in the Regenerating RGC Population.

Two weeks after ONT and an attachment of an SNG, the average proportion of c-jun⁺ RGCs in the regenerating RGC population increased to 85.8% ± 8.23% (n = 6) in the SNG group and 84.4% ± 7.3% (n = 5) in the group with SNG plus intravitreal implant of the PN segment (Fig. 3) . These two percentages are not significantly different from each other, but are both significantly (P < 0.01) higher than that in surviving RGCs at the same time point. 

Three weeks after ON injury, the proportion stayed at similar high levels in the regenerating RGC population in the SNG (81.5% ± 7.99%; n = 5; Figs. 1E 1F 3 ) and SNG+PN (90.5% ± 1.94%; n = 5; Figs. 1G 1H 3 ) groups. There is no significant difference between these two values at this time point (Fig. 3) , and they are not significantly different from the value in surviving RGCs at that time point (Figs. 2B 3) . 

At 4 weeks after ON injury, the average proportion of regenerating RGCs that expressed c-jun remained at high levels in the SNG (95.2% ± 4.08%; n = 6) and SNG+PN (93.6% ± 2.99%; n = 6) groups (Fig. 3) . These two percentages are significantly higher (P < 0.05) than that in the surviving RGCs at the same time interval (Figs. 2B 3) . In addition, the proportion in the SNG group at this time interval is significantly (P < 0.01) higher than that at 3 weeks. 

Six weeks after ONT, most of the regenerating RGCs in the SNG (88.8% ± 6.82%; n = 5) and SNG+PN (93.3% ± 2.34%; n = 5) groups were still c-jun⁺ (Fig. 3) . These data show that surviving RGCs reached the peak of c-jun expression at 3 weeks after ON injury, whereas regenerating RGCs stayed at a high c-jun⁺ proportion throughout the 6-week examination period (Figs. 2B 3) , indicating a close association of c-jun expression and axon regeneration. 

Intensity of Staining and Morphology of c-Jun⁺ Cells.

Apart from analysis of the percentage of c-jun⁺ RGCs in surviving and regenerating RGC populations, we also examined the intensity of c-jun immunohistochemical staining and the morphology of c-jun⁺ nuclei. Overall, c-jun expression was weak in surviving RGCs (Fig. 1D) , whereas it was markedly stronger in the regenerating RGCs (Figs. 1F 1H) . Most c-jun⁺ nuclei in the SNG and SNG+PN groups were round with a smooth boundary (Figs. 1F 1H) , whereas those in the ONT group were small, with some round and some irregularly shaped (Fig. 1D) . Of interest, the nuclei of 32 strongly c-jun⁺ regenerating RGCs randomly selected in the SNG group were 135.4 ± 31.9 μm² on average, and 29 c-jun⁺ regenerating RGCs in the SNG+PN group had an average nucleus size of 156.6 ± 23.5 μm². In contrast, the nuclei of 29 randomly selected surviving RGCs in the ONT group were of a much smaller size (73.9 ± 14.4 μm²). The difference in nucleus size between surviving and regenerating c-jun⁺ RGCs is statistically significant (P < 0.05). 

c-Jun Expression 3 Weeks after Neurotrophic Factor Treatment in Animals with ONT Only

In this part of the study, no SNG was attached to the ON stump, thus c-jun expression was examined only in surviving RGCs. The proportion of c-jun⁺ RGCs was 79.6% ± 6.83% in the vehicle-injected group (Fig. 4A) . This is similar to what was obtained (86.5% ± 4.87%) in a time course study in ONT-only (without eye injection) animals at the same time point, confirming that ON injury caused upregulation of c-jun expression. The proportion of c-jun⁺ surviving RGCs increased to 90.7% ± 2.04% in CNTF-treated and 86.2% ± 8.7% in BDNF-treated eyes (Fig. 4A) . The former is significantly higher than the vehicle control. Thus, CNTF upregulated c-jun expression in surviving RGCs (Fig. 4A) . Combined injection of BDNF and CNTF did not further increase the percentage (87.4% ± 7.37%) of surviving RGCs that expressed c-jun (Fig. 4A) . 

Measurement of the sizes of FG-labeled surviving RGC somata and c-jun⁺ RGC nuclei revealed that intravitreal CNTF treatment significantly (P < 0.05) increased both sizes compared with normal and vehicle controls, with the increase in RGC soma size the more obvious (Fig. 5) . Treatment of BDNF also significantly (P < 0.05) increased these sizes but the magnitudes were not as high as after CNTF treatment (Fig. 5) . No synergistic effect was seen after combined CNTF and BDNF application (Fig. 5) . Detailed analysis showed that about half or more of the surviving RGCs had a large soma size of more than 400 μm² after CNTF or CNTF+BDNF treatment whereas most of the surviving RGCs had smaller somata (between 100 and 250 μm²) with BDNF treatment (Fig. 6A) . Most c-jun⁺ RGC nuclei in surviving RGCs were 50 to 100 μm² in the vehicle control group. However, a substantial increase in the range to 125 to 150 μm² and a dramatic decrease in the range to 50 to 75 μm² were seen in CNTF- and CNTF/BDNF-treated animals, respectively (Fig. 6B) . 

c-Jun Expression 3 Weeks after Intravitreal Neurotrophic Factor Treatment in Animals with ONT plus SNG

The proportions of c-jun⁺ RGCs were similarly high in regenerating RGCs in all groups (with or without intravitreal neurotrophic factor treatment), ranging from 94.3% to 97.5% (Fig. 4B) . Thus, most of regenerating RGCs expressed c-jun. No significant difference was detected among these groups. Coupled with observations in the c-jun time-course study, these data suggest that c-jun, either as an effector or a consequence of the regenerative response, is closely associated with axonal regeneration in the visual system. 

Besides the differences in the proportion of c-jun⁺ RGCs between surviving and regenerating RGC populations after neurotrophic factor treatment, we noticed that the c-jun staining intensity was also different between them. Figure 7 shows the characteristics of FG-labeled RGCs and c-jun⁺ RGCs under these conditions. 

The c-jun⁺ nuclei of regenerating RGCs were round in shape with a smooth boundary, and the staining intensity was very strong. In surviving RGCs, the shape of c-jun⁺ nuclei varied substantially among different treatment groups. They were generally of smaller size and round or irregular in shape. Similar observations were obtained in animals that underwent ONT in the time-course study (Fig. 1) . 

Discussion

In this study, the expression of c-jun in surviving RGCs was upregulated after ON injury, whereas no c-jun expression was observed in normal RGCs. This expression was further upregulated in regenerating RGCs (Fig. 7) . Both c-jun expression intensity and the proportion of c-jun⁺ RGCs in regenerating the RGC population were higher. In addition, the average c-jun⁺ nucleus was much larger in regenerating than in surviving RGCs. CNTF, which was shown to enhance RGC axon regeneration, ¹⁷ also upregulated c-jun expression in surviving RGCs. The upregulation of c-jun expression in the regenerating RGC population suggests that c-jun is closely associated with axonal regeneration. 

c-Jun and RGC Axonal Regeneration and Neuronal Survival

Results in studies have suggested that c-jun may play a role in neurite outgrowth, as c-jun overexpression or inhibition of the c-jun pathway has been shown to enhance neurite outgrowth in vitro. ^{24 25} In this study, we found that 75.4% to 79% of surviving RGCs expressed c-jun 3 to 4 weeks after axotomy. However, in regenerating RGCs, there was a much higher proportion of c-jun⁺ RGCs (>93%). Similar close correlation of c-jun expression and axonal regeneration was also found in the goldfish. ²³ We also found that the immunoreactivity of c-jun was stronger and the size of c-jun⁺ nuclei was larger in regenerating than in surviving RGCs. These data indicate that the production of c-jun was increased in regenerating RGCs, another indication of a close association between c-jun and axonal regeneration. 

CNTF promoted RGC axon regeneration in our previous studies ^{14 15 17} and upregulated the expression of c-jun in surviving RGCs in this study. There are at least two distinct pathways that can mediate the signaling of CNTF. One is Janus kinase (JAK) and signal transducer and activator of transcription (STAT) pathway, and the other is the Ras and mitogen-activated protein kinase (MAPK) pathway. ²⁶ c-Jun is involved in both pathways. ²⁶ It has been reported that direct interaction between c-jun and a STAT protein (STAT3) results in synergistic transcriptional activation. ²⁷ It is thus possible that c-jun potentiates the promoting effect of CNTF on axon regrowth by interacting with its signal-transduction pathways. 

It has been shown that long-lasting c-jun is coexpressed with growth-associated protein (GAP)-43 in regenerating neurons in the thalamus and cerebellum and correlates well with the axonal regrowth potential. ¹¹ GAP-43 plays an important role in axonal regeneration. ²⁸ The coexpression of c-jun and GAP-43 has also been found in RGCs in rats. ²⁹ The highly conserved c-jun site, activator protein (AP)-1, has been detected in the promoter region of the GAP-43 gene, and the AP-1–containing promoter can strongly regulate the transcription of GAP-43. ^{30 31} These data suggest that c-jun is involved in regulating the expression of GAP-43. It has also been suggested that c-jun may play multiple roles in neuronal survival and axonal regeneration. The manifestation depends on which partners it uses to form an AP-1 complex and on its phosphorylation state and activity. ³⁰  

The significant difference in nucleus size between c-jun⁺ surviving and regenerating RGCs indicates the different expression level among the surviving and regenerating RGCs. It may also suggest that a specific RGC population has a preferentially expressed c-jun, better survived the ON injury and is more capable of regenerating their injured axons. It is interesting to see that CNTF and BDNF, to a lesser extent, increased the size of RGC somata and c-jun⁺ nuclei. Both CNTF and BDNF can enhance axon and neurite growth in RGCs. Whereas regenerating RGCs tend to have larger somata, CNTF and BDNF have different effects on RGC axonal regeneration into a PN graft. CNTF promotes axon regeneration into PN grafts, whereas BDNF does not. ¹⁷ However, BDNF enhances neurite outgrowth within the retina. ¹⁵ Although we also observed an increase in the total number of c-jun⁺ RGCs in BDNF-treated rats, the percentage of c-jun RGCs remained unaffected, and the increase may merely be a reflection of more RGCs that had survived the injury after BDNF treatment. 

c-Jun has been reported to involve in neuronal survival and axon growth in adults ³² and, although controversial, ³³ during development. ³⁴ c-Jun expression in RGCs can be induced by axotomy and is coincident with the apoptotic loss of RGCs. ⁹ Recently, c-jun has been shown to be directly involved in ON-lesion–induced apoptosis of RGCs, ¹³ and inhibition of c-jun activity reduces neurotoxicity-induced loss of nigrostriatal dopaminergic neurons. ³⁵ Therefore c-jun has been hypothesized to act as a death signal in degenerating neurons. However, c-jun overexpression fails to affect adult Purkinje cell survival after axotomy in vivo. ³⁶ Recently, it has been shown that the concentration of c-jun is not critical for neuronal survival but rather the concentration of active (phosphorylated) c-jun. ³⁷  

Because phosphorylated c-jun plays a critical role in neuronal survival, it will be interesting to see what the level of phosphorylated c-jun is in surviving and regenerating RGCs after ON injury. However, studies in whole retinas using conventional methods such as PCR and Western blot may not accurately reflect what occurs in RGCs, ³⁸ because RGCs account for only a small proportion among the total retinal neuronal populations. The exact role of c-jun in axonal regeneration is still unclear. Whereas c-jun overexpression promotes neurite outgrowth in vitro, it fails to do so in vivo. ³⁶ We are currently investigating its role in axonal regeneration in vivo and are trying to clarify the mechanism, if it does play a part in promoting axon regeneration. 

²  Contributed equally to the work and therefore should be considered equivalent senior authors.

 Supported by a grant from the Research Grant Council of Hong Kong and awards from the University of Hong Kong and Shantou University Medical College.

 Submitted for publication May 6, 2003; revised July 18, 2003; accepted August 20, 2003.

 Disclosure: Q. Lu, None; Q. Cui, None; H.K. Yip, None; K.-F. So, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement ” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 8.

 

No caption available.

Figure 1.

 

Fluorescence photomicrographs showing (A) FG- and (B) DY-labeled RGCs in normal retinas. Three weeks after ON injury (C) shows FG-labeled surviving RGCs, and (D) shows c-jun expression in the same field. There were many double-labeled RGCs (arrows), but often surviving RGCs were observed that were not c-jun⁺ (C, D, arrowheads). After SNG attachment, (E) shows FG-labeled surviving RGCs, whereas (F) shows c-jun expression in the same field. After SNG attachment plus placement of a PN segment in the vitreous, (G) shows FG-labeled surviving RGCs, and (H) shows c-jun expression in the same field. Most regenerating RGCs were c-jun⁺, but c-jun⁺ nuclei were sometimes present that were not regenerating RGC (E, F, arrowheads). Compared with surviving RGCs (D), the intensity of c-jun expression increased in regenerating RGCs (F, H), in which the intensity was slightly stronger after intravitreal insertion of the PN segment. Scale bar, 50 μm.

Figure 2.

 

Average number of surviving RGCs at different time points after ONT (A) and the average proportion of c-jun⁺ RGCs among them (B). Data indicated by empty triangle were published elsewhere. ²² Error bars, SD. Significance levels *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 3.

 

Average proportion and SD of c-jun⁺ RGCs in the regenerating population, with or without intravitreal insertion of a PN segment at different weeks after ONT. Most regenerating RGCs expressed c-jun throughout the examination period.

Figure 4.

 

Average proportion and SD of c-jun⁺ RGCs in surviving (A) and regenerating (B) populations after intravitreal applications of CNTF and/or BDNF 3 weeks after surgery. A significant increase (P < 0.05) in c-jun⁺ RGCs was seen in the surviving population after intravitreal injection of CNTF, indicating that CNTF upregulated c-jun expression.

Figure 5.

 

Average size and SD of FG-labeled RGC somata and c-jun⁺ nuclei 3 weeks after treatment of CNTF and/or BDNF. Comparisons were made against both normal and vehicle injection controls. The size of FG-labeled RGC somata or c-jun⁺ nuclei was significantly higher (*P < 0.05, Bonferroni test) after intravitreal CNTF and CNTF+BDNF treatments.

Figure 6.

 

Detailed analysis of distribution of FG-labeled RGC soma and c-jun⁺ RGC nuclear size in the surviving RGC population 3 weeks after ONT and intravitreal BDNF/CNTF intervention.

Figure 7.

 

Fluorescence photomicrographs showing FG-labeled regenerating RGCs (A, C, E, G) and c-jun⁺ nuclei (B, D, F, H). Each row represents the same fields under the various experimental conditions. Most regenerating RGCs expressed c-jun. Compared with surviving RGCs (Fig. 1D) and regenerating RGCs in vehicle-treated animals (B), the intensity of c-jun expression was higher in regenerating RGCs after neurotrophic factor treatments (D, F), especially after combined application of CNTF and BDNF (H). Scale bar, 50 μm.