Heme Oxygenase-1 Induced in Müller Cells Plays a Protective Role in Retinal Ischemia–Reperfusion Injury in Rats

Abstract

purpose. To investigate the protective roles played by heme oxygenase (HO)-1 and -2 in the rat retina after ischemia–reperfusion injury.

methods. Retinal ischemia was induced in rats by increasing the intraocular pressure to 110 mm Hg for 60 minutes. The expression of HO-1 and -2 in the retina was determined by Western blot, real-time polymerase chain reaction (PCR), and immunohistochemistry. To inhibit the upregulation of HO-1, short interfering (si)RNA of HO-1 was injected intravitreally before ischemia and that of green fluorescent protein (GFP) was used as the control. Müller cell damage was assessed by counting the number of S-100–positive cells. The number of macrophages invading the retina was determined by counting the number of ED-1–positive cells.

results. The expression of HO-1 mRNA and protein was upregulated at 6 hours after reperfusion and peaked at 12 to 24 hours, whereas that of HO-2 was not altered. HO-1 immunoreactivities were detected in Müller cells at 24 hours after reperfusion, and HO-2 immunoreactivities were detected in retinal cells. The HO-1 expression in the retina treated with siRNA of HO-1 was reduced at 12 and 24 hours after reperfusion compared with that injected with siRNA of GFP. The number of S-100–positive cells at 24 hours after reperfusion decreased significantly in retinas treated with HO-1 siRNA (P < 0.01). The number of macrophages that had infiltrated the retina was increased in retinas pretreated with the siRNA of HO-1 compared with those treated with siRNA of GFP. On day 14 after reperfusion, HO-1 siRNA-treated retinas showed severe retinal injury and destruction of the retinal architecture.

conclusions. HO-1 promotes the survival of Müller cells after ischemia–reperfusion injury. Because inhibition of the upregulation of HO-1 resulted in an infiltration of inflammatory cells and destruction of the retina, the authors conclude that HO-1 induced in Müller cells plays a protective role in retinal ischemia-reperfusion.

Müller cells are specialized glial cells found only in the retina. ¹ Their cell bodies extend from the internal limiting membrane to the outer limiting membrane, and they contact every type of neuronal cell with their processes. ² Therefore, Müller cells are thought to be central to the normal functioning of the retina. ³ Under pathologic conditions, Müller cells are activated and are strongly resistant to retinal injuries induced by ischemia and phototoxicity. ^{4 5 6 7} The retinal neuronal cells, in contrast, are more vulnerable to injuries, and they die by apoptosis after the injuries. ^{5 6 8}  

To investigate why Müller cells survive relatively severe retinal injuries, we used a rat model of retinal ischemia-reperfusion. Earlier, we screened the gene expression profiles after retinal ischemia–reperfusion injury by using a DNA microarray system and found that genes that are classified in the pro- and antiapoptotic groups are simultaneously upregulated after the injury. One of the antiapoptotic genes that was upregulated after ischemia–reperfusion injury is heme oxygenase (HO)-1. ⁹ HO is a rate-limiting enzyme in heme catabolization, and heme is oxidized to free iron, carbon monoxide (CO), and biliverdin by the enzyme. ¹⁰ The formed biliverdin is converted to bilirubin, a potent antioxidant, ¹¹ by biliverdin reductase. 

Among the three isoforms of HO, ¹² HO-1 is an inductive isoform and is induced by oxidative stress, UV irradiation, ischemia-reperfusion, heavy metals, cytokines, and nitric oxide. ^{13 14 15} HO-1 belongs to the heat-shock protein family protein and functions as an antioxidant, an antiapoptotic agent, cytoprotective agent, and anti-inflammatory agent, under different pathologic conditions. ^{13 16 17 18} In contrast HO-2 is an isoenzyme that is expressed constitutively in nervous and vascular tissues. ¹⁹ CO, formed by HO-2, is a putative neurotransmitter in the brain and peripheral autonomic nervous system. Although HO-1 and -2 are expressed in normal and injured retinas, ^{20 21 22} the functions of HO-1 and -2 in the retina have still not been completely determined. 

We investigated the possible roles of HO-1 and -2, especially those of HO-1, in Müller cells after ischemia–reperfusion injury. Short interfering (si)RNAs were used to inhibit the upregulation of HO-1 expression. 

Materials and Methods

Animals and Ischemia–Reperfusion Model

A total of 126 adult male Sprague–Dawley (SD) rats weighing 250 to 300 g were used. All studies were conducted in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Institutional Animal Care and Use Committee of Shinshu University. 

Rats were anesthetized by an intraperitoneal injection of pentobarbital (60 mg/kg) and the pupils dilated with topical phenylephrine hydrochloride and tropicamide. The anterior chamber of the right eye was cannulated with a 27-gauge infusion needle connected to a physiological saline reservoir. The intraocular pressure was increased to 110 mm Hg for 60 minutes by elevating the saline reservoir. ^{8 23 24 25} Retinal ischemia was confirmed by the collapse of the central retinal artery and the whitening of the iris. Sham-treated left eyes underwent similar procedures but without the elevation of the saline reservoir. 

Tissue Preparations

Rats were killed with an overdose of pentobarbital sodium, and eyes were immediately enucleated. Eyes for histologic studies were enucleated at 24 hours and 14 days after reperfusion, and fixed in either 4% paraformaldehyde in phosphate buffer (PB) or 4% paraformaldehyde in phosphate-buffered saline (PBS). 

Western Blot Analysis

Eyes that underwent the ischemia–reperfusion injury were enucleated at 3, 6, 12, 24, and 48 hours after reperfusion, and eyes from siRNA-treated rats were enucleated at 12 and 24 hours after reperfusion. Retinas from normal rats were also studied. The sensory retina was immediately removed after enucleation and homogenized in RIPA buffer (50 mM Tris-HCl, 150 mM NaCl, 1% NP-40, 0.25% Na-deoxycholate, and 1 mM EDTA [pH 7.4]) containing protease inhibitors (1 μg/mL aprotinin, 1 μg/mL leupeptin, 1 μg/mL pepstatin, and 1 mM phenylmethylsulfonyl fluoride [PMSF]) and a phosphatase inhibitor (1 mM Na₃VO₄, and 1 mM NaF). The supernatant was collected after centrifugation at 15,000 rpm for 15 minutes. Equal amounts of protein (75 μg) were loaded onto a 12% SDS polyacrylamide gel and then transferred to a nitrocellulose membrane. After incubation with 3% nonfat dried milk in 0.05% Tween 20 and Tris-buffered saline (TBST), the membranes were incubated with 1:500 monoclonal mouse anti-HO-1 antibody, or 1:2000 polyclonal rabbit anti-HO-2 antibody (Stressgen, Victoria, British Columbia, Canada), or 1:3000 monoclonal mouse anti-β-actin antibody (Sigma-Aldrich, St. Louis, MO) with 1% bovine serum albumin (BSA) in TBST at 4°C for 12 hours. After a wash in TBST, the membranes were incubated with horseradish-conjugated secondary antibodies at a dilution of 1:4000 for 1 hour. The chemiluminescence Western blot analysis system (ECL; Amersham Biosciences, Buckinghamshire, UK) was used for signal detection. 

TdT-dUTP Terminal Nick-end Labeling Method

Specimens fixed in 4% paraformaldehyde for 12 hours were dehydrated and embedded in paraffin, and 4-μm sections were cut. The TdT-dUTP terminal nick-end labeling (TUNEL) method was performed with an apoptosis-detection kit (Mebstain; MBL, Nagoya, Japan). ⁸ Deparaffinized and proteinase K–treated sections were incubated for 90 minutes at 37°C with TdT and biotinylated dUTP in TdT buffer (30 mM Tris-HCl [pH 7.2], 140 mM sodium cacodylate, and 1 mM cobalt chloride). After they were rinsed, the sections were processed for avidin-FITC activity and were examined with a scanning laser confocal microscope (LSM 410; Carl Zeiss Meditec, Oberkochen, Germany). The number of TUNEL-positive cells in the INL and ONL was counted in 20 areas per section. Data are expressed as the mean number of positive cells per millimeter in sections cut along the vertical meridian of the eye. 

Immunohistochemical Study

Frozen sections were used for immunohistochemical studies of HO-1, HO-2, and OX-42, ²⁶ and paraffin-embedded sections were used to detect S-100- and ED1-like immunoreactivities. ^{8 26 27} The sections were incubated with 2% normal goat serum for 30 minutes at room temperature. After rinsing, the sections were incubated overnight at 4°C with 1:500 polyclonal rabbit anti-HO-1 antibody (Stressgen), 1:1000 polyclonal rabbit anti-HO-2 antibody (Stressgen), polyclonal rabbit anti-S-100 protein antibody (Immunotech, Marseille, France), 1:70 mouse anti-rat ED1 antibody (Serotec, Oxford, UK), or 1:50 anti-rat OX-42 (Serotec). The optimum concentration of the primary antibodies was determined in preliminary experiments with different concentrations. 

The secondary antibodies were conjugated either with Alexa Fluor 488 or 543 (Molecular Probes, Eugene, OR). The nuclei were then counterstained with propidium iodide (PI; 0.5 μg/mL). Double staining with monoclonal anti-S-100 (Sigma-Aldrich) and anti-HO-1 antibodies was also performed to identify the retinal cell types that expressed HO-1. BSA (1%) was used instead of the primary antibody for the negative control. All specimens were examined with a scanning laser confocal microscope. For quantitative analysis, four to eight eyes were used to count the number of S-100–and ED1-positive cells. The number of positive cells was counted in 20 areas of each section. The data for the S-100–positive cells are expressed as the mean number of positive cells per millimeter retina counted in sections cut along the vertical meridian of the eye. 

Real-Time PCR

Total RNA was extracted by the acid guanidium thiocyanate-phenol-chloroform method, ²⁸ and the cDNAs were prepared by reverse transcription. Real-time PCR was conducted with a commercial system (Light Cycler-FasStart DNA Master SYBR Green I Kit; Roche Molecular Biochemicals, Basel, Switzerland), according to the manufacturer’s specifications. Primers used for HO-1 were, 5′-GGGTGACAGAAGAGGCTAAGACC-3′ (sense) and 5′-AGATTCTCCCCTGCAGAGAGAAG-3′ (antisense). The primers used for HO-2 were 5′-ACCAGGCTGGCTCCATGCTAAC-3′ (sense) and 5′-AGATGGGGATGGAAGAGGTCAC-3′ (antisense). 

The expression of each gene was normalized to GAPDH, a housekeeping gene, at each time point after reperfusion. The multiples of change (x-fold) in each target gene were calculated to compare with the control retina. 

siRNA Treatment

A 21-nucleotide, double-stranded RNA was synthesized by Japan Bioservices (Saitama, Japan). The targeted sequence of rat HO-1 was 5′-GGCUUUAAGCUGGUGAUGGTT-3′ (sense), and 5′-CCAUCACCAGCUUAAAGCCTT-3′ (antisense). The siRNA of GFP mRNA, purchased from Dahmacon (Lafayette, CO), served as the control. Five micrograms of siRNA was injected into the vitreous cavity before reperfusion with a 32-gauge needle on a Hamilton syringe. 

Statistical Analysis

The data on the number of S-100–and ED1-positive cells were analyzed by the Mann-Whitney test. P < 0.05 was considered statistically significant. 

Results

Expression of HO-1 and -2 after Retinal Ischemia–Reperfusion Injury

Western Blot Analysis for HO-1 and -2.

The expression of HO-1 protein was hardly detectable in control retinas. On the contrary, in experimental retinas, the anti-HO-1 antibody detected a 32-kDa band at 6 to 48 hours after reperfusion. The peak HO-1 expression was observed at 24 hours (Fig. 1A) . The anti-HO-2 antibody detected a 36-kDa band in the control retinas, and the expression level did not change from 3 to 48 hours after reperfusion. 

Immunohistochemical Detection of HO-1 and -2.

Very little HO-1 immunoreactivity was observed in normal retinas (Fig. 1Ba). However, at 12 and 24 hours after reperfusion, immunostaining was detected by the anti-HO-1 antibody in the cell bodies in the inner nuclear layer (INL), the inner plexiform layer (IPL), and the outer nuclear layer (ONL; Fig. 1Bb). Double staining for HO-1 and S-100 protein showed that cells expressing HO-1 were Müller cells (Figs. 1Be–g). 

Immunostaining with anti-HO-2 showed that almost all neuronal cell bodies in the ganglion cell layer (GCL), INL, and ONL were positive, and the expression level did not change at the different examination times after the ischemic insult (Figs. 1Bc, 1Bd). 

Real-Time PCR Analysis of HO-1 and -2 mRNA Expression.

The expression of HO-1 mRNA started at 6 hours after reperfusion, and the expression was upregulated at 6, 12, 24, and 48 hours compared with that in sham-treated retinas. The peak expression level was found at 12 hours after reperfusion, and the level was as much as 12 times more than that in control retinas. The HO-2 mRNA expression level remained unchanged during the entire follow-up period (Fig. 2) . 

Cell Death in Retinal Ischemia–Reperfusion Injury

Histologic Findings.

To determine the level of cell death in the retina, TUNEL staining and S-100 immunostaining were performed on normal control retinas and the retina that had had 60 minutes of ischemia-reperfusion. At 24 hours after reperfusion, TUNEL-positive cells were found mainly in the INL and ONL, as previously reported. ^{8 24} The number of TUNEL-positive cells in the INL and ONL at 24 hours after reperfusion is shown in Table 1 . In both the INL and ONL, the number of TUNEL-positive cells after an ischemic insult of 60 minutes was significantly higher than that in the control retina (P < 0.05; Mann-Whitney test). 

The number of S-100–positive cells in the INL at 24 hours after reperfusion is also shown in Table 1 . The results show that the number of S-100–positive cells in eyes after 60 minutes of ischemia did not differ significantly from that in the control eyes (P = 0.248, Table 1 ). 

Effects of siRNA Treatment on Retinal Ischemia–Reperfusion Injury

HO-1 Expression in Retinal Samples of siRNA-Treated Rats.

At both 12 and 24 hours after reperfusion, the expression of HO-1 protein was lower in rats injected with siRNA HO-1 than with siRNA GFP. The expressions of HO-2 and β-actin protein were not significantly altered by the injection of the siRNAs (Fig. 3A) . Densitometric analysis of the ratio of HO-1/β-actin showed that the expression level of HO-1 in HO-1 siRNA-treated retinas was approximately 50% of that in GFP siRNA-treated retina at 12 hours and approximately 70% at 24 hours. Furthermore, immunohistochemical studies of HO-1 at 24 hours after reperfusion showed that treatment with HO-1 siRNA reduced the HO-1 expression in Müller cells (Figs. 3B 3C) . 

Histologic Findings and Immunohistochemical Studies of S-100 and ED1.

Although the number of S-100–positive cells in the INL was not reduced by 60 minutes of ischemia, intravitreal injection of siRNA of HO-1 decreased the number of S-100–positive cells. In severely damaged areas, a very small numbers of S-100–positive cells were observed. The number of S-100–positive cells in the INL of GFP siRNA- and HO-1 siRNA-treated eyes at 24 hours after reperfusion was 82.4 ± 3.1 (n = 5) and 53.8 ± 2.4 (n = 8) cells/mm, respectively (Fig. 4) . Statistical analysis showed that inhibition of the upregulation of HO-1 by the siRNA reduced the number of S-100–positive Müller cells significantly (P < 0.01). 

Histologic studies of the HO-1 siRNA-treated retinas showed inner retinal edema and the invasion of many inflammatory cells and prompted us to study the nature of such invading cells (Fig. 5B) . Specific immunostaining of macrophages in the retina was performed by the anti-ED1 antibody (Fig. 6) , and specific immunostaining of granulocytes/polymorphonuclear leukocytes was determined by the anti-OX-42 antibody (graphic data not shown). The number of ED1-positive cells in the retina of GFP siRNA-treated eyes and HO-1 siRNA-treated eyes at 24 hours after reperfusion was 6.12 ± 3.72 (n = 4) and 31.72 ± 8.37 (n = 4) cells/mm, respectively (Fig. 6C) . The difference was statistically significant (P < 0.05). Also, on day 14 after reperfusion, histologic findings of the HO-1 siRNA-treated retinas showed more severe retinal injury: invasion of the vitreous cavity by a large number of inflammatory cells (graphic data not shown), loss of the retinal cells in the INL and the outer segments, and destruction of the architecture of the sensory retina (Fig. 5D) . Similar findings were noted in the retinas, especially in the posterior retinas, of five eyes. However, in GFP siRNA-treated eyes, the retinal architecture was preserved even though the sensory retina became extremely thin (Fig. 5C) . 

Discussion

Our results show that Müller cells were much more resistant to ischemia-reperfusion than retinal neuronal cells in the INL and the ONL (Table 1) . Ischemia-reperfusion led to an upregulation of HO-1 in Müller cells, and the inhibition of the upregulation by the siRNA of HO-1 resulted in more severe retinal damage, a decreased number of Müller cells, and a larger number of infiltrating inflammatory cells. HO-1 expression at the protein level was upregulated in the injured retina from 6 to 48 hours after the insult, with a peak at 24 hours after reperfusion (Fig. 1) . HO-1 gene expression also gradually increased after reperfusion with a peak at 12 hours after reperfusion. 

Immunohistochemical studies showed that HO-1 was expressed selectively in Müller cells, and expression was found in all parts of the Müller cell from the internal to the external limiting membrane (Fig. 1) . Earlier studies have shown that peak neuronal cell death and production of reactive oxygen species occurs at 12 to 24 hours after reperfusion, ^{8 24 25 29} which coincides with the peak of HO-1 expression in Müller cells. 

To address the significance of upregulation of HO-1 in the Müller cells of injured retina, siRNA of HO-1 was injected intravitreally. The use of interference RNA has become a powerful method for analyzing functions of a given gene quickly. ³⁰ This technique was chosen because siRNAs can specifically inhibit HO-1 expression, and no other specific inhibitors of HO-1 are known. In the lung ischemia–reperfusion model, intranasal administration of HO-1 siRNA was reported to suppressed HO-1 protein expression from 8 to 16 hours. ³¹ In our study, the time of the siRNA injection was determined by the time course of HO-1 expression after retinal ischemia–reperfusion injury. Also, our preliminary study showed that intravitreally administered siRNA was delivered to the retina and was taken up by the retinal cells (graphic data not shown). Even so, the lack of detailed pharmacokinetic studies of the intravitreally injected siRNA is a drawback of this study. However, the expression of HO-1 at the protein level was clearly downregulated (Fig. 3A) . The data indicate that there may be a more efficient time for siRNA administration than that used in our study. 

The reduction of HO-1 expression resulted in significant morphologic changes in the injured retina, the destruction of retinal structures, and the disruption of the outer limiting membrane. These morphologic changes are characteristic features of Müller cell damage. ³²  

To confirm the Müller cell injury, the number of S-100–positive cells in the INL was counted, and as expected, a 35% reduction in the number of Müller cells was found. Furthermore, many Müller cells were not present in some areas of the severely injured retina. Inhibition of HO-1 expression not only reduced the number of S-100–positive cells but also induced the migration of macrophages and other inflammatory cells into the retina and vitreous cavity (Figs. 4 6) . After 2 weeks, the vitreous was filled with numerous inflammatory cells and the retinal structure was partly destroyed (Fig. 5) . These findings agree with the recently suggested role of HO-1 as an immunomodulator. ^{13 33} Macrophages, in turn, can produce more free radicals and induce more severe retinal damage. ^{29 34} The loss of Müller cells may destroy the retinal architecture. 

By using a well-established model of ischemia–reperfusion injury, ^{5 6 8 23 24 25} Hegazy et al. ³⁵ have shown that transfection of the retinal ganglion cells with human HO-1 cDNA has a neuroprotective effect. HO-1, a strong cellular protective enzyme, is promptly induced in Müller cells after ischemic insult, and Müller cells survive this ischemic stress. Müller cells are known to withstand a variety of insults to the retina: local pH shifts, excitotoxic reactions, generation of oxygen free radicals, and hypoglycemia in the retinal ischemic injury, for example. ^{36 37 38 39 40} It has been shown that upregulation of HO-1 expression inhibits apoptosis in vitro and in vivo. ^{16 33 35 41 42 43} Regulation of cellular iron and CO production have been considered possible mechanisms by which HO-1 prevents cell death. ^{16 42 44 45 46} In addition, bilirubin, a catalytic product of HO-1, is a strong antioxidant. ^{11 14 41 47}  

HO-2, which is a constitutively expressed, catalyzes the same reactions as HO-1 and is abundantly expressed in retinal neuronal cells, ²⁰ albeit with a different K _m. However, neuronal cells were vulnerable to the retinal ischemia–reperfusion injury. Thus, the constitutively expressed HO-2 in retinal neuronal cells is apparently not sufficient to protect them from injury. Evidence from HO-2–knockout mice has shown that HO-2 plays a neuroprotective role in the mouse brain and cultured brain neuronal cells. ⁴⁸ Thus, the roles of HO-2 in the retina and brain may be different, and retinal HO-2 may function mainly to form CO, a neurotransmitter. 

In our study, the reduction of HO-1 expression by siRNA injection lead to Müller cell death and severe retinal damage after ischemia-reperfusion. The retina is believed to have special mechanisms to resist injury because slight injuries can result in severe loss of visual function. Many researchers have suggested that Müller cells support the retinal neurons by secreting nutrients, neurotrophic factors, and growth factors, and thus prevent neuronal cell death in pathologic conditions. ^{49 50 51 52} The HO-1 expressed in Müller cells may protect not only the Müller cells but also the whole retina from injuries. 

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

 Supported by Grants-in-Aid for Scientific Research 12877276 and 13470364 from the Ministry of Education, Culture, Sports, Science, and Technology of the Japanese Government.

 Submitted for publication April 19, 2004; revised June 29 and July 28, 2004; accepted July 29, 2004.

 Disclosure: S. Arai-Gaun, None; N. Katai, None; T. Kikuchi, None; T. Kurokawa, None; K. Ohta, None; N. Yoshimura, 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 1.

 

Western blot analysis and immunohistochemical studies of HO-1 and -2 expression after the ischemia-reperfusion. (A) Western blot analysis. HO-1 protein expression was detected at 6 hours and peaked at 24 hours after reperfusion. HO-2 protein was detected in all samples of the retina. (B) Immunohistochemical studies (frozen sections). (Ba) HO-1 immunostaining of normal retina. (Bb) HO-1 immunostaining of retina 24 hours after reperfusion. (Bc) HO-2 immunostaining of normal retina. (Bd) HO-2 immunostaining of retina 24 hours after reperfusion. Green: HO-1 immunostaining (Ba, Bb) and HO-2 immunostaining (Bc, Bd); red: PI staining. HO-1 immunoreactivity at 24 hours after reperfusion was detected in the cell bodies in the INL, in processes in the IPL, and in the ONL (Bb). HO-2 immunoreactivity was detected in almost all types of cell bodies in the GCL, INL, and ONL (Bc). (Be) Double staining with anti-S-100 and anti-HO-1 antibodies at 24 hours after reperfusion. Green: HO-1–positive cells (Be, Bf); red: S-100–positive cells (also in Be, Bg). Scale bars: (Ba–d) 50 μm; (Be–g) 25 μm.

Figure 2.

 

Quantitative analysis of HO-1 and -2 mRNA expression in the retina after an ischemic insult. Expression of HO-1 mRNA was upregulated in the retina at 6 to 48 hours after reperfusion compared with that in control rat retinas. Expression of HO-2 mRNA was stable. Results are expressed as the mean ± SEM; n = 4.

Table 1.

 

Quantitative Analysis of Cells at 24 Hours after Ischemia–Reperfusion Injury

	n	INL	ONL
TUNEL-positive cells
Control	4	0.486 ± 0.423*	0.673 ± 0.364*
60 minutes	4	52.50 ± 5.60*	11.61 ± 1.67*
S-100-positive cells
Control	4	84.8 ± 5.6
60 minutes	4	77.4 ± 5.0

 The number of TUNEL-positive cells after a 60 minute-ischemic insult increased significantly compared with that in the control retina. The number of S-100-positive cells in the ischemia-reperfusion injury was not significantly different from that in the normal control retina. Results are expressed as the mean ± SEM (cells per millimeter).

*   P < 0.05, Mann-Whitney test.

Figure 3.

 

Western blot analysis and immunohistochemical studies of HO-1 protein after ischemia-reperfusion with pretreatment with HO-1 siRNA or GFP siRNA (frozen sections). (A) At both 12 and 24 hours after reperfusion, expression of HO-1 protein was downregulated in HO-1 siRNA-treated retinas compared with GFP siRNA-treated retinas. (B) HO-1 immunostaining of GFP siRNA-treated retina at 24 hours after reperfusion. HO-1 immunoreactivity was clearly visible in some processes in the IPL and the ONL and in some cells in the INL. (C) HO-1 immunostaining of HO-1 siRNA-treated retina at 24 hours after reperfusion. Weak immunostaining was visible in the IPL and the INL. The expression of HO-1 in the HO-1 siRNA-treated retina was less than that after GFP siRNA injection. Green: HO-1 immunostaining; red: PI staining. Scale bar, 50 μm.

Figure 4.

 

Immunohistochemical studies and quantitative analysis of S-100–positive cells in the INL at 24 hours after 60 minutes of ischemia with HO-1 siRNA pretreatment or GFP siRNA pretreatment (paraffin-embedded sections). (A) GFP siRNA-treated retina at 24 hours after reperfusion. S-100–positive cells were observed in the INL. (B) HO-1 siRNA-treated retina at 24 hours after reperfusion. The number of S-100–positive cells in the INL of HO-1 siRNA-treated retina was less than that after GFP siRNA injection. Green: S-100 immunostaining; red: PI staining. Scale bar, 50 μm. (C) Quantitative analysis. The number of S-100–positive cells in HO-1 siRNA-treated retinas is significantly fewer in GFP siRNA-treated retinas (**P < 0.01, Mann-Whitney test). Results are the means ± SEM (n = 5–8).

Figure 5.

 

Histologic changes in the retina at 24 hours and 14 days after reperfusion after a 60-minute ischemic insult (paraffin-embedded sections, hematoxylin and eosin [H&E] staining). All figures show sections of posterior retina near the optic disc. (A) GFP siRNA-treated retina at 24 hours after reperfusion. Infiltrating cells were barely visible in the retina. (B) HO-1 siRNA-treated retina at 24 hours after reperfusion. Inner retinal edema was noted, and inflammatory cells infiltrated the retina. (C) GFP siRNA-treated retina on day 14 after reperfusion. The retina was extremely thin, but the retinal architecture was preserved. (D) HO-1 siRNA-treated retina on day 14 after reperfusion. The retinal architecture was in disarray and the disorganization in the sensory retina resulted in the formation of rosettelike structures in the ONL. Scale bars, 25 μm.

Figure 6.

 

Immunohistochemical studies and quantitative analysis of ED1-positive cells in the retina at 24 hours after 60 minutes of ischemia, with HO-1 siRNA or GFP siRNA pretreatment (paraffin-embedded sections). (A) GFP siRNA-treated retina at 24 hours after reperfusion. There were few ED-1–positive cells in the retina. (B) HO-1 siRNA-treated retina at 24 hours after reperfusion. Numerous ED-1–positive cells infiltrated the retina and caused partial destruction of the INL and ONL. Green: ED1 immunostaining; red: PI staining. Scale bars, 50 μm. (C) Quantitative analysis. The number of ED1-positive cells in HO-1 siRNA-pretreated retinas increased significantly compared with those in GFP siRNA-treated retinas (*P < 0.05, Mann-Whitney test). Results are expressed as the mean ± SEM; n = 4.