Role of PI3K/Akt and MEK/ERK in Mediating Hypoxia-Induced Expression of HIF-1α and VEGF in Laser-Induced Rat Choroidal Neovascularization

Abstract

purpose. The transcription factor hypoxia-inducible factor (HIF)-1 plays a central physiological role in oxygen and energy homeostasis and is activated during hypoxia by stabilization of the subunit HIF-1α. Hypoxia plays an important role in the development of choroidal neovascularization (CNV). Expression of HIF-1α has been demonstrated in CNV. Vascular endothelial growth factor (VEGF) is one of the most well-characterized angiogenic factors in CNV, which is under the regulation of HIF-1. The aim of the present study was to explore the upstream signaling pathways involved in regulating hypoxia-induced expression of HIF-1α and VEGF in laser-induced rat CNV.

methods. A well-established rat model of CNV and cultured human retinal pigment epithelium (hRPE) was used to investigate the role of PI3K/Akt and MEK/ERK pathways in regulating HIF-1α and VEGF expression in CNV in rat and hRPE under hypoxia by immunohistochemistry, Western blot analysis, real-time PCR, and ELISA.

results. pAkt, pERK, HIF-1α, and VEGF were upregulated in vivo and in vitro. PI3K inhibitor (Ly294002) significantly decreased pAkt activity and HIF-1α and VEGF expression in vivo and in vitro, whereas MEK inhibitor (PD98059) reduced ERK phosphorylation and the expression of VEGF but had no effect on HIF-1α. LY294002 and PD98059 severely inhibited the formation of CNV.

conclusions. The PI3K/Akt pathway was required for hypoxia-induced expression of HIF-1α and VEGF, whereas the MEK/ERK pathway was required only for VEGF in laser-induced rat CNV.

Choroidal neovascularization (CNV) complicating age-related macular degeneration (AMD) is the most common cause of severe visual loss in people older than 60 years in developed countries. ^{1 2} CNV is a complex process in which tissue ischemia is thought to be involved in the development of CNV. ³  

A master regulator of the hypoxic response is the transcription factor hypoxia-inducible factor (HIF)-1, which consists of an α-subunit whose proteasomal degradation and thus relative abundance are regulated by oxygen tension, and a constitutively expressed β-subunit. ⁴ HIF-1 transactivates the expression of proangiogenic genes in response to hypoxic conditions and plays important roles in vasculogenesis and angiogenesis. ^{5 6} Binding of HIF-1 to the hypoxia response element of the vascular endothelial growth factor (VEGF) promoter results in transcriptional activity. ⁷ VEGF, a potent and specific mitogen for vascular endothelial cells, is a critical mediator of CNV. Animal studies have shown that VEGF overexpression is sufficient to induce CNV in the eye, ^{8 9} whereas inhibition reduces this effect. ¹⁰ VEGF was also expressed in laser-induced CNV ¹¹ and surgically excised CNV membranes, ¹² and multiple preclinical and clinical trials have proved that anti-VEGF strategies were effective as potential therapeutic agents for the treatment of CNV. ¹³  

Because HIF-1α activates the transcription of VEGF, which is required for CNV, it is possible that hypoxia may mediate CNV through the induction of HIF-1α and VEGF. However, little is known about this; even less is known about the upstream signaling events that are activated by hypoxia and that mediate its effects in CNV. The extracellular signal-regulated kinase (ERK) is a subfamily member of mitogen-activated protein kinases (MAPKs) activated by an upstream kinase called MAPK/ERK kinase (MEK). The ERK pathway mediates a number of cellular fates, including growth, proliferation, and survival. ^{14 15} In addition, the serine/threonine kinase Akt, also known as protein kinase B (PKB), plays a pivotal role in cell proliferation, differentiation, and survival. It is activated by a phosphoinositide 3-kinase (PI3K)-dependent signaling pathway. ^{16 17} Increased phosphorylation of ERK1/2 was observed in retinal neovascularization and other ischemia diseases, created by retinal vein occlusion, or subjected to ischemia-reperfusion injury after ligation of the optic nerve. Furthermore, the inhibition of ERK1/2 could significantly retard retinal neovascularization. ¹⁸ The PI3K/Akt pathway has also been known to be critical for ischemia and angiogenesis. ^{19 20}  

We studied whether PI3K/Akt and MEK/ERK signaling pathways were involved in regulating the expression of HIF-1α and VEGF in laser-induced rat CNV. We further investigated the role of Akt and ERK in hypoxia-induced expression of HIF-1α and VEGF in cultured human retinal pigment epithelium (hRPE). Here we showed that PI3K/Akt was needed for the expression of HIF-1α and VEGF, whereas MEK/ERK was required only for the expression of VEGF in experimental CNV and hRPE under hypoxia. Both PI3K and MEK inhibitors severely inhibited the formation of CNV. These results suggest the potential therapeutic use of Akt or ERK inhibitor to block the effect of hypoxia on CNV formation. 

Materials and Methods

Induction of CNV

Brown Norway rats (Vital River Laboratory, Beijing, China), weighing 200 to 250 g each, were used in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Laser-induced CNV was performed as previously described. ²¹ Rats were anesthetized for all procedures with intraperitoneal injection of 1% sodium pentobarbital (45 mg/kg body weight), and the ocular surface was then anesthetized with topical instillation of amethocaine hydrochloride. Pupils were dilated with tropicamide (5 mg/mL) and phenylephrine hydrochloride (5 mg/mL). For each animal, one eye was randomly selected for laser treatment, and eight laser spots (140 mW, 0.1 second, 75 μm) were delivered with a diode-pumped frequency-doubled, 532-nm laser (Oculight GLx; Iridex, Mountain View, CA) between the retinal vessels in a peripapillary distribution in each fundus. Production of a subretinal bubble at the time of laser treatment confirmed the rupture of Bruch membrane. 

Treatment with PI3K or MEK Inhibitor

After photocoagulation, all animals were randomly divided into four groups. For the first two groups, the lasered eyes immediately received intravitreal injection, as previously described, ²¹ with a PI3K inhibitor (LY294002; 3 mM, 3 μL; Cell Signaling Technology, Beverly, MA) or a MEK inhibitor (PD98059; 5 mM, 3 μL; Sigma, St. Louis, MO) on days 0 and 7. For the last two groups, the lasered eyes were treated with dimethyl sulfoxide (DMSO; 3 μL) as vehicle control or without intravitreal injection as control. Animals were killed after photocoagulation on days 1, 3, 7, and 14. 

Histologic Analysis of CNV

Fourteen days after photocoagulation, five eyes of each group were enucleated, and serial 4-μm paraffin sections were prepared for staining with hematoxylin and eosin (H&E). The distance from the disrupted RPE layer to the top of the lesion was measured (Image-Pro Plus software). The largest distance was selected in each lesion for statistical comparison among the groups. 

Immunohistochemistry

Serial 4-μm paraffin sections of rat eyes (three eyes from each group) were obtained at 3, 7, and 14 days after photocoagulation. Slides were briefly washed in phosphate-buffered saline (PBS; pH 7.4) and blocked with 5% normal goat serum for 1 hour. Sections were incubated sequentially with mouse anti-HIF-1α (1:500; Chemicon, Temecula, CA) or mouse anti-VEGF (1:100; Santa Cruz, Fremont, CA) antibody, biotinylated secondary anti-mouse antibody (1:300; Sigma-Aldrich, St. Louis, MO), and streptavidin peroxidase (Vector Laboratories, Burlingame, CA, with three PBS washes in between. Specificity of staining was assessed by substitution of nonimmune serum for primary antibody. Immunoreactivity was visualized with the peroxidase substrate amino ethyl carbazole (AEC kit; Hao Yang Biological, TianJin, China). Slides were rinsed with tap water, counterstained with hematoxylin, and mounted with mounting medium. 

Fluorescein Angiography

CNV lesions were studied at 14 days after laser photocoagulation by fluorescein angiography (FA) with a confocal scanning laser ophthalmoscope (Retinal Angiography; Heidelberg Engineering, Heidelberg, Germany). Fluorescein sodium (10%; 0.1 mL/kg) was injected into the tail veins of the anesthetized rats. Late-phase angiograms were obtained 5 minutes after injection. The area of CNV on FA was measured (Image-Pro Plus software; Media Cybernetics, Bethesda, MD). The mean area of CNV was derived from measurements of all the CNV lesions (40 lesions from each group) by two masked specialists. 

Choroidal Flatmounts

Two weeks after treatment, RPE-choroid-sclera complexes (five eyes from each group) were fixed with 4% paraformaldehyde and incubated with HEPES-buffered saline containing 1:1000 rhodamine-conjugated Ricinus communis agglutinin (Vector Laboratories). CNV was visualized with 543-nm wavelength using a scanning laser confocal microscope (FV 1000; Olympus, Tokyo, Japan). The area of CNV-related fluorescence was measured (Image-Pro Plus; Media Cybernetics). 

Real-Time Polymerase Chain Reaction Analyses

Total RNA was prepared from three eyecups (RPE-choroid-sclera complex) from each group at 1, 3, 7, and 14 days after photocoagulation using reagent (Trizol; Invitrogen, Carlsbad, CA) and was reverse transcribed with a cDNA synthesis kit (RevertAid First-Strand cDNA Synthesis Kit; Fermentas, Burlington, ON, Canada). Real-time PCR was conducted with the ABI 7000 SDS (Applied Biosystems, Foster City, CA) using SYBR green (TaKaRa Bio, Otsu, Shiga, Japan). Each cDNA sample was analyzed in duplicate. Primer sequences were HIF-1α 5′-TCA AGT CAG CAA CGT GGA AG-3′ and 5′-TAT CGA GGC TGT GTC GAC TG-3′; VEGF 5′-AGA AAG CCC ATG AAG TGG TG-3′ and 5′-ACT CCA GGG CTT CAT CAT TG-3; β-actin 5′-GAA GTA CCC CAT TGA ACA CGG-3′ and 5′-TTA GGG TTC AGA GGG GCC TC-3′. 

Western Blot Analysis

Protein extracts were obtained from three homogenized eyecups from each group at each time point. Samples were then assessed for protein concentration (Bradford assay; Bio-Rad Laboratories, Munich, Germany). Electrophoresis of proteins was performed with 12% SDS-polyacrylamide gels. Fifty micrograms of protein was loaded on each lane. After the protein was electrotransferred to a polyvinylidene difluoride membrane (Millipore, Bedford, MA), the membrane was blocked in a solution of 5% (wt/vol) skimmed milk powder in PBS (pH 7.5) for 1 hour at room temperature and then probed overnight at 4°C with Akt/pAkt (1:500; Cell Signaling Technology), ERK/p-ERK(1:200; Santa Cruz), HIF-1α (1:500; Chemicon, Temecula, CA), or VEGF (1:200; Santa Cruz) antibodies. After they were washed with PBS, horseradish peroxidase-conjugated secondary antisera (Cell Signaling Technology) were incubated for 1 hour at room temperature. Immunoreactivity was visualized by enhanced chemiluminescence (ECL; Amersham Biosciences) reagent. For sequential blotting with additional antibodies, the membranes were stripped with a restored Western blot stripping buffer and reprobed with the indicated antibodies. Protein levels were quantitated by densitometry and normalized to the β-actin levels. 

In Vitro Assays

hRPE was isolated from keratoplasty donor eyes within 24 hours of death; their use was approved by the Ethics Committee of the Fourth Military Medical University and followed the tenets of the Declaration of Helsinki. The isolation and cultivation of hRPE were performed as described previously. ²² Briefly, the neurosensory retina was gently separated from the hRPE monolayer, and hRPE was immersed in a trypsin (0.05%)-EDTA (0.02%) solution at 37°C for 1 hour. Culture medium with 20% FBS was added, and hRPE was isolated and collected with a pipette under a dissecting microscope. hRPE was cultured in Dulbecco modified essential medium (DMEM) supplemented with 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin. hRPE exhibited uniform immunohistochemical staining for cytokeratin 8/18 (1:100; Santa Cruz). All experiments were performed on cells from passages 3 to 6. 

hRPE was cultured in DMEM with 200 μM CoCl₂ (Sigma-Aldrich, St. Louis, MO) to mimic chemical hypoxia for 1, 2, 4, 6, 8, 12, and 24 hours. PI3K (30 μM) or MEK (20 μM) inhibitor was pretreated for 1 hour and then was further cultured for the indicated time. At each time point, cells were collected and total cellular RNA was processed for real-time PCR analyses for HIF-1α and VEGF. Primer sequences were HIF-1α 5′-CCA TTA GAA AGC AGT TCC GC-3′ and 5′-TGG GTA GGA GAT GGA GAT GC-3′; VEGF 5′-CCG CAG ACG TGT AAA TGT TCC T-3′ and 5′-CGG CTT GTC ACA TCT GCA AGT A-3′; GAPDH 5′-GCG CTG AGT ACG TCG TGG AG-3′ and 5′-CAG TTG GTG GTG CAG GAG G-3′. For protein analyses, supernatant and cell lysates were collected. The concentration of VEGF was measured by ELISA (R&D Systems) according to the manufacturer’s instructions, whereas the protein expression of Akt/pAkt, ERK/pERK, HIF-1α were determined by Western blot. 

Statistical Analysis

All results were expressed as mean ± SD. Statistical analyses were made using Student t-test or 2-way ANOVA (SPSS, Chicago, IL) where appropriate. P < 0.05 was considered statistically significant. 

Results

Expression of Akt/pAkt, ERK/pERK, HIF-1α, and VEGF in Experimental CNV

We used an established rat model of CNV to confirm whether Akt/pAkt, ERK/pERK, HIF-1α, and VEGF were involved in CNV. Immunohistochemical staining revealed that HIF-1α (Fig. 1A)and VEGF (Fig. 1C)were positive in CNV lesions compared with control (Fig. 1D) . Double-stained immunofluorescence was also performed, and the results indicated that at least RPE, endothelial cells, smooth muscle cells, and glial cells in the lasered lesion expressed these two factors (data not shown). Based on Western blot results, laser treatment greatly increased the activation of pAkt (Fig. 2C)and pERK (Fig. 2D)and the expression of HIF-1α (Figs. 2C 2D)and VEGF (Figs. 2C 2D) . The peak expression for pAkt and pERK was at day 3 and then substantially returned to baseline. For HIF-1α, the peak expression was at day 3 and then decreased at day 7; it was still higher than at day 1 but returned to baseline at day 14. VEGF peaked at day 7 and remained at that level until day 14 (P < 0.05; n = 3). The total amount of Akt and ERK remained unchanged during the development of CNV (Figs. 2C 2D)(P > 0.05; n = 3). The result of real-time PCR analysis showed a similar pattern for HIF-1α and VEGF (Figs. 2A 2B) . There was no statistical difference in the cytokine expression level in rats treated with DMSO compared with control (P > 0.05; n = 3). 

We also studied whether the p110α isoform of PI3K was involved in CNV. Our results indicated that under normal conditions, the p110α isoform of PI3K already existed in the retina and that laser treatment greatly stimulated its expression, with the highest expression at day 3, and then dropped it to baseline (Fig. 2E ; P < 0.05; n = 3). 

Effect of Intravitreal Injection of PI3K- or MEK-Specific Inhibitor on the Activation of Akt or ERK and the Expression of HIF-1α and VEGF in CNV

We further investigated whether PI3K/Akt and MEK/ERK signaling pathways were involved in the expression of HIF-1α and VEGF in CNV. After photocoagulation, intravitreal injection of PI3K (LY294002) or MEK inhibitor (PD98059) was performed immediately. LY294002 greatly decreased pAkt activation (Fig. 2C)and the expression of HIF-1α and VEGF at mRNA (Figs. 2A 2B)and protein (Fig. 2C)levels (P < 0.05; n = 3), while PD98059, showing no effect on the expression of HIF-1α (P > 0.05; n = 3) (Figs. 2A 2D) , could reduce pERK activation (Fig. 2D)and the expression of VEGF at mRNA (Fig. 2B)and protein (Fig. 2D)levels (P < 0.05; n = 3). Immunohistochemical staining further proved that PD98059 treatment did not affect the HIF-1α level in CNV lesions (Fig. 1B) . 

Effect of Intravitreal Injection of PI3K- or MEK-Specific Inhibitor on the Severity of Experimental CNV

Fourteen days after laser, laser-induced photocoagulation reproducibly resulted in the development of CNV, as demonstrated by FA (Fig. 3A) , histology (Fig. 3D) , and flatmount (Fig. 3G) . Although CNV membranes in the DMSO-treated group showed moderate to severe fluorescein leakage, the rats treated with PD98059 or LY294002 showed decreased fluorescein leakage with 45.4% and 47.2% reductions, respectively (Figs. 3B 3C) . H&E-stained tissue sections of the CNV membranes were also evaluated and showed that specific inhibitor-treated eyes and DMSO-treated eyes contained true CNV membranes centered on a disrupted Bruch membrane with vascular channels, fibrous stroma, and macrophage infiltration. Lesions in the PD98059- or LY294002-injected eyes, however, appeared to have significantly less CNV thickness than DMSO-treated lesions, with 56.5% and 54.8% reductions, respectively (Figs. 3E 3F) . Laser-induced CNV membrane areas were significantly reduced in PD98059- and LY294002-treated eyes; reductions were 50.8% and 53.9%, respectively (Figs. 3H 3I) . Eyes treated with DMSO showed no difference in fluorescein leakage, CNV thickness, and CNV area compared with control (P > 0.05). 

Effect of Hypoxia on the Activation of Akt and ERK and the Expression of HIF-1α and VEGF in hRPE

To confirm the role of hypoxia and its subsequent effect on the development of CNV, we further detected the activation of Akt and ERK and the expression of HIF-1α and VEGF in hRPE (Fig. 4) . Western blot indicated that pAkt and pERK were upregulated after 1 hour of hypoxic stress and up to 24 hours (P < 0.05; n = 3), whereas the total amount of Akt and ERK remained unchanged (P > 0.05; n = 3; Figs. 4C 4D ). As expected, HIF-1α protein levels were rapidly increased in response to hypoxia after 1 hour and remained elevated for up to 24 hours (Figs. 4C 4D) , whereas HIF-1α mRNA levels were also increased in response to hypoxia, peaking at 1 hour and returning to basal levels after 4 hours (Fig. 4A) ; Similarly, VEGF mRNA (Fig. 4B)and protein (Figs. 4E 4F)expression was upregulated in 1 hour and up to 24 hours in a time-dependent manner (P < 0.05; n = 3). 

Effect of PI3K- or MEK-Specific Inhibitor on HIF-1α and VEGF Expression in hRPE under Hypoxia

We next examined the effect of PI3K/Akt and MEK/ERK pathway inhibition on HIF-1α and VEGF expression in hypoxic hRPE. The PI3K inhibitor LY294002 decreased the activation of pAkt (Fig. 4C)and the expression of HIF-1α and VEGF at mRNA (Figs. 4A 4B)and protein (Figs. 4C 4E)levels (P < 0.05; n = 3). The MEK inhibitor PD98059 severely reduced ERK phosphorylation (Fig. 4D) . It also blocked hypoxia-stimulated VEGF mRNA (Fig. 4B)and protein expression (Fig. 4F)(P < 0.05; n = 3). Pretreatment with PD98059 was not effective in reducing hypoxia-induced HIF-1α mRNA (Fig. 4A)and protein expression (Fig. 4D ; P > 0.05; n = 3). 

Discussion

A previous study ²³ suggested that foveolar choroidal blood flow decreased in patients with AMD and large drusen and reported a systematic decrease in choroidal circulatory parameters and an increase in the severity of AMD features associated with risk for CNV. Ross et al. ²⁴ showed an association between the locations of the macular choroidal watershed vascular filling zones detected by FA and CNV membranes. During angiography, watershed filling zones corresponded to the last areas of the choroid that fill with the dye. CNV occurs in proximity to these areas that are the most prone to the development of ischemia and hypoxia in cases of decreased choroidal blood flow. Furthermore, CNV always associated with the deposition of materials and the thickening of the RPE-Bruch membrane complex, which may impede the diffusion of substances and increase the distance oxygen must travel from the choriocapillaris to the photoreceptors, further reducing the availability of oxygen in the outer retina. ^{25 26} HIF-1α, the main reactor of ischemia, plays a pivotal role in angiogenesis. ²⁷ Previous results showed that HIF-1α is expressed in human CNV membranes, ²⁸ human retinal angiomatous proliferation specimens, ²⁹ and retinal neovascularization. ³⁰ The HIF-1α inhibitor YC-1 could also prohibit the formation of laser-induced CNV. ³¹ Our previous in vitro results also showed that mRNA and protein levels of HIF-1α in the RPE increased in response to hypoxia, followed by increasing expression of VEGF. mRNA and protein levels of HIF-1α and VEGF in the RPE were decreased dramatically after transfection with an HIF-1α-specific small interference RNA vector. ³² The proliferation, migration, and tube formation of choroidal endothelial cells were significantly inhibited by the HIF-1α knocked-down RPE compared with the control in the coculture system. ³³ Based on these results, we presumed that ischemia/hypoxia may mediate CNV through HIF-1α-regulated VEGF. To further elucidate the mechanism for this, we investigated whether PI3K/Akt and MEK/ERK signaling pathways were involved in regulating HIF-1α/VEGF. Our in vivo results illustrated that HIF-1α and VEGF were upregulated in laser-induced CNV lesions. We also concluded that the PI3K/Akt pathway was required for the expression of HIF-1α and VEGF, whereas MEK/ERK was needed only for VEGF expression. PI3K- and MEK-specific inhibitors could significantly suppress the development of CNV. Given that PI3K has several different isoforms (p110α, p110β, and p110δ) and only p110α was selectively required for angiogenesis, ³⁴ we proved that p110α was involved in the development of laser-induced CNV. Our group also found that hypoxia could stimulate the expression of p110α at mRNA and protein levels in cultured hRPE and that LY294002 could downregulate the expression of VEGF by hRPE under normoxia and hypoxia (data not shown). Further study is needed to determine whether p110β or p110δ also functions in laser-induced CNV. 

To clarify which cell type in the CNV lesion expresses these two factors, double-stained immunofluorescence was performed. Our results indicated that at least RPE, endothelial cells, smooth muscle cells, and glial cells expressed them. hRPE, one of the most important components of CNV, plays an important role in regulating CNV through releasing many angiogenesis-related factors. As a main producer of VEGF in CNV, hRPE was further focused in our research. Our results showed that in hRPE, the activation of Akt contributed to hypoxia-induced HIF-1α and VEGF expression whereas the activation of ERK contributed only to VEGF expression. 

Molecular regulation of the HIF-1α subunit is multifaceted and involves the control of mRNA expression, protein stability, and activity. Classical hypoxic regulation of HIF-1α is at the level of protein stability. Under normoxia, oxygen mediates posttranslational hydroxylation of two proline residues in the oxygen-dependent degradation (ODD) domain of HIF-1α. Hydroxylation occurs through the activities of three prolyl hydroxylase domain proteins and mediates binding of the tumor suppressor protein von Hippel-Lindau (VHL) to the N-terminal transactivation domain of HIF-1α. Binding of VHL, an E3 ubiquitin-protein ligase, targets HIF-1α for rapid proteasomal degradation. ³⁵ HIF-1α protein under hypoxia escapes degradation, enabling heterodimerization with HIF-1β. We found that hypoxia increased HIF-1α levels not only by protein stabilization but also by de novo transcription of the HIF-1α gene with a maximum at 1 hour, returning to basal levels within 8 hours in RPE. HIF-1α mRNA peaked at 3 days after laser in our in vivo CNV model. Belaiba et al. ³⁶ also found that short-term hypoxia resulted not only in an accumulation of HIF-1α protein but in a transient increase in HIF-1α mRNA levels in hypoxic pulmonary artery smooth muscle cells. 

Although our in vivo/in vitro experiments indicated that MEK/ERK had no effect on HIF-1α expression, we still could not exclude the effect of MEK/ERK on HIF-1α activation under hypoxia. Hypoxia promotes ERK phosphorylation and translocation to the nucleus, where ERK exerts part of its biological activity. ³⁷ Previous reports have suggested that ERK was known to promote the expression of HIF-1α and has been shown to increase its transactivation by direct phosphorylation within its C-terminal transactivation domain or by indirect phosphorylation of p300/CBP. ³⁸ The reason for the difference between our results and the previous results may be due in part to the different stimulus and specific cell type; further studies are needed. 

 Supported by the National Natural Science Foundation of China (Grants 30371516, 30672291) and the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry, China. Supported in part by the equipment donation from the Alexander Von Humboldt Foundation in Germany (Y-SW; V8151/02085).

 Submitted for publication July 20, 2008; revised October 19 and November 25, 2008; accepted February 23, 2009.

 Disclosure: X.-M. Yang, None; Y.-S. Wang, None; J. Zhang, None; Y. Li, None; J.-F. Xu, None; J. Zhu, None; W. Zhao, None; D.-K. Chu, None; P. Wiedemann, 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.

 

Immunohistochemical studies after laser photocoagulation. Sections were stained with aminoethyl carbazole as the chromogen (red) and counterstain with hematoxylin (blue). HIF-1α (A) and VEGF (C) were expressed in the lasered lesions 7 days after photocoagulation (black arrow; ×400). PD98059 treatment did not affect the expression of HIF-1α (B; black arrow; ×400). Control sections (D) stained in parallel with PBS showed no background staining (×400).

Figure 2.

 

Effects of MEK or PI3K inhibitor on both the activation of ERK1/2 or Akt and the expression of HIF-1α and VEGF in CNV analyzed by real-time PCR (A, B) and Western blot (C, D). The activation of Akt and ERK1/2 was elevated during the development of CNV (C, D). The expression of HIF-1α and VEGF was also induced at the mRNA (A, B) and protein (C, D) levels (*P < 0.05; n = 3). LY294002 reduced the pAkt level (C) and the expression of HIF-1α and VEGF at the mRNA (A, B) and protein (C, D) levels. PD98059, while showing no effect on the expression of HIF-1α (A, D), decreased pERK (D) and the expression of VEGF at the mRNA (B) and protein (D) levels. (#, ##P < 0.05; n = 3). PI3K p110α was involved in the development of laser-induced CNV, as shown by Western blot (E).

Figure 3.

 

Inhibitory effect of MEK- or PI3K-specific inhibitor on the severity of experimental CNV. FA was performed 14 days after laser surgery. PD98059 (B) and LY294002 (C) diminished fluorescein leakage compared with dimethyl sulfoxide (DMSO) treatment (A), shown by graph (J) (*P < 0.05; n = 40). Light micrograph of rat CNV at day 14 showed PD98059 (E) and LY294002 (F) decreased CNV thickness compared with DMSO treatment (D), shown by graph (K) (*P < 0.05; n = 20; H&E stain; ×400). CNV area was measured by confocal analysis of rhodamine-conjugated R. communis agglutinin-stained flatmounts on day 14. The reduction in CNV area was significant in PD98059 (H)- and LY294002 (I)-treated rats compared with DMSO-treated rats (G), shown by graph (L) (*P < 0.05; n = 20; ×200).

Figure 4.

 

Effects of MEK or PI3K inhibitor on hypoxia-induced activation of ERK1/2 or AKT and the expression of HIF-1α and VEGF in hRPE. Hypoxia increased the expression of HIF-1α and VEGF at the mRNA (A, B) and protein (C–F) levels (*P < 0.05; n = 3). pAKT (C) and pERK1/2 (D) were also activated. LY294002, decreasing pAKT level (C), severely reduced the expression of HIF-1α and VEGF at the mRNA (A, B) and protein (C, E) levels. PD98059, showing no effect on the expression of HIF-1α (A, D), downregulated pERK1/2 (D) and VEGF expression at the mRNA (B) and protein (F) levels (#, ## P < 0.05; n = 3).