Abstract

PURPOSE. To generate a profile of genes expressed in the retina, RPE, and choroid after laser treatment and to identify genes that may contribute to the beneficial effects of laser photocoagulation in the treatment of angiogenic retinal diseases. METHODS. Argon laser irradiation was delivered to the left eye of normal C57BL/6J mice (n ϭ 30), with the right eye serving as the control in each animal. Three days after laser treatment, mice were culled, eyes enucleated, and the retinas dissected and pooled into respective groups. The total RNA of replicate samples was extracted, and expression profiles were obtained by microarray analysis. Data comparisons between control and treated samples were performed and statistically analyzed. RESULTS. Data revealed that the expression of 265 known genes and expressed sequence tags (ESTs) changed after laser treatment. Of those, 25 were found to be upregulated. These genes represented a number of biological processes, including photoreceptor metabolism, synaptic function, structural proteins, and adhesion molecules. Thus angiotensin II type 2 receptor (Agtr2), a potential candidate in the inhibition of VEGF-induced angiogenesis, was upregulated, whereas potential modulators of endothelial cell function, permeability factors, and VEGF inducers, such as FGF-14, FGF-16, IL-1␤, calcitonin receptor-like receptor (CRLR), and plasminogen activator inhibitor-2 (PAI2), were downregulated. CONCLUSIONS. In this study, genes were identified that both explain and contribute to the beneficial effects of laser photocoagulation in the treatment of angiogenic retinal diseases. The molecular insights into the therapeutic effects of laser photocoagulation may provide a basis for future therapeutic strategies. (Invest Ophthalmol Vis Sci. 2003;44:1426 -1434) DOI:10.1167/iovs.02-0622 A ngiogenic retinal diseases such as diabetic retinopathy, age-related macular degeneration, retinopathy of prematurity, retinal ischemia, and choroidal neovascularization are the most common cause of vision loss in the developed world. Although the underlying cause of these angiogenic retinal diseases is different, the factors initiating the angiogenic process are likely to be similar. One of the most important factors in the initiation of an angiogenic response is the upregulation of the expression of vascular endothelial growth factor (VEGF), a potent angiogenic and permeability factor, the upregulation of which has been shown in all patients who have neovascular eye diseases. It has been proposed that the therapeutic effects of laser photocoagulation are due to the destruction of photoreceptors, the highest oxygen consumers in the retina. Subsequently, these photoreceptors are replaced by glial cells, allowing increased oxygen diffusion from the choroid to the inner retina and thereby relieving inner retinal hypoxia. (1) Constriction of the retinal arteries results in decreased hydrostatic pressure in capillaries and the constriction of capillaries and venules, 7 and (2) the cellular production of VEGF is inhibited. With the development of microarray technology, it is possible to monitor thousands of genes simultaneously, enabling the high throughput analysis of treatment methods, such as laser photocoagulation, on retinal gene expression. Such global investigations into altered gene expression can facilitate the identification of key regulatory factors and/or events that contribute to the therapeutic effects of laser photocoagulation in the inhibition of both neovascularization and the progression of retinal diseases. An examination of altered gene expression patterns in normal tissue also provides a baseline from which comparisons to the effects of laser photocoagulation in retinal disease models can then ensue. In this study, the effects of laser photocoagulation on gene expression in the retina, RPE, and choroid were examined by using microarray technology. To validate the methodology, the expression profiles of selected genes were confirmed by quantitative PCR techniques. The molecular insights into the therapeutic effects of laser photocoagulation will not only increase 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. our understanding of the mechanisms that underlie this treatment but will also identify genes for future gene therapy strategies. METHODS Retinal Laser Photocoagulation of Mouse Retina Female mice (C57BL/6J), aged 10 weeks, were used. The mice were treated and maintained in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the University of Western Australia Animal Ethics Committee. The mice were anesthetized by intraperitoneal injection of a mixture of ketamine (100 mg/kg; Warner Lambert Co., Morris Plains, NJ) and xylazine (10 mg/kg; Bayer AG, Leverkusen, Germany), and the pupils dilated with 2.5% tropicamide (Alcon, Fort Worth, TX). Argon laser irradiation (514 nm; Coherent Radiation Systems, Palo Alto, CA) was delivered through a slit lamp ophthalmoscope (Carl Zeiss, Oberkochen, Germany) with a handheld coverslip serving as a contact lens. The laser spots (20 spots/retina) were placed at a setting of 50-m diameter, 0.05-second duration, and 120-mW intensity and were scattered in the upper quadrant of the fundus. The left eye of each animal was laser treated, and the right eye remained untreated, serving as a control. Perfusion of Laser-Treated Eyes with Fluorescence-Labeled Dextran Two laser-treated mice were perfused with fluorescence-labeled dextran, as described previously. 10 Briefly, the animals were anesthetized as just described. PBS (4 mL) was perfused through the left ventricle into the aorta to wash away the circulating blood, followed by 2 mL fluorescence-labeled dextran (50 mg/mL, molecular weight, 2.0 ϫ 10 6 ; Sigma, St. Louis, MO) for perfusion. The eyes were enucleated, fixed in 2% paraformaldehyde for 30 minutes, and flatmounted for fluorescence microscopy, as described previously. 11 Histology of Laser-Treated Eyes Eyes (control, n ϭ 2; laser-treated, n ϭ2) were harvested for histology 3 days after laser photocoagulation. The enucleated eyes were fixed in 4% paraformaldehyde for 2 hours and embedded in paraffin. Sections (ϳ5 m in thickness) were cut and stained with hematoxylin and eosin (H&E). The sections were then viewed under a light microscope to confirm the anticipated structural changes at the site of the laser burns. Sample Preparation and Assessing RNA Quality Enucleated control-and laser-treated eyes were harvested 3 days after laser treatment and maintained in RNA stabilization solution (RNAlater; Ambion, Austin, TX) for a minimum of 2 hours. To maximize the solution's penetration into the inner parts of the eye and to minimize RNA degradation, the eyes were slit before storage. The tissue was then dissected from the anterior segment and the lens. The resultant eyecups (comprising the retina and adjoining RPE layer and choroid) were pooled (n ϭ5 eyes per pool) to both maximize the amount of RNA obtained and to minimize bias due to biological variation. Total RNA was isolated with extraction reagent (TRIzol; Invitrogen, Life Technologies, San Diego, CA) and further purified using a kit (RNeasy; Qiagen, Valencia, CA). The RNA concentration was determined spectrophotometrically. Microarray Hybridization and Analysis Biotinylated cRNA samples were prepared as described by the manufacturer (Affymetrix, Santa Clara, CA) and hybridized onto arrays (Test 3 Arrays; Affymetrix), which provided information on the quality of the cRNA product, background levels, and enabled maintenance of quality control of the hybridization technique and the scanning equipment. "Spiked" controls were added to the hybridization cocktail, to enable quality control of the hybridization process. Samples of sufficient quality were then hybridized onto the gene chip standard arrays (MGU74Av2 GeneChip; Affymetrix). The sequence source for the MGU74Av2 array was largely the C57BL/6J mouse strain, making it an appropriate array for this study. Replicate hybridizations of three independent pooled samples were performed by using separate chips for both control and laser-treated samples. Statistical Analysis The raw-image data were analyzed using the accompanying software (GeneChip Expression Analysis Software; Affymetrix) to produce perfect match (PM) and mismatch (MM) values to which we applied our own analysis. Normalization of each array was performed using a method that is intended to make the PM and MM quantiles of all arrays agree, referred to as quantile normalization. 12 Normalized PM data were log transformed (base 2) and corrected for the effects of nonspecific binding by a novel background correction (for more details of this analysis procedure, see Ref. 13). The three replicate arrays of treatment and control were combined at the probe level, and the difference between the combined treatment and control arrays was calculated. For treatment arrays T 1 , T 2 , and T 3 and control arrays C 1 , C 2 , and C 3 the difference (d) between treatment and control for probe i of a given gene was formed: ͪ whereas the mean difference (d ) and SE over all probes i ϭ 1. . . n for a given gene were calculated by A t statistic was formed for the difference between treatment and control for each gene: t ϭ d /SE. Genes were ranked according to the size of their mean difference and t statistic. Ranking is performed on both statistics in an attempt to produce a list of genes with large mean differences and reasonable SEs. Requiring that the t statistic be large removes genes with large SEs from the list, and requiring that the mean difference also be large removes genes for which the t statistic is large, only because of very small SEs. Those genes for which both statistics are large relative to most genes are most likely to be differentially expressed. Quantitative Real-Time PCR Real time PCR (performed on three independent pooled samples) was used to confirm the gene expression data obtained in the microarray experiment. The RNA from eyecups of pooled groups of control and laser-treated eyes was reverse transcribed and used as a template in a real-time PCR reaction, using specific oligonucleotide primers and a fluorescent gel dye (SYBR green I; Applied Biosystems, Foster City, CA). The reaction was performed on a commercial system (RotorGene 2000; Corbett Research, Sydney, Australia). A standard curve was established for each PCR reaction. This was derived from the serial dilutions (10 Ϫ3 -10 Ϫ7 pmol/L) of a single-stranded, synthetic cDNA standard designed to represent the fragments of the genes listed later, and was used to calculate mRNA concentrations in the RNA samples. A "melt-curve" analysis of the PCR products was performed after the amplification, to demonstrate that only a single product was amplified. Oligonucleotide primers were designed to span intronic sequences to exclude the possibility of amplifying genomic DNA during the limited extension time of the PCR protocol (15 seconds). Primer IOVS, April 2003, Vol. 44, No. , 2490-2510, 2588-2570). RESULTS Changes of the Fundus, Retina-Choroidal Microvasculature, and Retinal Structure after Laser Photocoagulation The laser photocoagulation protocol (estimated to affect Յ25% of the fundus) adopted in this study produced laser burns of moderate whiteness, but without central heating bubble formation and subretinal or retinal hemorrhage Microarray Analysis: Genes Modulated by Laser Photocoagulation Approximately 4 g of total RNA was extracted per eyecup and confirmed to be intact by the generation of cRNA and subsequent hybridization to the microarrays. The 5Ј-to-3Ј ratios of the mouse housekeeping genes GAPDH and ␤-actin as well as the spiked controls were all below 3, and background readings were below 130, demonstrating that high-quality RNA was obtained and that the cRNA synthesis worked efficiently