Role of Glyceraldehyde 3-Phosphate Dehydrogenase in the Development and Progression of Diabetic Retinopathy

Diabetes Publish Ahead of Print, published online October 13, 2008

Mamta Kanwar and Renu A. Kowluru*

Kresge Eye Institute, Wayne State University, Detroit, MI

*Corresponding Author: Renu A. Kowluru, Ph.D. Kresge Eye Institute Wayne State University Detroit, MI 48201 E-Mail: [email protected]

Key Words: Glycemic control, Glyceraldehyde-3-phosphate dehydrogenase, Metabolic memory, Retinopathy

Submitted 21 April 2008 and accepted 2 October 2008.

This is an uncopyedited electronic version of an article accepted for publication in Diabetes. The American Diabetes Association, publisher of Diabetes, is not responsible for any errors or omissions in this version of the manuscript or any version derived from it by third parties. The definitive publisher-authenticated version will be available in a future issue of Diabetes in print and online at http://diabetes.diabetesjournals.org.

ABSTRACT

Objective: Mitochondrial superoxide levels are elevated in the retina in diabetes, and MnSOD overexpression prevents the development of retinopathy. Superoxide inhibits glyceraldehyde-3- phosphate dehydrogenase (GAPDH), which activates major pathways implicated in diabetic complications, including AGEs, protein kinase C and hexosamine pathway. Our aim is to investigate the role of GAPDH in the development and progression of diabetic retinopathy, and to elucidate the mechanism.

Research Design: Streptozotocin diabetic rats were either in poor control (PC, glycated hemoglobin>11%) for 12 months, or good control (GC, GHb<7) soon after induction of diabetes, or six months PC with six months GC (PC-GC). Retinal GAPDH, its ribosylation and nitration and AGEs and PKC activation were determined, and correlated with microvascular histopathology.

Results: In PC rats retinal GAPDH activity and expressions were subnormal with increased ribosylation and nitration (25-30%). GAPDH activity was subnormal in both cytosol and nuclear fractions, but its protein expression and nitration were significantly elevated in nuclear fraction. Re-institution of good control failed to protect inactivation of GAPDH, its covalent modification and translocation to the nucleus. PKC, AGEs and hexosamine pathways remained activated, and microvascular histopathology unchanged. However, GAPDH and its translocation in GC rats were similar to those in normal rats.

Conclusions: GAPDH plays a significant role in the development of diabetic retinopathy and its progression after cessation of hyperglycemia. Thus, therapies targeted towards preventing its inhibition may inhibit development of diabetic retinopathy and arrest its progression.

GADPH and diabetic retinopathy

etinopathy is a multifactorial sight- diabetic patients, but its effects on the threatening complication of diabetes. progression of retinopathy are not immediate, RIt is a progressive disease associated and it takes years for retinopathy to halt with chronic hyperglycemia (1). Although progression after the re-establishment of GC. many glucose-induced retinal metabolic The imprinted effects of prior glycemic abnormalities are postulated to contribute to control either produce the long lasting its development, the exact mechanism benefits of GC, or resist the arrest of remains elusive (2-5). We have shown that in progression of diabetic retinopathy after re- diabetes retinal mitochondria experience institution of GC. Re-institution of GC after a increased oxidative damage and the profound period of poor glycemic control mitochondrial enzyme that scavenges (PC) does not immediately benefit the superoxide (MnSOD) prevents vascular progression of retinopathy. This suggests a histopathology that is characteristic of ‘metabolic memory’ phenomenon. (17-20). diabetic retinopathy (6-8). Increased Metabolic memory phenomenon is observed mitochondrial superoxide production also in animal models of diabetic retinopathy inactivates glyceraldehyde-3-phosphate (21-26), the formation of acellular capillaries, dehydrogenase (GAPDH) in vascular characteristic of early signs of diabetic endothelial cells, and inhibition of GAPDH is retinopathy, does not stop for at least six postulated to activate some of the key months when GC is initiated six months after pathways that are associated with the induction of diabetes in rats, and nitrotyrosine development of diabetic complications, levels and oxidative stress remain elevated including increased formation of advanced (24, 26). These abnormalities are, however, glycation end-products (AGEs) and activation partially inhibited if the duration of PC is of protein kinase C (PKC) and hexosamine reduced to two months, suggesting the role of pathway (9, 10). oxidative stress in the metabolic memory GAPDH is a glycolytic enzyme that phenomenon (24). The role of GAPDH in catalyzes the conversion of glyceraldehyde-3- metabolic imprinting remains to be phospahte to 1, 3-bis-phosphoglycerate. elucidated. Recent studies have shown that GAPDH is a In the present study we have investigated protein with multiple cytoplasmic, membrane how GAPDH inhibition contributes to the and nuclear functions, and is a major development of retinopathy in diabetes, and intracellular messenger mediating apoptosis the mechanism(s) that could result in its of cells (11, 12). GAPDH translocation to the inactivation. We have also explored the role nucleus is considered as an important step in of GAPDH in the metabolic memory glucose-induced apoptosis of retinal Muller phenomenon in diabetic rats by maintaining cells (13). The mechanism that initiates its them in PC before initiation of GC. translocation is not well understood; covalent modification by nitration/ribosylation is METHODS considered as a likely possibility (14-16). Animals and glycemia. Lewis rats (male, How GAPDH contributes to the pathogenesis 200g) were randomly assigned to normal or of diabetic retinopathy remains to be diabetic groups. Diabetes was induced with established. streptozotocin (55mg/kg BW), and rats were Good glycemic control (GC) attenuates divided at random among three groups the development/progression of retinopathy in according to intended degree of glycemic 3

GADPH and diabetic retinopathy control. Rats in group 1 remained in PC for 0.5mmoles/L dithiolthretol and protease 12 months, in group 2 rats were in PC for six inhibitors. The homogenate was centrifuged months followed by GC for six additional at 750xg for 5 minutes, and the supernatant months (PC-GC), and in group 3 rats were was centrifuged at 5000xg for 15 min. The subjected to GC soon after induction of nuclear pellet was resuspended in diabetes (GC group). Some of the same rats 50mmoles/L HEPES buffer (pH 7.5) have been used by us in our previous studies containing 1% triton X-100, 150mmoles/L (26, 27). sodium chloride, 1mmoles/L EDTA and The degree of glycemia was achieved by protease inhibitors. The supernatant was adjusting the dose and frequency of insulin further centrifuged at 105,000xg for 90 min to (NPH) administration. The rats in PC group separate the cytosolic fraction (13, 28). The received a single injection of insulin (1-2 purity of the fractions was determined by units) 4-5 times a week to prevent ketosis and measuring the expressions of Histone 2-B weight loss, and the rats in GC group received (H2-B, nuclear marker) and lactate insulin twice daily (7-8 units total) to dehydrogenase (LDH, cytosolic marker). maintain their blood glucose levels below 150 GAPDH enzyme activity. The enzyme mg/dl and a steady gain in body weight (24, activity was measured spectrophotometrically 26). The entire rat colony was housed in in a final assay volume of 100μl containing metabolism cages: 24 hour urine samples 50mM triethanolamine buffer (pH 7.6), were measured daily and tested for 50mmoles/L arsenate, 2.4mmoles/L GSH, glycosuria. Blood glucose was measured once 250μmoles/L M NAD and cytosolic/nuclear a week using Elite Glucometer (Bayer protein (0-10μg protein). The assay mixture Corporation, Tarrytown, NY), and glycated was pre-incubated for 5 min at 37oC, and the hemoglobin (GHb) every two months. The reaction was initiated by 100μg/mL entire rat colony received powdered diet glyceraldehyde-3-phosphate. Increase in (PURINA 5001, TestDiet, Richmond, IN); NADH production was monitored at 340nm. and their food consumption was measured GAPDH activity was expressed as difference once and body weights 2-3 times every week. in absorbance in the presence/absence of These experiments conformed to the ARVO glyceraldehyde-3-phosphate (28). Resolution on Treatment of Animals in Quantitative real-time PCR (Q-RT-PCR). Research, as well as to the specific RNA was isolated from the retina using Institutional Guidelines. The experiment was TRIzol reagent, and 1μg RNA was converted terminated 12 months after initiation of to single stranded cDNA and quantified diabetes and the animals were sacrificed by an spectrophotometerically. Gene expression overdose of pentobarbital. Eyes were was measured using 90-300ng cDNA enucleated immediately; one eye was fixed in templates in 96 well plates in ABI-7500 10% buffered formalin, and from the other sequence detection system (29, 30). Each eye the retina was immediately isolated by sample was analyzed in triplicate, and the gently separating sensory retina from choroid data normalized to β2-macroglobulin (B2-M) using a micro-spatula. expression in each sample. GenBank Preparation of subcellular fractions. Retina accession numbers for the ABI TaqMan was gently homogenized in a glass assays for GAPDH and B2-M used were NM- homogenizer in 50mmoles/L glycyl glycine 017008.3 and NM_012512.1 respectively. buffer (pH 7.0) containing 10mmoles/L The fold change in gene expression relative to EDTA, 100mmoles/L sodium fluoride, 4

GADPH and diabetic retinopathy

normal was calculated using the ddCT AGEs formation was determined by western method. blot using anti-AGE antibody (Wako Protein expression. Protein (15-30μg) was Chemicals, Richmond, VA). PKC activation separated by SDS-PAGE on a 4-16% gradient was determined by quantifying the expression gel and blotted to nitrocellulose membrane. of PKC βII, the isoform that is activated in The membranes were blocked in 5% nonfat diabetes, as previously described by us (23). milk, incubated with the target primary Since addition of single O-linked ß-N- antibody, washed, followed by incubation acetylglucosamine (O-GlcNAc) with appropriate horseradish peroxidase- monosaccharides to serine or threonine coupled secondary antibody. The target residues on proteins is one of the processes proteins were enhanced by ECL reagent, and coupled to the hexosamine pathway, we determined by autoradiography. The assessed hexosamine pathway by quantifying membranes then were stripped and re-probed O-linked N-acetylglucosamine modified with β-actin (Sigma-Aldrich, St Louis, MO). proteins in the retina by western blot using The band intensity was quantified using Un- monoclonal antibody against O-GlcNAc Scan-It Gel digitizing software (Silk (Covance Inc, Princeton, NJ). Bovine serum Scientific Inc, Orem, UT), and protein albumin was used as a blocking medium (30). expression levels were calculated relative to Isolation of retinal microvessels and β-actin in the same sample. Activity of quantification of acellular capillaries. poly(ADP-ribose) polymerase (PARP) was Microvessels were prepared from formalin determined by measuring poly(ADP- fixed eyes by trypsin digestion. The retina ribosyl)ation of retinal proteins by separating was isolated, rinsed in water overnight, and them on SDS-PAGE gel. incubated with 3% crude trypsin containing Ribosylation and nitration of GAPDH. 0.2mmol/L sodium fluoride for 90 minutes at Covalent modification of GAPDH was 37oC. Nonvascular cells were removed by determined by first immunoprecipitating gentle brushing, and the isolated vasculature protein (75-100μg) with polyclonal anti- was dried onto a microscope slide. The slides GAPDH antibody. Protein A/G plus agarose were stained with hematoxylin and periodic beads were used to collect GAPDH acid-schiff. The number of acellular complexed with the antibody, and analyzed capillaries (representing basement membrane by western blot technique using monoclonal tubes lacking cell nuclei), was counted in a antibodies against poly ADP-ribose (PAR; masked manner in multiple mid-retinal fields Alexis Biochemicals, San Diego, CA) and with one field adjacent to each of the 5 to 7 nitrotyrosine (Upstate Biotechnology, Lake retinal arterioles radiating out from the optic Placid, NY). To normalize for equal loading disc, and expressed as per mm2 of retinal area in each lane, the membranes were re-probed examined (5, 8, 26). for GAPDH. Results are presented as mean ± SD and Quantification of GAPDH-mediated analyzed statistically using the nonparametric pathways. Since inhibition of GAPDH Kruskal-Wallis test followed by Mann- activates major pathways that are implicated Whitney test for multiple group comparison. in the development of diabetic complications, Similar conclusions were achieved by using we investigated the effect of reversal of ANOVA with Fisher or Tukey. hyperglycemia on AGEs, PKC and hexosamine pathways in the retina from the RESULTS same set of animals used for GAPDH. Total 5

GADPH and diabetic retinopathy

Severity of hyperglycemia. Hyperglycemia, rats, despite increased expression of GAPDH, as reported earlier (26), was severe in the rats the glycolytic activity was decreased by in PC group; GHb values were over 11% almost 70% in the nuclear fraction (Figure throughout the entire duration of the 5a). GC group had similar expressions as experiment (12 months). The rats in GC normal control rats (data not shown). group maintained their GHb values similar to Ribosylation and nitration have been those in normal rats. In PC-GC group, GHb shown to inhibit GAPDH activity (15, 31). values before initiation of GC were not We determined the levels of its ribosylation different from the PC group (GHb>11%), but and nitration of retinal GAPDH. Twelve became similar to those in normal group after months of poor PC resulted in approximately initiation of GC (GHb<7%) (Table I). 25% increase in ribosylation of GAPDH Average body weight and 24 hour urine (Figure 3b) and 30% increase in nitration volumes were similar in GC rats and normal compared to the normal rat retina (Figure 3c). rats. The rats in GC group had similar expressions Effect of diabetes on retinal GAPDH. as normal control rats (data not shown). The Twelve months of PC had a marginal, but ratio of nitrated GAPDH was 2.5 to 1 in significant effect on the expression of normal rat retina, and increased to almost 1:1 GAPDH in the retina; protein expression was in PC group. However, in the nuclear fraction, decreased by 15-20% in diabetic rats compar nitrated GAPDH was about two fold higher in was decreased by about 20% (Figure 2). PC rats compared to normal rats (Figure 5b), Effect of diabetes on activation of retinal strongly suggesting that the enzyme in PARP. As shown in figure 3a, 12 months of nuclear fraction is mainly in its covalently PC significantly increased poly(ADP- modified and inactivated form. ribosyl)ation of retinal proteins compared to Effect of reversal of hyperglycemia on normal rats suggesting increase in PARP GAPDH. When diabetic rats were allowed to activity. We did not, however, identify these remain in PC for six months before institution ribosylated proteins (other than GAPDH, of GC, the protein and gene expressions of please see below), and that is beyond the retinal GAPDH remained subnormal, and focus of this study. PARP remained activated; values obtained in Subcellular translocation of GAPDH and PC and PC-GC groups were not different its covalent modification. Since retinal cell from each other (Figures 1, 2 and 3a). In apoptosis precedes the development of addition, both ribosylation and nitration of diabetic retinopathy (32, 33), we investigated GAPDH were also not different in the retina the effect of diabetes on subcellular from PC group and PC-GC group suggesting translocation of GAPDH in the retina. In that six months of GC had no beneficial effect normal rat retina the protein expression of on the covalent modification of the enzyme GAPDH was 35% higher in the cytosolic (Figure 3b and c). The expression and activity fraction compared to the nuclear fraction, but of GAPDH in nuclear fraction of retina from 12 months of PC in rats resulted in reduction PC-GC group were significantly different in its expression in the cytosol fraction with a from those in normal rats, and the enzyme concomitant increase in the nuclear fraction remained nitrated (Figures 4 and 5). However, (Figure 4). The ratio of GAPDH expression in these values were similar to those obtained cytosolic and nuclear fraction was almost 4 from PC group, suggesting that six months of to1 in normal rat retina, and decreased to 1.2 GC did not prevent the enzyme from to1 in diabetic rat retina. In the same diabetic translocating to the nucleus, and thus failed to 6

GADPH and diabetic retinopathy

protect the retina from apoptosis. But, when play a significant role in cell death (12), and the rats were maintained in GC soon after its inhibition is considered to activate major induction of diabetes (GC group), GAPDH pathways of endothelial cell damage gene expression was similar to that obtained including activation of PKC, hexosamine from normal rat retina (Figure 2). pathway flux, and AGEs formation (9, 10). Effect of reversal of hyperglycemia on Here we show that diabetes inhibits GAPDH GAPDH-mediated pathways. As shown in activity in the retina, and its expression figure 6a, multiple protein bands with becomes subnormal. The enzyme is increased AGEs were observed in the retina translocated from cytosol to the nucleus, and from rats in PC group compared with those GAPDH-mediated downstream and upstream from normal rats, and protein staining from signaling pathways (AGEs, PKC and the rats in PC-GC group was similar to those hexosamine pathway) are activated. Our data obtained from the rats in PC group. We, suggests that the enzyme translocated to the however, did not identify the retinal proteins nuclear fraction is covalently modified. that had increased AGEs. In the same retina Further, we provide exciting data samples, reversal of hyperglycemia had no demonstrating that re-institution of GC after effect on increased PKCβII (Figure 6b), the six months of PC does not produce any enzyme expression remained significantly beneficial effects on the inactivation of retinal elevated in both PC and PC-GC groups GAPDH. The enzyme remains inactive with compared to the normal group of rats its expression elevated in the nucleus (p<0.05). Similarly, six months of GC did not suggesting that inhibition of GAPDH and its produce any reduction in O-GlcNAcylation of subcellular translocation resist reversal after retinal proteins (data not shown), confirming re-institution of GC. Re-institution of GC also that reversal of hyperglycemia had no fails to provide any benefit to the covalent beneficial effect on GAPDH-mediated modification; the elevated levels of both downstream and upstream signaling ribosylation and nitration are sustained for at pathways. least six months after reversal of PC. In the Effect of reversal of hyperglycemia on the same animals the signaling pathways that are retinal histopathology. Poor glycemic in direct control of GAPDH remain activated control in rats for 12 months increased the in the retina, and the number of acellular number of acellular capillaries in the retinal capillaries elevated after re-establishment of vasculature (Figure 7) by about 4 fold GC. These novel and exciting observations compared with normal rats. Six months of GC strongly suggest a role for GAPDH in the that followed six months of PC failed to development and progression of diabetic provide any protection, the number of retinopathy. acellular capillaries was similar in PC and Although others have reported decreased PC-GC rats (average number of acellular retinal GAPDH activity at three months of capillaries/mm2 retina in rats in normal, PC diabetes in rats (34), ours is the first report and PC-GC groups=1.5, 6.1 and 6.8 showing that the enzyme remains inhibited at respectively). duration of diabetes in rats when signs of retinopathy can be detected. High glucose DISCUSSION decreases GAPDH in vascular cells, GAPDH, a classic glycolytic enzyme, is putatively due to overproduction of implicated in diverse cytoplasmic, membrane superoxide by mitochondrial electron and nuclear activities, and has been shown to transport chain (9). Retinal mitochondria are 7

GADPH and diabetic retinopathy

dysfunctional in diabetes and superoxide implying that nitration is associated with its levels are elevated (7, 8, 35), and complex III inhibition. In support, covalent modification is considered as one of the sources of of GAPDH by nitration is also observed in increased superoxide (8). Overexpression of other pathological conditions associated with MnSOD that is shown to prevent glucose- inflammation (44), and diabetic retinopathy is induced inhibition of GAPDH in vascular believed to be a low grade chronic cells (9) also prevents the development of inflammatory disease (45-48). Peroxynitrite retinopathy in diabetic mice (8). Thus, taken itself damages DNA and triggers activation of together, data strongly implicate the role for PARP (49); we clearly show that ribosylation GAPDH in the pathogenesis of retinopathy in of GAPDH is also increased in diabetes. diabetes. Further, reduction in retinal GAPDH Mitochondrial superoxide break DNA expression (gene and protein) suggests that, in strand and activate PARP, and PARP- addition to its covalent modification, the gene mediated poly (ADP-ribosyl)ation of GAPDH transcript of GAPDH is decreased in diabetes. is considered as one of the mechanisms in the GAPDH is also a major intracellular inhibition of GAPDH activity in messenger that mediates cell death via hyperglycemic conditions (14). Our data apoptosis. Covalent modification of GAPDH show that PARP activity is significantly is suggested to trigger its translocation from increased in PC group, and this is in cytosol to the nuclear fraction (12, 13). accordance with the other published report During nuclear translocation its activity is lost (36). Increased PARP activity plays an (50), and increase in hydrophobicity due to its important role in diabetes-induced retinal nitration is being postulated as one of the capillary cell death and inhibitors of PARP mechanisms (16). Nuclear translocation of prevent retinal leukostasis, oxidative stress GAPDH is suggested to play a role in retinal and retinopathy in diabetic rats (36-39). Here glial cell apoptosis in hyperglycemic we show that increased ribosylation of retinal conditions (13), and increased apoptosis of GAPDH could be one of the mechanisms retinal capillary cells precedes the responsible for its inactivation. development of retinal pathology associated GAPDH is susceptible to nitration by with diabetic retinopathy (33). Our results peroxynitrite (formed by reaction between show that the expression of GAPDH is nitric oxide and superoxide); steady-state increased in nuclear fraction in the retina from exposure of GAPDH to low doses of diabetic rats compared to normal rats; peroxynitrite in rat astrocytes results in its however, the enzyme in the nuclear fraction inhibition (15, 16). GAPDH is also a target appears to be in a more nitrated and for inactivation by nitric oxide in endothelial inactivated state compared to the cytosol cells (40), and these cells are one of the fraction. This implies that, although diabetes microvascular cells in the retina that present increases GAPDH translocation to the pathology of diabetic retinopathy. Nitric nucleus, it is covalently modified before being oxide and peroxynitrite levels are elevated in translocated from the cytosol, and is in the retina and its capillary cells, and these inactivated state. levels remain elevated at duration when Re-institution of GC after six months of vascular histopathology characteristic of PC failed to produce any significant retinopathy is developing in diabetic rats (5, beneficial effects on GAPDH, but if GC was 6, 25, 41-43). Our results clearly show that initiated soon after induction of diabetes in diabetes increases nitration of retinal GAPDH rats and allowed to continue for 12 months, 8

GADPH and diabetic retinopathy

GAPDH remained similar to that obtained months after hyperglycemia is terminated, from age-matched normal rats. This strongly further strengthening the role of GAPDH in supports that GAPDH, in addition to being a the development/progression of diabetic central player in the development of diabetic retinopathy. retinopathy, is important in the metabolic Data presented here clearly show that the memory phenomenon. Reversal of rats that presented no effect of reversal of hyperglycemia failed to provide any benefit to hyperglycemia on GAPDH and its the activation of PARP, the enzyme remained translocation to the nucleus also showed no activated even after six months of GC. on the development of retinopathy, the Sustained increases in nitration and number of acellular capillaries remains ribosylation of retinal GAPDH after reversal comparable in the rats in PC and in PC-GC of hyperglycemia in rats suggests that GC groups. This confirms that GAPDH-mediated fails to provide any benefit to the covalent pathways, and the process of apoptosis of modifications of the enzyme. Further, GC did retinal capillary cells that begins before not prevent translocation of the retinal histopathology can be detected in the retinal enzyme from cytosol to the nucleus, vasculature (25), continues to progress even suggesting that GAPDH remains proapoptotic after hyperglycemia is terminated. The even after GC is re-established. Due to tissue analyses of GAPDH, its translocation and availability, apoptosis of retinal capillary cells covalent modification, and the consequences was not measured in this study; however, in of its inhibition on the pathways were support of continued apoptosis our previous performed in the whole retina samples, and studies have shown that the apoptosis this approach did not allow us to identify the execution enzyme caspase-3 remains active in specific cell type. The failure to reverse the the retina even after reversal of development of retinal vascular hyperglycemia in rats. The process of histopathology by re-institution of GC, caspase-3 activation that starts before the however, strongly suggests that GAPDH has a appearance of retinal histopathology resists significant role in the development and reversal by re-institution of GC (25), and progression of diabetic retinopathy. nuclear accumulation of covalently modified In conclusion, we have provided strong GAPDH could be one of the important factors evidence demonstrating that GAPDH is associated with increased apoptosis and inhibited and its downstream and upstream histopathology of diabetic retinopathy. signaling pathway activated in the retina in Inhibition of GAPDH increases the levels diabetes at a duration when histopathology of glycolytic metabolite glyceraldehyde 3- characteristic of retinopathy can be observed phosphate, and this activates the major in rats. The covalent modification of the pathways implicated in diabetic complications enzyme is increased. Diabetes accelerated including AGEs formation, PKC activation translocation of retinal GAPDH into the and hexosamine pathway (10). Here we show nucleus, suggesting its pro-apoptotic nature that the reversal of hyperglycemia in rats, in possibly contributing to the development of addition to failing to provide any benefit to diabetic retinopathy. In addition to its role in the inhibition of retinal GAPDH and its the pathogenesis of diabetic retinopathy, nuclear translocation, has no significant effect GAPDH is also an important in metabolic on GAPDH-mediated pathways. Thus, both, memory phenomenon. Sustained nitration and the downstream and upstream consequences ribosylation are possible mechanisms for of GAPDH inhibition persist for at least six resistance of GAPDH to reverse inhibition 9

GADPH and diabetic retinopathy after reversal of hyperglycemia. Therapies Authors thank Dr. Bindu Menon for targeted towards preventing GAPDH technical assistance, and Yakov Shamailov inhibition by blocking its covalent and Divyesh Sarman for their help in modification should help in the development maintaining the rats. This study was and also in arresting the progression of supported in part by grants from the National diabetic retinopathy, a sight threatening Institutes of Health, Juvenile Diabetes complication of diabetes. Research Foundation, The Thomas Foundation, and Research to Prevent ACKNOWLEDGMENTS Blindness.

REFERENCES 1. Engerman RL, Kern TS. Hyperglycemia as a cause of diabetic retinopathy. Metabolism 1986;35:20-23. 2. Baynes JW. Role of oxidative stress in development of complications in diabetes. Diabetes 1991;40:405-412. 3. Xia P, Inoguchi T, Kern TS, Engerman RL, Oates PJ, King GL. Characterization of the mechanism for the chronic activation of DAG-PKC pathway in diabetes and hypergalactosemia. Diabetes 1994;43:1122-1129. 4. Aiello LP. Vascular endothelial growth factor and the eye: biochemical mechanisms of action and implications for novel therapies. Ophthalmic Res 1997;29:354-362. 5. Kowluru RA, Tang J, Kern TS. Abnormalities of retinal metabolism in diabetes and experimental galactosemia. VII. Effect of long-term administration of antioxidants on the development of retinopathy. Diabetes 2001;50:1938-1942. 6. Kowluru RA, Atasi L, Ho YS. Role of mitochondrial superoxide dismutase in the development of diabetic retinopathy. Invest Ophthalmol Vis Sci 2006;47:1594-1599. 7. Kowluru RA, Kowluru V, Ho YS, Xiong Y. Overexpression of mitochondrial superoxide dismutase in mice protects the retina from diabetes-induced oxidative stress. Free Rad Biol Med 2006;41:1191-1196. 8. Kanwar M, Chan PS, Kern TS, Kowluru RA. Oxidative damage in the retinal mtochondria of diabetic mice: possible protection by superoxide dismutase. Invest Ophthalmol Vis Sci 2007;48:3805-3811. 9. Du X, Matsumura T, Edelstein D, Rossetti L, Zsengeller Z, Szabo C, et al. Inhibition of GAPDH activity by poly(ADP-ribose) polymerase activates three major pathways of hyperglycemic damage in endothelial cells. J Clin Invest 2003;112:1049-1057. 10. Brownlee M. The pathobiology of diabetic complications: a unifying mechanism. Diabetes 2005;54:1615-1625. 11. Saunders PA, Chalecka-Franaszek E, Chuang DM. Subcellular distribution of glyceraldehyde-3-phosphate dehydrogenase in cerebellar granule cells undergoing cytosine arabinoside-induced apoptosis. J Neurochem 1997;69:1820-1828. 12. Hara MR, Cascio MB, Sawa A. GAPDH as a sensor of NO stress. Biochim Biophys Acta 2006;1762:502-509. 13. Kusner LL, Sarthy VP, Mohr S. Nuclear translocation of glyceraldehyde-3-phosphate dehydrogenase: a role in high glucose-induced apoptosis in retinal Muller cells. Invest Ophthalmol Vis Sci 2004;45:1553-1561. 10

GADPH and diabetic retinopathy

14. Deveze-Alvarez M, Garcia-Soto J, Martinez-Cadena G. Glyceraldehyde-3-phosphate dehydrogenase is negatively regulated by ADP-ribosylation in the fungus Phycomyces blakesleeanus. Microbiology 2001;147:2579-2584. 15. Buchczyk DP, Grune T, Sies H, Klotz LO. Modifications of glyceraldehyde-3-phosphate dehydrogenase induced by increasing concentrations of peroxynitrite: early recognition by 20S proteasome. Biol Chem 2003;38:237-241. 16. Batthyany C, Schopfer FJ, Baker PR, Durán R, Baker LM, Huang Y, et al. Reversible post- translational modification of proteins by nitrated fatty acids in vivo. J Biol Chem 2006;281:450- 463. 17. Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993;329:977-986. 18. Diabetes Control and Complications Trial Research Group. Early worsening of diabetic retinopathy in the diabetes control and complication trial. Arch. Ophthalm. 1998;116:874-886. 19. Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Research Group. Effect of intensive therapy on the microvascular complications of type 1 diabetes mellitus. JAMA 2002;287:2563-2569. 20. LeRoith D, Fonseca V, Vinik A. Metabolic memory in diabetes--focus on insulin. Diabetes Metab Res Rev 2005;21:85-90. 21. Engerman RL, Kern TS. Progression of incipient diabetic retinopathy during good glycemic control. Diabetes 1987;36:808-812. 22. Hammes H-P, Klinzing I, Wiegand S, Bretzel RG, Cohen AM, Federlin K. Islet transplantation inhibits diabetic retinopathy in the sucrose-fed diabetic Cohen diabetic rat. Invest Ophthalmol Vis Sci 1993;34:2092-2096. 23. Kowluru RA, Koppolu P. Termination of experimental galactosemia in rats, and progression of retinal metabolic abnormalities. Inves Ophthalmol Vis Sci 2002;43:3287-3291. 24. Kowluru RA. Effect of re-institution of good glycemic control on retinal oxidative stress and nitrative stress in diabetic rats. Diabetes 2003;52:818-823. 25. Kowluru RA, Chakrabarti S, Chen S. Re-Institution of good metabolic control in diabetic rats on the activation of caspase-3 and nuclear transcriptional factor (NF-kB) in the retina. Acta Diabetologica 2004;44:194-199. 26. Kowluru RA, Kanwar M, Kennedy A. Metabolic memory phenomenon and accumulation of peroxynitrite in retinal capillaries. Exp Diabetes Res 2007;2007:2196. 27. Chan PS, Kanwar M, Kowluru RA. Resistance of retinal inflammatory mediators to suppress after re-institution of good glycemic control: Novel mechanism for metabolic memory. J Diabetes & its Compl 2008; under editorial review. 28. Veluthakal R, Khan I, Tannous M, Kowluru A. Functional inactivation by interleukin-1beta of glyceraldehyde-3-phosphate dehydrogenase in insulin-secreting cells. Apoptosis 2000;7:241- 246. 29. Kowluru RA, Menon B, Gierhart D. Beneficial effect of zeaxanthin on retinal metabolic abnormalities in diabetic rat. Inves Ophthalmol Vis Sci 2008;49:1645-1651. 30. Zachara NE, Hart GW. O-GlcNAc a sensor of cellular state: the role of nucleocytoplasmic glycosylation in modulating cellular function in response to nutrition and stress. Biochim Biophys Acta 2004;1673:13–28..

GADPH and diabetic retinopathy

31. Kiss L, Szabo C. The pathogenesis of diabetic complications: the role of DNA injury and poly(ADP-ribose) polymerase activation in peroxynitrite-mediated cytotoxicity. Mem Inst Oswaldo Cruz 2005;100:29-37. 32. Mizutani M, Gerhardinger C, Lorenzi M. Muller cell changes in human diabetic retinopathy. Diabetes 1998;47:455-459. 33. Kern TS, Tang J, Mizutani M, Kowluru R, Nagraj R, Lorenzi M. Response of capillary cell death to aminoguanidine predicts the development of retinopathy: Comparison of diabetes and galactosemia. Invest Ophthalmol Vis Sci 2000;41:3972-3978. 34. Ola MS, Berkich DA, Xu Y, King MT, Gardner TW, Simpson I, et al. Analysis of glucose metabolism in diabetic rat retinas. Am J Physiol Endocrinol Metab 2006;290:E1057-E1067. 35. Kowluru RA, Abbas SN. Diabetes-induced mitochondrial dysfunction in the retina. Inves Ophthal Vis Sci 2003;44:5327-5334. 36. Zheng L, Szabo C, Kern TS. Poly(ADP-ribose) polymerase is involved in the development of diabetic retinopathy via regulation of nuclear factor-kappaB. Diabetes 2004;53:2960-2967. 37. Sugawara R, Hikichi T, Kitaya N, Mori F, Nagaoka T, Yoshida A, et al. Peroxynitrite decomposition catalyst, FP15, and poly(ADP-ribose) polymerase inhibitor, PJ34, inhibit leukocyte entrapment in the retinal microcirculation of diabetic rats. Curr Eye Res 2004;29(11- 16). 38. Obrosova IG, Minchenko AG, Frank RN, Seigel GM, Zsengeller Z, Pacher P, et al. Poly(ADP-ribose) polymerase inhibitors counteract diabetes- and hypoxia-induced retinal vascular endothelial growth factor overexpression. Int J Mol Med 2004;14:55-64. 39. Xu B, Chiu J, Feng B, Chen S, Chakrabarti S. PARP activation and the alteration of vasoactive factors and extracellular matrix protein in retina and kidney in diabetes. Diabetes Metab Res Rev 2008; 24:404-412. 40. Padgett CM, Whorton AR. S-nitrosoglutathione reversibly inhibits GAPDH by S- nitrosylation. Am J Physiol 1995;269:C739-749. 41. Du Y, Smith MA, Miller CM, Kern TS. Diabetes-induced nitrative stress in the retina, and correction by aminoguanidine. J Neurochem 2002;80:771-779. 42. Kowluru RA, Koppolu P, Chakrabarti S, Chen S. Diabetes-induced activation of nuclear transcriptional factor in the retina, and its inhibition by antioxidants. Free Radic Research 2003;37:1169-1180. 43. Kowluru RA, Odenbach S. Effect of long-term administration of alpha lipoic acid on retinal capillary cell death and the development of retinopathy in diabetic rats. Diabetes 2004;53:3233- 3238. 44. Kanski J, Behring A, Pelling J, Schöneich C. Proteomic identification of 3-nitrotyrosine- containing rat cardiac proteins: effects of biological aging. Am J Physiol Heart Circ Physiol 2005;288:H371-H381. 45. Joussen AM, Poulaki V, Le ML, Koizumi K, Esser C, Janicki H, et al. A central role for inflammation in the pathogenesis of diabetic retinopathy. FASEB J 2004;18:1450-1452. 46. Kowluru RA, Odenbach S. Role of interleukin-1beta in the development of retinopathy in rats: effect of antioxidants. Invest Ophthalmol Vis Sci 2004;45:4161-4166. 47. Kern TS. Contributions of inflammatory processes to the development of the early stages of diabetic retinopathy. Exp Diabetes Res 2007;2007:95103. 48. Adamis AP, Berman AJ. Immunological mechanisms in the pathogenesis of diabetic retinopathy. Semin Immunopathol 2008;30:65-84. 12

GADPH and diabetic retinopathy

49. Pacher P, Szabo C. Role of poly(ADP-ribose) polymerase-1 activation in the pathogenesis of diabetic complications: endothelial dysfunction, as a common underlying theme. Antioxid Redox Signal 2005;11-12:1568-1580. 50. Sawa A, Khan AA, Hester LD, Snyder SH. Glyceraldehyde-3-phosphate dehydrogenase: nuclear translocation participates in neuronal and non neuronal cell death. Proc Natl Acad Sci USA 1997;94:11669-11674.

TABLE 1 Degree of glycemia in rats assigned to different glycemic control

Duration Body Weight GHb Urine Vol (months) (g) (%) (ml/24 hrs)

Normal (9) 12 421±35 6.7±0.8 13±6

PC (13) 12 289±37 12.7±1.7 112±39

GC (7) 12 454±41 6.9±1.1 26±10

PC (11) 6 272±27 13.1±1.6 135±19 ↓ ↓ ↓ ↓ ↓ GC 6 409±26 7.1±0.9 19±11

The rats were weighed two times every week and their food consumption once very week. Body weight is the mean value during the entire duration of the intended metabolic control. GHb was quantified. The values are means ± SD, and the numbers in parentheses represent number of rats in each group.

GADPH and diabetic retinopathy

FIGURE LEGENDS

Figure 1: Protein expression of GAPDH in diabetes, and effect of reversal of glycemic control. GAPDH expression was determined in the retinal homogenate by western blot technique using rabbit polyclonal GAPDH antibody. Equal loading of the sample in each lane was ensured by determining the expression of β-actin. The western blots shown here are representative of at least 5 different rats in each group, and the bars represent the mean ± SD of the adjusted band intensities obtained from those rats. Normal PC PC-GC

GAPDH β-Actin

120

* * # 80

40 GAPDH:Actin band intensity

0 Normal PC PC-GC

Figure 2: Gene expression of GAPDH in the retina and effect of re-institution of good glycemic control. Retinal gene expression was measured with Q-RT-PCR and the values were normalized to the expression of the housekeeping gene B2M in the same sample. The value obtained from the retina of age-matched normal rats is considered 100%. Data represent the mean ± SD from 8- 10 rats in each of the 4 groups. *P<0.05 compared to normal, and #P>0.05 compared to PC. 120

* *# 80

40 mRNA expression (% normal)

0 Normal PC PC-GC GC

GADPH and diabetic retinopathy

Figure 3: PARP activity in the retina, and covalent modification of GAPDH: (a) PARP activity was determined in the retinal extract by western blot technique. Poly(ADP-ribosyl)ated proteins were detected using antibody obtained from Santa Cruz Biotechnology. To determine covalent modification of GAPDH, it was immunoprecipated from retinal proteins, and analyzed by western blot technique using monoclonal antibodies against either poly PAR or nitrotyrosine. To ensure equal loading, the membranes were re-probed for GAPDH. The histograms represent ribosylation (b) or nitration (c) of retinal GAPDH from 5-6 rats in each group, and the values from normal rat retina are considered as 100%. *P<0.05 compared to normal, and #P>0.05 compared to PC. A BCNorm PC PC-GC Norm PC PC-GC Ribos-GAPDH Ribos-GAPDH MW Norm PC PC-GC GAPDH GAPDH

131kD 140 140 * *# 86kD (ADP-ribose) group * *# 44kD } 105 105 33kD 70 70 β-actin Nitration (% normal)

Ribosylation (% normal) 35 35

0 0 Normal PC PC-GC Normal PC PC-GC

Figure 4: Effect of diabetes on subcellular localization of GAPDH: Subcellular fractionation was performed on retinal homogenate by centrifugation. GAPDH expression was determined by western blot. Histone 2B and lactate dehydrogenase (LDH) were used to determine the purity of nuclear and cytosolic fractions respectively. The histogram represents the relative expression of GAPDH in cytosolic and nuclear fractions, and the total expression of GAPDH in cytosol and nuclear fraction is considered as 100%. The values obtained are mean ± SD from 4 or more rats in each of the 4 groups. *P<0.05 compared to normal, and #P>0.05 compared to PC. ABCytosol Nucleus

Norm PC PCGC Norm PC PCGC

GAPDH GAPDH

LDH LDH

100

75 *# * * *# 50 GADPH band intensity 25

0 Cyt Nuc Cyt Nuc Cyt Nuc Normal PC PC-GC 15

GADPH and diabetic retinopathy

Figure 5: Glycolytic activity and covalent modification of GAPDH in cytosolic and nuclear fractions: a. The enzyme activity of GAPDH was measured in the cytosolic and nuclear fractions of the retina by measuring the increase in the production of NADH at 340nm. Each measurement was performed in duplicate and assay repeated three or more times. The activity obtained in the cytosol obtained from normal rat retina is considered as 100%. The values are mean ± SD from at least 6 rats in each of the 4 groups. *P<0.05 compared to normal, and #P>0.05 compared to PC. b. Nitration of GAPDH was performed by first immunoprecipitating GAPDH from the cytosolic and nuclear fractions of the retina, followed by separation on SDS-Gel. The nitrated GAPDH was identified using antibody against nitrotyrosine from Upstate Biotechnology. The blots are representative of 4-5 rats in each group. A 120

*# * 80

Enzyme activity (%) 40

* *# 0 Cyt Nuc Cyt Nuc Cyt Nuc Normal PC PC-GC

B Normal PC PC-GC Cytosol Nuclear Cytosol Nuclear Cytosol Nuclear Nitr-GAPDH

GADPH and diabetic retinopathy

Figure 6: Activation of AGEs and PKC in retina: Total AGEs and PKC activation were measured in the retinal homogenate by western blot technique using antibodies against anti-AGE and PKC βII, respectively. β-actin was used as a loading standard. These blots are representative of 4 or more rats in each group. Norm PC PC-GC A 218K

131K

86K

44K

β-actin

Norm PC PC-GC B PKC

β-actin

180 * *#

120

PKC:Actin expression (arbitrary units) 0 Normal PC PC-GC

GADPH and diabetic retinopathy

Figure 7: Histopathology in retinal microvasculature: Trypsin digested retinal microvasculature was stained with periodic acid-Schiff and hematoxylin. The number of acellular capillaries was counted in multiple mid-retinal fields and standardized to retinal area (per square millimeter). The arrows indicate acellular capillaries in the trypsin-digested microvessels obtained from a rat that was maintained in PC for 12 months. Normal PC