Animals

Adult (between 3 and 5 months of age), male C57BL/6 mice were purchased from Charles River Laboratories (Wilmington, MA) and were kept on a standard 12 h:12 h light-dark cycle with free access to food and water. All animals were treated using protocols approved by the Boston University Charles River Campus Institutional Animal Care and Use Committee. Mice were age matched and made diabetic by an intraperitoneal (IP) injection of 75 mg/kg streptozotocin (STZ; Alexis Biochemicals, San Diego, CA) in pH 4.5 citrate buffer; each injection was given after an 8-h fast, one injection per day for three consecutive days. Control animals received three IP injections of citrate buffer alone after an 8 h fast on the same days as the experimental animals. The animals were allowed to recover for two days before their blood glucose levels were tested with a FreeStyle Flash ^TM blood glucose meter (Therasense, Almeda, CA). Mice were considered diabetic if their fasting blood glucose level was at or over 250 mg/dl. Animals that did not become diabetic were not used for this study. Treatment with insulin, aminoguanidine (AG), or N _ω -nitro-L-arginine methyl ester (L-NAME) was given to subsets of animals starting one week after induction of diabetes. Thus, these treatments were given for a duration of 4 weeks up to the time of sacrifice as follows. Once a day, 2 U/kg insulin from bovine pancreas (in sterile saline; Sigma I5500) was administered via IP injection. Also once a day, NOS inhibitor L-NAME (37.5 mg/kg in sterile 0.1 M pH 7.4 phosphate buffer, PB) was administered by IP. Every day, 1 g/l AG, an inhibitor of both NOS and advanced glycation end product (AGE) formation, was administered to animals in their drinking water, which was changed daily. Animal weight and fasting blood glucose levels were recorded at the beginning and end of each experimental cycle. Final blood glucose levels and bodyweights were compared between each group using ANOVA. Post hoc analysis was performed using Tukey’s test. A p<0.05 was considered significant. Animals were euthanized in each experiment by exposing them to IsoFlow™ Isoflurane gas (Abbott Laboratories) until they were deeply anesthetized, and then they were immediately decapitated.

Protein extraction and western blotting

Following sacrifice, the eyes were rapidly removed and placed into balanced salt solution (BSS; 137 mM NaCl, 5 mM KCl, 2 mM CaCl ₂ , 15 mM D-glucose, 1 mM MgSO ₄ , 1 mM Na ₂ HPO ₄ , and 10mM HEPES, pH 7.4) and the retinas were then surgically isolated from the eyes. Specimens in solution were immediately placed into 1.5 ml centrifuge tubes, set on dry ice. Six retinas (three animals) were pooled together to ensure an adequate amount of protein for detection by western blotting. The retinas were then suspended in 200 μl of homogenization buffer containing 20 mM Tris-HCl, 0.05% SDS, and 1X Complete Mini Protease Inhibitor Cocktail with EDTA (Roche Applied Science, Indianapolis, IN) before being ultrasonically homogenized. The protein concentration in each sample was determined using the Bradford method. In each lane, 100 μg protein was loaded onto 8% SDS PAGE gels and transferred onto 0.4 μm Immobilon™ PVDF membranes (Millipore, Billerica, MA). A rapid detection method was used, whereby the blocking step was omitted [ 28 ]. The primary antiserum was a rabbit anti-nNOS (sc648; Santa Cruz Biotechnology, Santa Cruz, CA) that was diluted 1:5,000 in 2% BSA in Tris buffered saline w/ 0.25% Tween-20 (TBST). The secondary antibody was a horseradish peroxidase-conjugated goat polyclonal antiserum raised against rabbit IgG (Molecular Probes, Invitrogen, Carlsbad, CA) at a dilution of 1:100,000. The blots were then treated with Immobilon™ HRP substrate (Millipore) and exposed to blue X-ray film (F-BX57; Phenix, Asheville, NC). To confirm equal lane loading, we stripped the blots and reprobed them with mouse anti-β-tubulin. The β-tubulin monoclonal antibody used was developed by Dr. Willi Hafler and was obtained from the Developmental Studies Hybridoma Bank. This antibody was developed under the auspices of the NICHD and is maintained by The University of Iowa (Department of Biology, Iowa City, IA). Quantitative analysis was performed using Image J image analysis software (Wayne Rasband, National Institute of Mental Health, Bethesda, MD). The difference in protein levels was evaluated using the paired Student’s t -test. A p<0.05 was considered significant.

Immunocytochemistry

After euthanasia, the mouse eyes were rapidly enucleated and placed in ice-cold BSS. The anterior chamber, lens, and vitreous were then removed and the resultant eyecups were immediately placed into 4% paraformaldehyde in PB for 90 min at room temperature. Next, the eyecups were cryoprotected in 30% sucrose in PB, embedded and frozen in Optimal Cutting Temperature embedding media (OCT; Tissue-Tek, Miles, Inc., Elkhard, IN), and cut into 14 µm thick cross-sections using a cryostat and mounted on Superfrost /Plus ^® slides (Fisher Scientific). The eyecups from the control and diabetic mice were embedded into the same block and cryosectioned onto the same slide to reduce variability when comparing changes in immunoreactivity. The asymmetry provided by this embedding method allowed us to clearly distinguish each of the eyecups in the resultant cryosections. Thus we could embedded one eye from each of 4 separate mice with different treatments in the same block.

Immunocytochemistry was performed on the retinal sections using standard methods as previously described [ 25 , 29 ] on retinal cross-sections mounted on slides. Immunochemical procedures were replicated, using at least three different animals (n=3), where each "n" was the average of three trials for a given animal compared to the other retinas on the same slide as described in the previous paragraph. The primary antiserum used was rabbit anti-nNOS (sc648; Santa Cruz Biotechnology) at a dilution of 1:1,000 overnight in a humid chamber at 4 °C. The slides were then washed 3×10 min in 0.1 M PB at room temperature and then incubated for 2 h in an Alexa488-conjugated goat anti-rabbit IgG secondary antiserum (Molecular Probes, Carlsbad, CA) at a dilution of 1:500. Incubation with the primary antiserum omitted served as a control for nonspecific secondary antiserum staining. Slides were washed in 0.1 M PB as before, coverslipped with glycerol, and the fluorescent signal was visualized using an Olympus Fluoview™ 300 confocal microscope (Olympus, Melville, NY). The excitation was done using the 488 nm laser line from an argon laser and the emission was visualized using a 510–550 nm bandpass filter. Ten to 14 optical sections were taken at 1 μm intervals, with the same number of optical sections always being captured and compared for both the control and experimental retinas. The laser intensity and all confocal settings were kept consistent within replicates.

The “z project” function of Image J image analysis software was used to obtain a single image from a collapsed confocal optical stack. The images were then converted to inverted grayscale so that the immunoreactivity would appear black. Mean intensity values were then recorded for each treatment and compared to the control retina on the same slide. These values were averaged together for each animal and compared to the average control mean using the paired Student’s t -test. A p<0.05 was considered significant. The relative intensity of immunoreactivity within each retinal layer was quantified using Image J to produce a horizontal line-profile average of the intensity of the immunoreactivity in the different retinal layers, minimizing the effect of local regional differences.

RNA isolation and reverse transcription

The retinas were isolated in BSS as described in the immunocytochemistry methods section, although all reagents and materials were kept RNase-free. Total RNA was then isolated from the retinas using a standard Trizol™ reagent (Invitrogen, Carlsbad, CA) extraction. To obtain sufficient amounts of RNA to perform these reactions, we pooled three animals (six retinas) together for each extraction. The RNA was then treated with rDNase™ (Ambion, Applied Biosystems, Austin, TX), following the manufacturer’s instructions, to remove any DNA contaminants. The RNA was quantified using a Nanodrop™ spectrophotometer (ThermoFisher Scientific, Waltham, MA). RNA was then converted into cDNA using the Verso™ cDNA kit (ThermoFisher Scientific) and subsequently treated with 2 U of RNase H™ (ThermoFisher Scientific) at 37 °C for 20 min.

Quantitative real-time PCR analysis

Quantitative real-time PCR analysis (qPCR) was performed using cDNA converted from 1 µg of retinal RNA as described in the previous section, using an ABI Prism ^TM 7900HT Sequence Detection System (Applied Biosystems, Carlsbad, CA). We used a pre-designed TaqMan™ gene expression assay (Applied Biosystems) for nNOS (assay ID: Mm00435173_m1). Our normalizing control was an optimized 18s rRNA primer set that was kindly provided to us by Dr. Ulla Hansen (Boston University, Boston, MA), which was designed to work with SYBR Green™ (Applied Biosystems). All data was obtained in quadruplicate and analyzed using the Microsoft Excel ^TM qGene template [ 30 ].

NO increased after 5 weeks of STZID and was reduced by insulin, aminoguanidine, or L-NAME

To examine any changes in NO levels in the retina after 5 weeks of STZID, we measured the NO-induced fluorescence (NO-IF) using the NO sensitive dye DAF-FM. We detected basal levels of NO-IF in all retinal layers in the dark-adapted retinas ( Figure 1A ). We observed an overall increase in NO-IF in dark-adapted diabetic retinas. In comparison to control retinas, the intensity of NO-IF in diabetic retinas increased significantly in the outer plexiform layer (OPL) by 82±19% (p<0.0001) and in the IPL by 48±18% (p=0.016; Figure 1A,B ). NO-IF also appeared in the diabetic outer nuclear layer (ONL), which was usually not seen in the control ONL ( Figure 1A ). Interestingly, presumptive bipolar cells in the control inner nuclear layer (INL) often had NO-IF both in their somata and processes, but they appeared to have lost NO-IF in their processes in the diabetic condition. There was no significant change in the number or intensity of labeled cell bodies in the INL or ganglion cell layer (GCL) in diabetic retinas when compared to controls.

Open in a separate window

Figure 1

Changes in NO-induced fluorescence (NO-IF) in the diabetic retina. A : NO-IF increased in the diabetic retina compared to control. NO-IF in processes from somata in the inner nuclear layer (INL) were present in control (white arrowhead) but absent in the diabetic. Treatment with insulin or aminoguanidine (AG) reduced NO-IF and insulin restored NO-IF in some the processes in the INL (white arrowheads). Scale bars represent 25 μm. Quantitative analysis of NO-IF in retinal regions is shown in dark-adapted unstimulated ( B ) and light-stimulated retinas ( C ) as compared to control (100%). The outer and inner plexiform layers (OPL and IPL) depict the percent intensity, while the inner nuclear layer (INL) and ganglion cell layer (GCL) depict the percent of NO-IF labeled somata. B, C : Like symbols indicate statistically significant differences between groups (* represents diabetic versus control OPL, # represents diabetes versus control IPL; p<0.05). Error bars represent SD.

Treatment with insulin partially reduced the NO-IF in the diabetic retinas ( Figure 1A,B ). Insulin treatment lowered NO-IF levels in the OPL and IPL to about half that of the untreated diabetics, however this trend was only statistically significant in the OPL (p=0.01). Neither the number of cells exhibiting NO-IF nor the intensity of NO-IF in the INL and GCL were significantly different from controls. In addition, insulin restored NO-IF to some processes of somata in the INL and greatly decreased the levels of NO-IF in the ONL ( Figure 1A ). While insulin did help restore NO-IF to near normal levels, we did not observe a complete rescue. This was most likely because the insulin was only administered once daily. As a consequence, the blood glucose levels of the insulin treated mice were lower than the untreated diabetics, although their blood glucose levels were still elevated above 250 mg/dl ( Table 1 ).

Treatment with L-NAME reduced retinal NO-IF to levels too low to be detected by our NO imaging method (data not shown). Treatment with AG significantly returned NO-IF to near control levels in the OPL (86±14%, p<0.0001) and in the IPL (80±18%, p=0.004; Figure 1B ), but the NO-IF in the processes in the INL did not return with AG treatment as it did with insulin ( Figure 1A ). The number of cells exhibiting NO-IF and intensity in the INL and GCL levels were not significantly different from control.

While the results described were for dark-adapted, unstimulated retinas, we also performed the same experiments using retinas stimulated with flashing light ( Figure 1C ). The results were comparable to retinas in the dark, although the basal levels of NO-IF were higher in stimulated control retinas. In diabetic mice, the NO-IF in the OPL significantly increased by 49 ± 11% (p=0.0002), while in the IPL it went up by 23±13% (p=0.02). Interestingly, insulin seemed to have a stronger effect in the OPL and IPL of light-stimulated retinas, significantly lowering NO-IF to near control levels in the OPL (93±18%, p<0.0001) and in IPL (83±17%, p=0.0016). AG had a dramatic effect in the light-stimulated retinas, with intensities significantly down to 112±13% in the OPL (p=0.02) and lower (but not statistically significant) to 98±14% in the IPL ( Figure 1B,C ).

nNOS protein levels decrease after 5 weeks of STZID

While NO-IF increased, nNOS-like immunoreactivity (nNOS-LI) significantly decreased by 21±11% in the OPL (p=0.005), 15±9% in the IPL (p=0.012), 19±17% in the INL (p=0.001), and 26±14% in the GCL (p=0.006; Figure 2A ). These changes were not affected by treatment with insulin or AG (data not shown). There was nNOS-LI in the IPL in many apparent neuronal processes in control retinas. Interestingly, nNOS-LI was largely absent from such processes of the IPL in the diabetic condition and appeared more punctate in expression, suggesting nNOS-LI was in synaptic boutons ( Figure 2B ). Western blots probed with the nNOS antiserum confirmed an overall reduction of nNOS protein by 24±11% (p=0.005), consistent with our immunocytochemistry (ICC) findings ( Figure 2C ).

Open in a separate window

Figure 2

nNOS protein levels decreased after 5 weeks of diabetes. A : Mean fluorescence levels of nNOS immunoreactivity were significantly reduced in the outerplexiform layer (OPL), inner nuclear layer (INL), inner plexiform layer (IPL), and ganglion cell layer (GCL) when compared to controls (n=11). The graph on the right indicates the line profile average of the intensity of nNOS in each layer of the retina. B : nNOS immunoreactivity in the IPL no longer filled neuronal processes in the diabetic and was localized in structures which resembled synaptic boutons in size and location. C : western blots detected a single ~160 kDa band consistent with nNOS. Levels of β-tubulin are shown to confirm equal protein loading. There was a 24%±11 decrease in total nNOS protein (C is control, D is diabetic; n=8). Line on the left indicates the location of the 150 kDa molecular weight marker.

To determine if the change in nNOS protein was due to a decrease in nNOS gene expression, we performed qPCR using an Applied Biosystem’s TaqMan™ assay. We found no difference in nNOS mRNA expression levels between control and diabetic retinas ( Figure 3 ), an indication that changes in nNOS protein levels were not due to changes in its gene expression.

Open in a separate window

Figure 3

qPCR measurements of retinal nNOS transcripts using an Applied Biosystems TaqMan™ assay. No difference was seen between control and diabetic nNOS mRNA levels. The values are depicted as mean normalized expression (MNE) compared to the 18s rRNA (n=4). The data were analyzed using the qGene template [ 30 ], using the following equation:

MNE = [ ( E r e f ) C T r e f , r e p l i c a t e 1 ( E t a r g e t ) C T t a r g e t , r e p l i c a t e 1 + ( E r e f ) C T r e f , r e p l i c a t e 2 ( E t a r g e t ) C T t a r g e t , r e p l i c a t e 2 + ( E r e f ) C T r e f , r e p l i c a t e 3 ( E t a r g e t ) C T t a r g e t , r e p l i c a t e 3 + ( E r e f ) C T r e f , r e p l i c a t e 4 ( E t a r g e t ) C T t a r g e t , r e p l i c a t e 4 ] 4

where; E _target , PCR amplification efficiency of the target gene; E _ref , PCR amplification efficiency of the reference gene; CT _target , threshold cycle of the PCR amplification of the target gene; CT _ref , threshold cycle of the PCR amplification of the reference gene. The CT was defined as the cycle at which the fluorescence rises appreciably above the background fluorescence and determined automatically using Applied Biosystems SDS software version 2.3.

Discussion

Neuronal versus vascular complications

Several studies of DR find evidence for cellular and metabolic abnormalities in retinal neurons before vascular complications [ 4 , 8 , 32 ]. In both human and animal models of DR, retinal dysfunction can first be detected using ERGs, long before any vascular pathology becomes apparent [ 8 ]. However, permeability changes can occur as early as eight days after diabetic onset in rats [ 33 ], and future studies should consider this. Nonetheless, studies that have looked at the vasculature in greater detail did not find major vascular complications until 6 months to 21 months after the onset of diabetes in rodents [ 11 , 32 , 34 , 35 ].

Role of nNOS in diabetic retinopathy

We found a large increase in NO production, which mirrored the anatomic localization of nNOS-LI, despite an overall decrease in nNOS protein levels. There have been relatively few studies examining nNOS in retinal neurons of diabetic animals, and the results have varied. One study used a combination of western blots and ICC to conclude that the number of nNOS-positive bipolar cells increases in the INL of the diabetic rat after nine weeks [ 26 ]. More consistent with our findings, an earlier study using a combination of NADPH diaphorase histochemistry and ICC found that the total number of nNOS expressing neurons decreases, and this loss is prevented by AG [ 27 ].

nNOS is responsible for the increases in NO-IF in the OPL and IPL

It is theoretically possible that some of the NO-IF we observed was produced from inducible NOS (iNOS) or endothelial NOS (eNOS). Several studies have examined iNOS as a potential mediator of DR. However, these studies all indicated that iNOS plays a critical role in subsequent vascular damage and do not link iNOS with early neuronal dysfunction [ 36 - 38 ]. Du et al. [ 39 ], showed that high glucose caused an increase in iNOS in the Müller cell culture line rMC-1, but not in bovine retinal endothelial cells. Abu El-Asrar et al. [ 38 ] detect iNOS in Müller cells of human patients with DR, but not in any cells of the retinas of normal individuals. Therefore, although iNOS is clearly involved in the pathology of the diabetic retina, it is more closely associated with glial and vascular complications. To our knowledge, there is no definitive evidence linking iNOS to the early neuronal pathology seen in DR, and there has been no selective localization of iNOS within the IPL or OPL. In contrast, our NO imaging results from this and previous studies [ 24 , 25 ] clearly indicated that the NO-IF in retina is closely anatomically correlated with nNOS-LI—observations that support our belief that the increased NO was coming from nNOS and not iNOS.

Although we detected relative low levels of eNOS mRNA compared to nNOS mRNA in retina using qPCR and found eNOS-LI in retinal vasculature (data not shown), we did not detect eNOS in retinal neurons using several commercially available eNOS antisera (Santa Cruz Biotechnology, Inc., sc8311, sc653, and sc654; BD Transduction Labs, 610298). However, this was not surprising since previous work shows only weak immunocytochemical staining of eNOS protein in the INL and blood vessels, suggesting that it is not abundant in the retina [ 36 , 40 ]. Although there is evidence that eNOS is involved in vascular complications of DR, the vasculature only makes up about 3%–5% of the retina [ 36 , 41 ]. Most important, the increased NO-IF we observed in the plexiform layers is not likely to be produced by eNOS, because the NO-IF is clearly not associated with blood vessels.

If NO is contributing to various stages of DR, then it is extremely important to find usable inhibitors of NOS as potential therapeutic targets. Gross inhibition of NOS is not a realistic option, as we found that L-NAME almost completely eliminated NO-IF and this treatment was not well tolerated. In the present study, AG was well tolerated, and it was apparent that AG was able to inhibit nNOS. Other studies also show that AG is not exclusively an iNOS inhibitor, as it is often reported to be [ 42 , 43 ]. However, since AG is also an inhibitor of AGE formation [ 44 , 45 ], we cannot discount the role AGE formation may play in regulating NO levels.

Potential mechanisms for increased nNOS activity

In both diabetic human [ 46 ] and rat retinas [ 47 ], there is increased NMDA receptor NR1 subunit immunoreactivity. These NMDA receptor channels provide an important route for Ca ²⁺ entry, and subsequent activation of Ca ²⁺ -dependent intracellular enzymes such as nNOS. Neurotoxicity associated with excitatory amino acids is reported to be mediated to a large extent through the activation of NMDA receptors [ 48 ]. Since nNOS activity is Ca ²⁺ -dependent and calmodulin-dependent, it would be also be activated by these Ca ²⁺ increases [ 49 ]. Furthermore, a subset of nNOS resides at the post-synaptic density by virtue of its PDZ-domain; this may explain why the nNOS we find remaining in the diabetic IPL appears largely synaptic [ 50 ]. Any nNOS at the post-synaptic density would be in close proximity to the largest influxes of calcium. Thus the reduction of nNOS protein may be compensated for by large glutamate-induced influxes of calcium at synapses.

Another significant possibility is that there is a post-translational activation of nNOS. nNOS enzyme activity is strongly inhibited by a Ca ²⁺ -dependent and calmodulin-dependent protein kinase II that phosphorylates nNOS at Ser ⁸⁴⁷ [ 51 , 52 ]. Increased levels of Ca ²⁺ can activate calcineurin (PP2B) to dephosphorylate nNOS at Ser ⁸⁴⁷ , which in turn allows nNOS to increase NO production [ 53 ]. Thus it is possible that increased Ca ²⁺ levels can stimulate NO production in two distinct ways: by increasing intracellular Ca ²⁺ to directly activate nNOS, and by activating calcineurin to activate nNOS through dephosphorylation of an inhibitory site.

Consequences of increased NO production in diabetes

There are several potential molecular and cellular consequences of an increase in NO in synapses. One consequence of diabetes is hypoxia in the diabetic retina [ 54 ]; NO from nNOS has been shown to sensitize neurons to hypoxia-induced death via competitive inhibition of cytochrome oxidase [ 55 ]. Thus, the increased NO may contribute to the early neuronal cell death in DR. NO has also been shown to increase release of neurotransmitters through both cGMP-dependent mechanisms [ 56 ] and through Ca ²⁺ independent mechanisms involving synaptic vesicle docking and fusion reactions [ 57 ]. In particular, NO has been shown to increase release of glutamate in chick retina [ 58 ]. In contrast, NO has been shown to decrease glycine release in retina [ 59 ]. Such increased levels of retinal glutamate and decreased release of glycine may also be related to the increased cell death in DR. These neurochemical changes are consistent with studies using a combination of ICC, western blots, and qPCR that report that the presynaptic proteins synaptophysin, synapsin 1, VAMP2, SNAP25, and PSD95 all showed decreases after only one month of diabetes, especially when synaptosomal fractions were selectively examined [ 18 ].

Future studies will need to determine the exact mechanisms that led to the increase in nNOS activity. In addition, it will be important to separate the timing and relative contributions of nNOS versus iNOS to get an accurate picture of how each pool of NO affects the retina as a whole. However, the results of our study indicate that nNOS may play a major role in providing excess NO in early DR. Future studies should address how this pathologically increased NO can be selectively reduced without detrimentally affecting the many normal functions of NO in the retina. This approach could provide a significant new strategy for treating the early neuronal cell loss in DR.

Acknowledgments

We thank Dr. Todd A. Blute, Felicitas B. Eldred, and Jan Blom for their outstanding support, valuable discussions, and technical assistance. We also thank Dr. Amit Agrawal for creating the ImageJ plug-in that allowed the analysis of our NO-IF images. This research was supported by NIH NEI EY04785 to WDE.

References

1. Aiello LM. Perspectives on diabetic retinopathy. Am J Ophthalmol. 2003; 136 :122–35. [ PubMed ] [ Google Scholar ]

2. Antonetti DA, Barber AJ, Bronson SK, Freeman WM, Gardner TW, Jefferson LS, Kester M, Kimball SR, Krady JK, LaNoue KF, Norbury CC, Quinn PG, Sandirasegarane L, Simpson IA, JDRF Diabetic Retinopathy Center Group. Diabetic retinopathy: seeing beyond glucose-induced microvascular disease. Diabetes. 2006; 55 :2401–11. [ PubMed ] [ Google Scholar ]

3. Lieth E, Gardner TW, Barber AJ, Antonetti DA. Retinal neurodegeneration: early pathology in diabetes. Clin Experiment Ophthalmol. 2000; 28 :3–8. [ PubMed ] [ Google Scholar ]

4. Barber AJ. A new view of diabetic retinopathy: a neurodegenerative disease of the eye. Prog Neuropsychopharmacol Biol Psychiatry. 2003; 27 :283–90. [ PubMed ] [ Google Scholar ]

5. Lieth E, Gardner TW, Barber AJ, Antonetti DA. Retinal neurodegeneration: early pathology in diabetes. Clin Experiment Ophthalmol. 2000; 28 :3–8. [ PubMed ] [ Google Scholar ]

6. Di Leo MA, Caputo S, Falsini B, Porciatti V, Greco AV, Ghirlanda G. Presence and further development of retinal dysfunction after 3-year follow up in IDDM patients without angiographically documented vasculopathy. Diabetologia. 1994; 37 :911–6. [ PubMed ] [ Google Scholar ]

7. Parisi V, Uccioli L. Visual electrophysiological responses in persons with type 1 diabetes. Diabetes Metab Res Rev. 2001; 17 :12–8. [ PubMed ] [ Google Scholar ]

8. Shirao Y, Kawasaki K. Electrical responses from diabetic retina. Prog Retin Eye Res. 1998; 17 :59–76. [ PubMed ] [ Google Scholar ]

9. Bresnick GH, Palta M. Predicting progression to severe proliferative diabetic retinopathy. Arch Ophthalmol. 1987; 105 :810–4. [ PubMed ] [ Google Scholar ]

10. Li Q, Zemel E, Miller B, Perlman I. Early retinal damage in experimental diabetes: electroretinographical and morphological observations. Exp Eye Res. 2002; 74 :615–25. [ PubMed ] [ Google Scholar ]

11. Engerman RL, Kern TS. Retinopathy in animal models of diabetes. Diabetes Metab Rev. 1995; 11 :109–20. [ PubMed ] [ Google Scholar ]

12. Daley ML, Watzke RC, Riddle MC. Early loss of blue-sensitive color vision in patients with type I diabetes. Diabetes Care. 1987; 10 :777–81. [ PubMed ] [ Google Scholar ]

13. Roy MS, Gunkel RD, Podgor MJ. Color vision defects in early diabetic retinopathy. Arch Ophthalmol. 1986; 104 :225–8. [ PubMed ] [ Google Scholar ]

14. Sokol S, Moskowitz A, Skarf B, Evans R, Molitch M, Senior B. Contrast sensitivity in diabetics with and without background retinopathy. Arch Ophthalmol. 1985; 103 :51–4. [ PubMed ] [ Google Scholar ]

15. Lieth E, Barber AJ, Xu B, Dice C, Ratz MJ, Tanase D, Strother JM. Glial reactivity and impaired glutamate metabolism in short-term experimental diabetic retinopathy. Penn State Retina Research Group. Diabetes. 1998; 47 :815–20. [ PubMed ] [ Google Scholar ]

16. Abu-El-Asrar AM, Dralands L, Missotten L, Al Jadaan IA, Geboes K. Expression of apoptosis markers in the retinas of human subjects with diabetes. Invest Ophthalmol Vis Sci. 2004; 45 :2760–6. [ PubMed ] [ Google Scholar ]

17. Martin PM, Roon P, Van Ells TK, Ganapathy V, Smith SB. Death of retinal neurons in streptozotocin-induced diabetic mice. Invest Ophthalmol Vis Sci. 2004; 45 :3330–6. [ PubMed ] [ Google Scholar ]

18. VanGuilder HD, Brucklacher RM, Patel K, Ellis RW, Freeman WM, Barber AJ. Diabetes downregulates presynaptic proteins and reduces basal synapsin I phosphorylation in rat retina. Eur J Neurosci. 2008; 28 :1–11. [ PubMed ] [ Google Scholar ]

19. Kern TS, Miller CM, Du Y, Zheng L, Mohr S, Ball SL, Kim M, Jamison JA, Bingaman DP. Topical administration of nepafenac inhibits diabetes-induced retinal microvascular disease and underlying abnormalities of retinal metabolism and physiology. Diabetes. 2007; 56 :373–9. [ PubMed ] [ Google Scholar ]

20. Eldred WD. Nitric oxide in the retina. Functional neuroanatomy of the nitric oxide system. In: Steinbusch HWM, Devente J, Vincente SR, editors. Handbook of Chemical Neuroanatomy. Amsterdam: Elsevier; 2000. p. 111–145. [ Google Scholar ]

21. Yilmaz G, Esser P, Kociok N, Aydin P, Heimann K. Elevated vitreous nitric oxide levels in patients with proliferative diabetic retinopathy. Am J Ophthalmol. 2000; 130 :87–90. [ PubMed ] [ Google Scholar ]

22. Hattenbach LO, Allers A, Klais C, Koch F, Hecker M. L-Arginine-nitric oxide pathway-related metabolites in the aqueous humor of diabetic patients. Invest Ophthalmol Vis Sci. 2000; 41 :213–7. [ PubMed ] [ Google Scholar ]

23. Blute TA, Mayer B, Eldred WD. Immunocytochemical and histochemical localization of nitric oxide synthase in the turtle retina. Vis Neurosci. 1997; 14 :717–29. [ PubMed ] [ Google Scholar ]

24. Eldred WD, Blute TA. Imaging of nitric oxide in the retina. Vision Res. 2005; 45 :3469–86. [ PMC free article ] [ PubMed ] [ Google Scholar ]

25. Giove TJ, Deshpande MM, Eldred WD. Identification of alternate transcripts of neuronal nitric oxide synthase in the mouse retina. J Neurosci Res. 2009; 87 :3134–42. [ PubMed ] [ Google Scholar ]

26. Park JW, Park SJ, Park SH, Kim KY, Chung JW, Chun MH, Oh SJ. Up-regulated expression of neuronal nitric oxide synthase in experimental diabetic retina. Neurobiol Dis. 2006; 21 :43–9. [ PubMed ] [ Google Scholar ]

27. Roufail E, Soulis T, Boel E, Cooper ME, Rees S. Depletion of nitric oxide synthase-containing neurons in the diabetic retina: reversal by aminoguanidine. Diabetologia. 1998; 41 :1419–25. [ PubMed ] [ Google Scholar ]

28. Mansfield MA. Rapid immunodetection on polyvinylidene fluoride membrane blots without blocking. Anal Biochem. 1995; 229 :140–3. [ PubMed ] [ Google Scholar ]

29. Pong WW, Stouracova R, Frank N, Kraus JP, Eldred WD. Comparative localization of cystathionine beta-synthase and cystathionine gamma-lyase in retina: differences between amphibians and mammals. J Comp Neurol. 2007; 505 :158–65. [ PubMed ] [ Google Scholar ]

30. Muller PY, Janovjak H, Miserez AR, Dobbie Z. Processing of gene expression data generated by quantitative real-time RT-PCR. Biotechniques. 2002; 32 :1372–9. [ PubMed ] [ Google Scholar ]

31. Blute TA, Lee MR, Eldred WD. Direct imaging of NMDA-stimulated nitric oxide production in the retina. Vis Neurosci. 2000; 17 :557–66. [ PubMed ] [ Google Scholar ]

32. Feit-Leichman RA, Kinouchi R, Takeda M, Fan Z, Mohr S, Kern TS, Chen DF. Vascular damage in a mouse model of diabetic retinopathy: relation to neuronal and glial changes. Invest Ophthalmol Vis Sci. 2005; 46 :4281–7. [ PubMed ] [ Google Scholar ]

33. do CA Ramos P, Reis A, Proenca R, Cunha-Vaz JG. Breakdown of the inner and outer blood retinal barrier in streptozotocin-induced diabetes. Exp Eye Res. 1998; 67 :569–75. [ PubMed ] [ Google Scholar ]

34. Lieth E, Gardner TW, Barber AJ, Antonetti DA. Retinal neurodegeneration: early pathology in diabetes. Clin Experiment Ophthalmol. 2000; 28 :3–8. [ PubMed ] [ Google Scholar ]

35. Kern TS, Engerman RL. A mouse model of diabetic retinopathy. Arch Ophthalmol. 1996; 114 :986–90. [ PubMed ] [ Google Scholar ]

36. Kern TS. Contributions of inflammatory processes to the development of the early stages of diabetic retinopathy. Exp Diabetes Res. 2007; 2007 :95103. [ PMC free article ] [ PubMed ] [ Google Scholar ]

37. Zheng L, Du Y, Miller C, Gubitosi-Klug RA, Ball S, Berkowitz BA, Kern TS. Critical role of inducible nitric oxide synthase in degeneration of retinal capillaries in mice with streptozotocin-induced diabetes. Diabetologia. 2007; 50 :1987–96. [ PubMed ] [ Google Scholar ]

38. Abu El-Asrar AM, Desmet S, Meersschaert A, Dralands L, Missotten L, Geboes K. Expression of the inducible isoform of nitric oxide synthase in the retinas of human subjects with diabetes mellitus. Am J Ophthalmol. 2001; 132 :551–6. [ PubMed ] [ Google Scholar ]

39. Du Y, Sarthy VP, Kern TS. Interaction between NO and COX pathways in retinal cells exposed to elevated glucose and retina of diabetic rats. Am J Physiol Regul Integr Comp Physiol. 2004; 287 :R735–41. [ PubMed ] [ Google Scholar ]

40. Takeda M, Mori F, Yoshida A, Takamiya A, Nakagomi S, Sato E, Kiyama H. Constitutive nitric oxide synthase is associated with retinal vascular permeability in early diabetic rats. Diabetologia. 2001; 44 :1043–50. [ PubMed ] [ Google Scholar ]

41. Joussen AM, Poulaki V, Tsujikawa A, Qin W, Qaum T, Xu Q, Moromizato Y, Bursell SE, Wiegand SJ, Rudge J, Ioffe E, Yancopoulos GD, Adamis AP. Suppression of diabetic retinopathy with angiopoietin-1. Am J Pathol. 2002; 160 :1683–93. [ PMC free article ] [ PubMed ] [ Google Scholar ]

42. Southan GJ, Szabo C. Selective pharmacological inhibition of distinct nitric oxide synthase isoforms. Biochem Pharmacol. 1996; 51 :383–94. [ PubMed ] [ Google Scholar ]

43. Jianmongkol S, Vuletich JL, Bender AT, Demady DR, Osawa Y. Aminoguanidine-mediated inactivation and alteration of neuronal nitric-oxide synthase. J Biol Chem. 2000; 275 :13370–6. [ PubMed ] [ Google Scholar ]

44. Edelstein D, Brownlee M. Mechanistic studies of advanced glycosylation end product inhibition by aminoguanidine. Diabetes. 1992; 41 :26–9. [ PubMed ] [ Google Scholar ]

45. Cameron NE, Gibson TM, Nangle MR, Cotter MA. Inhibitors of advanced glycation end product formation and neurovascular dysfunction in experimental diabetes. Ann N Y Acad Sci. 2005; 1043 :784–92. [ PubMed ] [ Google Scholar ]

46. Santiago AR, Hughes JM, Kamphuis W, Schlingemann RO, Ambrosio AF. Diabetes changes ionotropic glutamate receptor subunit expression level in the human retina. Brain Res. 2008; 1198 :153–9. [ PubMed ] [ Google Scholar ]

47. Ng YK, Zeng XX, Ling EA. Expression of glutamate receptors and calcium-binding proteins in the retina of streptozotocin-induced diabetic rats. Brain Res. 2004; 1018 :66–72. [ PubMed ] [ Google Scholar ]

48. Romano C, Price MT, Olney JW. Delayed excitotoxic neurodegeneration induced by excitatory amino acid agonists in isolated retina. J Neurochem. 1995; 65 :59–67. [ PubMed ] [ Google Scholar ]

49. Bredt DS, Snyder SH. Isolation of nitric oxide synthetase, a calmodulin-requiring enzyme. Proc Natl Acad Sci USA. 1990; 87 :682–5. [ PMC free article ] [ PubMed ] [ Google Scholar ]

50. Brenman JE, Chao DS, Gee SH, McGee AW, Craven SE, Santillano DR, Wu Z, Huang F, Xia H, Peters MF, Froehner SC, Bredt DS. Interaction of nitric oxide synthase with the postsynaptic density protein PSD-95 and alpha1-syntrophin mediated by PDZ domains. Cell. 1996; 84 :757–67. [ PubMed ] [ Google Scholar ]

51. Nakane M, Mitchell J, Forstermann U, Murad F. Phosphorylation by calcium calmodulin-dependent protein kinase II and protein kinase C modulates the activity of nitric oxide synthase. Biochem Biophys Res Commun. 1991; 180 :1396–402. [ PubMed ] [ Google Scholar ]

52. Komeima K, Hayashi Y, Naito Y, Watanabe Y. Inhibition of neuronal nitric-oxide synthase by calcium/ calmodulin-dependent protein kinase IIalpha through Ser847 phosphorylation in NG108–15 neuronal cells. J Biol Chem. 2000; 275 :28139–43. [ PubMed ] [ Google Scholar ]

53. Xu Y, Krukoff TL. Adrenomedullin stimulates nitric oxide production from primary rat hypothalamic neurons: roles of calcium and phosphatases. Mol Pharmacol. 2007; 72 :112–20. [ PubMed ] [ Google Scholar ]

54. Crawford TN, Alfaro DV, III, Kerrison JB, Jablon EP. Diabetic retinopathy and angiogenesis. Curr Diabetes Rev. 2009; 5 :8–13. [ PubMed ] [ Google Scholar ]

55. Jekabsone A, Neher JJ, Borutaite V, Brown GC. Nitric oxide from neuronal nitric oxide synthase sensitises neurons to hypoxia-induced death via competitive inhibition of cytochrome oxidase. J Neurochem. 2007; 103 :346–56. [ PubMed ] [ Google Scholar ]

56. Sistiaga A, Miras-Portugal MT, Sanchez-Prieto J. Modulation of glutamate release by a nitric oxide/cyclic GMP-dependent pathway. Eur J Pharmacol. 1997; 321 :247–57. [ PubMed ] [ Google Scholar ]

57. Meffert MK, Calakos NC, Scheller RH, Schulman H. Nitric oxide modulates synaptic vesicle docking fusion reactions. Neuron. 1996; 16 :1229–36. [ PubMed ] [ Google Scholar ]

58. Ientile R, Picciurro V, Pedale S, Nucci C, Malecka B, Nistico G, Macaione S. Nitric oxide enhances amino acid release from immature chick embryo retina. Neurosci Lett. 1996; 219 :79–82. [ PubMed ] [ Google Scholar ]

59. Yu D, Eldred WD. Nitric oxide stimulates gamma-aminobutyric acid release and inhibits glycine release in retina. J Comp Neurol. 2005; 483 :278–91. [ PMC free article ] [ PubMed ] [ Google Scholar ]