Changes In Glutamate Homeostasis In The Retina During Diabetes

There is considerable evidence that Müller cells are functionally abnormal in diabetes. The most obvious change that has been reported in Müller cells during diabetes is reactive gliosis (34-38). However, functional abnormalities have also been observed, including anomalies in glutamate and GABA turnover (39,40). In support of the notion that glutamate turnover is abnormal in diabetes is the observation that glutamate levels are elevated in the vitreous of diabetic rats and human patients (37,41,42). In addition, the concentration of glutamate within the rat retina following three months of diabetes is elevated and the rate of glutamine formation is reduced, suggesting that the conversion of glutamate to glutamine by Müller cells might be abnormal (37,42). As noted earlier, showing that the glutamate concentration in the whole retina is altered during diabetes is not evidence per se of an abnormality in glutamate transport, but rather of a defect or defects in the glutamate-glutamine cycle as a whole. An elevation in glutamate within the retina could be explained by; (i) a change in glutamate transport, (ii) a reduction in the release of glutamate from glutamatergic neurons, or (iii) a reduction in the activity of glutamine synthetase. In relation to these three possible explanation, the activity of glutamine synthetase, the principal enzyme involved in converting glutamate to glutamine within Müller cells was found to be reduced following as little as two months duration of diabetes in the rat retina (42). In addition, there is evidence for a reduction in neuronal function from an early stage of diabetes (43,44). This change in neuronal function could cause a reduction in the release of glutamate from glutama-tergic neurons.

Two studies have considered, in detail, glutamate uptake by Müller cells during diabetes (45,46). Li and Puro (45) have shown that glutamate uptake was decreased in isolated rat

Müller cells, possibly through oxidative processes. These authors used patch-clamping techniques to measure currents in isolated Müller cells, following application of the glutamate analog 1-trans-pyrrolidine-2,4-dicarboxylate. Glutamate uptake into Müller cells via GLAST is electrogenic, because sodium ions are cotransported into Müller cells with glutamate, while potassium is transported out of the cell. This movement of sodium and potassium ions across Müller cell membranes can be measured as current using patch-clamping electrodes. Li and Puro (45) measured GLAST function by measuring the current induced in freshly isolated control and diabetic rat Müller cells. They observed that the current induced following application of the glutamate analog, 1-trans-pyrrolid-ine-2,4-dicarboxylate, was reduced by 36% following as little as four weeks of diabetes, suggesting that GLAST activity is reduced early in the course of diabetes.

The second study, by Ward et al. (46), examined the expression of GLAST and EAAT4 in the rat retina and measured glutamate uptake during diabetes in vivo. These authors used semiquantitative polymerase chain reaction (PCR) and immunocytochemistry to examine whether the expression of EAAT4 or GLAST was altered following 12 weeks of diabetes. No differences in messenger RNA or protein expression were noted. In order to probe the function of GLAST in vivo, these authors quantified the level of uptake of D-aspartate into Müller cells following 1, 4, and 12 weeks of diabetes. D-aspartate is taken up into Müller cells via GLAST, but is not degraded by glutamine synthetase and can therefore be visualized using immunocytochemical techniques. These authors showed that D-aspartate uptake was greater in diabetic rat retina following as little as one week of diabetes. Results from a similar experiment are shown in Fig. 6. Control and diabetic rats were injected intravitreally with D-aspartate. Rats were then sacrificed, and their eyes were removed, fixed and prepared for postembedding immunocytochemistry. Fig. 6 shows the ganglion cell layer, and nerve fiber layer, of a control and diabetic rat retina that has been labeled with an antibody directed against D-aspartate. The end-feet of Müller cells are prominently labeled in both the control and diabetic retina, and appear a little darker in the diabetic rat retina. Fig. 6C shows the intensity of D-aspartate labeling in control and diabetic Müller cell end-feet following 1, 4, or 12 weeks of diabetes. D-aspartate labeling was greater in the end-feet of Müller cells at all stages of diabetes (analysis of variance, p < 0.05).

How can the results of Li and Puro (45) be reconciled with the apparently opposite finding of Ward et al. (46)? First, experimentation on isolated cells may not always reflect function in vivo, because the external environment is artificially controlled. This could be particularly significant in studying the changes in cellular function during diabetes, because the external milieu can be altered by breakdown of the blood-retinal barrier (47).

It has been recently recognized that measuring the function of GLAST using patch-clamping electrophysiological tools can lead to different results compared to those generated by measuring D-aspartate uptake. A recent study by Sarthy et al. (32) showed that Müller cells isolated from GLAST-knockout mice displayed no electrophysiologi-cally recordable current when glutamate was applied, verifying that GLAST function did not exist in these cells. However, the retinas of these same animals showed extensive D-aspartate uptake into Müller cells. Therefore, it is likely that glutamate uptake into Müller cells does not rely solely on the function of GLAST, but requires that other nonelectrogenic transporters are present.

Weck 1 Week 4 Week 12

Duration of Diabetes

Weck 1 Week 4 Week 12

Duration of Diabetes

Fig. 6. Vertical sections of the inner retina of (A) a control retina, and (B) a diabetic rat retina, that were intravitreally injected with D-aspartate and then immunolabeled for D-aspartate. Müller cell end-feet within the nerve fiber layer are heavily labeled for D-aspartate. The level of D-aspartate immunoreactivity within Müller cell end-feet is higher during diabetes than in control retina, even as early as one week following the onset of diabetes (C) Graph shows the mean pixel intensity, plus or minus the standard error of the mean (±SEM), of D-aspartate labeling within Müller cell end-feet following 1, 4, and 12 weeks of diabetes. Uptake of D-aspartate was significantly greater in diabetic end-feet compared with control retinas, at all stages of diabetes examined. Abbreviations as in Fig. 1. Scale bar = 20 |M.

Therefore, when one considers the studies of Li and Puro (45) and Ward et al. (46) together, it is possible to conclude that GLAST function is reduced in diabetes (as shown by Li and Puro), but that another nonelectrogenic glutamate transporter is upregulated in diabetes. More importantly, overall, glutamate uptake by Müller cells during diabetes is greater than in controls and this may have a bearing on the function of retinal neurons during diabetes.

0 0

Post a comment