Glutathione (GSH), together with NADPH-producing pathways and glutathione reductase, provides a defense system against oxidants. Oxidation of GSH causes stimulation of the hexose monophosphate shunt and increased production of NADPH. We have asked if hexose monophosphate shunt activity is required for the recovery of GSH following exposure of the isolated rat retina to an oxidant. Hexose monophosphate shunt activity was decreased by depleting the retina of hexose stores, before exposing the tissue to diamide (0.04-1.0mM), an oxidant for GSH, for 30 min. After exposure, retinas were transferred to either glucose-containing or glucose-free recovery medium for an additional 30 min. Control retinas kept in glucose-free, oxygenated medium (no diamide) for 90-120 min maintained GSH at 90% of the value found in retinas incubated with glucose. After exposure of hexose-depleted retinas to 0.4 mM diamide, a nearly 90% decrease in GSH was observed. When the oxidant was removed, the level of GSH returned to more than 80% of the control value in the presence or absence of glucose. In contrast, no recovery of GSH was observed after diamide treatment if the retinas were transferred to ice-cold (1-5 degrees C) media with or without glucose or if the retinas were pre-treated with 2 mM 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) to inhibit glutathione reductase. Measurements of two NADPH-producing cytosolic enzymes, namely NADP+-dependent malic enzyme and NADP+-dependent isocitrate dehydrogenase, revealed high activities. Optimum production of NADPH from malic enzyme was 0.90 nmol NADPH produced min-1 per retina, while with isocitrate dehydrogenase the average rate was 6.9 nmol NADPH produced min-1 per retina. We suggest that these enzymes together with a long-lived endogenous substrate (probably glutamate) are responsible for the recovery of GSH in hexose-depleted retinas. The present results suggest that more than one NADPH-producing system is capable of controlling the GSH concentration in retina. Studies that have focused on the hexose monophosphate shunt pathway as the sole source of NADPH for glutathione reductase in retina and other tissues may require re-evaluation depending on the overall metabolic capacity and substrate utilization of the particular tissue. Thus, the present findings are significant not only with respect to the retina but also for other tissues whose metabolic characteristics are similar to those found in the retina.