Fig. 1. Panel A shows the effect of H₂O₂ on the viability of RGC-5 cells in culture as measured by the MTT assay. Viability of RGC-5 cells to H₂O₂ is concentration-dependent with 400 μM H₂O₂ causing death to approximately 60% of cells. Panel B shows the EGCG at 10 or 50 μM significantly attenuates RGC-5 cell death caused by 400 μM H₂O₂ in a concentration-dependent way. The results also show that trolox at 50 μM is less effective than EGCG in blunting the influence of H₂O₂. Results are mean values with error bars indicating ± SEMs (n = 6 in each experiment and **p < 0.01).

Fig. 2. RGC-5 cells subjected to the APOPercentage™ assay for apoptosis (A, C and E) or stained for the localisation of ROS (B, D and F). Panels A and B show control RGC-5 cells in culture. Panels C and D show that 400 μM H₂O₂ causes many of the cells to die by apoptosis indicated by brown nuclei (C) and a generation of ROS, observed as a white colouration of nuclei (D). EGCG clearly blunts the effect of 400 μM H₂O₂ with very few brown apoptotic cells detected (arrows, E) and no obvious staining for ROS (F). Scale bars = 40 μm.

Fig. 3. This figure shows that the a- and b-wave amplitudes of the ERGs are drastically reduced following ischemia/reperfusion and that this effect is counteracted significantly when animals are treated with EGCG. Individual characteristic ERGs recordings are shown in panel B where it can be seen that the ERG from a control eye taken before ischemia (baseline) and after ischemia to the other eye (non-ischemia) is almost identical. The form of the ERG is much affected by ischemia but in animals treated with EGCG this is not the case. Panel A shows data from 6 animals in each case where values are the means and error bars represent ± SEMs.

Fig. 4. Effect of ischemia/reperfusion on Thy-1 and ChAT immunoreactivities in vehicle and EGCG-treated rats. Normal Thy-1 immunoreactivity is associated with an immunoreactive band corresponding to ganglion cells and their dendrites (arrow heads). (A1) A thinning of Thy-1 immunoreactivity was observed after ischemia/reperfusion in vehicle-treated animals (A2) but this reduction was less apparent in EGCG-treated rats (A3). Normal ChAT immunoreactivity is seen as two bands (double arrows) of immunoreactivity in the inner plexiform layer and occasional cell bodies (single arrows) on either side of the bands (B1). In vehicle-treated rats, ischemia/reperfusion caused most of the two bands of ChAT immunoreactivity to disappear and also a reduction in the numbers of ChAT positive cell bodies (B2). Administration of EGCG significantly counteracted the effect of ischemia/reperfusion (B3). Scale bars = 40 μm.

Fig. 5. Thy-1 (A), NF-L (B), caspase-3 (C), caspase-8 (D), GFAP (E) and opsin (F) mRNA levels relative to cyclophilin mRNA in normal retina (control) and following ischemia/reperfusion with treatment of vehicle (saline) or EGCG. It can be seen that the changes caused by ischemia/reperfusion where all significant in vehicle-treated animals and these were all significantly blunted (except for GFAP) by treatment with EGCG. Data are means with error bars indicating ± SEMs. n = 6 in each case, *p < 0.05, **p < 0.01.

Fig. 6. NF-L (A), caspase-3 (B), GFAP (C), rhodopsin kinase (D), NF-L (E) and tubulin (F) proteins relative to actin in retina (A–D) and optic nerves (E, F) in the non-ischemic eyes (control) and following ischemia/reperfusion with vehicle-treatment (saline) or EGCG-treatment (EGCG). Ischemia/reperfusion caused a significant change in the amounts of all proteins and treatment with EGCG significantly counteracted this effect except in the case of rhodopsin kinase. Data are means with error bars indicating ± SEMs. n = 6 in each case, *p < 0.05, **p < 0.01.

Fig. 7. Comparative effect of 55 min (A and B) or 45 min (C and D) ischemia on NF-L and tubulin proteins in optic nerves (saline) relative to actin. Ischemia/reperfusion caused a significant reduction of NF-L and tubulin in optic nerves compared with controls (control). EGCG treatment by injection both into the eye and intraperitoneally significantly counteracted the changes caused by ischemia/reperfusion. Data are means with error bars indicating ± SEMs. n = 6 in each case, *p < 0.05, **p < 0.01.