*3.2. Retinal Explant Model Provides a Platform to Rapidly Test Candidate Neuroprotective Therapeutics*

Axotomy of the RGC axon results in RGC degeneration. In the retina explant model [9,11], severing of the optic nerve leads to RGC axotomy, and results in 30–50% RGC loss over 3–5 days in the mouse. We utilized this model to identify compounds for an RGC neuroprotective effect (Figure 2). Valproic acid, SB-431542, folic acid, and WY-14643 were selected for testing because an established literature already exists for progesterone, choline, and estradiol (see discussion). Following axotomy and maintenance in culture ex vivo for 3 days, retinas exhibited a marked loss of RGCs as identified by significant reduction in the number of CFP+ cells (40% loss, *p* < 0.05) and RBPMS+ cells (39% loss, *p* < 0.001). Nuclear density was variable and was not significantly reduced at this time point, likely reflecting a combination of neuronal loss and glial proliferation. Surviving RGCs had significantly reduced nuclear diameter (14% smaller, *p* < 0.001), indicating significant cellular stress.

Three of the tested compounds conferred significant neuroprotection when dissolved in the culture media. Survival of CFP+ RGCs was best promoted by folic acid (1.75-fold survival from untreated, *p* < 0.05), followed by valproic acid (1.59-fold survival from untreated, *p* < 0.05), and was significant, but highly variable with SB-341542 (1.65-fold survival from untreated, *p* < 0.05). RBPMS+ cell counts were highest in retinas treated with valproic acid (1.47-fold survival from untreated, *p* < 0.01; 11% loss compared with control, *p* > 0.05), followed by WY-14643 (1.4-fold survival from untreated, *p* < 0.01; 11% loss compared with control, *p* > 0.05). RGC loss assessed by RBPMS+ counts was reduced for folic acid treated retinas relative to control (21% loss compared with control, *p* > 0.05), but was not significantly different from untreated retinas, demonstrating variable protection (1.3 fold survival from untreated, *p* > 0.05). Only WY-14643 demonstrated significant protection against nuclear shrinkage (1.11-fold larger diameter compared with untreated, *p* < 0.05; 4.5% loss compared with control, *p* < 0.05). Folic acid and valproic acid demonstrated a less severe nuclear shrinkage relative to control (10% loss, *p* < 0.01; and 9.7% loss, *p* < 0.01, respectively). SB-341542, despite showing some protection to CFP+ RGCs, did not demonstrate significant neuroprotection to RBPMS+ RGCs (1.08-fold survival from untreated, *p* > 0.05; 35% loss compared with control, *p* > 0.001), or against nuclear shrinkage (0.98-fold survival from untreated, *p* > 0.05; 15% loss compared with control, *p* > 0.001), suggestive of a possible RGC subtype bias or preferential protection of healthier cells (i.e., those able to maintain CFP expression) [12].

As further validation of this approach, we selected an identified compound that should not enhance RGC survival (as the CTD returns interactions based only on literature link, irrespective of context). We tested the effects of supplementing the culture media with genistein as it has been demonstrated to influence a number of neuroprotective effects. Genistein had no effect on RGC survival compared with untreated retinas, as assessed by CFP+ counts (*p* > 0.05), RBPMS+ counts (*p* > 0.05), TOPRO-3+ counts (*p* > 0.05), or nuclear diameter (*p* > 0.05). Density of TOPRO-3+ nuclei was actually significantly reduced from control in only genistein-treated retinas (15% loss from control, *p* < 0.05).

**Figure 2. Sequencing data and drug database screening successfully identifies therapeutic candidates that provide retinal ganglion cell neuroprotection.** Retina from B6.Cg-Tg(Thy1-CFP)23Jrs/J mice were explanted into tissue culture and maintained for 3 days ex vivo (DEV) with candidate drugs supplemented in the media (in addition to controls: 0 DEV and 3 DEV untreated). (**A**) Retinas were labelled for CFP (anti-GFP), RBPMS, and TOPRO-3 before imaging. (**B**) CFP+ cell density was significantly reduced at 3 DEV in untreated retinas and this was significantly improved in folic acid, SB-431542, and valproic acid treated retinas. RBPMS+ RGC density was significantly reduced at 3 DEV in untreated retinas and this was significantly improved in valproic acid and WY-14643 treated retinas, with moderate protection from folic acid. TOPRO-3+ round nuclei were not significantly altered with the exception of genistein, supporting its lack of protection and possible neurotoxic effects. Mean nuclear diameter was significantly smaller at 3 DEV, and this was significantly improved by WY-14643. Overall, these data support valproic acid, folic acid, and WY-14643 as neuroprotective against acute RGC injury. This validates the approach of identifying potential neuroprotective therapeutics from existing -omics data. Scale bar in A = 20 μm. \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001.
