Fig 2: BclXL overexpression improved RGC somal survival but not axonal degeneration after CONC.A Transduction efficiency of AAV2.2-Pgk-mCherry-BclXL in RGCs as assessed by the percentage of mCherry+ (red) RGCs (RBPMS + cells, green) depicted in retinal flat mounts. On average, 76.1 ± 1.2% of RGCs were colabeled with mCherry. n = 4. Scale bar, 50 µm. B WT (n = 5) and BclXLAAV (n = 6) retinal flat mounts and quantification of cleaved caspase 3 (cCASP3)+ cells 5 days post-CONC. BclXLAAV retinas had 74.9 ± 10.5% fewer cCASP3+ cells compared to WT controls. *P = 0.030, two-tailed t-test. Scale bar, 50 µm. C WT and BclXLAAV retinal flat mounts and quantification of RGCs (RBPMS + cells) 14 days post-CONC. Both WT and BclXLAAV retinas had significant RBPMS + cell loss after CONC compared to Sham controls (85.4 ± 0.8% and 34.8 ± 4.2% loss respectively, *P < 0.001). However, BclXLAAV retinas had 59.2 ± 4.9% improved RGC survival after CONC compared to WT controls (*P < 0.001). n = 5, two-way ANOVA, Holm-Sidak’s post hoc test. Scale bar, 50 µm. D Quantification of RBPMS + RGC soma size from BclXLAAV retinas 14 days after Sham and CONC. After CONC, surviving RGCs from BclXLAAV retinas were 27.5 ± 2.2% smaller compared to Sham controls. n = 5, *P < 0.001, two-tailed t-test. E Representative PERG traces and quantification of PERG amplitudes from WT and BclXLAAV eyes 14 days post-Sham (n = 17, 18, respectively) and CONC (n = 18, 17, respectively). WT and BclXLAAV eyes had significant reductions in PERG amplitude after CONC relative to Sham (43.5 ± 5.8% and 36.2 ± 6.6% reductions respectively, *P < 0.05). BclXLAAV eyes did not have improved PERG amplitudes after CONC compared to WT controls (P = 0.816). Two-way ANOVA, Holm-Sidak’s post hoc test. Scale bar: X: 100 ms, Y: 5µV. F Longitudinal BclXLAAV optic nerve sections 5 days post-Sham and CONC. Sham BclXLAAV optic nerves had notable axonal mCherry labeling, which was markedly “beaded” and lost post-CONC. n = 4. Scale bar, 50 µm. G Longitudinal WT and BclXLAAV optic nerve sections 5 days post-Sham and CONC immunoassayed for neurofilament-H. BclXLAAV optic nerves had similar histological signs of degeneration after CONC compared to WT controls. n = 4. Scale bar, 50 µm. H Representative CAP traces and quantification of CAP amplitudes from WT and BclXLAAV optic nerves 5 days post-Sham and CONC. Both WT and BclXLAAV optic nerves had significantly decreased CAP amplitudes after CONC compared to Sham controls (59.7 ± 5.6% and 59.3 ± 3.7% amplitude reductions respectively, *P < 0.001). After CONC, BclXLAAV optic nerves did not have improved CAP amplitudes compared to WT controls (P = 0.582). n = 5, two-way ANOVA, Holm-Sidak’s post hoc test. Scale bar: X: 1 ms, Y: 1 mV. All numerical data are reported as mean ± standard error of the mean. For graphs, bars represent the mean, and error bars represent standard error of the mean.

Fig 3: RNA FISH showing localization of the lncRNA Gm35369 in the retina. (A) RNA fluorescence in situ hybridization showing the localization of Gm35369 in the retina, mostly in the GCL and INL. (B) Colocalization of the retinal ganglion cell markers RBPMS (red) and Gm35369 (green) in the GCL. (C) Colocalization of the horizontal cell markers CaBP (red) and Gm35369 (green). Signals were apparent in both the nucleus and cytoplasm. DAPI (blue).

Fig 4: BM88, BRN3A, ERR-β, NEUN NF-L, PGP9.5, RBPMS, γ-SYN, and TUJ1 expression in intact mouse retina

Fig 5: EDN1 caused JUN activation in RGCs.a Retinal flat mounts depicting JUN activation in RBPMS+ RGCs 24 h after PBS and EDN1 injection. b Quantification of RBPMS+ and RBPMS- pJUN+ cells 24 h after PBS and EDN1 injection. JUN was accumulated and activated (phosphorylated, pJUN, red) in 1203.8 ± 9.4 RBPMS+ cells per mm2 after EDN1 injection. A small portion of pJUN+ cells (113.0 ± 9.4 cells per mm2) were RBPMS−. n = 4 per condition, *P < 0.001, two-way ANOVA, Holm-Sidak’s post hoc. c Percentage of pJUN+ RBPMS+ cells after PBS and EDN1 injection. JUN was accumulated and activated (phosphorylated, pJUN, red) in a portion of RBPMS+ (green) RGCs 24 h post-EDN1, and was rarely activated in RBPMS+ RGCs post-PBS (%pJUN+ RBPMS+ cells; PBS: 0.1 ± 0.1; EDN1: 30.2 ± 7.2; n = 4 per condition, *P = 0.001, Welch’s t test). Note: in some EDN1-injured retinas, JUN activation in RGCs was localized to portions of retina and was not necessarily diffuse throughout the entire retina. Images represent the average cell count for each condition. d Retinal flat mounts depicting caspase 3 activation and JUN accumulation in RBPMS+ RGCs 5 days post- PBS and EDN1 injection. Arrows indicate cCASP3+ JUN+ cells, arrowheads indicate cCASP3+ JUN- cells. Merged images (far right) depict cCASP3+ (blue), JUN+ (red), and RBPMS+ (green) cells. e Quantification of cCASP3+ RBPMS+ cells that were JUN+ and JUN- after PBS and EDN1 injection. After PBS, 0.9 ± 0.1 cells/mm2 were RBPMS+ cCASP3+ JUN+, and 0.1 ± 0.1 cells/mm2 were RBPMS+ cCASP3+ JUN-. After EDN1, 15.7 ± 2.8 cells per mm2 were RBPMS+ cCASP3+ JUN+, and 7.4 ± 2.7 cells per mm2 were RBPMS+ cCASP3+ JUN−. (n = 6 per condition, *P < 0.05, two-way ANOVA with Holm-Sidak’s post hoc test). f Comparison of the percentage of cCASP3+ RBPMS+ cells that were JUN+ and JUN− between EDN1 injured retinas (n = 6) and controlled optic nerve crush-injured retinas (CONC; n = 4) 5 days post injury. Of all cCASP3+ RBPMS+ cells, 68.8 ± 8.7% were JUN+ after EDN1 insult, and 43.8 ± 9.6% were JUN+ after controlled optic nerve crush. (*P < 0.05, two-way ANOVA with Holm-Sidak’s post hoc). Scale bars, 50 μm.