Fig 1.
Identification of Myo/Nog cells in the rat retina.
Sections of adult rat retinas were double labeled with the G8 mAb and antibodies to Noggin and MyoD to localize Myo/Nog cells. Nuclei were stained with bisbenzimide dye (blue: A, D, E, H, I, L, M and P). Co-labeling with the G8 mAb (green: B and red: G) and antibodies to Noggin (red: C and D) and MyoD (green: F, H) identified Myo/Nog cells in the outer plexiform and inner nuclear layers. Incubation with a non-specific IgM primary antibody and the AlexaFluor 488 IgM secondary antibody showed no visible labeling in the “green channel” (green: J, L) or “red channel” (red: K, L). Incubation with only the AlexFluor 488 IgM secondary antibody showed no visible labeling in either channel (N-P). ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; GCL, ganglion cell layer.
Fig 2.
Endogenous Myo/Nog cells increase following light damage.
Sections from light damaged and control retinas were labeled with the G8 mAb (red) and bisbenzimide dye (blue). (A, B) Representative section of from a control retina showing the presence of Myo/Nog cells in the OPL. C/D: Representative section of a light damaged retina showing G8+ cells Myo/Nog cells in the OPL, INL, RPE (*) and GCL (˂). E: The numbers of G8+ cells per section of control (squares) and LD retinas More G8+ cells were present in the LD than control retinas. **p = 0.02 (n = 8). F: Distribution of Myo/Nog cells per retinal layer in control and LD retinas. (G-H) LD eyes were injected with Myo/Nog cells that had been prelabeled with bisbenzimide 33342 (blue). At 2 days post-injection, eyes were sectioned and labeled with the G8 mAb. Myo/Nog cells were identified by localizing bisbenzimide. (G) Brightfield image, (H) Fluorescent image showing the presence of a bisbenzimide 33342+/G8+ cell s (red arrow). (I) Merged image showing the location of the bisbenzimide labeled Myo/Nog cell within the OPL. LD, light damage; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; GCL, ganglion cell layer.
Fig 3.
Severity of light-induced cell death correlates with an increase in the number of Myo/Nog cells.
Light damage and control retinas were labeled using G8 mAb (green), a marker of Myo/Nog (cells detonated with a white˂), DAPI (blue) a marker of cell nuclei and TUNEL (red) a maker of dying cells. (A,B) Representative image of control retina. Myo/Nog cells (green) are rare, but will occasionally appear in locations like the OPL and CC. We rarely, if ever will observe dying photoreceptors in control animals. (C, D) Representative image of LD retina. In the light damaged retinas, we observed an increase in the TUNEL positive cells (red) and also an increase in the number of Myo/Nog cells in the all layers, but particularly in the OPL and RPE. (E) Correlation between G8+ (Myo/Nog cells) and TUNEL+ cells (dying photoreceptors) in LD retinal section (n = 8). r2 = 0.78, p < 0.01 (slope significantly non-zero). CC, choriocapillaris; ILM, inner limiting membrane; INL, inner nuclear layer; ONL, outer nuclear layer; OLM, outer limiting membrane.
Fig 4.
Intravitreal injection of Myo/Nog cells partially preserved visual function after light damage.
Visual function was assessed using the ERG 7 days after light damage. The a-wave and b-wave amplitudes in response to bright flash were normalised to baseline (before light damage) bright flash responses and averaged for each treatment group. Control animals suffered no injection and no light damage. G8+ animals suffered no light damage but were injected with purified G8+ cells and assessed 7 days after injection. LD stands for Light damage. (A) Representative ERG traces from eyes with LD alone, LD eyes injected with PBS (LD/sham) and LD eyes injected with G8+ (LD/G8+). (B) Mean change in a-wave amplitudes for retinas from the 5 experimental groups, from before treatment. Error bars show SEMs. Control rats that did not receive light damage showed no significant change in the a-wave between the two ERG measurements (n = 4). Control rats that did not receive light damage, but were injected with G8+ cells also showed no significant change in the a-wave between the two ERG measurements (n = 3). Visual function, as measured by the a-wave, was significantly improved following light damage with injection of G8+ cells (n = 18) compared to LD alone (n = 18) and LD/sham (n = 18) (LD/G8+: vs. LD ** p < 0.01, vs. LD/Sham * p < 0.05).
Fig 5.
Myo/Nog cell injections mitigated photoreceptor loss.
(A-F) Representative sections from superior and inferior retina of non-light damaged control (control), light damage + injection of PBS (LD/Sham) and light damage + injection of g8+ cells (LD/G8+). Dying cells were labeled with TUNEL (red) and cell nuclei were stained with bisbenzimide dye (blue). The arrows at the left of D show how the thickness of the ONL (shorter arrow) and of the nuclear layers of the retina (longer arrow) was measured. Retinas without light damage showed no labeled cells with TUNEL reagents (A and D). TUNEL+ cells were mostly present in the ONL of retinas with light damage in the superior retina when injected with PBS or G8+ cells (B, C and E). TUNEL+ cells were present in the sham injected eye (E) but not present in these control and G8+ injected sections (D and F). The ONL was disorganized in the LD/Sham retina, but not in the control and LD/G8+ retinas. (G) Normalized ONL thickness for the non-light damaged) control (black line), LD/Sham (Gray line) and LD/G8+ (blue line) groups from superior to inferior areas of the retina. Error bars show SEMs. Significant difference in retinal thickness were seen between LD/Sham and LD/G8+; * p < 0.05; **p<0.01. (H) Normalised ONL thickness for control, LD/Sham and LD/G8+ groups. Error bars show SEMs. The ONL is significantly thicker in LD retinas treated with G8+ cells than those injected with PBS. No difference was seen in ONL thickness between control and LD/G8+ sections. **p<0.01. (I) Number of TUNEL+ cells in the ONL/section for the same three groups. Error bars show SEMs. Significantly more TUNEL+ cells were present in the LD retinas than control retina (asterisk p value). Fewer TUNEL+ cells were present in the retinas injected with G8+ cells than those injected with PBS (asterisk p value). (G-I: Control = 5, LD/Sham = 4, LD/G8+ = 5). ***p < 0.001; ** p < 0.01. ILM, inner limiting membrane; INL, inner nuclear layer; ONL, outer nuclear layer; OLM, outer limiting membrane.
Fig 6.
Myo/Nog cell injections mitigated retinal stress following LD, as measured by GFAP expression.
(A-F) Representative sections from superior and inferior retina of control (non-light damaged) and light damaged eyes injected with PBS (LD/Sham) or G8+ cells (LD/G8+). Sections were labeled with an antibody to GFAP (white). ILM–inner limiting membrane; INL- inner nuclear layer; ONL–outer nuclear layer; OLM–outer limiting membrane. (G) Normalised length of GFAP labeling for control as measured by the length of the labeled Muller cells normalized to the length of the nuclear retina. Error bars show SEMs. Non-LD retinas had significantly shorter GFAP+ labeled Muller cells than LD retinas (asterisk with p value). Injection of G8+ cells following LD had significantly shorter Muller cells than LD retinas injected with PBS. (Control, n = 5; LD/Sham, n = 4; LD/G8+ = 9). *** indicates p < 0.001; ** indicates p < 0.01.