Blue Light Promotes Neurite Outgrowth of Retinal Explants in Postnatal ChR2 Mice

The apparatus and temporal patterns of blue light stimulation used during retinal explant culture. A, The blue light LEDs (∼680 cd/m², 470 nm) were powered and driven by an Arduino circuit to provide light stimulation. Retinal explants were cultured in a 12-well plate and stimulated by the LED array from below for only 1 h at the beginning of each experiment. B, The temporal patterns of the 5, 20, and 100 Hz square wave light stimulations.

Blue light stimulation promotes neurite outgrowth of retinal explants from P5 and P11 mice with ChR2-expressed RGCs. A, A vertical section through the P11 ChR2 mouse retina. DAPI was used to indicate nuclei in the retinal slice (blue), and ChR2-eYFP cells were expressed exclusively in the ganglion cell layer (GCL) with their dendrites in the inner plexiform layer (IPL). Scale bar, 100 μm. B, ChR2-expressed RGCs in ChR2 mice were distributed randomly across the entire retina. Brn3a was used to label pan-RGCs (red), and ChR2-eYFP indicated ChR2-RGCs. White arrows show examples of RGCs expressing both Brn3a and ChR2. Scale bar, 20 μm. C–F, Confocal images of P5 retinal explants at DIV 5 with no light stimulation (Ctrl), and with 5, 20, and 100 Hz light stimulations for 1 h at the beginning of the experiment, respectively. G–J, Confocal images of P11 retinal explants at DIV 5 with the same treatments as the P5 retinal explants. All morphologically recognized neurites were TUJ1-positive (red), and DAPI was used to label the nuclei (blue). Scale bar, 100 μm. K, All three light stimulation protocols promoted neurite outgrowth in the P5 retinal explants (n = 5 for each condition). L, Similar results were obtained with the P11 retinal explants, except for the 100 Hz condition (n = 7 for the control; n = 5 for 5 Hz; n = 9 for 20 Hz; n = 6 for 100 Hz); *p < 0.05, **p < 0.01, ***p < 0.001. Error bars, mean ± SEM.

Neural developments of WT and ChR2 retinas are relatively normal in P11 mice. A–F, A vertical section through P11 WT and ChR2 mouse retinas. DAPI was used to indicate nuclei in the retinal slice (blue), and GluA2 was used to label AMPA receptors (red). Scale bar, 100 μm. G, GluA2 showed a similar expression level in both WT and ChR2 retinas. H–M, A vertical section through P11 WT and ChR2 mouse retinas. Similarly, DAPI was used to indicate nuclei in the retinal slice (blue), and ChAT was used to label cholinergic amacrine cells (red). Scale bar, 100 μm. N, ChAT also showed a similar expression level in both WT and ChR2 retinas. Taken together, these findings demonstrate that the effect of blue light stimulation on neurite outgrowth in ChR2 mice is not a result of maturation defect (n = 4 for WT; n = 5 for ChR2). n.s, p > 0.05. Error bars, mean ± SEM.

Retinal angiogenesis is accelerated in P11 MKO mice. A–F, Confocal images without or with analytical labels of the superficial vascular layer of P11 retina in ChR2, WT, and MKO mice, respectively. All morphologically recognized vessels were CD31-positive, the red lines represent the skeleton of vessels, the yellow lines represent the outline of vessels, and the blue dots represent the branching points of vessels. Scale bar, 200 μm. G, H, Compared to ChR2 and WT mice, MKO mice showed larger junction density and vessels percentage area in the superficial vascular layer (n = 5 for ChR2; n = 11 for WT; n = 11for MKO). I–N, Confocal images without or with analytical labels of the intermediate vascular layer of P11 retina in ChR2, WT, and MKO mice, respectively. All morphologically recognized vessels were CD31-positive. The analytical labels are the same as in panels above. Scale bar, 200 μm. O, P, The junction density and the percentage of vessel area significantly increased in the intermediate vascular layer of MKO mice. These results indicate that vertical angiogenic sprouting into the deeper layer of the retina occurred earlier in the absence of melanopsin (n = 5 for ChR2; n = 11 for WT; n = 11 for MKO). n.s, p > 0.05; ***p < 0.001. Error bars, mean ± SEM.

Blue light stimulation induces a low level of apoptosis in P11 ChR2 retinal explants. A, B, Confocal images of P11 retinal explants at DIV 5 with no light stimulation (Ctrl) and 20 Hz light stimulation for 1 h at the beginning of the experiment. DAPI was used to label cell nuclei (blue), and casepase-3 antibody was used to detect cell apoptosis (green). C, Blue light stimulation did not significantly increase the expression of caspase-3 in the retinal explants (n = 4 for the control; n = 4 for 20 Hz). Scale bar, 20 μm. n.s, p > 0.05. Error bars, mean ± SEM.

Blue light facilitates neurite outgrowth of retinal explants from P5 WT mice. A–C, Confocal images of whole-mount retinas from P5, P11, and adult mice with ChR2-expressed RGCs. The fluorescence intensity indicates the eYFP-ChR2 expression level. Scale bar, 100 μm. D–G, Confocal images of P5 retinal explants of WT mice at DIV 5 with no stimulation (Ctrl), and with 5, 20, and 100 Hz light stimulation for 1 h at the beginning of the experiment. Recognized neurites were TUJ1-positive (red), and DAPI was used to label the nuclei (blue). Scale bar, 100 μm. H, In the absence of ChR2 expression in RGCs, the neurite outgrowth of retinal explants was still enhanced by blue light stimulation (n = 5 for the control; n = 5 for 5 Hz; n = 7 for 20 Hz; n = 5 for 100 Hz). This observation suggests that blue light was likely to activate ipRGCs within the P5 retinas at which time the retina's rods and cones have not fully developed; **p < 0.01. Error bars, mean ± SEM.

Activation of ipRGCs via blue light stimulation further enhances neurite outgrowth of retinal explants from P11 mice. A–C, Confocal images of the P11 retinal explants from ChR2 mice at DIV 5 with no stimulation (Ctrl), and with 20 and 100 Hz light stimulation for 1 h at the beginning of the experiment. D–F, Confocal images of P11 retinal explants from WT mice at DIV 5 with the same treatments as above. G–I, Confocal images of P11 retinal explants from MKO mice at DIV 5 with the same treatments as above. Recognized neurites were TUJ1-positive (red), and DAPI was used to label the nuclei (blue). Scale bar, 100 μm. J, Neurite outgrowth of P11 retinal explants was significantly different when mice with three different genetic backgrounds (ChR2, WT, and MKO) were compared, even without blue light stimulation (n = 7 for ChR2; n = 5 for WT; n = 6 for MKO). K, Neurite outgrowth of P11 retinal explants was significantly better for ChR2 mouse retinas than for WT mouse retinas when there was 20 Hz blue light stimulation; furthermore, the neurite outgrowth was significantly reduced in MKO mice (n = 9 for ChR2; n = 5 for WT; n = 6 for MKO). L, A similar trend was found for P11 retinal explants with 100 Hz light stimulation (n = 6 for ChR2; n = 5 for WT; n = 6 for MKO). M, When ipRGCs and ChR2-RGCs were absent in the MKO mice, neurite outgrowth of P11 retinal explants was significantly poorer under all three conditions (no stimulation, 20, and 100 Hz blue light stimulations). N, O, Similar comparison schemes indicate that 20 Hz blue light stimulation enhanced neurite outgrowth more significantly than the ones without light stimulation for both WT and ChR2 mice, respectively. n.s, p > 0.05; *p < 0.05; ***p < 0.001. Error bars, mean ± SEM.

MEA recording of P11 retinas with ChR2-expressed RGCs in response to blue light stimulation. A, Responses of the RGCs from the three different strains of mice (ChR2, WT, and MKO) on a 200-ms blue light stimulation. The dotted line represents the start of light stimulation. B, Responses of the RGCs from ChR2 mice on 5, 20, and 100 Hz continuous blue light stimulation. When the retina was activated by 5 and 20 Hz blue light, the recorded RGC showed phase-locking responses. However, the same RGC, when stimulated by 100 Hz blue light, was not able to elicit reliable spiking responses after each light stimulation. C–E, the spiking rates of the RGCs from ChR2 mouse retinas in response to 5 Hz (n = 35), 20 Hz (n = 37), and 100 Hz (n = 20) blue light stimulation continuously for 1 h. The spiking rate of individual RGCs at each time point was based on a 1 min measurement every 10 min. The thick semitransparent lines represent the average spiking rates. F, Comparison of the average responses of the RGCs on blue light stimulation using the three temporal patterns. The RGCs responses at 5 and 20 Hz blue light were relatively sustained throughout the 1-h stimulation, while the 100 Hz blue light evoked mostly only transient responses.

Blocking the gap junctions in the explant retinas significantly reduces the response strength and number of blue light evoked RGCs from P11 retinas of ChR2 mice. A, MEA recording of four neighbor channels showing the responses of RGCs on 5 Hz blue light stimulation. All channels showed strong spiking responses before the addition of MFA (100 μM) to block the gap junctions. After blocking gap junction activity, the light responses of all channels were decreased drastically, and only one channel showed a reliable spiking response on each blue light stimulation. B, Application of MFA significantly decreased the spiking rates of RGCs in response to 5, 20, and 100 Hz blue light stimulation (n = 52 for each condition). C, An MEA recording showing a representative RGC responses on multiple 1-s blue light stimulations. After blocking the retina’s gap junctions, most channels showed no light response (only #32, #52, and #64 are shown here). However, a few channels did show a reduced yet reliable response on each blue light stimulation (and #47), and these are likely to be ChR2-expressed RGCs. Occasionally, one or two channels showed a long-lasting response to blue light stimulation, which suggests that these are one of ipRGCs present in the developing retina. D–F, Confocal images of P11 ChR2 retinal explants at DIV 5 with 20 Hz light stimulation for 1 h, 20 Hz light stimulation for 1 h with 100 μM MFA, and no light stimulation with 100 μM MFA, respectively. All morphologically recognized neurites were TUJ1-positive (red), and DAPI was used to label nuclei (blue). Scale bar, 100 μm. G, MFA application which blocks gap junction coupling significantly reduced neurite outgrowth of P11 ChR2 retinal explants, even under 20 Hz light stimulation for 1 h (n = 9 for 20 Hz; n = 5 for 20 Hz with MFA; n = 3 for the control with MFA); **p < 0.01, ***p < 0.001. Error bars, mean ± SEM.