Figure 1.
Morphology of M1 and M2 ipRGCs in the retina of Arvicanthis.
A–I) Wholemounts of the retina from adult (9 months old) Arvicanthis immunostained for melanopsin (A, D, G; green in C, F, I) and ChAT (B, E, H; magenta in C, F, I), a cholinergic amacrine cell marker. Pictures were taken from focal planes on the GCL (A–C), IPL (D–F) and IPL/INL border (G–I). M1 cells show an intense melanopsin staining (arrow in A and C). Strongly melanopsin-positive dendrites of M1 cells pass the IPL (arrow in D and F) and stratify in the outer portion of the IPL (OFF sublamina) in proximity to the INL (in G and I the arrow shows the branching point of the main dendrite); M2 cells show a less intense melanopsin staining (arrowhead in A and C). Weakly melanopsin-positive M2 dendrites stratify in the inner portion of IPL (ON sublamina) near to the GCL (A–C). J–L) Side view and stratification level of M1 cell (soma indicated by an arrow). M–O) Side view and stratification level of the M2 cell (soma indicated by an arrowhead); note that in (M) and (O) also the intense OFF plexus of processes from melanopsin cells is visible. ChAT, choline acetyl transferase; GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer; Off, OFF sublamina of IPL; On, ON sublamina of IPL. Scale bar: 50 µm.
Figure 2.
Morphology of orthopic and displaced ipRGCs in the retina of Arvicanthis.
A–H) Retinal wholemounts of adult (9 months old) Arvicanthis immunostained for melanopsin (A, C, E, G; green in B, D, F, H) and ChAT (magenta in B, D, F, H), a cholinergic amacrine cell marker. Pictures were taken from focal planes on the GCL (A–B), IPL (C–D), IPL/INL border (E–F) and the INL (G–H). I–K) Side view of the cells. Orthotopic M1 cell (arrow in A and C) have cell bodies located in the GCL and stratify in the outer portion of the IPL (OFF sublamina), close to the INL (E–H and I). These cells show stronger melanopsin-immunoreaction than M2 cells (arrowhead in A), including cell body and dendrites. Displaced M1 (M1-d) cells show a comparable intensity of melanopsin staining to the orthotopic M1 cells, nevertheless their cell bodies are located in the INL (two headed arrow in E and G). M1-d cells stratify in the outer portion of the IPL, similar to their orthotopic counterparts (E–H and J–K). Figure 2 I shows only one M1 cell, Figure 2 J a M1 and a M1-d cell, and Figure 2 K only a single M1-d cell. ChAT, choline acetyl transferase; GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer. Scale bar: 50 µm.
Figure 3.
Density of melanopsin-positive ipRGCs in the retina of Arvicanthis at different ages.
A) Representative photomicrographs showing the change of the density of melanopsin-positive ipRGCs at different stages of the postnatal development. Scale bar: 100 µm. B) Plot showing a change of mean density of ipRGCs in the course of postnatal development. In the new-born retina (P0, postnatal day 0) almost 300 cells/mm2 were observed. The values considerably declined during the first 2 postnatal weeks reaching a density of 20–40 cells/mm2 in adult animals and stayed at nearly constant levels until the late adulthood. Values are presented as mean density (cells/mm2) ± SEM (1–6 retinae per age group).
Figure 4.
Distribution of ipRGCs in the retina of Arvicanthis immunopositive to specific RGC markers.
Melanopsin-positive ipRGCs of both types (M1 and M2) show high expression of MAP2, γ-Syncl and Isl1. Brn3 distinguishes between M2 and M1 cells. Almost 60% of the M2 type expresses Brn3, but only 5% of the M1 type. Another clear difference between M1 and M2 cells is observed in the expression of different neurofilaments (NF200, NF160 and NF68) that label a small population of the M2 cells, but virtually none of the M1 ipRGCs. Expression of NeuN is higher in the M2 type than in the M1 cells.
Figure 5.
Expression of different RGC markers in retinal ipRGCs of Arvicanthis.
Double immunolabeling of retina wholemounts with an anti-melanopsin antibody (left column) and a specific RGC marker (middle column). The right column depicts the superposition of the two previous pictures (melanopsin-positive ipRGCs in magenta, specific markers in green). Cells showing a colocalization of melanopsin and the given marker are indicated by yellow arrowheads. Scale bar: 50 µm. A–C) Colocalization of Brn3 and melanopsin in ipRGCs. Brn3 is expressed only in some ipRGCs, mostly of the M2 group. The picture shows one M2 cell expressing Brn3, the other ipRGC (M1) is negative to Brn3. D–F) Colocalization of NF200 and melanopsin in ipRGCs. Only very few ipRGCs express NF200 and all of them belong to the M2 cells. The NF200+-ipRGCs have large somata and show very faint melanopsin-immunoreactivity. G–I) Colocalization of NeuN and melanopsin in ipRGCs. NeuN is expressed only in some M1 and M2 ipRGCs. One NeuN-positive M1 ipRGC is shown in the picture, the other melanopsin+-RGC (M2) is negative to NeuN. J–L) Colocalization of MAP2 and melanopsin in ipRGCs. Nearly all melanopsin+-RGCs coexpress MAP2. The picture shows one M1 and one M2 ipRGCs coexpressing MAP2. M–O) Colocalization of γ-Synuclein (γ-Sncl) and melanopsin in ipRGCs. γ-Synuclein is expressed in the majority of M1 and M2 ipRGCs. Two γ-Synuclein+ M1 cells are shown in the picture. P–S) Colocalization of Isl1 and melanopsin in ipRGCs. The great majority of M1 and M2 ipRGCs express Isl1. In the picture two M2 and one M1 cells are visible, all show Isl1-immunoreactivity.
Figure 6.
Physiological heterogeneity of ipRGCs in newborn Arvicanthis retinas.
Light evoked responses of the three ipRGC types to 30 s stimuli (505 nm, full-field illumination) of increasing intensity recorded from a P0 Arvicanthis retina. Left column shows representative examples of spike trains recorded with multielectrode arrays. Each vertical line represents a single action potential. The inset in the left column in the lower graph shows a magnification of such an action potential. Right column depicts the corresponding histograms of the firing rate. The asterisks in the right column point to retinal waves that interfere with the spike responses of ipRGCs. Upper graphs ipRGCs type I, middle graphs type II, and lower graphs type III.
Figure 7.
Response characteristics of the three ipRGC types recorded from newborn Arvicanthis and mouse retinas.
A–B) response latency and C–D) light-evoked spike rate. (A) and (C) depicts the mean values (± SEM) and (C) and (D) the normalized response fitted with a sigmoidal Michaelis-Menten function. Light responses of single units measured from P0 – P3 retinas (4 retinas, n = 21, 51 and 21 for types I, II and III, respectively). Stimulus: 505 nm, 30 s. E–J) Response characteristics of the different ipRGCs types from newborn Arvicanthis (open circles, black curves) in comparison to ipRGCs from retinas of newborn C57BL/6 mice (filled circles, dashed curves). Arvicanthis: n = 93 cells (4 retinas; P0 – P3); mice: n = 39 cells (3 retinas; P0 – P2). E–G) Normalized response latency for cells of type I (E), II (F) and III (G). H–J) Normalized spike rate of type I (H), II (I) and III (J). All values are mean ± SEM. Stimulus: 505 nm, 30 s.
Figure 8.
Responses of ipRGCs from Arvicanthis and mouse retina to stimuli of different duration.
A) Light-induced responses of ipRGCs in Arvicanthis (left) and in the mouse retina (right) to short flashes of constant irradiance (14.3 log NQ * cm-2 * s-1). The most sensitive ipRGC type (type I) respond to very short light flashes of up to 10 ms in Arvicanthis and 50 ms in the mouse, respectively, whereas type II and type III ipRGCs were insensitive to flashes shorter than 1 s. B) Response characteristics of ipRGCs type I of Arvicanthis to light flashes of different duration. The left graph shows that the mean spike rate (filled circles) and peak firing rate (open circles) reached a maximum with stimuli lengths of 1 s. The response duration (middle graph) continuously increased with the stimulus duration, whereas response latencies (right graph; filled circles: latency to the first spike, open circles: latency to peak firing rate) rapidly declined with increasing stimulus duration. Data presented as mean values ± SEM.
Figure 9.
Type I ipRGCs’ response to flash stimuli (10 ms) of different frequency.
Light responses of a type I ipRGCs of Arvicanthis to brief, repeated light stimulation of constant irradiance (505 nm; 10 ms; 14.3 log NQ * cm-2 * s-1). A) Spike trains in response to repeated light flashes, and (B) corresponding spike histograms. Light stimulation is shown below the spike trains and the histograms as an upward deflection of the traces. Flashes were applied with interstimulus intervals (ISI) between 50 ms and 50 s (20–0.02 Hz) indicated on the left side. For each ISI flashes were repeated 10x. However, for the ISI of 50 s only the first three light flashes are shown. The Lomb-Scargle periodograms revealed significant periods for the 5 s, 10 s and 50 s ISI; when the time between two stimuli is below 5 s, single responses showed a fusion with the preceding response.