Necl-1/CADM3 regulates cone synapse formation in the mouse retina

Abstract

In vertebrates, retinal neural circuitry for visual perception is organized in specific layers. The outer plexiform layer is the first synaptic region in the visual pathway, where photoreceptor synaptic terminals connect with bipolar and horizontal cell processes. However, molecular mechanisms underlying cone synapse formation to mediate OFF pathways remain unknown. This study reveals that Necl-1/CADM3 is localized at S- and S/M-opsin-containing cones and dendrites of type 4 OFF cone bipolar cells (CBCs). In Necl-1^-/- mouse retina, synapses between cones and type 4 OFF CBCs were dislocated, horizontal cell distribution became abnormal, and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors were dislocated. Necl-1^-/- mice exhibited aberrant short-wavelength-light-elicited signal transmission from cones to OFF CBCs, which was rescued by AMPA receptor potentiator. Additionally, Necl-1^-/- mice showed impaired optokinetic responses. These findings suggest that Necl-1 regulates cone synapse formation to mediate OFF cone pathways elicited by short-wavelength light in mouse retina.

Keywords: Biological sciences; Molecular neuroscience; Neuroscience;.

Graphical abstract

Figure 1

Necl-1 is abundantly expressed in mouse retina Retinas from 8-week WT mice were stained with an anti-Necl-1 pAb, which recognizes the intracellular region of Necl-1 protein. (A–C) Schematic drawings of an adult mouse retina and Necl-1 localization. (A) Transverse sectional view of the retina. Light pink, retinal ganglion cells; light orange, AII BCs; light gray, amacrine cells. Orange, dark blue, and magenta arrows show signal transduction between cells. (B) Transverse sectional view of an S or S/M-opsin-containing cone pedicle and the dense plexus near the basal synapse in WT mouse retina. Expression of Necl-1 is shown in red. White circles, synaptic vesicles; light yellow, HCs; light green, ON CBCs; orange, type 4 OFF CBCs; brown, OFF CBCs; and green oval, GluA1. (C) Schematic showing Necl-1-mediated cell adhesion at the cone pedicle (left and middle) and the dense plexus near basal synapses (right). (D and E) In situ hybridization for Necl-1. (D) Antisense probe; (E) sense probe. (F) Expression of Necl-1 in mouse retina. Lysates obtained from 8-week WT and Necl-1^−/− mouse retinas were subjected to western blotting using an anti-Necl-1 pAb and an anti-β-actin pAb. (G and H) Immunohistochemistry in Necl-1^−/− mouse retina. Retinas of 8-week WT and Necl-1^−/− mice from the same litter were subjected to immunohistochemistry using the anti-Necl-1 pAb. (I) Quantification of the Necl-1 signal detected in each stratum of the IPL. NFL, nerve fiber layer; GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer; OPL, outer plexiform layer; ONL, outer nuclear layer; and PC, photoreceptor cell. Scale bars in D, E, G, and H, 20 μm. Error bars in I represent ±SD from WT mouse (n = 4). See also Figure S1.

Figure 2

Necl-1 is expressed in retinal cone photoreceptor cells (A–D) Localization of the Necl-1 signal in the OPL and ONL. The retina was double-immunostained with the Necl-1 pAb (green) plus one of the following Abs for various retinal photoreceptor cell marker proteins (red): (A) an anti-cone somata mAb (a marker for cones); (B) an anti-S-opsin pAb (a marker for S-opsin-positive cones); (C) an anti-M-opsin pAb (a marker for M-opsin-positive cones); (D) an anti-rhodopsin mAb (a marker for rods); (E) Observed percentage of Necl-1-positive cells among various opsin-expressing photoreceptor cells. INL, inner nuclear layer; OPL, outer plexiform layer; and ONL, outer nuclear layer. Scale bars in A–D, 15 μm. Error bars in E represent ±SD from means of n = 6 to 8 retina sections per WT mouse (n = 3).

Figure 3

Necl-1 is expressed in OFF cone bipolar cells (A–E) Localization of the Necl-1 signal in the INL, OPL, and ONL. The retina was double-immunostained with the Necl-1 pAb (green) plus one of the following Abs for various retinal bipolar cell marker proteins (red): (A) an anti-HCN4 mAb for type 3a OFF CBCs; (B) an anti-PKARIIβ mAb for type 3b OFF CBCs; (C) an anti-calsenilin mAb for type 4 OFF CBCs; (D) an anti-SYT2 mAb for type 2 OFF CBCs and 6 ON CBCs; (E) an anti-PKCα mAb for rod BCs. For nuclear counter staining (blue), 4′,6-diamidino-2-phenylindole (DAPI) was used. IPL, inner plexiform layer; INL, inner nuclear layer; OPL, outer plexiform layer; and ONL, outer nuclear layer. Scale bars in A–E, 15 μm. See also Figure S2.

Figure 4

Role of Necl-1 in formation of synapses between cones and type 4 OFF CBCs Retinas from 8-week WT and Necl-1^−/− mice were used. (A and B) Density of S-opsin-positive cones. (A) S-opsin-positive cones in the ONL. Red, S-opsin; and blue, DAPI. Scale bars, 15 μm. (B) Density of S-opsin-positive cones. The average density of S-opsin-positive cones was calculated. Error bars represent ±SD from means of n = 7 to 9 retina sections collected from individual animals (n = 5; WT and Necl-1^−/−, respectively). n.s., not significant (p = 0.71) by paired Student’s t test. (C and D) Density of bipolar cells. (C) Bipolar cells in the ONL. Red, Chx10; and blue, DAPI. Scale bars, 15 μm. (D) Density of bipolar cells. The average density of Chx10-positive cells was calculated. Error bars represent ±SD from means of n = 5 to 7 retina sections collected from individual animals (n = 4; WT and Necl-1^−/−, respectively). n.s., not significant (p = 0.50) by paired Student’s t test. (E–M) Density and abnormal extension of dendrites of OFF CBCs. (E, H, and K) Retinal sections stained with OFF CBC markers and peanut agglutinin (PNA). Green, HCN4 (E), PKARIIβ (H), and calsenilin (K); red, PNA; and blue, DAPI. Arrows indicate abnormal dendrites of type 4 OFF CBCs to the photoreceptor cell side in Necl-1^−/−mouse retina. Scale bars, 15 μm. (F, I, and L) Average density of OFF CBCs was calculated. F, HCN4; I, PKARIIβ; L, calsenilin. Error bars represent ±SD from means of n = 5 to 8 retina sections collected from individual animals (n = 4; WT and Necl-1^−/−, respectively). n.s., not significant (F, p = 0.84; I, p = 0.97; L, p = 0.92) by paired Student’s t test. (G, J, and M) Number of dendrites of OFF CBCs extending into the ONL per OFF CBC. G, HCN4; J, PKARIIβ; M, calsenilin. Error bars represent ±SD from means of n = 5 to 8 retina sections collected from individual animals (n = 4 or 5; WT and Necl-1^−/−, respectively). n.s., not significant (G, p = 0.06; J, p = 0.44). ∗∗, p = 0.001 by paired Student’s t test. See also Figure S3.

Figure 5

Role of Necl-1 in signal transduction from short-wavelength-sensitive cones to OFF CBCs (A–C) Photopic flicker ERGs in response to UV-LED flashes. Light-adapted photopic flicker ERGs were elicited by six frequencies. (A) Representative ERGs; Black broken lines, a peak in a cycle; and lower gray line markers, UV-LED stimuli. (B) Amplitude of flicker ERGs; (C) Implicit time from trough to peak. Error bars represent ±SEM from means of n = 7 (WT) or n = 7 (Necl-1^−/−). ∗, p = 0.03; and ∗∗, p = 0.002 by paired Student’s t test. (D–F) Long-duration flash ERGs after intravitreous injection of L-AP4. (D) Representative long-duration flash ERGs; (E) Peak-to-trough amplitude of each long-duration ERGs; (F) Implicit time from trough to peak. Error bars represent ±SEM from means of n = 4 (WT) or n = 5 (Necl-1^−/−). ∗, p = 0.03; and ∗∗, p = 0.0003 by paired Student’s t test. See also Figures S4–S6.

Figure 6

Role of Necl-1 in localization of AMPA receptors at synapses between cones and type 4 OFF CBCs 8-week WT and Necl-1^−/− mice were used for experiments. (A) Localization of GluK1. Retinal sections were stained with an anti-GluK1 Ab and PNA. Green, GluK1; red, PNA. Scale bars, 5 μm. (B and C) Abnormal localization of GluA1 in Necl-1^−/− mouse retina. (B) Retinal sections stained with an anti-GluA1 Ab and PNA. Green, GluA1; and red, PNA. Solid arrows and dotted arrows indicate the punctate GluA1 signal. Scale bars, 5 μm. (C) The punctate GluA1 signal at the inner side of the cone pedicle. Localization of GluA1 was analyzed. Error bars represent ±SD from means of n = 7 retina sections collected from individual animals (n = 5; WT and Necl-1^−/−, respectively). ∗∗, p < 0.001 by paired Student’s t test. (D and E) Schematic drawings of localizations of Necl-1, GluA1, and GluK1 in adult mouse retina. Transverse sectional view of the S- or S/M-opsin-containing cone pedicles in WT and Necl-1^−/− mouse retinas. Red circles, Necl-1; white circles, synaptic vesicles; light yellow, HCs; light green, ON CBCs; orange, type 4 OFF CBCs; brown, OFF CBCs; green oval, GluA1; and dark blue oval, GluK1. (D) WT; (E) Necl-1^−/−.

Figure 7

Role of Necl-1 in the S-opsin-mediated signaling OFF pathway via AMPA receptors (A and B) Light-adapted photopic 30-Hz flicker ERGs in 8-week WT mice. Gray line, control saline injection; and black line, 13.15-mM PDA injection. (A) Representative light-adapted photopic 30-Hz flicker ERGs. Black broken lines, a peak in a cycle; and lower gray line markers, UV-LED stimuli. (B and C) Photopic 30-Hz flicker ERGs. Gray lines, control saline injection; and black lines, PDA injection. (B) Amplitude. (C) Implicit time. Error bars show ±SD (n = 6, for the saline and PDA mice). ∗, p = 0.03 (amplitude at 15 min) and ∗∗, p = 0.002 and 0.001 (implicit time at 30 min and 45 min, respectively) by Student’s t test. (D–F) Time course of photopic 30-Hz flicker ERGs in response to UV-LED flashes after an intravitreous injection of the AMPA receptor potentiator, LY404187, in Necl-1^−/− mice. Gray line, saline injection; and black line, LY404187 injection. (D) Representative response trace showing cycle peaks in photopic 30-Hz flicker ERGs. Both cycles were under UV-LED stimuli. (E) Amplitude of flicker ERGs. Gray line, Necl-1^−/− mice with saline injection (n = 6); black line, Necl-1^−/− mice with LY404187 injection (n = 5); gray broken line, WT mice with saline injection (n = 6), and black broken line, WT mice with LY404187 injection (n = 4). ∗, p = 0.01 (LY404187 vs. saline in Necl-1^−/− mice at 15 min); and ∗∗, p = 0.003 (Necl-1^−/− vs. WT mice with saline injection at 15 min) by Wilcoxon signed-rank test. (F) Implicit time. ∗, p = 0.01 (LY404187 in Necl-1^−/− mice vs. saline in WT mice at 15 min); ∗, p = 0.01 (LY404187 vs. saline in Necl-1^−/− mice at 15 min); ∗, p = 0.03 (Necl-1^−/− vs. WT mice with saline injection at 60 min); ∗, p = 0.04 (Necl-1^−/− vs. WT mice with LY404187 injection at 60 min); and ∗∗, p = 0.005 (Necl-1^−/− vs. WT mice with saline injection at 15 min) by Wilcoxon signed-rank test.

Figure 8

Abnormal distribution of HCs caused by Necl-1 ablation Immunohistochemical analysis of 8-week WT and Necl-1^−/− mouse retinas. (A–C) Abnormal distribution of calbindin-positive HCs in the Necl-1^−/− mouse retina. (A) Calbindin-positive HCs in the OPL. The anti-calbindin mAb/pAb was used. Red, calbindin; and blue, DAPI. The white line indicates the distance between the uppermost border of the soma of a calbindin-positive HC and the position of the OPL. (B) The average distance between the uppermost border of somas of calbindin-positive HCs and the position of the OPL in WT and Necl-1^−/− mouse retinas. A total of 36 horizontal cells were measured in WT or Necl-1^−/−, respectively. ∗∗, p < 0.001 by paired Student’s t test. Error bars represent ±SD from means of n = 4 (WT and Necl-1^−/−, respectively). (C) Numbers of HC somata per retinal section. Error bars represent ±SD from means of n = 6 to 11 retina sections collected from individual animals (n = 3 and 4; WT and Necl-1^−/−, respectively). n.s., not significant (p = 0.95) by paired Student’s t test. (D–G) Increase in numbers of HC somata surrounded by BC somata in Necl-1^−/− mouse retina. Numbers of HC somata surrounded by Chx10-positive or PKCα-positive BC somata were counted. (D and E) HCs surrounded by somata of Chx10-positive BCs. (D) Calbindin-positive HCs and Chx10-positive BCs in the INL and OPL. The anti-calbindin mAb and the anti-Chx10 pAb were used. Green, Chx10; and red, calbindin. (E) Percentage of HC somata surrounded by Chx10-positive BC somata in the WT and Necl-1^−/− mouse retinas. Error bars represent ±SD from means of n = 36 horizontal cells collected from individual animals n = 3 (WT and Necl-1^−/−, respectively). ∗, p = 0.02 by paired Student’s t test. (F and G) HCs surrounded by somata of PKCα-positive BCs. (F) Calbindin-positive HCs and PKCα-positive BCs in the INL and OPL. The anti-calbindin pAb and the anti-PKCα mAb were used. Green, PKCα; and red, calbindin. (G) Percentage of HC somata surrounded by PKCα-positive BC somata in WT and Necl-1^−/− mouse retinas. Error bars represent ±SD from means of n = 36 horizontal cells collected from individual animals n = 3 (WT and Necl-1^−/−, respectively). ∗, p = 0.04 by paired Student’s t test. (H) Whole-mount immunostaining of WT and Necl-1^−/− mouse retinas. The anti-calbindin mAb was used. Each immunofluorescence signal was collected horizontally from the plane containing HC somata to the plane containing the OPL. Scale bars, 5 μm in A, D, and F; and 50 μm in H. See also Figure S7.

Figure 9

Optokinetic responses to motion of sinusoidal gratings 8-week WT and Necl-1^−/− mice were used for these experiments. (A and B) Diagrams showing experimental equipment used to record mouse eye movements. Infrared images of the right eye were collected using a charge-coupled device (CCD) camera. To expose the entire visual field of mice to short-wavelength light (blue color) stimuli, three UV band-pass filters were positioned in front of mice. (A) Experimental setup without filters. (B) Setup with UV band-pass filters. (C) Experimental scheme. First, a stationary visual pattern was presented, followed by presentation of a counterclockwise (temporal-nasal motion for the right eye) or clockwise (nasal-temporal motion for the right eye) moving visual pattern at a constant speed. After a defined period, the pattern was removed. (D and E) Eye velocity profiles. Gray solid line, WT mice (n = 6); and black solid line, Necl-1^−/− mice (n = 5). (D) Without filters; (E) With UV band-pass filters. Broken lines indicate the standard error of the mean. “Time zero” denotes the onset time of visual motion. (F and G) Mean eye velocity during the open-loop phase of visual stimuli under the normal or UV condition. (F) During presentation of visual stimuli without filters; or (G) with UV band-pass filters. Error bars represent ±SD from means of n = 6 (WT) or n = 5 (Necl-1^−/−). ∗, p = 0.015; and ∗∗, p = 0.0074 by paired Student’s t test.