Increased phosphorylation of Cx36 gap junctions in the AII amacrine cells of RD retina

Abstract

Retinal degeneration (RD) encompasses a family of diseases that lead to photoreceptor death and visual impairment. Visual decline due to photoreceptor cell loss is further compromised by emerging spontaneous hyperactivity in inner retinal cells. This aberrant activity acts as a barrier to signals from the remaining photoreceptors, hindering therapeutic strategies to restore light sensitivity in RD. Gap junctions, particularly those expressed in AII amacrine cells, have been shown to be integral to the generation of aberrant activity. It is unclear whether gap junction expression and coupling are altered in RD. To test this, we evaluated the expression and phosphorylation state of connexin36 (Cx36), the gap junction subunit predominantly expressed in AII amacrine cells, in two mouse models of RD, rd10 (slow degeneration) and rd1 (fast degeneration). Using Ser293-P antibody, which recognizes a phosphorylated form of connexin36, we found that phosphorylation of connexin36 in both slow and fast RD models was significantly greater than in wildtype controls. This elevated phosphorylation may underlie the increased gap junction coupling of AII amacrine cells exhibited by RD retina.

Keywords: AII amacrine cells; gap junctions; hyperactivity; oscillations; phosphorylation; retinal degeneration; spontaneous activity.

Figure 1

Ser293-P antibody specifically recognizes Cx36 in the mouse retinal wholemounts. Confocal images show the lamina of the inner plexiform layer (IPL) adjacent to the retinal ganglion cell bodies. (A) mCx36 (blue) staining in wt retina (left). Magnified region shows strong punctate labeling (middle). No punctate labeling was present in Cx36KO retina, though there was some non-specific labeling of blood vessels by secondary anti-mouse antibodies (arrows). Their stereotypic localization of vascular elements in the IPL was used as a marker for specific retinal strata. The brightness of mCx36 puncta was increased for presentation purposes only to reveal all Cx36-positive plaques. (B) In the same region of wt retina as in (A), Ser293-P staining was present in wt retina, with similar punctate labeling. No Ser293-P staining was present in Cx36KO retina. The degree of phosphorylation, reflected by the intensity of the Ser293-P labeling, varied across Cx36 plaques. (C) The magnified region (middle) shows that the vast majority of Ser293-P-positive puncta colocalized with mCx36-postive puncta in wt retina. 97.8 ± 0.3% were colocalized with mCx36 (3 retinas × 3 samples = 976 detectably phosphorylated plaques). Scale bars: 10 μm, left and right columns; 2 μm, middle column.

Figure 2

The majority of Cx36 gap junctions in AII amacrine cells are on the arboreal dendrites. (A) Confocal z-stack reconstruction of AII amacrine cell in retina wholemount selectively labeled by recombinant adeno-associated virus serotype 2 (rAAV2)-green fluorescent protein (GFP). In the vertical retinal view, stratification of a single AII amacrine cell (white) is shown through the five IPL strata. (B) In the 1.5 um thick vertical cryostat section, matched to the IPL area in (A), Cx36 puncta were predominantly present in the On-sublamina (strata 3–5). Cx36 were indirectly labeled by the mCx36 antibody (blue) and phosphorylated Cx36 at Ser293 were recognized by Ser293-P antibody (red). (C,D) In a retinal wholemount, a rAAV2-GFP-labeled AII amacrine cells expressed more Cx36 in stratum 5 (D, marked by dots) than in stratum 2 (C). Single confocal sections from the boxed areas were magnified (×6.2) and are shown on the right. Among all Cx36-positive puncta, the puncta colocalized with the processes of the AII amacrine cell are marked by arrows. Within the processes of the same AII amacrine cell different amount of phosphorylation was detected. Scale bars: 5 μm.

Figure 3

Phosphorylation of Cx36 gap junctions is elevated in RD retina. (A–C) Single confocal sections through the stratum 5 of the retinal IPL labeled for Cx36 (mCx36, blue) and its phosphorylated form (Ser293-P, red) are shown for wt (left column), rd10 (middle), and rd1 (right). (D) The magnified boxed areas from the merged images are shown in the bottom row. High amount of phosphorylation, reflected by pink color in the merged images, was characteristic for both rd10 and rd1 retinas. Scale bars: 10 μm.

Figure 4

Quantification of Cx36 gap junctions in AII amacrine cells in wt, rd10 and rd1 retinas. (A) Cx36 was heterogeneously phosphorylated in the retinas of all mouse models, which is reflected by a low correlation between the intensity of Cx36 and its phosphorylation (wt r² = 0.096; rd10 r² = 0.131; rd1 r² = 0.299). (B) Relative phosphorylation of Cx36, extracted from the data in (A), shows overall more phosphorylation in retina of RD models. (C) The absolute amount of phosphorylation, reflected by the percentage of Cx36 plaques that show detectable Ser293-P labeling, was also elevated in RD retinas. (D,E) The density of Cx36-positive plaques (D), their size (E), and intensity of mCx36 labeling (F) were not significantly altered in RD models. The data are presented as average; error bars are SEM. All significances are based on ANOVA test, where ***p < 0.001, **0.001 < p < 0.01, *0.01 < p < 0.05, n.s. p > 0.05.

Figure 5

Illustration of the events leading to opening and closing of gap junctions in AII amacrine cell. (A) Glutamate, released from bipolar cells, induces Ca²⁺ influx via N-methyl-D-aspartate (NMDA) receptors which in turn mediates CaMKII-dependent phosphorylation of Cx36 and opens gap junctions. Dopamine, released by dopaminergic amacrine cells, binds to D1-like receptors activating phosphatase protein phosphatase 2A (PP2A) which in turn dephosphorylates Cx36 and closes gap junction. (B) The diagram illustrating a mechanism by which changes in glutamate and dopaminergic signaling may lead to increased phosphorylation of Cx36-containing gap junctions promoting coupling of AII amacrine cells.