Autophagy-mediated catabolism of visual transduction proteins prevents retinal degeneration

Abstract

Autophagy is a lysosomal degradation pathway critical to preventing the accumulation of cytotoxic proteins. Deletion of the essential autophagy gene Atg5 from the rod photoreceptors of the retina (atg5^Δrod mouse) results in the accumulation of the phototransduction protein transducin and the degeneration of these neurons. The purpose of this study is to test the hypothesis that autophagic degradation of visual transduction proteins prevents retinal degeneration. Targeted deletion of both Gnat1 (a gene encoding the α subunit of the heterotrimeric G-protein transducin) and Atg5 in the rod photoreceptors resulted in a significantly decreased rate of rod cell degeneration as compared to the atg5^Δrod mouse retina, and considerable preservation of photoreceptors. Supporting this we used a novel technique to immunoprecipitate green fluorescent protein (GFP)-tagged autophagosomes from the retinas of the GFP-LC3 mice and demonstrated that the visual transduction proteins transducin and ARR/arrestin are associated with autophagosome-specific proteins. Altogether, this study shows that degradation of phototransduction proteins by autophagy is necessary to prevent retinal degeneration. In addition, we demonstrate a simple and easily reproducible immunoisolation technique for enrichment of autophagosomes from the GFP-LC3 mouse retina, providing a novel application to the study of autophagosome contents across different organs and specific cell types in vivo.

Keywords: autophagy; photoreceptor; retina; retinal degeneration; transducin.

Figure 1.

Crossing of the atg5^Δrod mouse with the gnat1^−/− mouse resulted in a mouse in which the rod photoreceptors both lacked the ability to activate autophagy as well as the α-subunit of the transducin molecule. (A) Immunohistochemical staining confirms that the atg5^Δrod mouse contains the GNAT1 protein (Tα). Cre-recombinase is detected in the outer nuclear layer, as expected, due to the derivation of the atg5^Δrod mouse from the cross between the Atg5-floxed mouse and a mouse with Rho-promoter-driven expression of CRE (the [Rho]-Cre mouse). (B, C) Western blot analysis confirms that the atg5^Δrod and the gnat1^−/− atg5^Δrod mouse strains both lack ATG5, whereas only the gnat1^−/− atg5^Δrod mouse lacks Tα. Both strains show markedly diminished autophagy activation, as demonstrated by the significant reduction in LC3 lipidation and increased accumulation of SQSTM1, as compared to the (Rho)-Cre mouse. (D) Densitometry analysis of the western blot bands confirms the accumulation of SQSTM1 and reduction in LC3-II formation. Normalization for each of the measures was to the value for the (Rho)-Cre retina. The asterisk signifies statistical significance at a P < 0.05; the double asterisk at P < 0.01. GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; IS, inner segment; OS, outer segment; RPE, retinal pigment epithelium.

Figure 2.

The gnat1^−/− atg5^Δrod mouse shows reduced photoreceptor degeneration as compared to the atg5^Δrod mouse. (A) Optical coherence tomography (OCT) was used to measure the outer nuclear layer (ONL) thickness (red bar) in vivo. All measurements were at 500 µm from the optic nerve head. The top, middle and bottom panels are from mice at 2, 6 and 10 mo of age, respectively. (B) Line graph showing the ONL thickness, as measured by OCT, in the atg5^Δrod mouse versus the gnat1^−/− atg5^Δrod mouse. (C) Heat maps of the retinal thickness in false colors from OCT scan and the matching grids (diameter = 1.5 mm) with center of the grid on the optic nerve (ON) showed that retina of the gnat1^−/− atg5^Δrod mouse had significantly thicker measurements at age of 10 mo compared with atg5^Δrod mouse. As can be seen from the heat maps, the preservation of the photoreceptors and the retinal thickness is generalized across the whole retina, and not confined to just one quadrant or focus. (D) Ex vivo histological analysis confirmed the in vivo ONL thickness measurements by OCT.

Figure 3.

Immunohistochemical and western blot analyses demonstrate that the gnat1^−/− atg5^Δrod mice retain more robust photoreceptor morphology and the capacity to synthesize RHO. (A) The histology samples were from 4- and 10-mo-old mice. (B) Samples for the western blots were from 10-mo-old mice. Note that the amount of GNAT/Tα in the atg5^Δrod mouse at this time point is reduced compared to the (Rho)-Cre mouse due to the significant degeneration of the retina and fewer remaining photoreceptor cells. GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; IS, inner segment; OS, outer segment; RPE, retinal pigment epithelium.

Figure 4.

Cone morphology and function is preserved in the gnat1^−/− atg5^Δrod mouse as compared to the atg5^Δrod mouse. (A) Representative photopic electroretinogram (ERG) traces (single flash and 20-Hz flicker) and quantification of peak photopic b-wave amplitude at 2 and 10 mo of age. While significantly reduced amplitudes of photopic b-wave (right panel) were observed in atg5^Δrod mice at the age of 10 mo, gnat1^−/− atg5^Δrod mice displayed significantly higher photopic ERG responses. (B) Optokinetic tracking responses were recorded from atg5^Δrod, gnat1^−/− atg5^Δrod and age-matched (Rho)-Cre mice (n = 6 each) at 2 and 10 mo of age. Spatial frequencies (expressed as cycles/degree) were recorded at 100% contrast. Although lower than (Rho)-Cre controls, the tracking response of the gnat1^−/− atg5^Δrod mice was significantly higher than that of the atg5^Δrod mice at the age of 10 mo, indicating preserved cone-mediated visual behavior. (C) Representative immunohistochemistry images of the comparable region of the retina of atg5^Δrod and gnat1^−/− atg5^Δrod mice at 2 and 10 mo of age, stained for Cre (red), OPN1MW/M-opsin (green) and DAPI (blue). Diminished numbers of cone photoreceptors were observed in the atg5^Δrod retina at the age of 10 mo, whereas in the gnat1^−/− atg5^Δrod retina, the number and the morphology of the cones were preserved. All the values in the bar graphs were represented as mean ± SEM and were compared by 2-tailed paired Student t test. *, P < 0.05; **, P < 0.01.

Figure 5.

Transfection of 661W cells with GFP-LC3 allows for the enrichment of autophagosomes by immunoprecipitation. (A) Fluorescence microscopy of the GFP-LC3 cells treated with bafilomycin A₁ and starved for 2 h shows the formation of green puncta, consistent with autophagosome formation. (B) Western blot analysis of the contents of the immunoprecipitate (IP) compared to the cell lysate of the GFP-LC3-transfected 661W cells confirms enrichment of GFP-LC3 and SQSTM1 but not ATG5 or GAPDH, consistent with enrichment of autophagosomes.

Figure 6.

Immunoprecipitation with anti-GFP antibody results in enrichment of photoreceptor autophagosomes from the retina of the GFP-LC3 mouse. (A) Green puncta corresponding to autophagosomes localize primarily to the inner segment of the photoreceptor in the GFP-LC3 mouse. Some puncta are seen in the cytoplasm surrounding the photoreceptor nuclei in the outer nuclear layer. (B) Fluorescence microscopy of the immunoprecipitant revealed enrichment of vacuoles with green fluorescent membranes, consistent with GFP-LC3 coating these structures. (C, D) Transmission electron microscopy of these vacuoles confirmed that they had double membranes. GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; IS, inner segment; OS, outer segment.

Figure 7.

Validation of the organelle content consistent with autophagosome enrichment. (A) Western blot analysis of the immunoprecipitate (IP) showed enrichment of proteins consistent with the autophagosome, and exclusion of proteins not associated with the autophagosome. (B) Western blot shows that the enriched autophagosome fraction contained more GNAT/transducin than ARR when collected from light-adapted (LA) retinas, consistent with the fact that transducin is in the inner segment at that time. When the autophagosomes were enriched from dark-adapted (DA) retinas, there was more ARR than transducin, consistent with the translocation of ARR into the inner segment and transducin into the outer segment. Quantification of the western blot bands for ARR and GNAT, both normalized to the GFP-LC3 levels. **,P < 0.01. (C) Histogram showing percentage of proteins associated with a particular cellular compartment that was detected on proteomic analysis of the enriched autophagosome fraction.