Guanylate cyclase 1 relies on rhodopsin for intracellular stability and ciliary trafficking

Abstract

Sensory cilia are populated by a select group of signaling proteins that detect environmental stimuli. How these molecules are delivered to the sensory cilium and whether they rely on one another for specific transport remains poorly understood. Here, we investigated whether the visual pigment, rhodopsin, is critical for delivering other signaling proteins to the sensory cilium of photoreceptor cells, the outer segment. Rhodopsin is the most abundant outer segment protein and its proper transport is essential for formation of this organelle, suggesting that such a dependency might exist. Indeed, we demonstrated that guanylate cyclase-1, producing the cGMP second messenger in photoreceptors, requires rhodopsin for intracellular stability and outer segment delivery. We elucidated this dependency by showing that guanylate cyclase-1 is a novel rhodopsin-binding protein. These findings expand rhodopsin's role in vision from being a visual pigment and major outer segment building block to directing trafficking of another key signaling protein.

Keywords: guanylate cyclase; membrane protein trafficking; mouse; neuroscience; photoreceptor; rhodopsin; sensory cilium.

Figure 1.. Localization of outer segment membrane proteins in wild-type (WT) and Rho^-/- retinas.

(A) Electron micrographs showing the outer segment and connecting cilium in WT and Rho^-/- rods (scale bar 500 nm). (B–K) Immunofluorescent localization of individual outer segment proteins in WT and Rho^-/- retinal cross-sections: (B) Rom-1; (C) ABCA4; (D) guanylate cyclase 2 (GC-2); (E) cyclic nucleotide gated (CNG) α1; (F) CNGβ1; (G) prominin; (H) protocadherin 21 (PCDH21); (I) peripherin; (J) R9AP; and (K) GC-1. (L) Double labeling of GC-1 (green) and the cone maker, PNA (magenta). Here and in the following figures, the identity of antibodies used in each panel is indicated in ‘Materials and methods’. Scale bars, 10 μm. Nuclei are stained by Hoechst (blue). DOI: http://dx.doi.org/10.7554/eLife.12058.003

Figure 2.. Rhodopsin expression and outer segment localization do not rely on GC-1.

Rhodopsin (magenta) and GC-1 (green) were co-immunostained in retinal cross-sections from wild-type (WT), GC-1^-/-, and Rho^-/- mice. Scale bar, 10 μm. Nuclei stained in blue. DOI: http://dx.doi.org/10.7554/eLife.12058.004

Figure 3.. Quantification of outer segment transmembrane proteins in Rho^-/- retinas at P21.

(A) Representative Western blots show serial dilutions of wild-type (WT) and Rho^-/- retinal lysates for three proteins (guanylate cyclase 1 [GC-1], GC-2, and peripherin). The fluorescent signal produced by each band in the serial dilution was plotted and used to calculate the amount of each protein in Rho^-/- lysate. In these examples, GC-1 was to 10% of its WT content, GC-2 to 38%, and peripherin to 57%. (B) Expression levels of outer segment transmembrane proteins in Rho^-/- retinal lysates calculated as %WT. A minimum of four independent experiments was performed for each protein. Error bars represent SEM. DOI: http://dx.doi.org/10.7554/eLife.12058.005

Figure 3—figure supplement 1.. Transcript levels of GC-1 in the retinas of WT and Rho^-/- mice.

Quantitative RT-PCR of each transcript was performed on four mice of each genotype at P21. The relative mRNA expression level in each sample was normalized to the inner retina marker, Thy1. Error bars represent SEM. DOI: http://dx.doi.org/10.7554/eLife.12058.006

Figure 4.. Guanylate cyclase 1 (GC-1) rescue in Rho-/- rods requires the seven-helical core structure of rhodopsin.

Wild-type (WT) and Rho^-/- rods were transfected with (A) full-length Htr6, (B) seven-helical Htr6 core fused to rhodopsin’s C-terminus, (C) seven-helical rhodopsin core fused to the C-terminus of Htr6, (D) full-length rhodopsin. Sections from WT retinas were stained for each recombinant chimera using anti-green fluorescent protein (GFP), anti-myc, or anti-FLAG antibodies (magenta, each chimera’s tag is depicted in the construct diagram). Sections from Rho^-/- retinas were co-stained for GFP, myc, or FLAG (magenta, left panel) and endogenous GC-1 using anti-GC1 antibodies (green, middle panel). The merged images from Rho^-/- sections are shown in the right panel. Scale bar, 10 μm. Nuclei are stained by Hoechst (blue). DOI: http://dx.doi.org/10.7554/eLife.12058.007

Figure 5.. Guanylate cyclase 1 (GC-1) co-precipitation with rhodopsin from mouse retinal lysate.

(A) GC-1 and rhodopsin co-precipitation by monoclonal anti-rhodopsin antibody 1D4. Wild-type (WT) mouse retinal lysate (Input) was incubated with 1D4 antibody and then bound to protein A/G beads. After the unbound material in flow through (FT) was removed, the beads were washed and bound proteins were eluted (Eluate) and analyzed by Western blotting for GC-1, rhodopsin, DnaJB6, and peripherin. Non-specific protein binding was probed using either non-immune mouse IgG or 1D4 antibody treated with its epitope blocking peptide. (B) Co-precipitation of GC-1 and rhodopsin by the 1D4 antibody from retinal membranes solubilized under different detergent conditions. (C) Rhodopsin and GC-1 co-precipitation by monoclonal anti-GC-1 antibody 1S4. Rho^+/- mouse retinal lysate (Input) was incubated with 1S4 antibody bound to protein A/G beads. After the unbound material in flow through (FT) was removed, bound proteins were eluted from the beads (Eluate) and analyzed by Western blotting for GC-1 and rhodopsin. Non-specific rhodopsin binding was probed using non-immune mouse IgG. Protein loading on each lane was normalized to input in all panels. DOI: http://dx.doi.org/10.7554/eLife.12058.008