Identification of PKCα-dependent phosphoproteins in mouse retina

Abstract

Adjusting to a wide range of light intensities is an essential feature of retinal rod bipolar cell (RBC) function. While persuasive evidence suggests this modulation involves phosphorylation by protein kinase C-alpha (PKCα), the targets of PKCα phosphorylation in the retina have not been identified. PKCα activity and phosphorylation in RBCs was examined by immunofluorescence confocal microscopy using a conformation-specific PKCα antibody and antibodies to phosphorylated PKC motifs. PKCα activity was dependent on light and expression of TRPM1, and RBC dendrites were the primary sites of light-dependent phosphorylation. PKCα-dependent retinal phosphoproteins were identified using a phosphoproteomics approach to compare total protein and phosphopeptide abundance between phorbol ester-treated wild type and PKCα knockout (PKCα-KO) mouse retinas. Phosphopeptide mass spectrometry identified over 1100 phosphopeptides in mouse retina, with 12 displaying significantly greater phosphorylation in WT compared to PKCα-KO samples. The differentially phosphorylated proteins fall into the following functional groups: cytoskeleton/trafficking (4 proteins), ECM/adhesion (2 proteins), signaling (2 proteins), transcriptional regulation (3 proteins), and homeostasis/metabolism (1 protein). Two strongly differentially expressed phosphoproteins, BORG4 and TPBG, were localized to the synaptic layers of the retina, and may play a role in PKCα-dependent modulation of RBC physiology. Data are available via ProteomeXchange with identifier PXD012906. SIGNIFICANCE: Retinal rod bipolar cells (RBCs), the second-order neurons of the mammalian rod visual pathway, are able to modulate their sensitivity to remain functional across a wide range of light intensities, from starlight to daylight. Evidence suggests that this modulation requires the serine/threonine kinase, PKCα, though the specific mechanism by which PKCα modulates RBC physiology is unknown. This study examined PKCα phosophorylation patterns in mouse rod bipolar cells and then used a phosphoproteomics approach to identify PKCα-dependent phosphoproteins in the mouse retina. A small number of retinal proteins showed significant PKCα-dependent phosphorylation, including BORG4 and TPBG, suggesting a potential contribution to PKCα-dependent modulation of RBC physiology.

Keywords: BORG4; Protein kinase C-alpha; Quantitative phosphoproteomics; Retina; Rod bipolar cell; TPBG.

Fig. 1.

Phosphoserine labeling in the OPL is reduced in PKCα-KO retina. (A) Immunofluorescence confocal images of mouse retina sections from wild type and PKCα-KO retinas labeled with an antibody against phosphoserine residues within canonical PKC motif phosphoserine (PKC motif p-serine). (B) Images of PKC motif p-serine immunofluorescence in the outer plexiform layer of wild type and PKCα knockout (KO) retinas in obliquely cut sections. In both A and B, white arrows indicate labeling associated with presumptive cone pedicles. Scale bars: 20 μm. OPL: outer plexiform layer; IPL: inner plexiform layer.

Fig. 2.

PKCα is active in light-adapted RBCs. (A) Light- and dark-adapted mouse retina sections double-labeled with two antibodies against PKCα; conformation-specific PKCα-A (green) binds only active PKCα, while PKCα-B (magenta) binds to both active and inactive PKCα. (B) Sections from light-adapted and dark-adapted mouse retina labeled with an antibody mixture against PKC motif phosphoserines (PKC motif p-serine). (C) WT and TRPM1-KO mouse retina sections double-labeled with conformation-specific PKCα-A (green) and conformation non-specific PKCα-B (magenta). Scale bars: 10 μm. OPL: outer plexiform layer.

Fig. 3.

PMA increases phosphorylation by PKC isoforms in the mouse retina. (A) Wild type (WT) and PKCα knockout (PKCα-KO) retinas were incubated in PMA for 0, 15, 30, and 60 min followed by western blotting with an antibody against PKC motif phosphoserines. Arrows indicate candidate PKCα phosphorylation targets. (B) Immunofluorescent PKC motif phosphoserine labeling of wild type mouse OPL from retinas that were incubated with and without PMA for 1 h. Scale bar: 10 μm. PMA: phorbol 12-myristate 13-acetate; OPL: outer plexiform layer.

Fig. 4.

Experimental workflow of total protein and phosphopeptide identification. Wild type (n = 4) and PKCα knockout (KO) (n = 5) retinas were extracted and treated with PMA before lysis and trypsin digestion. A small fraction of each sample was removed for total protein analysis, while the rest of the samples underwent phosphopeptide enrichment. Following TMT labeling, samples were combined and analyzed by LCMS/MS. Tandem mass spectrometry data was collected on an Orbitrap Fusion and proteins were identified using Proteome Discoverer (SEQUEST and Percolator). Phosphorylation site localization was scored using phosphoRS, and reporter ion intensities were filtered and aggregated with an in-house Python script. TMT reporter ion intensities from total proteins or from phosphopeptides were tested for differential expression using the Bioconductor package edgeR. The presence of representative phosphoproteins was validated in the retina by western blot and confocal immunofluorescence microscopy. PMA: phorbol 12-myristate 13-acetate.

Fig. 5.

Identification of differentially expressed total proteins. (A) Scatter plot of peak reporter ion intensities from wild type (WT) and PKCα knockout (KO) total protein abundance samples. The dotted line corresponds to an FC (WT/KO) of 1, with FC > 1 corresponding to increased abundance in WT vs KO, and FC < 1 corresponding to increased abundance in KO vs WT. (B) Volcano plot of log₂ FC and −log₁₀ FDR. The dotted lines correspond to FC = 1 and FDR = 0.1. FDR < 0.1 corresponds to proteins passing the low significance threshold, and FDR > 0.1 corresponds to proteins failing the low significance threshold. The low significance (orange) threshold was 0.1, the med significance threshold (green) was 0.05, and the high significance threshold (blue) was 0.01. (C) Table of all differentially abundant proteins with an FDR lower than the low significance threshold of 0.1. Fold change was calculated by dividing the mean reporter ion intensities of each protein from the genotype with higher abundance by that with lower abundance. Colors correspond to DE FDR thresholds. DE: differential expression; FC: fold change; FDR: false discovery rate.

Fig. 6.

Identification of differentially expressed phosphopeptides. (A) Scatter plot of peak reporter ion intensities from wild type (WT) and PKCα knockout (KO) phosphopeptide abundance samples. The dotted line corresponds to an FC (WT / KO) of 1, with FC > 1 corresponding to increased abundance in WT vs KO, and FC < 1 corresponding to increased abundance in KO vs WT. (B) Volcano plot of log₂ FC and −log₁₀ FDR. The dotted lines correspond to FC = 1 and FDR = 0.1. FDR < 0.1 corresponds to proteins passing the low significance threshold, and FDR > 0.1 corresponds to proteins failing the low significance threshold. The low significance (orange) threshold was 0.1, the med significance threshold (green) was 0.05, and the high significance threshold (blue) was 0.01. (C) Table of all differentially abundance phosphopeptides with an FDR lower than the low significance threshold of 0.1. In the phosphopeptide sequence column, phosphorylated residues are in red. In the Total Protein Fold Change column, negative values indicate an increased abundance in the KO samples. Fold change was calculated by dividing the mean reporter ion intensities of each protein from the genotype with higher abundance by that with lower abundance. Fold change and DE FDR values were taken from the phosphopeptide abundance experiment and the total protein abundance experiment. Colors correspond to DE FDR thresholds. DE: differential expression; FC: fold change; FDR: false discovery rate.

Fig. 7.

TMT data from representative phosphoproteins. Reporter ion intensity values from each TMT channel for the five phosphopeptide fragments with the largest differential expression between WT (n = 4) and PKCα-KO (n = 5): two from BORG4 (orange), one from NHERF1 (green), and two from TPBG (blue). For statistical significance of differential expression analysis of mean WT (black) and mean KO (red) reporter ion intensities, see S1 – Total Protein and Phosphopeptide Abundance Analysis. (A) Two phosphopeptide fragments from an overlapping region of BORG4, each containing a phosphorylated serine corresponding to S64 on the full-length protein. (B) One phosphopeptide fragment from NHERF1 with a phosphorylated serine corresponding to S275 on the full-length protein. (C) Two phosphopeptide fragments from the C-terminal tail of TPBG with two similar phosphorylation patterns: one with two phosphoserines corresponding to S422 and S424, and one with a single phosphoserine corresponding to S424 of the full-length protein.

Fig. 8.

Significant total proteins and phosphoproteins grouped by biological function. Table of genes of identified total proteins (A) and phosphoproteins (B) with significant differential abundance between wild type and PKCα-KO samples grouped into broad categories based on general biological function gathered from Uniprot protein annotations.

Fig. 9.

Validation of representative phosphoproteins in the mouse retina. (A) Immunoblot of retinal lysate labeled with rabbit anti-BORG4 shows a band corresponding to BORG4 at ~38 kDa. (B) Confocal microscopy analysis of BORG4 (also called Cdc42EP4) and PKCα immunoreactivity in the retina using mouse anti-Cdc42EP2. (C) Immunoblot of retinal lysate shows a band corresponding to NHERF1 at ~50 kDa. (D) Confocal microscopy analysis of NHERF1 immunoreactivity in the retina with retinal layers labeled with DAPI. (E) Immunoblot of retinal lysate shows a smear corresponding to glycosylated TPBG at ~72 kDa. (F) Confocal microscopy analysis of TPBG and PKCα immunoreactivity in the retina. Scale bars: 20 μm. RPE: retinal pigment epithelium; OS: outer segments; IS: inner segments; ONL: outer nuclear layer; OPL: outer plexiform layer; INL: inner nuclear layer; IPL: inner plexiform layer; GCL: ganglion cell layer.