Differential effects of myopathy-associated caveolin-3 mutants on growth factor signaling

Abstract

Caveolin-3 is an important scaffold protein of cholesterol-rich caveolae. Mutations of caveolin-3 cause hereditary myopathies that comprise remarkably different pathologies. Growth factor signaling plays an important role in muscle physiology; it is influenced by caveolins and cholesterol-rich rafts and might thus be affected by caveolin-3 dysfunction. Prompted by the observation of a marked chronic peripheral neuropathy in a patient suffering from rippling muscle disease due to the R26Q caveolin-3 mutation and because TrkA is expressed by neuronal cells and skeletal muscle fibers, we performed a detailed comparative study on the effect of pathogenic caveolin-3 mutants on the signaling and trafficking of the TrkA nerve growth factor receptor and, for comparison, of the epidermal growth factor receptor. We found that the R26Q mutant slightly and the P28L strongly reduced nerve growth factor signaling in TrkA-transfected cells. Surface biotinylation experiments revealed that the R26Q caveolin-3 mutation markedly reduced the internalization of TrkA, whereas the P28L did not. Moreover, P28L expression led to increased, whereas R26Q expression decreased, epidermal growth factor signaling. Taken together, we found differential effects of the R26Q and P28L caveolin-3 mutants on growth factor signaling. Our findings are of clinical interest because they might help explain the remarkable differences in the degree of muscle lesions caused by caveolin-3 mutations and also the co-occurrence of peripheral neuropathy in the R26Q caveolinopathy case presented.

Figure 1

Sural nerve and gastrocnemius muscle biopsy features are shown. A: Sural nerve: Severe reduction in large myelinated nerve fiber density. Several small groups of regenerating axons (black arrows) and few large axons with disproportionately thin myelin sheaths (white arrows). Semithin section, toluidine blue. Scale bar = 30 μm. B: Marked muscle fiber atrophy with numerous flat/angular or rounded atrophic fibers, clumps of pycnotic nuclei in completely atrophic fibers (white arrow), and muscle fiber hypertrophy. Black arrows: nonsubsarcolemmal nuclei. In histochemical ATPase preparations, prominent fiber type grouping was present (not shown). Cryostat section, hematoxilin-eosin stain. Scale bar = 60 μm. C: Caveolin-3 immunoreactivity is absent in muscle fibers of normal size and in hypertrophic fibers. However, there is peculiar, strong caveolin-3 immunostaining (red) of the sarcolemma and/or the sarcoplasm of atrophic fibers (black arrows). Cryostat section, hematoxilin counterstain. Scale bar = 50 μm. D: Higher magnification of the region indicated in C, demonstrating that the labeled structures are, in fact, atrophic skeletal muscle fibers. Scale bar = 30 μm. E: Immunoblot demonstrating absent caveolin-3 immunoreactivity in lane 1 loaded with patient muscle lysate. Lane 2 is loaded with normal human control muscle lysate.

Figure 2

Schematic view of caveolin-3 domains and mutants. The point mutations R26Q, P28L, and A45T are localized in the N-terminal domain of the protein, whereas the amino acid change G55S is located in the scaffolding region. The point mutation A45T lies in a stretch of eight amino acids (FEDVIAEP) highly conserved in all three isoforms of caveolin.

Figure 3

Expression of wild-type-caveolin-3 (wt-caveolin-3) and caveolin-3 mutants in HEK293 cells. HEK293 cells were transiently transfected with constructs encoding either the wild-type-caveolin-3 or the constructs carrying the mutations R26Q, P28L, A45T, G55S, or GFP as a mock-transfected negative control. The next day the cells were lysed and subjected to immunoblot analysis. Using a polyclonal antibody against caveolin-3, we evaluated the expression levels of the mutated proteins. The proteins carrying the mutations R26Q or P28L showed a surprising upward or downward mobility shift, respectively. IB, immunoblot.

Figure 4

Analysis of NIH3T3 cells cotransfected with wild-type-caveolin-3 (wt-caveolin-3) and caveolin-3 mutants. NIH3T3 cells were transiently transfected either with both wild-type-caveolin-3 constructs carrying an HA- or GFP-tag or with wild-type-caveolin-3-GFP or mutated HA-constructs carrying the mutations R26Q, P28L, A45T, or G55S. One day after transfection, the cells were fixed with 4% paraformaldehyde and immunostained with a monoclonal mouse antibody against the HA-tag. In case of cotransfection of the wild-type-caveolin-3-GFP (green) and the constructs carrying the mutations P28L or A45T, less wild-type-caveolin-3 protein localizes to the plasma membrane than in cells co-expressing wild-type-caveolin-3-GFP and R26Q or G55S. White arrows indicate the localization of the wild-type-caveolin-3 or mutant cav-3 proteins. Scale bar = 10 um.

Figure 5

Analysis of NGF induced TrkA phosphorylation after transfection of cav-3 mutants in PC12 cells. PC12 cells were transiently transfected either with the wild-type-caveolin03 (wt-caveolin-3) construct or the constructs carrying the mutations R26Q, P28L, or GFP as a mock-transfected negative control. One day after transfection, the cells were prestarved for three hours and then stimulated with 50 ng/ml of NGF for five minutes. Subsequently, the cells were lysed, and lysates were subjected to Western blot analysis by using antibodies against TrkA (anti-TrkA) or antibodies against the Y490 phosphorylation site within the activation loop of TrkA (anti-pTrkA). After transfection of the caveolin-3 construct carrying the mutation P28L, TrkA phosphorylation levels are decreased compared with cells transfected with the wild-type-construct or the construct carrying the mutation R26Q. Loading controls were performed by probing with antibodies directed against TrkA. IB, immunoblot.

Figure 6

Analysis of the phosphorylation levels of Akt and ERK1/2. PC12 cells were transiently transfected either with the wild-type-caveolin-3 (wt-caveolin-3) construct or the constructs carrying the mutations R26Q, P28L, or GFP as a mock-transfected negative control. One day after transfection, the cells were prestarved for three hours and then stimulated with 50 ng/ml of NGF for five minutes. Subsequently the cells were lysed and subjected to Western blot analysis. Use of a polyclonal antibody against phosphorylated (pAkt) and unphosphorylated Akt (Akt) or phosphorylated ERK1/2 (pERK1/2) and ERK1/2 (ERK1/2) revealed lower phosphorylation levels of Akt after transfection of the construct carrying the mutation P28L than after transfection of the wild-type-construct or the construct carrying the R26Q mutation. IB, immunoblot.

Figure 7

Surface biotinylation assay with PC12 cells transfected either with wild-type-caveolin-3 (wt-caveolin-3) or mutants. PC12 cells were transiently transfected either with the wild-type-caveolin-3 protein or the constructs carrying the mutation R26Q or P28L. Twenty-four hours later, the cells were prestarved for three hours and stimulated with 50 ng/ml of NGF. After stimulation, a surface biotinylation assay was performed to investigate if the caveolin-3 mutations influence the amount of surface TrkA levels. Western blot analysis of streptavidin precipitates revealed that higher levels of TrkA are present at the plasma membrane of stimulated PC12 cells expressing caveolin-3 R26Q than in cells expressing P28L. We also noticed that expression of wild-type-caveolin-3 increases the surface levels of TrkA compared with mock transfected cells (GFP). SB, surface biotinylation; SL, straight lysates; IB, immunoblot.

Figure 8

Analysis of the phosphorylation levels of EGFR after cotransfection of EGFR and caveolin-3 constructs in HEK293 cells. HEK293 cells were transiently cotransfected with the wild-type-caveolin-3 (wt-caveolin-3) construct and an EGFR construct, or EGFR and the constructs carrying the point mutations R26Q, P28L, or GFP as a negative control (GFP). Five hours after transfection, cells were prestarved overnight and then stimulated with 50 ng/ml EGF for ten minutes. Immunoblotting of same amounts of protein lysates by using antibodies against tyrosine-phosphorylated EGFR (pEGFR) revealed that transfection of R26Q slightly decreased, whereas P28L clearly increased the levels of pEGFR. As a loading control, we probed with anti-EGFR (lower panel).

Figure 9

Analysis of the phosphorylation levels of ERK1/2 and Akt after cotransfection of EGFR and caveolin-3 constructs in HEK293 cells. HEK293 cells were transiently cotransfected with the wild-type-caveolin-3 construct and an EGFR construct, or EGFR and the constructs carrying the point mutations R26Q, P28L, or GFP as a negative control. Five hours after transfection, cells were prestarved overnight and then stimulated with 50 ng/ml EGF for ten minutes. Immunoblotting of same amounts of protein lysates by using antibodies against phosphorylated ERK1/2 (pERK1/2) and phosphorylated Akt (pAkt) revealed that transfection of both R26Q and P28L did not lead to major alterations. IB, immunoblot.

Figure 10

Surface biotinylation assay of HEK293 cells cotransfected with EGFR and either wild-type-caveolin-3 (wt-caveolin-3) or the mutants. HEK293 cells were transiently transfected with the EGFR-construct and either with the wild-type-caveolin-3 protein or the constructs carrying the mutation R26Q or P28L (or GFP as a mock-transfected negative control). The transfected cells were prestarved overnight and stimulated with 100 ng/ml EGF for ten minutes. After stimulation, a surface biotinylation assay was performed to investigate if the mutations in caveolin-3 influence the amount of surface EGFR levels. Western blot analysis revealed that slightly lower levels of EGFR are present at the plasma membrane of EGF-stimulated HEK293 cells expressing caveolin-3 R26Q than in cells expressing the wild-type-construct. SB, surface biotinylation; SL, straight lysate; IB, immunoblot.