Loss of caveolin-3 induced by the dystrophy-associated P104L mutation impairs L-type calcium channel function in mouse skeletal muscle cells

Abstract

Caveolins are membrane scaffolding proteins that associate with and regulate a variety of signalling proteins, including ion channels. A deficiency in caveolin-3 (Cav-3), the major striated muscle isoform, is responsible for skeletal muscle disorders, such as limb-girdle muscular dystrophy 1C (LGMD 1C). The molecular mechanisms leading to the muscle wasting that characterizes this pathology are poorly understood. Here we show that a loss of Cav-3 induced by the expression of the LGMD 1C-associated mutant P104L (Cav-3(P104L)) provokes a reduction by half of the maximal conductance of the voltage-dependent L-type Ca(2+) channel in mouse primary cultured myotubes and fetal skeletal muscle fibres. Confocal immunomiscrocopy indicated a colocalization of Cav-3 and Ca(v)1.1, the pore-forming subunit of the L-type Ca(2+) channel, at the surface membrane and in the developing T-tubule network in control myotubes and fetal fibres. In myotubes expressing Cav-3(P104L), the loss of Cav-3 was accompanied by a 66% reduction in Ca(v)1.1 mean labelling intensity. Our results suggest that Cav-3 is involved in L-type Ca(2+) channel membrane function and localization in skeletal muscle cells and that an alteration of L-type Ca(2+) channels could be involved in the physiopathological mechanisms of caveolinopathies.

Figure 1. Cav-3^P104L–YFP expression leads to the loss of Cav-3 in primary cultured myotubes

A, in control myotubes, Cav-3 immunolabelling strongly outlines the surface plasma membrane. It also localizes intracellularly in small dot-shaped clusters (arrowheads) as best seen on the partial 2 × enlargement (inset). B, Cav-3^P104L–YFP accumulates preferentially around the nuclei in vesicular structures (top panel). Cav-3 immunolabelling is greatly decreased, if not totally absent, in the Cav-3^P104L–YFP-expressing myotube (arrow) whereas membrane staining is seen at the extremity of a nearby non-transfected myotube (bottom panel). Scale bars, 10 μm.

Figure 2. I_Ca,L is reduced in Cav-3^P104L-expressing primary cultured myotubes

A, representative currents elicited by applying 500 ms duration voltage pulses from a holding potential of −80 mV to the indicated values in the two types of cells. B, mean current–voltage relationships established for the fast transient current (I_Ca,T) and for the slow activating current (I_Ca,L) in 40 control and 21 Cav-3^P104L-expressing myotubes. C, mean current–voltage relationships of isolated I_Ca,L on the same control and mutant myotubes. The superimposed curves were calculated using the mean parameters obtained from fitting the voltage dependence in each fibre (see text for details).

Figure 3. Double immunolabelling reveals a strong colocalization of Cav-3 and Ca_v1.1 in primary cultured myotubes

Cav-3 (red) and Ca_v1.1 (green) strongly colocalize as clearly shown on the merged image. The colocalization is particularly pronounced at the surface plasma membrane, where both proteins accumulate (arrows). Additionally, the Ca_v1.1 labelling appears characteristically concentrated within internal punctate structures distributed almost uniformly throughout the cell and probably belonging to the developing T-tubule network. Cav-3 labelling appears generally weaker in this internal compartment except for specific areas, where it colocalizes with Ca_v1.1 (arrowheads). Scale bar, 10 μm.

Figure 4. Localization and labelling intensity of Ca_v1.1 are altered in Cav-3^P104L-expressing myotubes

A, in this representative Cav-3^P104L–YFP expressing myotube (green), Ca_v1.1 labelling (red) appears much dimmer than in the non-transfected neighbouring myotubes and is also partly redistributed to the perinuclear region where it colocalizes with Cav-3^P104L, as shown on the merged image (arrow). Scale bar, 10 μm. B, numerical image analysis of Ca_v1.1 labelling intensity. Field-specific analysis was performed on confocal Z-stacks by measuring Ca_v1.1 immunoreactivity of control myotubes and Cav-3^P104L-expressing myotubes present on the same image. Mean labelling intensity appeared to be reduced by 66% in Cav-3^P104L-expressing myotubes (P < 0.0001).

Figure 5. I_Ca,L is altered by Cav-3^P104L expression but is not affected by wild-type Cav-3 expression in fetal fibres

A, immunolocalization of Cav-3 in fetal fibres expressing Cav-3^P104L–YFP (top panel) and YFP-tagged wild-type Cav-3 (WT-Cav-3, bottom panel). YFP-tagged Cav-3^P104L preferentially accumulates around nuclei (*), as shown in this representative transfected fibre in which almost no endogenous Cav-3 is revealed by anti-Cav-3 antibody immunolabelling. Conversely, immunolabelling clearly revealed endogenous Cav-3 in the nearby non-transfected fibre. YFP-tagged wild-type Cav-3 appears concentrated at the surface plasma membrane as well as in discrete intracellular clusters, where it is also detected by immunolabelling. Scale bar, 20 μm, applies to all images. B, representative I_Ca,L recordings, elicited by applying 500 ms duration voltage pulses from a holding potential of −80 mV to a potential of +30 mV. C, left panel, mean current–voltage relationships established for the current in 31 control, eight Cav-3^P104L and 16 wild-type Cav-3-expressing fetal fibres. The superimposed curves were calculated using the mean parameters obtained from fitting the voltage dependence in each fibre (see text for details). The right panel presents the mean normalized conductance–voltage relationship in control (n = 31) and Cav-3^P104L-expressing (n = 8) fetal fibres.