Caveolae act as membrane reserves which limit mechanosensitive I(Cl,swell) channel activation during swelling in the rat ventricular myocyte

Abstract

Background: Many ion channels are preferentially located in caveolae where compartmentalisation/scaffolding with signal transduction components regulates their activity. Channels that are mechanosensitive may be additionally dependent on caveolar control of the mechanical state of the membrane. Here we test which mechanism underlies caveolar-regulation of the mechanosensitive I(Cl,swell) channel in the adult cardiac myocyte.

Methodology/principal findings: Rat ventricular myocytes were exposed to solution of 0.02 tonicity (T; until lysis), 0.64T for 10-15 min (swelling), and/or methyl-beta-cyclodextrin (MBCD; to disrupt caveolae). MBCD and 0.64T swelling reduced the number of caveolae visualised by electron microscopy by 75 and 50% respectively. MBCD stimulated translocation of caveolin 3 from caveolae-enriched buoyant membrane fractions, but both caveolin 1 and 3 remained in buoyant fractions after swelling. I(Cl,swell) inhibition in control cells decreased time to half-maximal volume (t(0.5,vol); 0.64T), consistent with a role for I(Cl,swell) in volume regulation. MBCD-treated cells showed reduced time to lysis (0.02T) and t(0.5,vol) (0.64T) compared with controls. The negative inotropic response to swelling (an index of I(Cl,swell) activation) was enhanced by MBCD.

Conclusions/significance: These data show that disrupting caveolae removes essential membrane reserves, which speeds swelling in hyposmotic conditions, and thereby promotes activation of I(Cl,swell). They illustrate a general principle whereby caveolae as a membrane reserve limit increases in membrane tension during stretch/swelling thereby restricting mechanosensitive channel activation.

Figure 1. Time-course of myocyte swelling in response to 0.64T hypotonic solution.

Myoycte volume was estimated from a video image of the cell, assuming the cell is an elliptical cylinder. Data are fitted with a logistic sigmoidal curve from which values of maximum cell volume and time to half-maximal volume were obtained. Mean±S.E.M. from n = 16 cells.

Figure 2. Hyposmotic swelling reduces the density of caveolae in the closed configuration.

Ventricular myocytes were exposed to isotonic or hypotonic solutions for 15 min. Representative electron micrographs of membrane sections from myocytes in isotonic (A.) and hypotonic (B.) conditions. Caveolae were recorded in both open (O) and closed (C) configurations (see Results). C. Mean data showing that exposure to hypotonic solution reduced the total number of caveolae through a reduction in the closed, but not open, state (* P<0.05 unpaired Student's t-test, n = 9–11 cells).

Figure 3. The cholesterol-depleting agent methyl β cyclodextrin (MBCD) reduces the density of open and closed caveolae.

Representative electron micrographs of membrane sections from control (A.) and MBCD-treated (B.) myocytes. C. Mean data showing that MBCD reduced the total number of caveolae through a reduction in both open and closed configurations (* P<0.05 unpaired Student's t-test; n = 11 cells).

Figure 4. Hyposmotic swelling does not cause translocation of Cav 1 or Cav 3 from caveolar membrane fractions.

Ventricular myocytes were exposed to isotonic or hypotonic solutions for 15 min, then subject to Na₂CO₃ extraction and discontinuous sucrose density gradient fractionation. A. Mean data showing Cav 1 in each fraction normalised to the sum of Cav 1 in all fractions for each sample (n = 6 hearts). Representative immunoblots, with equal volume loading of fractions, is shown below. B. Mean data showing Cav 3 in each fraction with representative immunoblots below. In both isotonic and hypotonic conditions, Cav 1 and 3 were enriched in the buoyant lipid raft fractions (5,6) of the myocyte. Hyposmotic swelling had no effect on the location of either Cav isoform.

Figure 5. Effect of caveolar disruption on maximum cell volume, surface area and time to lysis in 5 mOsM solution.

Quiescent ventricular myocytes were exposed to a 0.02T hypotonic solution and the maximum cell volume and surface area achieved before lysis was estimated from a video image of the cell. Disruption of caveolae with MBCD did not significantly alter the maximum cell volume (A.) or surface area (B.) prior to lysis. C. The time to lysis was significantly reduced in cells treated with MBCD (** P<0.01 versus control; Student's t-test; n = 11 cells).

Figure 6. Comparison of the effect of caveolar disruption and I _Cl,swell inhibition on myocyte volume regulation.

Values for maximum volume increase (A.) and time to half-maximum volume (B.) were obtained by fitting a sigmoidal relationship to the time-course of cell volume changes during 9 min exposure to 0.64T solution (see Figure 1). Inhibition of I _Cl,swell with 10 µM tamoxifen (Tmf) and/or disruption of caveolae with MBCD had no significant effect on the maximum volume achieved during swelling. By contrast, Tmf and MBCD both significantly reduced the time to half-maximal swelling, although the effects were not additive (** P<0.01 vs. control group (Con); ANOVA; n = 16–19 cells).

Figure 7. Effect of caveolar disruption and I _Cl,swell inhibition on the contractile response to hyposmotic swelling.

Field-stimulated myocytes were exposed to 0.64T hypotonic solution for 10 min and the change in contractility (expressed as a % of resting cell length) measured. The negative inotropic response recorded at 6 and 10 min in control cells was enhanced in cells in which caveolae were disrupted with MBCD, and reversed in cells in which I _Cl,swell was inhibited with 50 µM DIDS. * P<0.05 vs time-paired control (ANOVA; n = 19–21 cells).

Figure 8. A simple model summarising the effect of cholesterol depletion and/or swelling on caveolae configuration, Cav 3 distribution, membrane tension and consequences for I _Cl,swell activation.

A. Under control conditions, caveolae exist in open and closed configurations, Cav 3 is found predominantly in buoyant caveolae-containing membrane fractions (BF; fractions 5,6 of the sucrose gradient, see Fig. 4), and the I _Cl,swell channel is inactive. B. MBCD reduces the number of open and closed caveolae in association with translocation of Cav 3 from the BF to non-buoyant fractions (nBF; fractions 7–12), which contain heavy membranes and cytosolic proteins. The I _Cl,swell channel remains inactive. C. By contrast, swelling reduces the number of caveolae in the closed configuration only. We propose that this is through flattening of open caveolae in tandem with sarcolemmal incorporation of closed caveolae. No Cav 3 translocation from BF to nBF is seen during swelling. Membrane tension increases with swelling and this activates I _Cl,swell which abbreviates the action potential. D. In MBCD-treated cells, lack of available membrane reserves (open and closed caveolae) means that swelling increases membrane tension more than in control cells, causing enhanced I _Cl,swell activation, action potential shortening and negative inotropy. Cav 3-containing cholesterol-enriched membranes are shown in red: open caveolae (sarcolemmal red crescents); closed caveloae (intracellular red circles); flattened caveolae (sarcolemmal red lines). A representation of relative Cav 3 band density in BF and nBF, an index of membrane tension, and a representative action potential under each condition is shown to the right.