Two Small Molecules Restore Stability to a Subpopulation of the Cystic Fibrosis Transmembrane Conductance Regulator with the Predominant Disease-causing Mutation

Abstract

Cystic fibrosis (CF) is caused by mutations that disrupt the plasma membrane expression, stability, and function of the cystic fibrosis transmembrane conductance regulator (CFTR) Cl^- channel. Two small molecules, the CFTR corrector lumacaftor and the potentiator ivacaftor, are now used clinically to treat CF, although some studies suggest that they have counteracting effects on CFTR stability. Here, we investigated the impact of these compounds on the instability of F508del-CFTR, the most common CF mutation. To study individual CFTR Cl^- channels, we performed single-channel recording, whereas to assess entire CFTR populations, we used purified CFTR proteins and macroscopic CFTR Cl^- currents. At 37 °C, low temperature-rescued F508del-CFTR more rapidly lost function in cell-free membrane patches and showed altered channel gating and current flow through open channels. Compared with purified wild-type CFTR, the full-length F508del-CFTR was about 10 °C less thermostable. Lumacaftor partially stabilized purified full-length F508del-CFTR and slightly delayed deactivation of individual F508del-CFTR Cl^- channels. By contrast, ivacaftor further destabilized full-length F508del-CFTR and accelerated channel deactivation. Chronic (prolonged) co-incubation of F508del-CFTR-expressing cells with lumacaftor and ivacaftor deactivated macroscopic F508del-CFTR Cl^- currents. However, at the single-channel level, chronic co-incubation greatly increased F508del-CFTR channel activity and temporal stability in most, but not all, cell-free membrane patches. We conclude that chronic lumacaftor and ivacaftor co-treatment restores stability in a small subpopulation of F508del-CFTR Cl^- channels but that the majority remain destabilized. A fuller understanding of these effects and the characterization of the small F508del-CFTR subpopulation might be crucial for CF therapy development.

Keywords: ABC transporter; chloride channel; cystic fibrosis; cystic fibrosis transmembrane conductance regulator (CFTR); ivacaftor; lumacaftor; protein purification; protein stability.

FIGURE 1.

The single-channel activity of wild-type CFTR, F508del-CFTR, and G551D-CFTR. A, representative CFTR single-channel recordings in excised inside-out membrane patches from BHK cells expressing wild-type and F508del-CFTR and FRT cells expressing G551D-CFTR. In this and subsequent figures using the patch clamp technique, ATP (1 mm) and PKA (75 nm) were continuously present in the intracellular solution, temperature was 37 °C, voltage was −50 mV, and there was a large Cl⁻ concentration gradient across the membrane patch (internal [Cl⁻], 147 mm; external [Cl⁻], 10 mm). The dotted lines indicate where channels are closed, and downward deflections of the traces correspond to channel openings. B, representative current amplitude histograms of single CFTR Cl⁻ channels in excised membrane patches from cells expressing wild-type (WT) CFTR, F508del-CFTR, and G551D-CFTR after filtering digitally at 50 Hz. The continuous lines represent the fit of Gaussian distributions to the data. The closed (C) and open (O) channel amplitudes are indicated; S-CS, a subconductance state of F508del-CFTR. For F508del-CFTR, a small leak current shifted the closed channel amplitude to ∼0.4 pA. (Current amplitude is calculated from the difference between the closed and open current amplitudes determined from the fit of Gaussian distributions to the data). C and D, summary single-channel current amplitude (i) and P_o or P_o(app) data. Symbols, individual values; columns, means ± S.E. (error bars) (wild type, n = 10; F508del-CFTR, n = 21; G551D-CFTR, n = 15); *, p < 0.05 versus wild-type CFTR; #, p < 0.05 versus F508del-CFTR fully open state; ND, not determined.

FIGURE 2.

Impact of ivacaftor and lumacaftor on the temporal instability of F508del-CFTR in excised inside-out membrane patches. A, C, E, G, I, K, and M, time courses of P_o; B, D, F, H, J, L, and N, representative recordings of wild-type CFTR, F508del-CFTR, and G551D-CFTR in excised inside-out membrane patches made in the continuous presence of ATP (1 mm) and PKA (75 nm) at 37 °C once channel activation was complete. For wild-type CFTR and G551D-CFTR, membrane patches were excised, and channels were activated and studied, all at 37 °C. For F508del-CFTR, membrane patches were excised, and channels were activated and potentiated at 27 °C to delay temperature-dependent channel deactivation. Only after F508del-CFTR was fully activated and potentiated was temperature increased to 37 °C and temporal stability evaluated. For F508del-CFTR, plasma membrane expression was rescued by incubation at 27 °C for 48–72 h (E, F, I, and J) or by treatment with lumacaftor (VX-809; 3 μm) at 37 °C for 24 h (G, H, K, L, M, and N). F508del-CFTR Cl⁻ channels were either acutely treated with ivacaftor (aVX-770; 10 μm) (I–L) or chronically incubated with ivacaftor (cVX-770; 1 μm) together with lumacaftor (VX-809; 3 μm) at 37 °C for 24 h (M and N). In E, G, I, K, and M, the left and right ordinates show P_o (bars) and normalized P_o (circles), respectively; P_o values were normalized to that measured immediately when temperature reached 37 °C (t = 0–30 s); horizontal dotted lines indicate 50% normalized P_o. For wild-type and F508del-CFTR, P_o values were calculated for each 30-s interval, whereas for G551D-CFTR, P_o(app) values were calculated. Data are means ± S.E. (error bars) (wild-type CFTR, n = 8; G551D-CFTR, n = 8; F508del-CFTR, n = 10; F508del-CFTR, VX-809, n = 7; F508del-CFTR, aVX-770, n = 4; F508del-CFTR, VX-809 + aVX-770, n = 4; F508del-CFTR, VX-809 + cVX-770, n = 8). In B, D, F, H, J, L, and N, arrows denote the closed channel state, and downward deflections correspond to channel openings. In A, E, and I, some data were originally published in Wang et al. (32) (A, n = 4; E, n = 5; I, n = 4); other data are newly acquired.

FIGURE 3.

Stability of wild-type and F508del-CFTR in excised inside-out membrane patches at 37 °C. Shown are representative 9-min single-channel recordings (A and B) and corresponding P_o time courses (C) of wild-type and low temperature-rescued F508del-CFTR in excised inside-out membrane patches following full channel activation. ATP (1 mm) and PKA (75 nm) were continuously present in the intracellular solution; temperature was 37 °C. In A and B, four 2-s single-channel records labeled 1–4 indicated by bars are displayed on an expanded time scale below the 9-min recording to indicate channel activity at the beginning, in the middle, and at the end of the experiment. The arrows and dotted lines indicate where channels are closed, and downward deflections correspond to channel openings. In C, P_o values were calculated in 20-s intervals. For mean data, see Fig. 2, and for further information, see Figs. 1 and 2.

FIGURE 4.

Thermal stability of full-length CFTR protein in membranes and the effects of ivacaftor and lumacaftor. A, representative gel showing the thermal stability of G551D-CFTR in solubilized microsomes probed by the formation of SDS-resistant aggregates (top arrow) and the concomitant disappearance of the monomeric CFTR band (bottom arrow). The position of the 250 kDa molecular mass marker is shown. CFTR was detected by the fluorescence of tagged GFP, which is stable up to 80 °C under the conditions used. B–D, summary data showing the thermal stability of wild-type CFTR, F508del-CFTR, and G551D-CFTR, the effects of ivacaftor (VX-770; 2 μm) on the thermal stability of G551D-CFTR and lumacaftor (VX-809; 2 μm) on the thermal stability of F508del-CFTR. Data are means ± S.E. (error bars) (n = 3).

FIGURE 5.

F508del accelerates the thermal denaturation of purified CFTR protein. A, unfolding of purified wild-type CFTR protein by thermal and chemical denaturation detected by changes in CPM fluorescence. The continuous and dashed lines show purified wild-type CFTR protein in DDM-containing buffer in the absence and presence of 4 m guanidine HCl. The experiment was initiated by injecting CPM dye into a cuvette containing purified wild-type CFTR protein at 10 °C and monitoring CPM fluorescence for 30 min (period A). At the end of period A, the purified wild-type CFTR protein was heated to 78 °C over a 30-min interval (period B). Numbers indicate changes in CPM fluorescence as follows: 1, background fluorescence of CPM in buffer unbound to protein; 2, kinetics of CPM binding to solvent-exposed cysteine residues; 3, initial thermal quenching of CPM fluorescence as purified wild-type CFTR protein is heated; 4, increase of CPM fluorescence as purified wild-type CFTR protein unfolds, exposing more cysteine residues to solvent; and 5, continued thermal quenching of CPM fluorescence after complete unfolding of wild-type CFTR protein. B, silver-stained SDS-PAGE of purified protein samples for WT CFTR, F508del-CFTR, and G551D-CFTR. Wild-type CFTR and G551D-CFTR were purified by two chromatography steps, and F508del-CFTR was purified by one step. The arrow indicates the expected position of the main contaminating protein, ribosomal protein L3, which was removed by the second chromatography step. C–E, unfolding transitions of purified CFTR protein determined by changes in CPM fluorescence show the effects of F508del and G551D and the actions of ivacaftor (VX-770; 2 μm) and lumacaftor (VX-809; 2 μm). Samples of purified CFTR protein were heated at a rate of 1.2 °C/min. Data are means ± S.E. (error bars) (n = 3). F, concentration dependence of lumacaftor-induced stabilization of purified F508del-CFTR protein. Filled circles, data from individual experiments.

FIGURE 6.

Thermostability of purified WT CFTR and F508del-CFTR after reconstitution with lipid. A and B, assessment of the stability of purified wild-type CFTR and F508del-CFTR using the thermal gel assay and detection of the GFP reporter tag fluorescence. An aggregate band A appears at higher temperatures with the concomitant disappearance of the monomer band M and an intermediate band. The positions of the 250 and 130 kDa molecular mass markers are indicated by the yellow bars for WT CFTR; in all other panels, only the 250 kDa marker is indicated. C, evaluation of the stability of purified wild-type and F508del-CFTR using CPM fluorescence. Data are means ± S.E. (error bars) (n = 3); for clarity, only one-half of the error bars are shown.

FIGURE 7.

Instability of F508del-CFTR Cl⁻ currents in excised inside-out membrane patches. A and B, representative long duration recordings of F508del-CFTR Cl⁻ currents and channels in excised inside-out membrane patches following F508del-CFTR rescue by treatment of cells with lumacaftor (3 μm) and ivacaftor (1 μm) for 24 h at 37 °C. F508del-CFTR Cl⁻ channels were activated fully with ATP (1 mm) and PKA (75 nm) at 27 °C before the temperature was increased to 37 °C and the recordings commenced. ATP (1 mm) and PKA (75 nm) were continuously present in the intracellular solution for the duration of the recordings. Beneath each 20-min recording, the indicated 10-s period is shown on an expanded time scale. The dotted lines indicate where channels are closed, and downward deflections of the traces correspond to channel openings. In A, the continuous gray line shows the fit of a single exponential function to determine the time constant for the deactivation of F508del-CFTR Cl⁻ currents. Similar results were observed in other experiments (F508del-CFTR Cl⁻ currents, n = 3; F508del-CFTR Cl⁻ channels, n = 5).

FIGURE 8.

Deactivation of CFTR-mediated transepithelial Cl⁻ currents after acute or chronic ivacaftor treatment. Shown are representative Ussing chamber recordings of WT CFTR (A and B), lumacaftor-rescued (VX-809; 3 μm for 24 h at 37 °C) F508del-CFTR (E and F), and G551D-CFTR (I and J). FRT epithelia were incubated for 24 h at 37 °C in the absence (A, E, and I; Acute) or presence (B, F,and J; Chronic) of ivacaftor (VX-770; 1 μm); F508del-CFTR-expressing FRT epithelia were co-incubated with lumacaftor and ivacaftor. Fifteen minutes before t = 0 h, FRT epithelia were treated with cycloheximide (50 μg/ml), added to both the apical and basolateral solutions. At the indicated times, FRT epithelia were mounted in Ussing chambers, and CFTR Cl⁻ currents were activated with forskolin (Fk; 10 μm), potentiated with ivacaftor (VX-770 (770); 1 μm), and inhibited by CFTR_inh-172 (C172; 10 μm); continuous lines indicate the presence of different compounds in the apical solution; cycloheximide (50 μg/ml) was present in the apical and basolateral solutions during I_sc recordings. Data are normalized to baseline current so that ΔI_sc represents the change in transepithelial current after CFTR activation by forskolin. C, G, and K (Summary), magnitude of ivacaftor-potentiated ΔI_sc for WT CFTR, lumacaftor-rescued F508del-CFTR, and G551D-CFTR at different times after cycloheximide (50 μg/ml) treatment. Data are means ± S.E. (error bars) (n = 5–10); the continuous and dotted lines show the fits of linear and single exponential functions, respectively, to WT CFTR, F508del-CFTR, and G551D-CFTR Cl⁻ currents acutely and chronically treated with ivacaftor. G, as a control, the magnitude of lumacaftor-rescued F508del-CFTR Cl⁻ current chronically treated with ivacaftor (1 μm for 24 h at 37 °C) but untreated with cycloheximide is shown (0.1% (v/v) DMSO; CHX Ctrl). D, H, and L (Baseline), magnitude of absolute I_sc before CFTR activation for WT CFTR-, lumacaftor-rescued F508del-CFTR-, and G551D-CFTR-expressing FRT epithelia acutely or chronically treated with ivacaftor. Symbols, individual values; columns, means ± S.E. (n = 5–10).