Skip to main content
NCBI home page
Molecular Pharmacology logoLink to Molecular Pharmacology
. 2016 Sep;90(3):275–285. doi: 10.1124/mol.116.104570

Synergistic Potentiation of Cystic Fibrosis Transmembrane Conductance Regulator Gating by Two Chemically Distinct Potentiators, Ivacaftor (VX-770) and 5-Nitro-2-(3-Phenylpropylamino) Benzoate

Wen-Ying Lin 1, Yoshiro Sohma 1, Tzyh-Chang Hwang 1,✉
PMCID: PMC4998663  PMID: 27413118

Abstract

Cystic fibrosis (CF) is caused by loss-of-function mutations of the cystic fibrosis transmembrane conductance regulator (CFTR) gene encoding a phosphorylation-activated but ATP-gated chloride channel. Previous studies suggested that VX-770 [ivacaftor, N-(2,4-di-tert-butyl-5-hydroxyphenyl)-4-oxo-1,4-dihydroquinoline-3-carboxamide], a CFTR potentiator now used in clinics, increases the open probability of CFTR by shifting the gating conformational changes to favor the open channel configuration. Recently the chloride channel blocker and CFTR potentiator 5-nitro-2-(3-phenylpropylamino) benzoate (NPPB) has been reported to enhance CFTR activity by a mechanism that exploits the ATP hydrolysis-driven, nonequilibrium gating mechanism unique to CFTR. Surprisingly however, NPPB increased the activity of nonhydrolytic G551D-CFTR, the third most common disease-associated mutation. Here, we further investigated the mechanism of NPPB’s effects on CFTR gating by assessing its interaction with well-studied VX-770. Interestingly, once G551D-CFTR was maximally potentiated by VX-770, NPPB further increased its activity. However, quantitative analysis of this drug–drug interaction suggests that this pharmacologic synergism is not due to independent actions of NPPB and VX-770 on CFTR gating; instead, our data support a dependent mechanism involving two distinct binding sites. This latter idea is further supported by the observation that the locked-open time of a hydrolysis-deficient mutant K1250A was shortened by NPPB but prolonged by VX-770. In addition, the effectiveness of NPPB, but not of VX-770, was greatly diminished in a mutant whose second nucleotide-binding domain was completely removed. Interpreting these results under the framework of current understanding of CFTR gating not only reveals insights into the mechanism of action for different CFTR potentiators but also brings us one step forward to a more complete schematic for CFTR gating.

graphic file with name mol.116.104570absf1.jpg

Open in a new tab

Introduction

Cystic fibrosis (CF), caused by loss-of-function mutations of the CFTR (cystic fibrosis transmembrane conductance regulator) gene, is one of the most common life-shortening genetic diseases, affecting 1 in every 2500 newborns in Caucasian populations (Zielenski and Tsui, 1995; Rowe et al., 2005). The CFTR protein, functioning as a chloride channel that plays critical roles in water and salt transport in epithelium-lining tissues (Riordan et al., 1989; Quinton and Reddy, 1991; Bear et al., 1992), is classified as an ATP-binding cassette protein because of its characteristic topologic features: two cytosolic nucleotide-binding domains (NBDs) each preceded by a transmembrane domain (TMD). Unique to CFTR, however, is a regulatory (R) domain that connects two TMD/NBD complexes and contains multiple consensus serine/threonine residues for phosphorylation dependent on protein kinase A (PKA) (Ostedgaard et al., 2001).

Once the regulatory domain is phosphorylated, the opening and closing (gating) of the CFTR chloride channel are controlled, respectively, by ATP binding-induced dimerization and hydrolysis-catalyzed separation of the NBDs. (Vergani et al., 2003, 2005; Hwang and Sheppard, 2009; Tsai et al., 2009, 2010). A CFTR gating model (Fig. 1A) depicting energetic coupling between dimerization of CFTR’s two NBDs and gate opening/closing in the TMDs was recently proposed (Jih and Hwang, 2012). In theory, mutations that disrupt any critical steps in this gating process may lead to CFTR dysfunction and hence cause CF. For example, the G551D mutation results in >100-fold reduction of the open probability (Po) (Cai et al., 2006; Bompadre et al., 2007; Miki et al., 2010) due to a prevention of ATP-induced NBD dimerization (Lin et al., 2014), two rightmost horizontal transitions in Fig. 1A. A recent study suggests defects in the conformational changes of CFTR’s TMDs (three vertical transitions in Fig. 1A) may account for the much-reduced Po observed in the R117H mutation associated with a mild form of CF (Yu et al., 2016).

Fig. 1.

Fig. 1.

Open in a new tab

Dual effects of NPPB on CFTR. (A) The energetic coupling model (Jih and Hwang, 2012) for WT-CFTR gating illustrating the relationship between opening/closing of the gate (three vertical transitions), ATP binding/unbinding (two horizontal transitions on the left), NBD dimerization/dissociation (two horizontal transitions on the right), and ATP hydrolysis and dissociation of hydrolytic products (oblique transitions below). The yellow trapezoids represent CFTR’s two TMDs, which craft a gated pore. The red and blue semicircles depict CFTR’s NBD1 and NBD2, respectively. They jointly form the ATP binding site 1 (lower binding pocket), which consists of the head subdomain of NBD1 and the tail subdomain of NBD2, and catalysis-competent site 2 (upper binding pocket), composed of the head subdomain of NBD2 and the tail subdomain of NBD1. The different lengths of the arrows marking the transitions represent relatively different rate constants. The top three states represent the closed states, whereas the lower four states are open channel conformations. (B) Effect of NPPB on macroscopic WT-CFTR currents. In an inside-out patch containing multiple WT-CFTR channels, the application of 200 µM NPPB results in a net current reduction that is readily reversed upon removal of NPPB. (C) Voltage-dependent block of E1371S-CFTR by NPPB. Ramped I-V curves over ± 100 mV membrane potential show a linear I-V relationship (red) in the absence of NPPB and an outwardly rectified I-V curve (blue) in the presence of 200 µM NPPB. (D) Effect of NPPB on G551D-CFTR, a mutant with defective ATP-induced NBD dimerization. Macroscopic G551D-CFTR currents increase immediately upon applying NPPB. The inset highlights a biphasic change of the currents upon removal of NPPB, a telltale sign for a fast release of the pore block followed by a slower abolition of the potentiation effect of NPPB. Fitting the decay phase in this representative trace with a single exponential function (red line) yields a time constant of 1.55 seconds (mean value of 1.67 ± 0.28, n = 6). Inh-172 (4-[[4-oxo-2-sulfanylidene-3-[3-(trifluoromethyl)phenyl]-1,3-thiazolidin-5-ylidene]methyl]benzoic acid) is a thiazolidinone CFTR inhibitor (Kopeikin et al., 2010; Ma et al., 2002); 10 µM Inh-172 is used here and throughout the other figures.

High-throughput screening (Van Goor et al., 2009) identified a high-affinity CFTR potentiator, VX-770 (ivacaftor or Kalydeco, N-(2,4-di-tert-butyl-5-hydroxyphenyl)-4-oxo-1,4-dihydroquinoline-3-carboxamide), that increases the Po of G551D by 8- to 10-fold in vitro. Subsequent successful clinical trials (Accurso et al., 2010; Ramsey et al., 2011) led to U.S. Food and Drug Administration approval for the first CF therapy targeting mutant CFTR. However, how VX-770 potentiates a wide spectrum of gating mutants (Yu et al., 2012; Van Goor et al., 2014) remained unknown until recent reports revealed the mechanisms for VX-770’s effects on gating kinetics and energetics (Eckford et al., 2012; Jih and Hwang, 2013; Yeh et al., 2015). These latest studies support the idea that VX-770 acts at the interface between CFTR’s TMDs and the lipid core of the membrane bilayer, and increases the Po by shifting the three vertical gating transitions depicted in Fig. 1A to favor the open states.

Of note, these detailed mechanistic studies provide a theoretical framework to help design second-generation drugs that can complement the action of VX-770. This strategy is important because, however encouraging, the VX-770–rectified Po of G551D channels is still less than 10% of that of wild-type (WT) CFTR (Jih and Hwang, 2013; but compare Van Goor et al., 2009). In addition, VX-770 diminishes the effects of VX-809 (3-[6-[[1-(2,2-difluoro-1,3-benzodioxol-5-yl)cyclopropanecarbonyl]amino]-3-methylpyridin-2-yl]benzoic acid) (Cholon et al., 2014; Veit et al., 2014), a compound designed to correct the trafficking defects of the most common pathogenic mutation ∆F508 (Van Goor et al., 2011). Although this negative impact of VX-770 on the action of VX-809 may not be as severe when the VX-770 concentration is in the nanomolar range (Matthes et al., 2016), new CFTR potentiators with few drug–drug interactions seem warranted.

Among the plethora of CFTR potentiators (Rowe and Verkman, 2013; Barry et al., 2015) that may serve as candidates to complement or supplant VX-770, the arylaminobenzoate 5-nitro-2-(3-phenylpropylamino) benzoate or 5-nitro-2-(3-phenylpropylamino) benzoate (NPPB) (Wang et al., 2005), which has long been known to the anion transport community as a chloride channel blocker (Wangemann et al., 1986), is particularly appealing. The action of NPPB is distinct from that of many other dual-action potentiators that increase the activity of CFTR at low concentrations but inhibit CFTR channel gating at high concentrations (Wang et al., 1998; Lansdell et al., 2000), as NPPB, when used at micromolar concentrations, potentiates CFTR channel gating as well as blocking the pore.

In our current study, we show that although both NPPB and VX-770 effectively increase the Po of WT and G551D CFTR to a similar extent, their effects differ markedly in other CFTR mutants tested. To investigate how NPPB and VX-770 work differently on CFTR gating, we quantify the combined effects of NPPB and VX-770, and compare them to the individual action of each compound. Our data suggest that these two reagents work not only through different binding sites, but also at different kinetic steps. The pharmacologic and physiologic implications of our results will be discussed.

Materials and Methods

Cell Culture and Transient Expression System.

We used Chinese hamster ovary cells for all our patch-clamp experiments. Cells were grown at 37°C in Dulbecco’s modified Eagle’s medium and supplemented with 10% fetal bovine serum. We first transfected cells with pcDNA plasmids containing various CFTR constructs and pEGFP-C3 (Takara Bio, Otsu, Japan) by using PolyFect (QIAGEN, Valencia, CA). The transfected Chinese hamster ovary cells were then trypsinized, transferred to new 35-mm tissue dishes containing sterile glass chips, and incubated at 27°C before electrophysiologic experiments done 2 to 3 days after transfection for microscopic current recordings or 4 to 6 days for macroscopic current recordings.

Mutagenesis.

For site-directed mutagenesis, QuikChange XL kit (Agilent, Santa Clara, CA) was used according to the manufacturer’s protocols. All the DNA constructs were sequenced by the DNA Core (University of Missouri) to confirm mutation identity.

Electrophysiologic Recordings.

We performed all the electrophysiologic experiments at room temperature using an EPC9 amplifier (HEKA Instruments, Holliston, MA). Recording electrodes were made from borosilicate capillary glass with a two-step micropipette puller (Narishige, Tokyo, Japan). The tips of micropipettes were then polished using a homemade microforge before experiments. The resistance of polished pipettes was 1.5–3.0 MΩ for macroscopic current recordings and 4–5 MΩ for microscopic current recordings.

After observing the seal resistance >40 GΩ, we excised membrane patches into an inside-out mode and applied 2 mM ATP with 25 IU PKA to activate CFTR until the current reached a steady state. The time course of this phosphorylation-dependent activation varies from patch to patch, but usually requires >5 minutes to reach the maximal level. To minimize effects of dephosphorylation by membrane-associated phosphatases and to assess the degree of phosphorylation-independent rundown, 10 IU PKA was added in all other ATP-containing solutions applied thereafter.

For the macroscopic current recording, the membrane potential was kept at −30 mV, whereas single-channel recordings were made at −50 mV. To obtain macroscopic I-V relationships, voltage ramps (−100 to +100 mV over a duration of 1 second) were applied. The data were filtered with an eight-pole Bessel filter (LPF-8; Warner Instruments) with a 100-Hz cutoff frequency and digitized to a computer at a sampling rate of 500 Hz. For a better visual effect, we invert the inward current so that upward deflections represent channel openings. For recordings that demand fast solution changes, a perfusion system (SF-77B; Warner Instruments, Hamden, CT) with a dead time of ∼30 milliseconds was used.

Chemicals and Solution Composition.

The pipette solution contained (mM) 140 NMDG (N-methyl-d-glucamine)-Cl (Fisher Biotec), 2 MgCl2 (Fisher Biotec, Perth, Australia), 5 CaCl2 (Fisher Biotec), and 10 HEPES (Fisher Biotec), pH 7.4, with NMDG for all of the patch-clamp experiments. This pipette solution was used routinely for all the experiments except those shown in Supplemental Fig. 3, where the pipettes were filled with NPPB-containing pipette solution. Cells were perfused with a bath solution having (mM) 145 NaCl (Fisher Biotec), 5 KCl (Fisher Biotec), 2 MgCl2, 1 CaCl2, 5 glucose (Fisher Biotec), 5 HEPES, and 20 sucrose (Fisher Biotec), pH 7.4, with NaOH (Fisher Biotec). After establishing an inside-out configuration, the patch was perfused with a standard perfusion solution (i.e., intracellular solution) containing (mM) 150 NMDG-Cl, 2 MgCl2, 10 EGTA (Fisher Biotec), and 8 Tris (Fisher Biotec), pH 7.4, with NMDG.

MgATP and PKA were purchased from Sigma-Aldrich (St. Louis, MO). MgATP was stored in 500 mM stock solutions at −20°C prepared to working concentration, and the pH was adjusted to 7.4 with NMDG. VX-770 [N-(2,4-di-tert-butyl-5-hydroxyphenyl)-4-oxo-1,4-dihydroquinoline-3-carboxamide, ivacaftor], provided by Prof. Robert Bridges (Rosalind Franklin University, North Chicago, IL), was stored as a 100 µM stock in dimethylsulfoxide at −70°C and diluted to 200 nM. NPPB (5-nitro-2-(3-phenylpropylamino)benzoate), purchased from Sigma-Aldrich, was stored as a 100 mM stock in dimethylsulfoxide at −20°C and diluted to 200 µM.

Data Analysis and Statistics.

We used Igor Pro program (WaveMetrics, Portland, OR) to measure the steady-state mean current amplitude. The current relaxation phase after NPPB removal was fitted with single exponential functions in G551D-, WT-, and ΔNBD2-CFTR using a built-in Levenberg–Marquardt-based algorithm. Two-tailed t tests assuming equal variance were conducted with Excel (Microsoft, Redmond, WA) for all the comparisons showing statistical probability (P was expressed as *P < 0.05, **P < 0.01, ***P < 0.001 in figures). P < 0.05 was considered statistically significant. Single-channel kinetic analysis for WT-CFTR was done with a program developed by Csanády (2000) (see Fig. 6).

Fig. 6.

Fig. 6.

Open in a new tab

A modified gating model of CFTR. This gating scheme modified from Jih and Hwang (2012) incorporates the mechanism for ATP-dependent inhibition of G551D gating in our previous study (Lin et al., 2014) as well as data in the current report. Two states, O0 and C0AD, with NBD-dimerized but a vacant ATP binding site 2, are added to accommodate the mechanism of NPPB in G551D and to explain the biphasic current decay after ATP washout in G551D in previous study (Lin et al., 2014). These two newly added states are simple mirror images of O1 and C2 AD. The idea of energetic coupling between NBD dimerization and gate opening predicts that O1 and O0 are two stable open states.

NPPB’s effect on gating was isolated by first quantifying its pure blocking effect using a hydrolysis-deficient mutant, E1371S-CFTR (Moody et al., 2002; Vergani et al., 2003), whose Po is close to unity whereas the single-channel conductance is unchanged suggesting that the pore was not affected by the mutation (Vergani et al., 2003). A similar strategy was employed in Csanády and Töröcsik (2014). Ramped I-V relationships of macroscopic E1371S-CFTR currents in the presence or absence of 200 µM NPPB were obtained (see Fig. 1C).

Fractional block of 73% was estimated for currents at −30 mV membrane potential, at which voltage all the experiments with NPPB were performed. This value was then used to calculate the net gating effect of NPPB with the following mathematic derivation:

graphic file with name mol.116.104570uneq1.jpg

N: number of active channels in the patch, Po (NPPB): Po in the presence of NPPB, iNPPB: single-channel amplitude in the presence of NPPB.

graphic file with name mol.116.104570uneq2.jpg

N: number of active channels in the patch, Po (control): Po in the absence of NPPB, icontrol: single-channel amplitude in the absence of NPPB.Then, INPPB / Icontrol = Po(NPPB) / Po(control) × iNPPB / icontrol, assuming the number of active channels does not change over the experimental duration.

Because iNPPB / icontrol = 0.27 based on 73% fractional block described previously, assuming this fractional block is the same among all CFTR variants tested,

graphic file with name mol.116.104570uneq3.jpg

Because our experimental protocol is always designed to include a control before and after NPPB treatment, our final control current amplitude will be the average of these two measurements:

graphic file with name mol.116.104570uneq4.jpg

where INPPB represents macroscopic currents in the presence of NPPB, and Icontrol before and Icontrol after are, respectively, macroscopic current before the application of NPPB and that after washing out of NPPB. By averaging Icontrol before and Icontrol after for the control, we effectively minimized the effect of time-dependent changes of the current amplitude.

Of note, the formula employed here to calculate the fold increase of the Po is based on two assumptions. First, the number of active channels in the membrane patch remains unchanged over the experimental duration. Second, the blocking effect of NPPB is the same among all CFTR variants tested. Although the first assumption is a common practice in patch-clamp experiments, the second one is presumably also safe as all the CFTR constructs used in the current study exhibit the same single-channel conductance, suggesting an unperturbed pore where NPPB may reside.

For experiments involving VX-770, as noted elsewhere (Jih and Hwang 2013), the “stickiness” of VX-770 makes it impossible to bracket our experiments with controls. Thus, the fold increase of the current was calculated based on the one control immediately before the application of VX-770.

Computer Simulations.

We developed an in-house computer program for simulating the drug effects on the gating of G551D-CFTR based on our proposed gating scheme (Supplemental Fig. 4) using Intel Visual Fortran Compiler on the Windows 7 operation system (Cai et al., 2011) (see the supplements for details).

Results

NPPB, best known as a chloride channel blocker (Wangemann et al., 1986), can be driven into the CFTR pore by hyperpolarizing voltages, hence disrupting the flow of chloride ions (Zhang et al., 2000; Zhou et al., 2010; Csanády and Töröcsik, 2014). Thus, when NPPB was applied as a CFTR potentiator to an inside-patch that yields macroscopic currents of WT-CFTR at a negative membrane potential, a net decrease of the current was observed because the potentiation of CFTR gating is effectively masked by pore blockade (Fig. 1B).

To quantitatively assess the gating effects of NPPB, one has to first quantify its blocking effect and use this information to isolate NPPB’s effects on gating. A simple way to measure pure block of the CFTR pore by NPPB is to apply the reagent to a channel with a Po close to unity. The hydrolysis-deficient mutant E1371S-CFTR (Vergani et al., 2003) offers just that, as demonstrated recently (Csanády and Töröcsik, 2014). Figure 1C shows ramped I-V curves of macroscopic E1371S-CFTR currents in the presence or absence of 200 µM NPPB.

As reported elsewhere (Csanády and Töröcsik, 2014), an NPPB-induced block exhibits a strong voltage-dependence with more block at hyperpolarizing membrane potentials. From these I-V relationships, we calculated the percentages of blockade at given membrane potentials and used these values to assess the net gating effect of NPPB, as described in the Materials and Methods. Once the magnitude of the block was corrected for the data shown in Fig. 1B, we deduced that NPPB increases the Po of WT-CFTR by 1.81 ± 0.14-fold (n = 8).

We next tested effects of NPPB on G551D-CFTR (Fig. 1D). In contrast to the result with WT channels (Fig. 1B), 200 µM NPPB increased the net G551D-CFTR currents, indicating that the gating effects outweighed the block for this mutant channel. Nonetheless, the existence of the block of G551D-CFTR currents by NPPB was evidenced by the transient current increase upon sudden washout of NPPB (inset in Fig. 1D). The biphasic time course of current changes upon removal of NPPB was interpreted as a fast release of NPPB’s blocking effect followed by a slow disappearance of its gating effects. Fitting the falling phase with a single exponential function yielded a time constant of 1.67 ± 0.28 seconds (n = 6). The lack of this transient increase of currents in WT-CFTR might be due to a high Po at 2 mM ATP such that the smaller transient was masked by the low bandwidth of recording.

Again, once the NPPB-induced block was corrected, we observed an increase of the Po by 12.1 ± 1-fold (n = 20) for G551D-CFTR. These results intrigued us not only because of the differences in fold increase between WT- and G551D-CFTR but also because NPPB could work on G551D-CFTR at all, as previous studies (Csanády and Töröcsik, 2014) suggested a mechanism for NPPB’s action that should not be applicable for this mutant whose gating does not involve ATP hydrolysis (Lewis et al., 2004; Lin et al., 2014; Xu et al., 2014).

To further understand how NPPB modulates G551D-CFTR gating, we first reviewed briefly our previous gating scheme for G551D-CFTR (Lin et al., 2014). As the lack of a side chain at position 551 in WT-CFTR exposes the peptide backbone amide to form a hydrogen bond with the γ-phosphate of ATP upon NBD dimerization, we proposed that with a negative charge introduced at this position, NBD dimerization is likely prohibited as a result of simple electrostatic repulsion between two negatively charged entities. Of note, this idea is supported by the observation that G551E, but not G551K, exhibits a similar gating abnormality as G551D (Lin et al., 2014). We therefore modified the gating scheme (Fig. 1A) originally proposed for WT-CFTR (Jih et al., 2012a) by eliminating NBD-dimerized states to explain G551D-induced gating defects (Fig. 2A).

Fig. 2.

Fig. 2.

Open in a new tab

Effect of ATP occupancy at site 2 of NBDs on NPPB’s potentiation in G551D-CFTR. (A) A modified gating scheme for G551D-CFTR. Two NBD-dimerized states from Fig. 1A are eliminated based on our previous report (Lin et al., 2014). (B) Potentiation of G551D-CFTR currents by NPPB in the presence of 20 µM ATP. Note an increase of the G551D-CFTR currents upon changing [ATP] from 2 mM to 20 µM. By lowering [ATP], we shift the channel to states with a vacant site 2 (i.e., C2 and O2 in A). (C) Effect of NPPB in the presence of 20 µM ATP on G551D/Y1219G-CFTR, a double mutant whose ATP affinity at site 2 is drastically reduced (Zhou et al., 2006). (D) Summary of increases in Po by NPPB under conditions marked. Fold increase in Po: 12.1 ± 1-fold (n = 20) for G551D-CFTR in the presence of 2 mM ATP, 7.31 ± 0.81-fold (n = 3, *P < 0.05) in the presence of 20 µM ATP, and 6.83 ± 1.53-fold (n = 8, **P < 0.01) in the presence of 20 µM ATP for G551D/Y1219G.

The resulting scheme contains two open and two closed states, “monoliganded” and “biliganded,” in equilibrium. Unlike classic ligand-gated channels, however, binding of the ligand ATP by itself does not alter gating parameters of CFTR (Vergani et al., 2005). Thus, once the NBD dimerization is removed from the scheme, gating with only ATP binding/unbinding is inefficient as happened in G551D-CFTR. Nonetheless, this four-state gating scheme predicts that altering [ATP] can change the distribution of the states.

As our previous study (Lin et al., 2014) had suggested that 20 µM ATP can effectively decrease ATP binding to site 2 in G551D-CFTR, we examined the effects of 200 µM NPPB in the presence of 20 µM ATP (Fig. 2B). Again, NPPB increased the net current, resulting in an increase of the Po by 7.31 ± 0.81-fold (n = 3), which was significantly lower than that at 2 mM ATP (P < 0.005)(Fig. 2D).

To further lower the probability of occupancy in site 2, we mutated Y1219, a residue involved directly in binding of ATP at the head subdomain of NBD2 (Zhou et al., 2006), to glycine (Y1219G). We then applied 200 µM NPPB to G551D/Y1219G-CFTR in the presence of 20 µM ATP (Fig. 2C). The calculated increase of the Po turned out to be 6.83 ± 1.53-fold (n = 8), very similar to the magnitude seen with G551D-CFTR at 20 µM ATP (Fig. 2D). If we accept the premise that our strategies have effectively eliminated the states with ATP bound in site 2, this result leads to the propositions that NPPB potentiates C2 ↔ O2 transition and that NPPB works on this transition not as well as it affects the C2 ATP ↔ O2 ATP transition. Regardless, this idea dictates that NPPB affects the same kinetic steps potentiated by VX-770, a well-studied CFTR potentiator (Eckford et al., 2012; Jih and Hwang, 2013; Yeh et al., 2015).

If the hypothesis that NPPB and VX-770 act on the same kinetic step is correct, we can consider three possible scenarios. First, they bind to the same site (i.e., a competitive relationship). Second, they bind to different sites and work independently, a scenario demonstrated for nitrate and VX-770 (Yeh et al., 2015). Third, although they bind to two different sites, the effect of one drug may influence that of the other (i.e., a dependent action or drug–drug interactions).

For the first scenario, when both drugs are used at saturating concentrations in our experiments, we would predict a lack of additive effects in gating potentiation upon applications of NPPB and VX-770 together. However, the experimental result contradicted this prediction (Fig. 3, A and B). The current of G551D-CFTR was further enhanced as we added VX-770 to the bath in the presence of NPPB and vice versa.

Fig. 3.

Fig. 3.

Open in a new tab

Synergistic effects of NPPB and VX-770 on G551D-CFTR. (A) A continuous recording of G551D-CFTR currents shows that the application of VX-770 further enhances G551D-CFTR current in the presence of NPPB. (B) Increase of VX-770-activated G551D-CFTR current by NPPB. (C) Synergistic effects of VX-770 and NPPB in the presence of 20 µM ATP. A continuous recording of G551D-CFTR current shows a sequential increase of the current by 20 µM ATP, VX-770, and NPPB. (D) A bar graph summarizes the mean ± S.E. of the Po increase in the presence of 2 mM ATP (solid bars) or 20 µM ATP (open bars), respectively, by NPPB or VX-770 or both under conditions marked. At 2 mM ATP in the presence of NPPB, VX-770 increases the Po of G551D-CFTR by 4.76 ± 1.75-fold (n = 5), significantly smaller than the effect of VX-770 in the absence of NPPB (*P < 0.05). Similarly, the presence of VX-770 significantly dampens the effect of NPPB (6.27 ± 1.0-fold, n = 9, **P < 0.01). Thus, an overall 46.6 ± 8.5-fold (n = 18) increase in Po is seen with VX-770 and NPPB applied together. At 20 µM ATP, the effect of VX-770 was reduced (4.79 ± 1.04-fold, n = 14; 9.32-fold at 2 mM ATP in Fig. 4F). In the presence of VX-770, NPPB yields a 5.00 ± 0.40-fold increase of G551D currents (n = 13, *P < 0.05 compared with Fig. 2D); whereas VX-770 enhances the Po of NPPB-treated G551D-CFTR by only 2.53 ± 0.31-fold (n = 3). When the two reagents were applied together, we obtained 18.0 ± 5.0-fold increase of the Po for G551D-CFTR (n = 7).

For the second scenario to be valid, one would expect that the fold increase of Po by one reagent should be independent of the presence of the other compound. Yet quantitative analyses of our results indicated that at 2 mM ATP the fold increase of NPPB did depend on the presence of VX-770 and vice versa (Fig. 3D). In the presence of NPPB, VX-770 increased the Po of G551D-CFTR by 4.76 ± 1.75-fold (n = 5), significantly smaller than the effect of VX-770 in the absence of NPPB (9.32 ± 1.42-fold, n = 21, P < 0.05). Similarly, the presence of VX-770 significantly dampened the effect of NPPB (6.27 ± 1.0-fold, n = 9, versus 12.1-fold, P < 0.005). Thus, the overall 46-fold increase of G551D-CFTR currents when NPPB and VX-770 were applied together (46.6 ± 8.5-fold, n = 18) was larger than the sum of the effects from individual reagents (∼12-fold), an indication of pharmacologic synergism supporting the notion that NPPB and VX-770 bind to different sites. However, the fact that the combined effects of NPPB and VX-770 are smaller than the product of individual effects (∼110-fold) suggests a mechanism by which these two potentiators work in a dependent manner (i.e., the third scenario).

Although the mechanism responsible for this drug–drug interaction is unclear, we can at least test one simple idea that the binding of one drug lowers the effect of the other by reducing its binding affinity. This hypothesis predicts that this “interference” can be overcome by raising the concentration of the compound of interest. Due to the limited solubility of NPPB, we were only able to increase the concentration of VX-770. As shown in Supplemental Fig. 1, no further increase in G551D-CFTR currents was seen when we switched from 200 nM VX-770 to 1 μM VX-770 (1.02 ± 0.04-fold, n = 7, Supplemental Fig. 1B). Thus, binding of NPPB does not appear to lower the binding affinity of VX-770.

As described earlier, by lowering [ATP] to 20 μM, we can shift the distribution of G551D-CFTR toward monoliganded states. Similar to the result with NPPB (Fig. 2B), the effect of VX-770 was retained but reduced (4.79 ± 1.04-fold, n = 14, Fig. 3, C and D; versus 7.31 ± 0.81-fold in Fig. 2D). Furthermore, the interdependency of the effects of VX-770 and NPPB on G551D-CFTR could still be observed. In the presence of VX-770, NPPB gave only 5.0 ± 0.4-fold increase (n = 13, Fig. 3C) whereas VX-770 enhanced the Po of NPPB-treated G551D-CFTR by 2.53 ± 0.31-fold (n = 3). When the two reagents were applied together at 20 μM ATP we obtained an 18.0 ± 5.0-fold increase of the Po for G551D-CFTR (n = 7, Fig. 3, C and D). Again, 18 is larger than the sum of individual effects (4.79- and 7.31-fold increase, respectively), but smaller than the product of 4.79 and 7.31, affirming pharmacologic synergism between VX-770 and NPPB through a “dependent” mechanism.

This dependent mechanism between VX-770 and NPPB could also be observed in the most common pathogenic mutation ∆F508. As shown in Supplemental Fig. 2A, macroscopic ∆F508-CFTR currents increase immediately upon applying NPPB, with a calculated fold increase in Po of 5.52 ± 2.18-fold (n = 10) in the presence of 2 mM ATP. However, once the activity is enhanced by VX-770, the application of NPPB causes a net decrease of the macroscopic currents (Supplemental Fig. 2B), with a calculated fold increase in Po of 2.29 ± 0.17-fold (n = 14). Previous reports (Van Goor et al., 2009; Kopeikin et al., 2014) showed a 5- to 10-fold increase of the Po of ∆F508-CFTR by VX-770, but here, in the presence of NPPB, the effect of VX-770 is decreased to 3.51 ± 1.27-fold (n = 3). In these three patches, an overall 19.4 ± 7.0-fold (n = 3) increase in Po is seen with VX-770 and NPPB applied together (Supplemental Fig. 2C).

Previous studies (Jih and Hwang, 2013; Csanády and Töröcsik, 2014) demonstrated that NPPB and VX-770 share similar kinetic effects on WT-CFTR gating: an increase of the opening rate but a decrease of the closing rate. The data we have presented here also show a similar magnitude of potentiation on G551D channels for both reagents. However, the fact that the effect of VX-770 and NPPB together on G551D-CFTR is larger than the sum of individual effects from each reagent when applied alone argues that VX-770 and NPPB bind to different sites in CFTR.

The results shown in Fig. 4 further support the proposition that NPPB and VX-770 work differently. Although NPPB is considered a CFTR potentiator, Csanády and Töröcsik (2014) showed that the locked-open time of K1250A-CFTR, a hydrolysis deficient mutant, is paradoxically shortened by NPPB. In contrast, VX-770 prolongs the lock-open time of a different hydrolysis-deficient mutant E1371S-CFTR (Jih et al., 2011). To rule out the possibility of mutation-specific effect, we compared NPPB and VX-770 on K1250A-CFTR.

Fig. 4.

Fig. 4.

Open in a new tab

Different mechanisms of action for NPPB and VX-770 on K1250A- and ΔNBD2-CFTR. (A) Slow current relaxation of K1250A-CFTR upon removal of ATP. The red line represents the single-exponential fit of the current decay with a time constant of 54.1 seconds. (B) Effect of VX-770 on current relaxation of K1250A-CFTR upon removal of ATP. The red line indicates the single-exponential fit of the current decay with a time constant of 96.4 seconds. (C) Effect of NPPB on current relaxation of K1250A-CFTR upon removal of ATP. Note an immediate decrease of the K1250A-CFTR currents upon the application of NPPB due to pore block. The current decay upon removal of ATP can be fitted with a single-exponential function (red line) with a time constant of 22.9 seconds. (D) A bar graph summarizes the current relaxation time constant of K1250A-CFTR upon ATP washout in the presence of VX-770 (solid bar, 93.2 ± 12.4 seconds, n = 9, ***P < 0.001 compared with control), NPPB (open bar, 26.9 ± 3.01 seconds, n = 9, ***P < 0.001 compared with control), and control (hatched bar, 55.6 ± 4.31 seconds, n = 19). (E) Effect of NPPB on ΔNBD2-CFTR. Macroscopic ΔNBD2-CFTR currents decrease immediately upon applying NPPB. The calculated gating effect of NPPB on ΔNBD2-CFTR was only 2.56 ± 0.15-fold (n = 5, ***P < 0.001). A biphasic change of the ΔNBD2-CFTR currents is shown upon NPPB removal due to a fast release of blockade followed by a slower abolition of the potentiation effect. The slow decay phase can be fitted by a single exponential curve with time course of 39.7 seconds (mean value: 39.65 ± 6.68 seconds, n = 9). (F) Summary of effects of VX-770 (solid bar) and NPPB (open bar) on G551D- and ΔNBD2-CFTR.

Figure 4A shows a representative current relaxation of K1250A-CFTR upon removal of ATP. The current decay phase could be well fitted with a single exponential function with a time constant of 54.1 seconds (mean data: 55.6 ± 4.3 seconds, n = 19). A similar experiment was performed with the addition of VX-770 (Fig. 4B), which significantly (P < 0.001) prolonged the time course of current relaxation upon ATP removal (time constant = 93.2 ± 12.4 seconds, n = 9). By contrast, Fig. 4C shows that NPPB indeed shortened the relaxation time constant of K1250A-CFTR to 26.9 ± 3.0 seconds (n = 9, P < 0.001). Figure 4D summarizes the current relaxation time constants for the indicated experimental conditions.

The other mutation ΔNBD2-CFTR, a construct with the NBD2 deleted and hence unresponsive to ATP (Cui et al., 2007) was also tested. As demonstrated previously, the Po of this mutant is extremely low (Yeh et al., 2015). But because the mutation eliminates NBD dimerization and ATP hydrolysis, the gating of ΔNBD2-CFTR can be described by an open-closed, two-state scheme in equilibrium. As seen in Fig. 4E, the application of 200 µM NPPB caused a net decrease of ΔNBD2-CFTR currents. The calculated gating effect of NPPB on ΔNBD2-CFTR was only 2.56 ± 0.15-fold (n = 5), which was not only much smaller than that of G551D (P < 0.01) but also far from the ∼10-fold increase of the Po of ΔNBD2-CFTR by VX-770 (Yeh et al., 2015).

Upon removal of NPPB, a rapid current rise followed by a slow current decay was observed. This biphasic response to NPPB removal is reminiscent of the result seen with G551D-CFTR (Fig. 1D). However, careful inspections of these current changes provided some insights regarding the interaction of NPPB with ΔNBD2-CFTR. First, fitting the decay phase with a single exponential function yielded a time course of 39.7 ± 6.7 seconds (n = 9), which was much longer than that of G551D. If we assume that this decay phase reflects the dissociation of NPPB from the channel, a slower decay rate of ΔNBD2-CFTR indicates a tighter binding of NPPB to ΔNBD2-CFTR. Therefore, the reduced effect of NPPB on ∆NBD2-CFTR is not due to a lower binding affinity of the reagent with the channel.

Second, because of the slow decay of the current, a stable peak current could be obtained just seconds after NPPB removal. This peak amplitude reflects the gating-enhanced current with the block eliminated. Because the ratio of this peak current and the steady-state current in the absence of NPPB is also around 2-fold, we can conclude that the calculated potentiation effect is accurate and also exclude the possibility that the ΔNBD2 mutation may somehow alter NPPB block.

Figure 4F summarizes comparisons of the gating effects between NPPB and VX-770 on two different constructs, G551D- and ΔNBD2-CFTR. Thus, while these two CFTR potentiators bear similar effects on G551D-CFTR, deletion of NBD2 dramatically dampens the effects of NPPB but not that of VX-770, a result further reinforcing the conclusion that they work through different mechanisms.

The results presented so far suggest that NPPB and VX-770 may bind to two different sites. Our previous study (Jih and Hwang 2013) proposed that VX-770 binds to CFTR’s transmembrane segments because external application of VX-770 obliterates the effect of internal VX-770. Here, we tested whether NPPB exhibits similar properties as VX-770 by loading NPPB into the pipette in experiments using inside-out membrane patches. Further application of NPPB from the intracellular side still yielded a 8.89 ± 2.43-fold (n = 5) increase in the G551D-CFTR macroscopic current (Supplemental Fig. 3A). This result suggests that most, if not all, of NPPB’s gating effects come from NPPB binding to a binding site that is readily accessible from the cytoplasmic but not the extracellular side of the membrane.

Before we proceed to the discussion of our results, we shall ask: Are the different magnitudes of potentiation by NPPB between WT- and G551D-CFTR (Fig. 1) due to the mutation itself? In other words, although we posit that NPPB’s effects on G551D-CFTR can be well explained under the same kinetic framework used for WT channels with the only required removal of post-NBD-dimerized states in Fig. 1, it remains plausible that modification of G551D-CFTR gating by NPPB is unique for G551D itself and not related to the behavior of WT channels. This question is relevant because many of the experiments shown in the current report were performed with G551D-CFTR, and the results were interpreted under a unified gating scheme.

As a general limitation for all mutational approaches, one can seldom absolutely rule out the possibility that the observed effects are specific to the mutation itself, but in this particular case we can at least examine whether NPPB’s effects on WT-CFTR gating can be magnified by maneuvers that minimize ATP-induced NBD dimerization, hence mimicking that of G551D-CFTR.

One such experimental manipulation is to lower the probability of ATP occupancy at site 2 by decreasing [ATP] and/or mutations that reduce ATP affinity at site 2. Figure 5 shows that indeed for WT-CFTR, as [ATP] was decreased, the potentiation effect of NPPB was increased. For example, in the presence of 100 µM ATP (Fig. 5A), 20 µM ATP (Fig. 5B), and 5 µM ATP (Fig. 5C), the calculated fold increases of Po were 2.37 ± 0.17-fold (n = 7), 3.40 ± 0.24-fold (n = 5), and 5.19 ± 0.60-fold (n = 8), respectively. Moreover, in the presence of 5 µM ATP, a net increase of the current upon applying NPPB was seen, a result very similar to that of G551D.

Fig. 5.

Fig. 5.

Open in a new tab

ATP concentrations influence the effects of NPPB on WT-CFTR gating. (A) Effect of NPPB on WT-CFTR at 100 µM ATP. A continuous recording of WT-CFTR shows a decrease of macroscopic currents by lowering [ATP] from 2 mM to 100 µM. Subsequent application of NPPB results in a net decrease of the currents, indicating blocking effects outweigh the gating effect. The calculated gating effect of NPPB was 2.37 ± 0.17-fold (n = 7). (B) Similar experiments as that in panel A except [ATP] = 20 µM ATP when NPPB was added. The calculated gating effect was 3.40 ± 0.24-fold (n = 5). (C) Effect of NPPB on WT-CFTR gating at 5 µM ATP, with a calculated gating effect of 5.19 ± 0.60-fold (n = 8). (D) Effect of NPPB on Y1219F at 100 µM ATP, with a calculated gating effect of 2.95 ± 0.43-fold (n = 5). (E) Effect of NPPB on Y1219F-CFTR at 5 µM ATP, with a calculated gating effect of 9.10 ± 1.93-fold (n = 9). (F) A bar graph summarizes the gating effects of NPPB under different ATP concentrations for WT- (solid bars) or Y1219F-CFTR (open bars). Calculated gating effects of NPPB on Y1219F-CFTR at 2 mM and 20 µM ATP were 1.92 ± 0.23-fold (n = 8) and 5.41 ± 0.98-fold (n = 6), respectively.

We also employed a second approach to lower the probability of occupancy in site 2 by mutating Y1219 to phenylalanine (Y1219F) and performing the same set of experiments. In the presence of 2 mM ATP (data not shown), 100 µM ATP (Fig. 5D), 20 µM ATP (data not shown), and 5 µM ATP (Fig. 5E) in Y1219F-CFTR, the fold increases of Po were 1.71 ± 0.17-fold (n = 9), 2.95 ± 0.43-fold (n = 5), 5.41 ± 0.98-fold (n = 6), and 9.10 ± 1.93-fold (n = 9), respectively (Fig. 5F). Of note, we chose not to lower [ATP] further lest a significant number of channels would lose the ATP molecule bound in site 1.

Although the results did not completely replicate the same magnitude of fold increase of Po in G551D because some of the WT channels might still go through the NBD dimerization-hydrolysis pathway, these results are qualitatively consistent with the notion that the same mechanism for NPPB’s effects on G551D-CFTR may be applicable to WT channels.

Discussion

The data we present in the current study support the conclusion that NPPB and VX-770 increase the activity of G551D-CFTR through different binding sites and their mechanisms of action likely differ from each other. It should be noted that this conclusion, based simply on quantitative comparisons between combined effects of these two reagents and their individual effects, is independent of any kinetic model for CFTR gating. There is further evidence supporting this conclusion. First, relative to the magnitudes of gating potentiation by NPPB and VX-770 on G551D-CFTR, deletion of NBD2 drastically reduces the efficacy of NPPB but not that of VX-770 (Fig. 4F). Second, NPPB and VX-770 pose opposite effects on the closing rate of K1250A-CFTR (Fig. 4D). Third, although the binding site for VX-770 is equally accessible from either side of the membrane (Jih and Hwang, 2013), the one for NPPB is preferentially accessible from the cytoplasmic side of the channel. To elucidate the potential mechanism responsible for this interesting drug–drug interaction between two CFTR potentiators, it is helpful to understand our results in the context of a gating model.

The modified gating model depicted in Fig. 2A for G551D-CFTR was used to see whether we can gain some insights into the mechanism of action for NPPB, which may then shed light on how NPPB acts on WT-CFTR. As proposed originally by Vergani et al. (2003), ATP binding to the catalysis-competent site (or site 2) is not coupled to gate opening; opening of CFTR’s gate is controlled instead by NBD dimerization, which is facilitated by ligand binding. Applying this concept to our gating model, we reckon that the two transitions in Fig. 2A, C2 ATP ↔ O2 ATP and C2 ↔ O2, should share the same parameters. This lack of energetic coupling between ATP binding and gating predicts that NPPB should exert a similar effect on these two kinetic reactions and hence the potentiation effect of NPPB should not depend on [ATP], changes of which shift the distribution of the channel between monoliganded and biliganded states. Our experimental data (Fig. 2B–D) apparently contradict this prediction. To keep the scheme unchanged, one has to propose that NPPB works differently on these two transitions: C2 ATP ↔ O2 ATP and C2 ↔ O2.

Suppose, indeed, that NPPB affects the C2 ↔ O2 step, the same transition proposed to be modulated by VX-770; we expect either a competitive relationship (i.e., binding to the same site) or an independent action from two distinct binding sites (e.g., nitrate and VX-770 in Yeh et al. (2015)) when NPPB is applied together with VX-770. However, neither was observed. Instead, a “reduction” of one reagent’s effect was seen in the presence of the other reagent as though they work in a dependent manner. Our results also make it unlikely that this apparent negative impact of one drug over the other is due to a change of binding affinity (Supplemental Fig. 1). Thus, although we cannot rule out some sort of nonspecific “allosteric” interference between NPPB and VX-770, here we attempt to find a more testable proposition based on a relatively new idea of the gating mechanism of CFTR.

As shown in Fig. 1A, the energetic coupling model depicts a probabilistic relationship between NBD dimerization and gate opening. ATP, the ligand for CFTR gating, serves as a catalyst that facilitates NBD dimerization. The sheer existence of this molecular motion predicts NBD dimerization with an unoccupied site 2 (i.e., C2 and O2 states). We thus propose to add C0AD and O0 states representing a closed state and an open state, respectively, with dimerized NBDs but an empty site 2 (Fig. 6). In the absence of the catalyst ATP in site 2, transitions from C2 and O2 to these two newly added states are slow, but the same principle of energetic coupling between NBD dimerization and gate opening should apply. That is, C0AD ↔ O0 should assume a higher Po than C2 ↔ O2. This modification of CFTR gating scheme, albeit speculative, not only explains data presented in our previous report (Lin et al., 2014), but also opens a new possibility to explain how VX-770 and NPPB can interact with each other.

We hypothesize that NPPB, by binding at the NBD dimer interface, promotes NBD dimerization, a mechanism distinct from that of VX-770. Qualitatively, this mechanism explains how NPPB increases the opening rate of WT-CFTR (increasing C2ATP → C2AD in Fig. 1A), shortens the lock-open time of K1250A-CFTR presumably due to destabilization of the NBD dimer because of molecular crowding, and inhibits ATP hydrolysis (Csanády and Töröcsik, 2014). This hypothesis also offers a straightforward explanation for the sensitivity of NPPB’s effects to the deletion of NBD2.

Furthermore, this proposed mechanism also qualitatively explains an intriguing, fortuitous observation noted in Fig. 3A: the presence of NPPB reduced the initial current rising phase upon ATP washout with G551D-CFTR (46% ± 5%, n = 4, versus 70% in the presence of VX-770). As reported elsewhere (Lin et al., 2014), this rapid current increase phase is due to G551D-CFTR channels being trapped by millimolar ATP in the C2ATP states depicted in Fig. 6; removal of ATP allows the CFTR protein to undergo conformational changes to the more stable dimerized open states, O0 in the modified gating scheme. Thus, NPPB, through the proposed mechanism, shifts the gating equilibrium of G551D-CFTR channels to dimerized-O0 and C0AD states, and hence diminishes the apparent “inhibitory” action of ATP. Finally, because of the coupling between NBD dimerization and gate opening, a dependent interaction is expected for NPPB acting on NBD dimerization and VX-770, which modulates directly gate opening/closing step. In Supplemental Fig. 4, we provide results based on computer simulations to collaborate this idea of a dependent mechanism for these two CFTR potentiators.

The results from our computer simulations are consistent with the idea that NPPB acts on NBD dimerization steps for G551D-CFTR (Supplemental Fig. 4), but the question is whether the same mechanism can be applied to WT-CFTR. This question is hard to answer, as one can never completely exclude the possibility of a global effect caused by the mutation on the protein structure and hence gating function. Nevertheless, as we showed in Fig. 5, the potentiation effect of NPPB at very low [ATP] in WT-CFTR, which eliminates most ATP-induced NBD dimerization events to mimic G551D-CFTR, is similar to that of G551D.

Although our computer simulations using the modified gating model seem to fit some experimental data (see Supplemental Fig. 3), we were unable to explain a reduction of the effect of VX-770 on G551D-CFTR when [ATP] is decreased to 20 µM (10-fold based on simulations, but 5-fold observed). Here we speculate on several possible reasons responsible for this discrepancy. First, we must accept the possibility that our model is wrong. Although our model does explain many of the current data as well as the published data, these inconsistent results do provide an opportunity to reexamine the model in the future. Second, because the Po of G551D-CFTR is extremely low, most of our recordings yield microscopic currents, resulting in imprecise measurements. However, we did not see the expected large variation of our data predicted by this scenario. Third, it is a generally held principle that modeling should be performed with as few assumptions as possible. Here, the simplest assumption for our simulations is that VX-770 affects each vertical transition to the same extent. However, it was recently shown that conformational changes in the NBDs will cause reorientation of transmembrane domains (Ehrhardt et al., 2016), which may in theory alter the effect of VX-770 as VX-770 is proposed to bind to the transmembrane domains. Therefore, although modifications of our assumption may better suit the experimental data, we decided not to increase complexity lacking supporting evidence.

One caveat of our study we recognize is that our functional assay does not allow direct visualization of NPPB’s binding site; therefore, much of what have been deliberated here remains speculative. It may also be interesting to consider the possible binding sites identified by previous studies. Moran et al. (2005) argue that the NBD1:NBD2 interface is one binding site for CFTR potentiators. Kalid et al.(2010) raise the possibility that the interface between the NBDs and the intracellular loops is another potential binding site.

Indeed, much more about how potentiators work on CFTR channels awaits future studies using diverse tools. Nonetheless, the current work—demonstrating synergistic effects of two well-studied CFTR potentiators and a logical gating scheme, supported by numerous functional data—does lay a sound foundation to guide future drug design. The unique dependent gating mechanism revealed here underscores the possibility of developing drugs that can complement the action of VX-770, and hence holds the promise of a major therapeutical advancement for at least a subset of CF patients.

Supplementary Material

Data Supplement
supp_90_3_275__index.html (1.7KB, html)

Acknowledgments

The authors thank Cindy Chu and Shenghui Hu for technical assistance, Dr. Robert Bridges for providing VX-770, and Kang-Yang Jih for efforts in performing some pilot experiments.

Abbreviations

CF

cystic fibrosis

CFTR

cystic fibrosis transmembrane conductance regulator

Inh-172

4-[[4-oxo-2-sulfanylidene-3-[3-(trifluoromethyl)phenyl]-1,3-thiazolidin-5-ylidene]methyl]benzoic acid

NBD

nucleotide-binding domain

NMDG

(N-methyl-d-glucamine)-Cl

NPPB

5-nitro-2-(3-phenylpropylamino) benzoate

PKA

protein kinase A

Po

open probability

TMD

transmembrane domain

VX-770

ivacaftor, Kalydeco, N-(2,4-di-tert-butyl-5-hydroxyphenyl)-4-oxo-1,4-dihydroquinoline-3-carboxamide

VX-809

3-[6-[[1-(2,2-difluoro-1,3-benzodioxol-5-yl)cyclopropanecarbonyl]amino]-3-methylpyridin-2-yl]benzoic acid

WT

wild-type

Authorship Contributions

Participated in research design: Lin, Sohma, Hwang.

Conducted experiments: Lin.

Contributed new reagents or analytic tools: Sohma.

Performed data analysis: Lin.

Wrote or contributed to the writing of the manuscript: Lin, Sohma, Hwang.

Footnotes

This work was supported by a scholarship from Taipei Veterans General Hospital/National Yang-Ming University Excellent Physician Scientists Cultivation Program in Taiwan [No. 102-Y-A-001 (to W.-Y.L.)], and the National Institutes of Health National Institute of Diabetes and Digestive and Kidney Diseases [Grant R01DK55835], the Cystic Fibrosis Foundation [Grant Hwang11P0 (to T.-C. H.)], and JSPS KAKENHI [Grants 25293049, 15K15035, 16H05122 (to Y.S.)]. During the course of the study, 5% of the salary support for T.-C. H. was provided by Vertex Pharmaceuticals Inc.

Inline graphicThis article has supplemental material available at molpharm.aspetjournals.org.

References

  1. Accurso FJ, Rowe SM, Clancy JP, Boyle MP, Dunitz JM, Durie PR, Sagel SD, Hornick DB, Konstan MW, Donaldson SH, et al. (2010) Effect of VX-770 in persons with cystic fibrosis and the G551D-CFTR mutation. N Engl J Med 363:1991–2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Barry PJ, Ronan N, Plant BJ. (2015) Cystic fibrosis transmembrane conductance regulator modulators: the end of the beginning. Semin Respir Crit Care Med 36:287–298. [DOI] [PubMed] [Google Scholar]
  3. Bear CE, Li CH, Kartner N, Bridges RJ, Jensen TJ, Ramjeesingh M, Riordan JR. (1992) Purification and functional reconstitution of the cystic fibrosis transmembrane conductance regulator (CFTR). Cell 68:809–818. [DOI] [PubMed] [Google Scholar]
  4. Bompadre SG, Sohma Y, Li M, Hwang T-C. (2007) G551D and G1349D, two CF-associated mutations in the signature sequences of CFTR, exhibit distinct gating defects. J Gen Physiol 129:285–298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cai Z, Sohma Y, Bompadre SG, Sheppard DN, Hwang T-C. (2011) Application of high-resolution single-channel recording to functional studies of cystic fibrosis mutants. Methods Mol Biol 741:419–441. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cai Z, Taddei A, Sheppard DN. (2006) Differential sensitivity of the cystic fibrosis (CF)-associated mutants G551D and G1349D to potentiators of the cystic fibrosis transmembrane conductance regulator (CFTR) Cl− channel. J Biol Chem 281:1970–1977. [DOI] [PubMed] [Google Scholar]
  7. Cholon DM, Quinney NL, Fulcher ML, Esther JCR, Das J, Dokholyan NV, Randell SH, Boucher RC, Gentzsch M. (2014) Potentiator ivacaftor abrogates pharmacological correction of ΔF508 CFTR in cystic fibrosis. Sci Transl Med 6:246ra296 DOI: 10.1126/scitranslmed.3008680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Csanády L. (2000) Rapid kinetic analysis of multichannel records by a simultaneous fit to all dwell-time histograms. Biophys j 78:785–799. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Csanády L, Töröcsik B. (2014) Catalyst-like modulation of transition states for CFTR channel opening and closing: new stimulation strategy exploits nonequilibrium gating. J Gen Physiol 143:269–287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cui L, Aleksandrov L, Chang XB, Hou YX, He L, Hegedus T, Gentzsch M, Aleksandrov A, Balch WE, Riordan JR. (2007) Domain interdependence in the biosynthetic assembly of CFTR. J Mol Biol 365:981–994. [DOI] [PubMed] [Google Scholar]
  11. Eckford PDW, Li C, Ramjeesingh M, Bear CE. (2012) Cystic fibrosis transmembrane conductance regulator (CFTR) potentiator VX-770 (ivacaftor) opens the defective channel gate of mutant CFTR in a phosphorylation-dependent but ATP-independent manner. J Biol Chem 287:36639–36649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ehrhardt A, Chung WJ, Pyle LC, Wang W, Nowotarski K, Mulvihill CM, Ramjeesingh M, Hong J, Velu SE, Lewis HA, et al. (2016) Channel gating regulation by the cystic fibrosis transmembrane conductance regulator (CFTR) first cytosolic loop. J Biol Chem 291:1854–1865. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hwang T-C, Sheppard DN. (2009) Gating of the CFTR Cl− channel by ATP-driven nucleotide-binding domain dimerisation. J Physiol 587:2151–2161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Jih K-Y, Hwang T-C. (2012) Nonequilibrium gating of CFTR on an equilibrium theme. Physiology (Bethesda) 27:351–361. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Jih K-Y, Hwang T-C. (2013) Vx-770 potentiates CFTR function by promoting decoupling between the gating cycle and ATP hydrolysis cycle. Proc Natl Acad Sci USA 110:4404–4409. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Jih K-Y, Li M, Hwang T-C, Bompadre SG. (2011) The most common cystic fibrosis-associated mutation destabilizes the dimeric state of the nucleotide-binding domains of CFTR. J Physiol 589:2719–2731. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Jih K-Y, Sohma Y, Hwang T-C. (2012a) Nonintegral stoichiometry in CFTR gating revealed by a pore-lining mutation. J Gen Physiol 140:347–359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kalid O, Mense M, Fischman S, Shitrit A, Bihler H, Ben-Zeev E, Schutz N, Pedemonte N, Thomas PJ, Bridges RJ, et al. (2010) Small molecule correctors of F508del-CFTR discovered by structure-based virtual screening. J Comput Aided Mol Des 24:971–991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kopeikin Z, Sohma Y, Li M, Hwang T-C. (2010) On the mechanism of CFTR inhibition by a thiazolidinone derivative. J Gen Physiol 136:659–671. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kopeikin Z, Yuksek Z, Yang HY, Bompadre SG. (2014) Combined effects of VX-770 and VX-809 on several functional abnormalities of F508del-CFTR channels. J Cyst Fibros 13:508–514. [DOI] [PubMed] [Google Scholar]
  21. Lansdell KA, Cai Z, Kidd JF, Sheppard DN. (2000) Two mechanisms of genistein inhibition of cystic fibrosis transmembrane conductance regulator Cl− channels expressed in murine cell line. J Physiol 524:317–330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lewis HA, Buchanan SG, Burley SK, Conners K, Dickey M, Dorwart M, Fowler R, Gao X, Guggino WB, Hendrickson WA, et al. (2004) Structure of nucleotide-binding domain 1 of the cystic fibrosis transmembrane conductance regulator. EMBO J 23:282–293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lin W-Y, Jih K-Y, Hwang T-C. (2014) A single amino acid substitution in CFTR converts ATP to an inhibitory ligand. J Gen Physiol 144:311–320. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Ma T, Thiagarajah JR, Yang H, Sonawane ND, Folli C, Galietta LJV, Verkman AS. (2002) Thiazolidinone CFTR inhibitor identified by high-throughput screening blocks cholera toxin-induced intestinal fluid secretion. J Clin Invest 110:1651–1658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Matthes E, Goepp J, Carlile GW, Luo Y, Dejgaard K, Billet A, Robert R, Thomas DY, Hanrahan JW. (2016) Low free drug concentration prevents inhibition of F508del CFTR functional expression by the potentiator VX-770 (ivacaftor). Br J Pharmacol 173:459–470. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Miki H, Zhou Z, Li M, Hwang T-C, Bompadre SG. (2010) Potentiation of disease-associated cystic fibrosis transmembrane conductance regulator mutants by hydrolyzable ATP analogs. J Biol Chem 285:19967–19975. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Moody JE, Millen L, Binns D, Hunt JF, Thomas PJ. (2002) Cooperative, ATP-dependent association of the nucleotide binding cassettes during the catalytic cycle of ATP-binding cassette transporters. J Biol Chem 277:21111–21114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Moran O, Galietta LJV, Zegarra-Moran O. (2005) Binding site of activators of the cystic fibrosis transmembrane conductance regulator in the nucleotide binding domains. Cell Mol Life Sci 62:446–460. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ostedgaard LS, Baldursson O, Welsh MJ. (2001) Regulation of the cystic fibrosis transmembrane conductance regulator Cl− channel by its R domain. J Biol Chem 276:7689–7692. [DOI] [PubMed] [Google Scholar]
  30. Quinton PM, Reddy MM. (1991) Regulation of absorption in the human sweat duct. Adv Exp Med Biol 290:159–170. [DOI] [PubMed] [Google Scholar]
  31. Ramsey BW, Davies J, McElvaney NG, Tullis E, Bell SC, Dřevínek P, Griese M, McKone EF, Wainwright CE, Konstan MW, et al. VX08-770-102 Study Group (2011) A CFTR potentiator in patients with cystic fibrosis and the G551D mutation. N Engl J Med 365:1663–1672. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Riordan JR, Rommens JM, Kerem B, Alon N, Rozmahel R, Grzelczak Z, Zielenski J, Lok S, Plavsic N, Chou J-L, et al. (1989) Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science 245:1066–1073. [DOI] [PubMed] [Google Scholar]
  33. Rowe SM, Verkman AS. (2013) Cystic fibrosis transmembrane regulator correctors and potentiators, in Cystic Fibrosis: A Trilogy of Biochemistry, Physiology, and Therapy (Riordan JR, Boucher RC, Quinton PM, eds) pp 287–301, Cold Spring Harbor Publishers, New York. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Rowe SM, Miller S, Sorscher EJ. (2005) Cystic fibrosis. N Engl J Med 352:1992–2001. [DOI] [PubMed] [Google Scholar]
  35. Tsai M-F, Li M, Hwang T-C. (2010) Stable ATP binding mediated by a partial NBD dimer of the CFTR chloride channel. J Gen Physiol 135:399–414. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Tsai M-F, Shimizu H, Sohma Y, Li M, Hwang T-C. (2009) State-dependent modulation of CFTR gating by pyrophosphate. J Gen Physiol 133:405–419. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Van Goor F, Hadida S, Grootenhuis PD, Burton B, Cao D, Neuberger T, Turnbull A, Singh A, Joubran J, Hazlewood A, et al. (2009) Rescue of CF airway epithelial cell function in vitro by a CFTR potentiator, VX-770. Proc Natl Acad Sci USA 106:18825–18830. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Van Goor F, Hadida S, Grootenhuis PD, Burton B, Stack JH, Straley KS, Decker CJ, Miller M, McCartney J, Olson ER, et al. (2011) Correction of the F508del-CFTR protein processing defect in vitro by the investigational drug VX-809. Proc Natl Acad Sci USA 108:18843–18848. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Van Goor F, Yu H, Burton B, Hoffman BJ. (2014) Effect of ivacaftor on CFTR forms with missense mutations associated with defects in protein processing or function. J Cyst Fibros 13:29–36. [DOI] [PubMed] [Google Scholar]
  40. Veit G, Avramescu RG, Perdomo D, Phuan PW, Bagdany M, Apaja PM, Borot F, Szollosi D, Wu YS, Finkbeiner WE, et al. (2014) Some gating potentiators, including VX-770, diminish ΔF508-CFTR functional expression. Sci Transl Med 6:246ra97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Vergani P, Lockless SW, Nairn AC, Gadsby DC. (2005) CFTR channel opening by ATP-driven tight dimerization of its nucleotide-binding domains. Nature 433:876–880. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Vergani P, Nairn AC, Gadsby DC. (2003) On the mechanism of MgATP-dependent gating of CFTR Cl− channels. J Gen Physiol 121:17–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Wang F, Zeltwanger S, Yang ICH, Nairn AC, Hwang T-C. (1998) Actions of genistein on cystic fibrosis transmembrane conductance regulator channel gating. Evidence for two binding sites with opposite effects. J Gen Physiol 111:477–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Wang W, Li G, Clancy JP, Kirk KL. (2005) Activating cystic fibrosis transmembrane conductance regulator channels with pore blocker analogs. J Biol Chem 280:23622–23630. [DOI] [PubMed] [Google Scholar]
  45. Wangemann P, Wittner M, Di Stefano A, Englert HC, Lang HJ, Schlatter E, Greger R. (1986) Cl−-channel blockers in the thick ascending limb of the loop of Henle. Structure activity relationship. Pflugers Arch 407 (Suppl 2):S128–S141. [DOI] [PubMed] [Google Scholar]
  46. Xu Z, Pissarra LS, Farinha CM, Liu J, Cai Z, Thibodeau PH, Amaral MD, Sheppard DN. (2014) Revertant mutants modify, but do not rescue, the gating defect of the cystic fibrosis mutant G551D-CFTR. J Physiol 592:1931–1947. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Yeh HI, Yeh JT, Hwang TC. (2015) Modulation of CFTR gating by permeant ions. J Gen Physiol 145:47–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Yu H, Burton B, Huang CJ, Worley J, Cao D, Johnson JP, Jr, Urrutia A, Joubran J, Seepersaud S, Sussky K, et al. (2012) Ivacaftor potentiation of multiple CFTR channels with gating mutations. J Cyst Fibros 11:237–245. [DOI] [PubMed] [Google Scholar]
  49. Yu Y-C, Sohma Y, Hwang T-C. (2016) On the mechanism of gating defects caused by the R117H mutation in cystic fibrosis transmembrane conductance regulator. J Physiol 594:3227–3244 10.1113/JP271723. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Zhang ZR, Zeltwanger S, McCarty NA. (2000) Direct comparison of NPPB and DPC as probes of CFTR expressed in Xenopus oocytes. J Membr Biol 175:35–52. [DOI] [PubMed] [Google Scholar]
  51. Zhou J-J, Li M-S, Qi J, Linsdell P. (2010) Regulation of conductance by the number of fixed positive charges in the intracellular vestibule of the CFTR chloride channel pore. J Gen Physiol 135:229–245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zhou Z, Wang X, Liu H-Y, Zou X, Li M, Hwang T-C. (2006) The two ATP binding sites of cystic fibrosis transmembrane conductance regulator (CFTR) play distinct roles in gating kinetics and energetics. J Gen Physiol 128:413–422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Zielenski J, Tsui LC. (1995) Cystic fibrosis: genotypic and phenotypic variations. Annu Rev Genet 29:777–807. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data Supplement
supp_90_3_275__index.html (1.7KB, html)
supp_mol.116.104570_Supplemental_104570.pdf (904.9KB, pdf)

Articles from Molecular Pharmacology are provided here courtesy of American Society for Pharmacology and Experimental Therapeutics

RESOURCES