(R)-vanzacaftor potentiates BK_Ca channels in the absence of CFTR correction or potentiation

Abstract

We previously demonstrated the CFTR correctors VX-445 (elexacaftor) and S-VX-121 (vanzacaftor) potentiate heterologously-expressed BK_Ca channels, as well as in primary human bronchial epithelial cells (HBEs). This potentiation of BK_Ca resulted in altered vasoreactivity and neuronal excitability. We postulated novel compounds could be identified that would potentiate BK_Ca while not affecting CFTR. Herein, we demonstrate the enantiomer of vanzacaftor, R-VX-121, possesses these attributes. Using Fisher rat thyroid (FRT) cells expressing F508del CFTR, we demonstrate S-VX-121 corrects F508del CFTR, when incubated overnight, as assessed by an increase in transepithelial Cl⁻ current (I_Cl) in response to forskolin, as well as the appearance of band C upon immunoblot (IB). In contrast, R-VX-121 failed to increase I_Cl and induce band C. Importantly, R-VX-121 competed with S-VX-121 to eliminate the correction of F508del CFTR observed during both I_Cl measurements and IB, indicating it associates with CFTR. Neither S- nor R-VX-121 potentiated CFTR, as assessed by changes in I_Cl. Distinct from our CFTR results, both S- and R-VX-121 potentiated BK_Ca in primary HBEs as well as during whole-cell patch-clamp recording of heterologously expressed α-BK_Ca. Using wire myography, we demonstrate both S- and R-VX-12 vasodilate preconstricted mouse mesenteric arteries in a paxilline-dependent manner, confirming a role for BK_Ca. In contrast, the CFTR inhibitor, CFTR_inh172 did not alter the effects of S- and R-VX-121 on vasoreactivity, confirming CFTR is not involved in this response. These data demonstrate R-VX-121 represents a novel BK_Ca potentiator that does not modulate CFTR function, suggesting R-VX-121 may be clinically useful as a BK_Ca agonist.

Graphical Abstract

NEW AND NOTEWORTHY

We previously demonstrated the CFTR correctors, VX-445 and S-VX-121 are BK_Ca channel potentiators. These CFTR correctors altered vasoreactivity and action potential firing frequency – effects which may explain the adverse events (AE) reported in cystic fibrosis (CF). We now demonstrate the enantiomer of vanzacaftor, R-VX-121 potentiates BK_Ca, while not correcting or potentiating CFTR. Thus, we have identified a novel BK_Ca potentiator that may be useful in diseases where BK_Ca modulation is therapeutically proposed.

INTRODUCTION

The large conductance, voltage- and Ca²⁺-activated K⁺ channel (BK_Ca) is widely expressed in human tissues. In vascular (1) and bladder (2) smooth muscle, potentiation of BK_Ca induces relaxation. In the central nervous system, BK_Ca is one of the intrinsic determinants of action potential firing (3). Indeed, BK_Ca channelopathies, involving both gain-of-function and loss-of-function, have been described (3-5). In epithelia, BK_Ca is instrumental in regulating K⁺ secretion (6). In human airway, BK_Ca plays a key role in regulating airway surface liquid (ASL) volume (7).

While pharmacological activation of BK_Ca may be therapeutically beneficial in numerous diseases, clinical modulators have not been forthcoming (8). In the airway, we proposed potentiation of a distinct basolateral K⁺ channel, KCa3.1 may be beneficial in airway disease, including cystic fibrosis (CF) and chronic obstructive pulmonary disease (COPD). We identified novel KCa3.1 openers, and demonstrated they stimulate transepithelial Cl⁻ secretion across human bronchial epithelial cells (HBEs) (9-11). More recently, BK_Ca channels have been proposed as therapeutic targets for CF (7). We previously demonstrated the CFTR correctors, VX-445 and S-VX-121 are novel BK_Ca potentiators (12), stimulating K⁺ secretion across WT and F508del CFTR-expressing HBEs. Potentiation of apical BK_Ca would be expected to promote Cl⁻ secretion across the apical membrane. Unfortunately, we further demonstrated VX-445 directly inhibits basolateral KCa3.1, resulting in the paradoxical inhibition of forskolin-mediated Cl⁻ secretion across WT and F508del HBEs (13). Adverse events (AE) have been reported in a subset of CF patients taking Trikafta, including headache and mental status changes (14-19). As we demonstrated VX-445 and VX-121 altered vasoreactivity and neuronal excitability (12), we proposed this potentiation of BK_Ca may underlie the AE reported and postulated this effect could be separated from CFTR correction (12).

Enantiomers of VX-445 display differences in their ability to correct F508del CFTR (20). Additionally, VX-445 has been shown to potentiate CFTR (21). Herein, we demonstrate a differential effect of VX-121 enantiomers on CFTR correction and BK_Ca potentiation. That is, S-VX-121 corrects F508del CFTR, whereas R-VX-121 does not. In contrast, both S- and R-VX-121 directly potentiate BK_Ca. This potentiation results in the vasodilation of preconstricted mouse mesenteric arteries. Thus, our results demonstrate that R-VX-121 allows the separation of BK_Ca and CFTR effects, suggesting this enantiomer may prove beneficial in diseases where potentiation of BK_Ca may be therapeutically useful.

METHODS

HEK293 cells expressing α-BK_Ca were kindly provided by Heike Wulff (University of California, Davis, Davis, California, USA) and cultured as described (22). Fisher Rat Thyroid (FRT cells) were grown as described (23, 24). FRT cells were transfected with WT or F508del CFTR in pcDNA3.1 (kindly provided by Dr. Sanjay Mishra, University of Pittsburgh) using Lipofectamine 2000. Cells were then seeded onto 0.33cm² Transwell filters and allowed to polarize for at least 48hrs before performing electrophysiological experiments.

Culture of primary human bronchial epithelial cells (HBEs) were carried out as described (12, 13). HBEs were obtained from the Center for Organ Recovery and Education via the University of Pittsburgh Pulmonary, Allergy, Critical Care, and Sleep Medicine Lung Biobank and Repository. All experimental procedures involving the use of HBEs were approved by the institutional review board (IRB STUDY19090084) and the Committee for Oversight of Research and Clinical Training Involving Decedents (CORID number 451) at the University of Pittsburgh. At all times, authors were blind to the identity and sex of HBE donors. Primary HBEs expressing WT CFTR from 9 donors were cultured as described (12, 13). HBEs were plated on Costar Transwell^® permeable polyester membrane supports (0.4 μM pore size, 6.5 mm insert, Corning) and grown at an air-liquid interface (ALI) for 5+ weeks prior to electrophysiological recordings.

Ussing chamber I_sc measurements

Short circuit current (I_sc) measurements were carried out as described (12, 13) in a modified Ussing chamber (P2300, Physiologic Instruments) and the monolayers continuously short-circuited (VCC MC8, Physiologic Instruments) by forcing the transepithelial voltage to 0 mV. Transepithelial resistance (R_te) was monitored by applying a 2 mV pulse every 90 s. Our standard solution contains (in mM): 120 NaCl, 25 NaHCO₃, 3.3 KH₂PO₄, 0.8 K₂HPO₄, 4 CaCl₂, 1.2 MgCl₂, and 10 glucose. For measurements of transepithelial K⁺ current (I_K) NaCl in the basolateral solution was replaced with 120 K-gluconate, whereas in the apical solution, NaCl was replaced with Na-gluconate, thereby creating a 125:5 mM basolateral-to-apical K⁺ concentration gradient. For measurements of transepithelial Cl⁻ current (I_Cl) NaCl in the apical solution was replaced with Na-gluconate, thereby creating a 125:5 mM basolateral-to-apical Cl⁻ concentration gradient. CaCl₂ was used at 4 mM to account for the Ca²⁺ buffering capacity of gluconate. The pH of the solution is 7.4 when gassed with 95% O₂ – 5% CO₂. Experiments were carried out at 37 °C, and compounds are added cumulatively following establishment of a new stable current response. Paxilline, CFTR_inh172 and amiloride were added to the apical membrane, while VX-770, VX-445, as well as S- and R-VX-121 were added to both membranes. ΔI_K is calculated as the difference between the baseline current after amiloride inhibition and the peak response to the agonist. ΔI_Cl is calculated as the difference between the baseline current and the peak response to the agonist. To calculate the inhibition of S-VX-121-mediated correction of F508del CFTR by R-VX-121 we carried out 4 separate transfections. For each experimental value reported we carried out 3 separate current measurements for a given concentration of R-VX-121 plus S-VX-121 and determined the average I_Cl response. This response was compared to 3 separate I_Cl measurements carried out with S-VX-121 at the same time. This was then done for each concentration of R-VX-121 plus S-VX-121 to construct the inhibition curve presented.

Mesenteric Arteries

Wire Myography experiments were conducted as reported (12). Male or female C57Bl/6 mice between 10-12 weeks of age were purchased from Jackson Labs. Mice were euthanized by CO₂ asphyxiation and mesenteric arteries (MAs) isolated and cut into 2 mm segments. MAs were placed in a physiological salt solution (PSS) containing (in mM): 0.026 EDTA, 119 NaCl, 5.5 d-glucose, 25 NaHCO₃, 4.7 KCl, 1.17 MgSO₄, 1.18 KH₂PO₄ and 2.5 CaCl₂. PSS was brought to a pH of 7.4 by bubbling with 95% O₂ - 5% CO₂ at 37°C. MAs were mounted on a wire myograph (Multiple Myograph Model 620 M, Danish Myotechnology) and allowed to rest for 30 minutes. Subsequently, MAs were incrementally stretched to a tension equivalent to 80 mm Hg of physiological pressure. After a 30 minute rest period, MAs were incubated for 15 minutes with either paxilline (10 μM), CFTR_inh172 (3 μM) or control buffer (0.1% DMSO). The MAs were then constricted with the prostaglandin mimetic U46619 to induce maximal constriction before drug treatment. Subsequently, a cumulative concentration response curve to S- or R-VX-121 was conducted (10⁻⁹ – 10⁻⁵ M for 15 minutes/concentration). Ca²⁺-free PSS containing 1 × 10⁻⁶ sodium nitroprusside was used to determine maximal relaxation. Percent relaxation was normalized to the change in maximal constriction (U46619) and maximal dilation (Ca²⁺-free PSS). Mouse studies were approved by the University of Pittsburgh Institutional Animal Care and Use Committee (Protocol # 23063078).

Whole-Cell Patch-Clamp Electrophysiology

Our whole-cell patch-clamp protocol for assessing the effects of CFTR correctors on α-BK_Ca channel function has recently been reported (12). HEK cells expressing the αBK_Ca channel were plated onto poly-l-lysine (0.01%, Sigma-Aldrich) coated glass coverslips 1 day prior to patch-clamp analysis. Pipettes were filled with a solution containing (in mM): 145 K-gluconate, 10 EGTA, 7.5 CaCl₂, 2 MgCl₂, 3 mM NaATP, and 10 Hepes. pH was adjusted to 7.2 with KOH. The bath solution contained (in mM): 140 K-gluconate, 5 KCl, 1.0 MgCl₂, 10 Hepes and 2 CaCl₂. pH was adjusted to 7.4 with KOH. Recordings were carried out at room temperature. Currents were recorded using an Axon 200B amplifier (Axon Instruments) in conjunction with Clampex data acquisition software (Axon Instruments, version 9.2). The data were filtered using lowpass Bessel filter set at 2 kHz and a digitization rate of 10 kHz (1.48 MB/min). Borosilicate glass electrodes (1.65 mm outer diameter, World Precision Instruments) were pulled with a Narishige puller (model PP-830). After fire polishing with a World Precision Instruments micro-forge (MF-200), pipettes had a resistance of 2-3 MΩ. Current-voltage (I-V) relationships were determined via a pulse protocol involving a 400 ms voltage pulse from −80 to +200 mV in 40 mV increments, flanked by a 50 ms pulse to a holding potential of −80 mV. The middle 300 ms of each trace was selected to avoid any influence of capacitive transients, and average currents were calculated for each voltage via the I-V tool in Clampfit. Whole-cell currents were normalized to cell capacitance. In all experiments, compounds were added under constant perfusion (2.3 mls/min). For determining the concentration-dependence of S- and R-VX-121 potentiation of αBK_Ca in HEK cells, we utilized each concentration of VX-121 (0.3, 1, 3 and 10 μM) consecutively applied for 3 min to determine the maximum response at +80 mV. The resulting responses for each concentration were then normalized to 10 μM VX-121 (100% response) and the data fit to the Hill equation to determine the apparent K_s and Hill coefficient.

Immunoblots

FRT cell lysates were prepared as described (23, 24). Lysates were heated to 55°C for 15 min, separated by SDS-PAGE (4-15% gradient gel, BioRad mini-protean precast gels), transferred to Nitrocellulose (Bio-Rad), and subjected to IB analysis using antibodies as indicated below (primary antibodies at a dilution 1:1000, secondary at 1:5000). Densitometric quantification of protein band intensities was carried out in Adobe Photoshop 2025 (release 26.9.0 Adobe Systems, Inc., San Jose, CA), and values were expressed as a percentage or fold change of control signal, following background subtraction, and normalization to total protein expression (β-actin).

Antibodies

Antibodies used were as follows: anti-β-Actin (mouse, monoclonal, Sigma-Aldrich. Cat # A5316), goat anti-mouse-HRP (BioRad, Cat#STAR117P), anti-CFTR (mAb 596, mouse monoclonal, University of North Carolina Cystic Fibrosis Antibody Distribution Program).

Chemicals

Paxilline (HY-N6778), VX-445 (HY-111772), S-VX-121 (HY-145603), R-VX-121 (HY-145603A) and CFTR_inh172 (HY-16671) were obtained from MedChemExpress. Amiloride (A7410) was obtained from Sigma-Aldrich. Forskolin (F-9929) was obtained from LC Laboratories. Ultroser-G (NC1700979) was obtained from Pall Life Sciences (Cergy-Saint-Christophe, France). All other unspecified reagents are from Sigma-Aldrich.

Statistics

All data are presented as means ± SEM, where n indicates the number of filters or patch-clamp recordings. We assessed whether the data were normally distributed using both the D’Agostino & Pearson omnibus normality test and the Shapiro-Wilk normality test in GraphPad Prism (v. 10.1.0). Comparisons between two experimental maneuvers within an experiment were assessed for significance using a paired Student’s t-test. Significance between experiments were determined by an un-paired t-test. Significance between multiple experimental maneuvers within an experiment were determined by an ANOVA followed by a Tukey’s HSD (Honestly Significant Different) post-hoc test. Comparison between control and VX-121-induced vasoreactivity was evaluated by a 2-way ANOVA followed by a post-hoc Sidak multiple comparison test. All statistical analysis was carried out using GraphPad Prism (v. 10.1.0). A value of p<0.05 is considered statistically significant and is reported.

RESULTS

To determine the effects of VX-121 enantiomers on CFTR correction, F508del CFTR was expressed in FRT cells and transepithelial I_Cl measured using a basolateral-to-apical Cl⁻ gradient (see Methods). This method has been extensively utilized to screen CFTR mutations for responses to CFTR correctors and potentiators (25, 26). As shown in Fig 1A, following overnight exposure (18 hrs) to S-VX-121 (1 μM), forskolin induced an increase in I_Cl and this was further increased by the CFTR potentiator, VX-770 (1 μM). In contrast, forskolin induced a small increase in I_Cl after overnight exposure to R-VX-121 (1 μM) (Fig 1B). The average data from 5 separate transfections, as well as the DMSO (0.1%) control, are shown in Fig 1C. S-VX-121 resulted in a significantly larger increase in I_Cl, relative to R-VX-121 (P<0.01). These results are supported by our IB studies, where WT CFTR was used as a control. As expected, WT CFTR was fully glycosylated, as indicated by band C (Fig 1D, lane 1), whereas band C was not observed in F508del CFTR-expressing FRTs (lane 2). Consistent with our I_Cl measurements, overnight incubation of F508del CFTR expressing FRT cells in S-VX-121 (1 μM) resulted in the appearance of band C (lane 4), indicative of a partial correction of F508del CFTR (P<0.01). In contrast, overnight incubation in R-VX-121 (1 μM) did not increase band C (lane 3). Actin was used as a loading control (bottom blot). The average data from three separate transfections is shown in Fig 1E. The chemical structures of S- and R-VX-121 and VX-770 are shown in Fig 1M.

Figure 1. Effect of S- and R-VX-121 on F508del CFTR correction and potentiation in FRT cells.

Following overnight incubation in S-VX-121 (A; 1 μM) or R-VX-121 (B; 1 μM), forskolin (10 μM) and VX-770 (1 μM) induced an increase in I_Cl that was inhibited by CFTR_inh172 (3 μM). C. Average responses to forskolin + VX-770 for experiments in A and B. DMSO (0.1%) was used as a control. D. Immunoblot of CFTR showing correction of F508del CFTR by S-VX-121, but not R-VX-121. E. Average results for 3 experiments. F. Cl⁻ currents in response to forskolin following overnight incubation in S-VX-121 (1 μM; blue trace) or S-VX-121 (1 μM) plus R-VX-121 (3 μM) (red trace). G. Concentration-response relationship for inhibition of S-VX-121-induced correction of I_Cl by R-VX-121. The data were fit to the Hill equation with an apparent IC₅₀ of 3.5 μM. H. Cl⁻ currents in response to forskolin following overnight incubation in VX-445 (1 μM; blue trace) or VX-445 (1 μM) plus R-VX-121 (6 μM) (red trace). I. Average response to forskolin for the experiments shown in H (n=23). J, K, L. F508del CFTR was corrected with VX-445 (1 μM) plus VX-661 (3 μM). Forskolin was used to stimulate I_Cl and the effect of VX-770 (1 μM; J), S-VX-121 (10 μM) followed by VX-770 (K) or R-VX-121 (10 μM) followed by VX-770 (L) determined. CFTR_inh172 (3 μM) was used to confirm I_Cl was due to CFTR. M. Chemical structures of VX-770, S-VX-121 and R-VX-121. #, P<0.01.

Our I_Cl measurements suggest a small, but significant, correction of the F508del CFTR response to R-VX-121, suggesting R-VX-121 binds to F508del CFTR – likely at the same binding site occupied by S-VX-121. Thus, we determined whether R-VX-121 would compete with S-VX-121 for F508del CFTR correction. As above, overnight incubation in S-VX-121 (1 μM) resulted in correction of F508del CFTR, as indicated by forskolin-dependent increase in I_Cl (Fig 1F, blue trace). Importantly, overnight incubation in 1 μM S-VX-121 plus 3 μM R-VX-121 resulted in a significant decrease in the forskolin response (Fig 1F, red trace). The concentration-dependence of this inhibition was determined by incubating separate filters in 1 μM S-VX-121 plus varying concentrations of R-VX-121. The results of experiments from 4 separate transfections are shown in Fig 1G. The data were fit to the Hill equation, resulting in an apparent IC₅₀ of 3.5 μM for inhibition of the S-VX-121 correction of F508del CFTR by R-VX-121. These data are supported by our IB studies, as shown in Figs 1D and E. That is, when FRT cells expressing F508del CFTR were incubated overnight in a combination of S-VX-121 (1 μM) plus R-VX-121 (6 μM) we were unable to detect band C (Fig 1D, lane 5, P<0.01). These results suggest R-VX-121 competes for binding on F508del CFTR with S-VX-121, thereby eliminating the corrector effect of S-VX-121.

Given these results, we determined whether R-VX-121 would similarly alter the correction of F508del CFTR by VX-445. As above, we incubated F508del CFTR-expressing FRT cells overnight in either VX-445 (1 μM) alone or VX-445 (1 μM) plus R-VX-121 (6 μM) and determined the I_Cl response to forskolin. As shown in Fig 1H (blue trace), incubation in VX-445 resulted in a forskolin-dependent increase in I_Cl. However, addition of R-VX-121 significantly decreased this I_Cl response (red trace). As shown in Fig 1I, R-VX-121 significantly reduced the correction of F508del CFTR by VX-445 (n=23; P<0.01).

VX-445 has dual effects on CFTR, acting as both a corrector and potentiator (21). To determine whether S- and R-VX-121 potentiate F508del CFTR, FRT cells were incubated in VX-445 (1 μM) plus VX-661 (3 μM) to correct F508del CFTR (11). We then determined the I_Cl response to forskolin plus S- or R-VX-121, where VX-770 was used as a positive control. In control cells (Fig 1J), forskolin stimulated I_Cl and this was further potentiated by VX-770 (1 μM). In 6 filters from 3 separate transfections, VX-770 increased I_Cl an average of 11±0.5 μA/cm². In contrast, neither S-VX-121 (10 μM; Fig 1K) nor R-VX-121 (10 μM; Fig 1L) increased I_Cl subsequent to forskolin, whereas the further addition of VX-770 (1 μM) increased I_Cl, as above. Similar results were seen in 6 filters. These results demonstrate enantiomers of VX-121 do not potentiate CFTR.

We demonstrated S-VX-121 directly potentiates BK_Ca (12). As shown above, R-VX-121 fails to correct or potentiate CFTR, raising the possibility that R-VX-121 may be selective for BK_Ca potentiation. Initially, we determined whether R-VX-121 potentiates BK_Ca in WT CFTR HBEs. Potassium current (I_K) across the apical membrane was measured using a basolateral-to-apical K⁺ gradient ((12), see Methods). As our solutions result in an apical-to-basolateral Na⁺ gradient, we initially inhibited Na⁺ absorption with amiloride (10 μM). As shown in Fig 2A, S-VX-121 (10 μM) stimulated a large inward current that was inhibited by the specific BK_Ca blocker, paxilline (10 μM), confirming this current is due to BK_Ca activation, as reported (12). The average peak I_K response for 24 filters from 6 separate donors is shown in Fig 2C. We previously demonstrated this response was not altered by permeabilization of the basolateral membrane with nystatin, confirming this is due to potentiation of apical BK_Ca (12). Importantly, as shown in Fig 2B, R-VX-121 similarly stimulated an increase in paxilline-sensitive I_K. However, as shown in Fig 2C, the response to R-VX-121 is significantly smaller than to S-VX-121 (P<0.01). This reduced response to R-VX-121 could be due to either a lower affinity for activation of BK_Ca or a decrease in the maximal possible activation of BK_Ca. As such, we determined the concentration-dependence of these responses. The average concentration-response curves for 11 experiments, from 3 separate donors, are shown in Fig 2D. S-VX-121 increased I_K with an affinity of 5.8 μM (blue line). While R-VX-121 increased I_K with only a small difference in affinity (11.3 μM; red line), the maximal response obtained was reduced compared to S-VX-121, being 97±10 μA/cm² compared to 138±11 μA/cm².

Figure. 2. Enantiomer-dependent effect of VX-121 on BK_Ca currents across WT CFTR HBEs.

Both (S)-VX-121 (A, 10 μM) and (R)-VX-121 (B, 10 μM) stimulated I_K that was inhibited by paxilline. C. Average responses to (S)- and (R)-VX-121 for 24 experiments from 6 separate donors (#, P<0.01). D. Average concentration-response curves for 11 experiments from 3 separate donors - (S)-VX-121 (blue) and (R)-VX-121 (red). The curves were fit to a Hill equation with a K_s of 5.8 μM and a Hill slope of 2.1 for (S)-VX-121 and a K_s of 11.3 μM with a Hill slope of 2.1 for (R)-VX-121.

To evaluate the ability of R-VX-121 to directly activate BK_Ca, we carried out whole-cell patch-clamp recordings on α-BK_Ca expressing HEK cells, as we previously demonstrated VX-445 and S-VX-121 do not stimulate any currents in these cells in the absence of BK_Ca expression (12). HEK cells were clamped to −80 mV and pulsed in 40 mV increments to +200 mV. As shown for a single cell (Fig 3 A and B), R-VX-121 (10 μM) induced a significant increase in outward current. This was completely inhibited by paxilline (not shown), as previously described (12). The average percent increase in current density (pA/pF) for both S-VX-121 (n=11) and R-VX-121 (n=9) at 3 μM and 10 μM is shown in Fig 3C. R-VX-121 resulted in a slightly larger BK_Ca current response at both concentrations of VX-121 (P<0.05). We next determined the concentration dependence of this response. As shown in Fig 3D (n=9), R-VX-121 induced a concentration-dependent increase in BK_Ca current density (pA/pF). The average concentration dependence for S- (blue line, n=11) and R- (red line, n=9) VX-121 are shown in Fig 3E. The data were fitted to the Hill equation yielding an apparent K_s of 2.8 μM (Hill coeff. = 2.2) for R-VX-121 and 2.3 μM for S-VX-121 (Hill coeff. = 1.4).

Figure 3. Activation of BK_Ca by R- and S-VX-121 heterologously expressed in HEK cells.

A. Control whole-cell recording. Voltage was stepped from −80 to +200 mV in 40 mV increments. B. Recording from the same cell following stimulation with 10 μM R-VX-121. C. Average response to R- and S-VX-121 at +80 mV at both 3 μM and 10 μM. R-VX-121 increased BK_Ca current density an average of 105±11% (n=9) at 3 μM, whereas S-VX-121 increased current density 37 ±11% (n=11; P<0.05). At 10 μM, R-VX-121 increased current density an average of 212±23% (n=9), whereas S-VX-121 increased current density 76±2% (n=11; P<0.05). D. Average I-V relationships demonstrating the concentration-dependent increase in BK_Ca current density in response to increasing concentrations of R-VX-121 (n=9). E. Average concentration-response curves for (S)-VX-121 (blue, n=11) and (R)-VX-121 (red, n=9). The curves were fit to a Hill equation with a K_s of 2.38 μM and a Hill slope of 1.4 for (S)-VX-121 and an K_s of 2.8 μM with a Hill slope of 2.2 for (R)-VX-121. *, P<0.05.

We previously demonstrated VX-445 induced vasodilation in mesenteric arteries in a BK_Ca-dependent manner (12). As targeting BK_Ca outside of the airway could potentially lead to the AE reported by CF patients (14-19), we assessed the effect of VX-121 on vasoreactivity in mouse mesenteric arteries pre-constricted with the prostaglandin mimetic U46619 (1 × 10⁻⁷ to 5 × 10⁻⁷ M). As shown in Fig 4A for a single mesenteric artery from a male mouse, following U46619-induced vasoconstriction, R-VX-121 induced a concentration-dependent vasodilation. The average responses from 7 animals for S-VX-121 (blue line) and R-VX-121 (red line) are shown in Fig 4B. S-VX-121 induced vasorelaxation with a K_s of 3 μM, whereas R-VX-121 induced vasorelaxation with a K_s of 0.4 μM. As shown, the effects of R-VX-121 (Fig 4C) and S-VX-121 (Fig 4D) were partially reversed by paxilline (P<0.01 for both), demonstrating both S- and R-VX-121 alter vasoreactivity in a BK_Ca-dependent manner. As shown in Fig 4E, identical results were observed in mesenteric arteries from female mice. In 7 animals, both S- (K_s=1.5 μM) and R-VX-121 (K_s=0.3 μM) induced a concentration-dependent vasorelaxation.

Figure 4. Effect of S- and R-VX-121 on vasoreactivity in mouse mesenteric artery.

A. Concentration-dependent vasodilation of mesenteric artery by R-VX-121 from a male mouse. B. Average responses in male mice to S- (blue line, n=7) and R- (red line, n=7) VX-121. C and D. Average responses in male mice to R-VX-121 (C) and S-VX-121 (D) under control conditions (solid lines) and in the presence of 10 μM paxilline (Pax; dashed lines). Paxilline inhibited the response to both R- and S-VX-121 (n=8 for all conditions, P<0.01). E. Average responses in female mice to S- (blue line, n=7) and R- (red line, n=7) VX-121. F and G. Average responses in male mice to S-VX-121 (F) and R-VX-121 (G) under control conditions (solid lines) and in the presence of 3 μM CFTR_inh172 (dashed lines). CFTR_inh172 had no effect (n=8 for all conditions).

CFTR is expressed in the vasculature, where it plays a role in controlling vascular tone (27). Thus, we determined whether CFTR plays a role in the VX-121-induced vasorelaxation observed. Mesenteric arteries from male mice were preincubated in CFTR_inh172 (3 μM) for 15 minutes prior to determining the concentration-dependence of S- and R-VX-121-induced vasorelaxation. Importantly, CFTR_inh172 had no effect on the vasodilation induced by either S-VX-121 (Fig 4E) or R-VX-121 (Fig 4F), demonstrating the vasorelaxing effects induced by VX-121 are independent of CFTR.

DISCUSSION

We recently demonstrated the CFTR correctors, VX-445 and S-VX-121 act as novel BK_Ca channel potentiators (12). We further demonstrated VX-445 induced vasorelaxation in mesenteric arteries and both VX-445 and S-VX-121 significantly decreased neuronal action potential firing (12). Thus, we speculated the AE reported by CF patients (14-19) may be related to these off-target effects. Herein, we demonstrate the enantiomer of vanzacaftor, R-VX-121 selectively potentiate BK_Ca in the absence of CFTR corrector effects. That is, S-VX-121 corrects F508del CFTR, as assessed by correction of the I_Cl response (Figs 1A and F) as well as the appearance of band C on IB (Figs 1D and E). In contrast, R-VX-121 produces only a very small increase in I_Cl (Fig 1B) with no detectable increase in band C (Figs 1D and E). However, both S- and R-VX-121 potentiate BK_Ca in WT CFTR-expressing HBEs (Fig 2) and when heterologously expressed in HEK cells (Fig 3). We further demonstrate S- and R-VX-121 induce vasorelaxation in mesenteric arteries in a BK_Ca-dependent manner (Fig 4). Importantly, the effects on the vasculature were independent of CFTR, consistent with our observation that S- and R-VX-121 do not potentiate CFTR (Figs 1 K and L).

Given the effects of R-VX-121 on BK_Ca in HBEs, it may be tempting to propose adding this compound to the CF therapeutic armament. This is based on the fact that potentiation of K⁺ channels increases the electrochemical driving force for Cl⁻ secretion across the apical membrane. Indeed, our work showing potentiators of basolateral membrane KCa3.1 stimulate Cl⁻ secretion across airway cells supports such a notion (9-11). As a direct confirmation, we demonstrated the KCa3.1 potentiator, DCEBIO induced a significant increase in Cl⁻ secretion following forskolin plus VX-770 in F508del CFTR-expressing HBEs corrected with VX-445 plus VX-661 (11). As BK_Ca plays a critical role in maintaining ASL height in HBE cultures (7), potentiation of BK_Ca may would be beneficial. Our current results suggest two caveats to such a proposal. First, BK_Ca is widely expressed where it may be targeted by compounds such as R-VX-121. Indeed, we show R-VX-121 induces vasorelaxation of mesenteric arteries in a BK_Ca-dependent manner (Fig 4). Targeting BK_Ca in the vasculature may affect blood pressure as well as induce headaches or migraines as was reported for the clinically-evaluated BK_Ca agonist MaxiPost (28). Further, we previously demonstrated that S-VX-121 directly affects neuronal action potential firing in primary cortical and hippocampal neurons (12). While not evaluated for R-VX-121 in our present study, given the expression of BK_Ca in neurons, and our demonstration that R-VX-121 directly potentiates BK_Ca, it seems likely R-VX-121 would also affect neuronal function. These potential off-target effects could theoretically be mitigated by developing an R-VX-121 analogue that remains in the airway following aerosolization. A second, and perhaps more concerning, caveat is our demonstration that R-VX-121 directly competes with S-VX-121 and VX-445 to mitigate their F508del CFTR corrector effects. This was confirmed by inhibition of the S-VX-121- and VX-445-induced I_Cl response to forskolin (Figs 1 F-I), as well as the loss of band C upon IB (Fig 1 D, E). This is a critical observation, as our results would suggest R-VX-121 is capable of reverting corrected F508del CFTR-expressing cells to a CF phenotype. Importantly, the IC₅₀ for this R-VX-121 effect (3.5 μM, Fig 1G) overlaps with the affinities we demonstrate for R-VX-121 potentiation of I_K in HBEs (11 μM, Fig 2D) and potentiation of α-BK_Ca expressed in HEK cells (2.8 μM, Fig 3E). Thus, our studies suggest the concentration required to potentiate BK_Ca would be detrimental to the correction of F508del CFTR in CF airway.

While our data suggests R-VX-121 has the potential for unwanted effects in CF airway, our results demonstrating an increased specificity for BK_Ca over CFTR suggests it may be beneficial in other diseases where potentiation of BK_Ca would be therapeutically useful. For example, similar to CF, COPD is characterized by impaired mucociliary clearance suggesting the stimulation of BK_Ca by R-VX-121, and the potential for increased ASL volume, may be beneficial. Indeed, BK_Ca modulators were developed for the possible treatment of bronchial asthma, allergic rhinitis and COPD, although they did not end up in clinics (8). The pharmaceutical industry has also invested significant efforts in developing BK_Ca modulators for ischemic stroke, overactive bladder, urinary incontinence, erectile dysfunction and bowel disorders (8). Finally, there are a number of BK_Ca channelopathies that have been described (29), suggesting BK_Ca modulators such as VX-121 may have additional therapeutic potential in the associated diseases.

Limitations

While we demonstrate S- and R-VX-121 do not potentiate CFTR using transepithelial measurements, we cannot rule out that patch-clamp studies may reveal a subtle effect of these compounds. Additionally, we only evaluated 10 μM VX-121 as this is the concentration that produces a maximal activation of BK_Ca and thus we cannot address whether other concentrations may potentiate CFTR. We demonstrate a partial reversal of vasorelaxation with paxilline, confirming a role for BK_Ca. However, we cannot rule out VX-121 effects on other channels playing a role in this response, including KCa3.1 and KCa2.3, which we have shown are inhibited by VX-445 and VX-121 (13).

Perspectives and Significance.

Our results provide the first demonstration that the R-enantiomer of VX-121 directly potentiates BK_Ca. Our results are also the first to demonstrate R-VX-121 fails to correct or potentiate CFTR. Importantly, our results demonstrate R-VX-121 prevents both S-VX-121 and VX-445 from correcting the folding and trafficking of F508del CFTR. As BK_Ca channels have long been a target for drug discovery efforts, furthering our understanding of the mechanism by which CFTR correctors potentiate BK_Ca channel activity is a critical step in advancing their therapeutic potential

Data availability:

Data will be made available upon reasonable request.