Additive Potentiation of R334W-CFTR Function by Novel Small Molecules

The R334W (c.1000C>T, p.Arg334Trp) is a rare cystic fibrosis (CF)-causing mutation for which no causal therapy is currently approved. This mutation leads to a significant reduction of CF transmembrane conductance regulator (CFTR) channel conductance that still allows for residual function. Potentiators are small molecules that interact with CFTR protein at the plasma membrane to enhance CFTR-dependent chloride secretion, representing thus pharmacotherapies targeting the root cause of the disease. Here, we generated a new CF bronchial epithelial (CFBE) cell line to screen a collection of compounds and identify novel potentiators for R334W-CFTR. The active compounds were then validated by electrophysiological assays and their additive effects in combination with VX-770, genistein, or VX-445 were exploited in this cell line and further confirmed in intestinal organoids. Four compounds (LSO-24, LSO-25, LSO-38, and LSO-77) were active in the functional primary screen and their ability to enhance R334W-CFTR-dependent chloride secretion was confirmed using electrophysiological measurements. In silico ADME analyses demonstrated that these compounds follow Lipinski’s rule of five and are thus suggested to be orally bioavailable. Dose–response relationships revealed nevertheless suboptimal efficacy and weak potency exerted by these compounds. VX-770 and genistein also displayed a small potentiation of R334W-CFTR function, while VX-445 demonstrated no potentiator activity for this mutation. In the R334W-expressing cell line, CFTR function was further enhanced by the combination of LSO-24, LSO-25, LSO-38, or LSO-77 with VX-770, but not with genistein. The efficacy of potentiator VX-770 combined with active LSO compounds was further confirmed in intestinal organoids (R334W/R334W genotype). Taken together, these molecules were demonstrated to potentiate R334W-CFTR function by a different mechanism than that of VX-770. They may provide a feasible starting point for the design of analogs with improved CFTR-potentiator activity.


Introduction
Cystic fibrosis (CF) is a life-shortening autosomal recessive inherited disease that leads to a multiorgan pathology [1,2]. It is caused by mutations in the gene encoding the CF transmembrane conductance regulator (CFTR) protein, an ATP-gated chloride channel activated by protein kinase A (PKA)-dependent phosphorylation [2,3]. CFTR is expressed at the apical membrane of secretory epithelia and plays a fundamental role in fluid and electrolyte movements, thus controlling the composition and amount of epithelial secretions [3,4]. In the airways, loss of CFTR function causes depletion of periciliary liquid and mucus accumulation that along with chronic inflammation and recurrent infections progressively impair lung function [3,4].
Over the last decade, CF therapies have been transformed by the approval of orally bioavailable drugs targeting the root cause of disease-so-called CFTR modulators [2]. The

Generation of a New Cell Line
A new CFBE cell line stably co-expressing the HS-YFP (F46L/H148Q/I152L) and R334W-CFTR was generated. Briefly, the HS-YFP cloned into the pcDNA3.1 expression vector was re-cloned into the lentiviral expression vector pLVX-Puro and then transfected into the HEK-293T cells to produce the lentiviral particles [39]. After 48 h, these particles were harvested and transduced into CFBE cells expressing R334W-CFTR [40]. The efficiency of transduction was assessed by fluorescence microscopy (Zeiss Axiovert 200, Jena, Germany) and then cells were sorted in a FACSAria III cell sorter (BD Biosciences, Franklin Lakes, NJ, USA) to select a homogeneous population with the highest expression of the HS-YFP.

HS-YFP Assay on a Plate Reader
A microplate reader (Tecan Infinite 200 Pro) equipped with high-quality excitation (485 ± 20 ηm) and emission (535 ± 25 ηm) was used for compound screening [40]. Briefly, CFBE cells co-expressing the HS-YFP and R334W-CFTR were plated in 96 well black-walled, clear-bottom microplates (#655090, Greiner Bio-One, Kremsmünster, Austria) at a density of 50,000 cell/well. On the following day, cells were washed twice with phosphate-buffered saline (PBS) and incubated for 30 min with 60 µL PBS with Fsk (5 µM) and test compounds. All plates in the potentiator screen contained wells with Fsk (5 µM) plus positive (1 µM VX-770 or 50 µM Gen) or negative (DMSO) controls. For the co-potentiator analysis, CFBE cells co-expressing CFTR variants were incubated with Fsk (5 µM) plus VX-770 (1 µM), Gen (50 µM), or VX-445 (3 µM) and active LSO compounds for 30 min. The assay consisted of 14-s fluorescence readings: 2-s of the initial fluorescence intensity and 12-s after injection of 100 µL of an iodide-containing solution (PBS with NaCl substituted by NaI, final iodide concentration: 100 mM). The initial iodide influx was computed from fluorescence intensity by single exponential regression [42]. All conditions were performed in triplicate in each microplate.
2.5. In Silico Absorption, Distribution, Metabolism, and Excretion (ADME) Analyses ADME analyses were carried out using the free online software SwissADME (http: //www.swissadme.ch, accessed on 10 October 2022) as previously described [43]. The following physicochemical variables were assessed according to Lipinski's rule of five [44]: molecular weight (MW); number of H-bond donors (nHBD); number of H-bond acceptors (nHBA); number of rotatable bonds (nRB); the calculated logarithm of the octanol-water partition coefficient (LogP); in addition to the topological polar surface area (TPSA). Active compounds were also subjected to filters for the identification of pan-assay interference compounds (PAINS), i.e., "promiscuous" substructures that demonstrate potential response in bioassays independent of the target [45].

Statistical Analyses
All conditions were carried out in at least three independent experiments. Data are represented as mean + standard deviation (SD), and dots depict value for each replicate. GraphPad Prism software version 8.3 (GraphPad Inc., San Diego, CA, USA) was used for all statistical analyses. Statistical comparisons between the conditions tested were assessed using one-way ANOVA followed by Dunnett's or Tukey's posthoc tests, and values of p < 0.05 were considered significant.

Characterization of a New Cell Line Co-Expressing the HS-YFP and R334W-CFTR
The R334 residue is located on transmembrane segment 6 at the outer mouth of the CFTR channel pore ( Figure S1). Compared to WT-CFTR, the R334W mutation similarly exhibited the presence of both core-glycosylated immature (~140 kDa, band C) and fullyglycosylated mature (~180 kDa, band C) forms of CFTR. Furthermore, R334W-CFTR processing was increased by the incubation of cells at low temperature (1.35-fold increase) ( Figure 1A,B).
The R334 residue is located on transmembrane segment 6 at the outer mouth of the CFTR channel pore ( Figure S1). Compared to WT-CFTR, the R334W mutation similarly exhibited the presence of both core-glycosylated immature (~140 kDa, band C) and fullyglycosylated mature (~180 kDa, band C) forms of CFTR. Furthermore, R334W-CFTR processing was increased by the incubation of cells at low temperature (1.35-fold increase) ( Figure 1A  Bright-field, fluorescence, and merged images were acquired to confirm the transduction of the YFP sensor in CFBE cells stably expressing R334W-CFTR ( Figure 1C). These cells were then used for the functional assay to measure the HS-YFP quenching following the addition of iodide-containing solution to the apical surface of cells ( Figure 1D). A decay in cell fluorescence was observed upon Fsk stimulation, which was further increased Bright-field, fluorescence, and merged images were acquired to confirm the transduction of the YFP sensor in CFBE cells stably expressing R334W-CFTR ( Figure 1C). These cells were then used for the functional assay to measure the HS-YFP quenching following the addition of iodide-containing solution to the apical surface of cells ( Figure 1D). A decay in cell fluorescence was observed upon Fsk stimulation, which was further increased by Fsk stimulation together with VX-770 or genistein (1.37-and 1.38-fold increase, respectively) ( Figure 1E,F). Cells incubated at 27 • C for 24 h also demonstrated an increase in cell fluorescence quenching upon Fsk stimulation compared to those cultured at 37 • C (1.76-fold increase).

Identification of Novel Potentiators for the Rescue of R334W-CFTR Function
To assess the potential utility of the investigational compounds as novel potentiators for R334W-CFTR, cells were acutely incubated (30 min) with a test compound together with Fsk before assay ( Figure 2A). Among 46 compounds, four demonstrated a small increase, albeit significant, in the HS-YFP quenching rate compared to DMSO (1.21-to 1.42-fold increase) ( Figure 2B). Furthermore, the effect of LSO-24, LSO-25, LSO-38, and LSO-77 was closely similar to that of VX-770 or genistein. Figure 2C depicts the chemical structure of the active compounds in the screening. The efficacy and potency of these compounds were further assessed by the incubation of cells with various concentrations in the range of 0.03 to 30 µM and the HS-YFP quenching rate was measured ( Figure S2). Table 1 depicts EC 50 and E max for these compounds.
fluorescence quenching upon Fsk stimulation compared to those cultured at 37ºC (1.76fold increase).

Identification of Novel Potentiators for the Rescue of R334W-CFTR Function
To assess the potential utility of the investigational compounds as novel potentiators for R334W-CFTR, cells were acutely incubated (30 min) with a test compound together with Fsk before assay ( Figure 2A). Among 46 compounds, four demonstrated a small increase, albeit significant, in the HS-YFP quenching rate compared to DMSO (1.21-to 1.42fold increase) ( Figure 2B). Furthermore, the effect of LSO-24, LSO-25, LSO-38, and LSO-77 was closely similar to that of VX-770 or genistein. Figure 2C depicts the chemical structure of the active compounds in the screening. The efficacy and potency of these compounds were further assessed by the incubation of cells with various concentrations in the range of 0.03 to 30 µ M and the HS-YFP quenching rate was measured ( Figure S2). Table 1 depicts EC50 and Emax for these compounds.  The theoretical ADME properties of the active LSO compounds were determined by in silico analysis according to Lipinski's rule of five, which states that a small molecule should have MW ≤ 500 Da, nHBD ≤ 5, nHBA ≤ 5, nRB ≤ 10, and LogP ≤ 5 to be orally active. TPSA should also be ≤ 140 Å 2 , since it influences absorption and membrane permeability.  The theoretical ADME properties of the active LSO compounds were determined by in silico analysis according to Lipinski's rule of five, which states that a small molecule should have MW ≤ 500 Da, nHBD ≤ 5, nHBA ≤ 5, nRB ≤ 10, and LogP ≤ 5 to be orally active. TPSA should also be ≤ 140 Å 2 , since it influences absorption and membrane permeability. Table 2 depicts the values for compounds LSO-24, LSO-25, LSO-38, and LSO-77, in addition to those for the reference compound VX-770. All four LSO compounds obey Lipinski's rule of five and have a TPSA < 140 Å 2 , suggesting they can be well absorbed in the gastrointestinal tract. Moreover, during the in silico analysis, these compounds were not recognized as PAINS. To validate the results obtained by the HS-YFP assay on a plate reader, short-circuit current was measured in polarized monolayers of CFBE cells expressing R334W-CFTR to quantitatively assess CFTR-dependent chloride current (Figure 3). Activation of cAMPdependent CFTR-mediated chloride secretion by Fsk promoted a small increase in equivalent short-circuit current (Isc eq ), which was further increased by the subsequent addition of the potentiator LSO-24, LSO-25, LSO-38, LSO-77, or VX-770 (4.1, 4.3, 4.1, 2.6 and 4.7 µA/cm 2 , respectively). These responses were thus inhibited by the addition of CFTRinh-172, indicating they are CFTR-specific. Table 2 depicts the values for compounds LSO-24, LSO-25, LSO-38, and LSO-77, in addition to those for the reference compound VX-770. All four LSO compounds obey Lipinski's rule of five and have a TPSA < 140 Å 2 , suggesting they can be well absorbed in the gastrointestinal tract. Moreover, during the in silico analysis, these compounds were not recognized as PAINS. To validate the results obtained by the HS-YFP assay on a plate reader, short-circuit current was measured in polarized monolayers of CFBE cells expressing R334W-CFTR to quantitatively assess CFTR-dependent chloride current (Figure 3). Activation of cAMPdependent CFTR-mediated chloride secretion by Fsk promoted a small increase in equivalent short-circuit current (Isceq), which was further increased by the subsequent addition of the potentiator LSO-24, LSO-25, LSO-38, LSO-77, or VX-770 (4.1, 4.3, 4.1, 2.6 and 4.7 µ A/cm 2 , respectively). These responses were thus inhibited by the addition of CFTRinh-172, indicating they are CFTR-specific.

Assessment of Potentiator Activity for WT-CFTR
Next, we assessed the ability of active LSO compounds in potentiating WT-CFTR function. In the HS-YFP assay on a plate reader, a further increase in cell-fluorescence quenching rate was observed in WT-CFTR-expressing cells incubated with LSO-77 or VX-770 (1.30-and 1.31-fold increase, respectively) compared to those incubated with DMSO ( Figure 4A). These results were also confirmed in polarized monolayers of CFBE cells expressing WT-CFTR in the Ussing chamber. A negative deflection was observed upon Fsk stimulation that was further increased following the acute application of LSO-77 or VX-770 ( Figure 4B-E). However, LSO-24, LSO-25, and LSO-38 were unable to potentiate WT-CFTR function either in the HS-YFP assay on a plate reader ( Figure 4A) or in Ussing-chamber measurements ( Figure S3). (G) Data are represented as mean (+ SD) increase in Ieq promoted by test potentiator: *P < 0.05 and **P < 0.01 vs. DMSO.

Assessment of Potentiator Activity for WT-CFTR
Next, we assessed the ability of active LSO compounds in potentiating WT-CFTR function. In the HS-YFP assay on a plate reader, a further increase in cell-fluorescence quenching rate was observed in WT-CFTR-expressing cells incubated with LSO-77 or VX-770 (1.30-and 1.31-fold increase, respectively) compared to those incubated with DMSO ( Figure 4A). These results were also confirmed in polarized monolayers of CFBE cells expressing WT-CFTR in the Ussing chamber. A negative deflection was observed upon Fsk stimulation that was further increased following the acute application of LSO-77 or VX-770 ( Figure 4B-E). However, LSO-24, LSO-25, and LSO-38 were unable to potentiate WT-CFTR function either in the HS-YFP assay on a plate reader ( Figure 4A) or in Ussingchamber measurements ( Figure S3).

Assessment of Co-Potentiator Activity for R334W-CFTR in a Cell Line and in Intestinal Organoids
In order to investigate the additivity of potentiator activity for the LSO compounds, cells were acutely incubated (30 min) with LSO-24, LSO-25, LSO-38, and LSO-77 together with VX-770, genistein or VX-445 and the HS-YFP assay on a plate reader was carried out (similarly to the scheme presented in Figure 2A). Upon Fsk stimulation, an increase in fluorescence quenching rate was observed in cells acutely treated with VX-770 in combination with LSO-24, LSO-25, LSO-38, or LSO-77 compared to DMSO (1.91-to 2.15-fold increase) or any compound individually ( Figure 5). However, the combination of LSO compounds with genistein demonstrated an equivalent cell-fluorescence quenching rate to that of genistein alone. VX-445 was unable to increase the cell-fluorescence quenching rate compared to DMSO, and its combination with LSO-24, LSO-25, LSO-38, or LSO-77 displayed effects similar to those of each compound individually ( Figure 5).

Assessment of Co-Potentiator Activity for R334W-CFTR in a Cell Line and in Intestinal Organoids
In order to investigate the additivity of potentiator activity for the LSO compounds, cells were acutely incubated (30 min) with LSO-24, LSO-25, LSO-38, and LSO-77 together with VX-770, genistein or VX-445 and the HS-YFP assay on a plate reader was carried out (similarly to the scheme presented in Figure 2A). Upon Fsk stimulation, an increase in fluorescence quenching rate was observed in cells acutely treated with VX-770 in combination with LSO-24, LSO-25, LSO-38, or LSO-77 compared to DMSO (1.91-to 2.15-fold increase) or any compound individually ( Figure 5). However, the combination of LSO compounds with genistein demonstrated an equivalent cell-fluorescence quenching rate to that of genistein alone. VX-445 was unable to increase the cell-fluorescence quenching rate compared to DMSO, and its combination with LSO-24, LSO-25, LSO-38, or LSO-77 displayed effects similar to those of each compound individually ( Figure 5).

Discussion
This study aimed to screen a collection of compounds to identify novel potentiators for R334W-CFTR, a residual function mutation that is reported to have no to suboptimal susceptibility to VX-770 monotherapy [7,22,23]. Because the four active compounds identified in the functional primary screening only demonstrated a small potentiation of R334W-CFTR function, their additivity with VX-770, genistein, or VX-445 was assessed in a new cell line and then further validated in intestinal organoids. Moreover, LSO-24, LSO-25, LSO-38, and LSO-77 were found to obey Lipinski's rule of five for drug-like molecules [44] and are thus suggested to have good bioavailability.
Similar to WT-CFTR, the R334W mutation enables CFTR to be fully glycosylated and traffic to the plasma membrane. However, in contrast to WT-CFTR, which is a functional protein, R334W-CFTR presents a significantly reduced channel conductance [20,21]. When R334W-expressing cells were shifted from 37ºC to 27ºC, there was an increase in CFTR processing, indicating that this mutant is temperature sensitive, similar to results previously reported for WT-and F508del-CFTR [49][50][51]. Furthermore, the increased cell-

Discussion
This study aimed to screen a collection of compounds to identify novel potentiators for R334W-CFTR, a residual function mutation that is reported to have no to suboptimal susceptibility to VX-770 monotherapy [7,22,23]. Because the four active compounds identified in the functional primary screening only demonstrated a small potentiation of R334W-CFTR function, their additivity with VX-770, genistein, or VX-445 was assessed in a new cell line and then further validated in intestinal organoids. Moreover, LSO-24, LSO-25, LSO-38, and LSO-77 were found to obey Lipinski's rule of five for drug-like molecules [44] and are thus suggested to have good bioavailability.
Similar to WT-CFTR, the R334W mutation enables CFTR to be fully glycosylated and traffic to the plasma membrane. However, in contrast to WT-CFTR, which is a functional protein, R334W-CFTR presents a significantly reduced channel conductance [20,21]. When R334W-expressing cells were shifted from 37 • C to 27 • C, there was an increase in CFTR processing, indicating that this mutant is temperature sensitive, similar to results previously reported for WT-and F508del-CFTR [49][50][51]. Furthermore, the increased cell-fluorescence quenching rate suggests that low-temperature incubation resulted in higher R334W-CFTRdependent anion transport, most likely by increasing the number of CFTR channels at the plasma membrane rather than enhancing channel gating.
Numerous screening campaigns have been carried out to identify CFTR potentiators in libraries of drug-like small molecules, which have resulted in the discovery of several compounds with distinct chemical structures that are able to potentiate CFTR function [11,25,27,29,52]. Nevertheless, few potentiators achieved clinical investigation and up to recently, only the potentiator VX-770 was clinically approved for PwCF carrying specific mutations [5][6][7][8][9][10]. ABBV-974 (formerly GLPG-1837) has the same binding site as VX-770 in the transmembrane domains of CFTR [53,54] and acts by facilitating the channel gating in a phosphorylation-dependent and ATP-independent fashion [12,55]. GLPG-2451 has improved potency [52] but is expected to share the same mechanism as VX-770 and ABBV-974 [27]. A previous study attempted to substitute the amide bioisostere in the VX-770 structure for a 1,2,3-triazole scaffold and demonstrated reduced efficacy and potency in potentiating WT-, F508del-and G551D-CFTR function by this compound series [56]. The active triazole compounds identified in our functional screening (LSO-24, LSO-25, and LSO-38) also demonstrated a suboptimal efficacy and weak potency, but they appear to potentiate R334W-CFTR function by a different mechanism than that of VX-770. Intriguingly, LSO-77, but not LSO-24, LSO-25, or LSO-38, was able to potentiate WT-CFTR function. This compound has an N-tosylpropanimidate instead of a triazole in the chemical structure, which might confer some different pharmacological properties compared to the other three active compounds. Genistein and its analog apigenin also potentiate CFTR channel gating by a distinct mechanism to that of VX-770 [25,29]. They are presumed to bind at the nucleotide-binding domain 1 and 2 interface and promote dimerization, thus accelerating channel opening and slowing down its closure [29]. ASP-11 is another compound that is suggested to potentiate CFTR function by this mechanism [25]. Furthermore, VX-445 (also termed elexacaftor) was initially identified as a CFTR corrector and approved for clinical use in the triple-combo therapy VX-445/VX-661/VX-770 [17], but is now recognized as a compound with dual corrector and potentiator activity [26,57]. Although the mechanism by which VX-445 promotes CFTR potentiation has not been elucidated, it is suggested to be different from that exerted by VX-770 and apigenin [26,58].
Despite the therapeutic success of VX-770 monotherapy [5], it achieves only a partial restoration of the CFTR-gating activity for several mutations, including G551D and F508del [11,24,26], and PwCF still experience a progressive decline of their lung function and pulmonary exacerbations albeit at reduced frequency [14,59]. As different potentiators may promote distinct conformational alterations in the CFTR protein to increase channel gating/conductance [53,54], co-potentiators have emerged as a strategy to enhance CFTR-dependent chloride secretion for mutations that do not respond well to a single potentiator [22,23,25,26]. Combinatorial profiling has shed some light on the mechanism of co-potentiation by clustering compounds into three mechanistic classes [22,26]. Nevertheless, CFTR mutations are not equally responsive to combined potentiators and some mutations were demonstrated to be sensitive (e.g., G551D, G1244E, and S1251N) while others were unresponsive (e.g., R347P, V520F, and L1077P) [22,23,26]. N1303K-CFTR function was also demonstrated to synergistically respond to a triple potentiator combo (VX-770, apigenin, and VX-445) [58]. Our investigational compounds (LSO-24, LSO-25, LSO-38, and LSO-77) demonstrated an additive potentiation of R334W-CFTR function in combination with VX-770, but not with genistein. These results suggest that they may share a common mechanism with genistein. ASP-11 is also suggested to share a common mechanism with genistein, but it was not additive to potentiate R334W-CFTR function in combination with VX-770 [22]. This probably occurred because this mutation is associated with residual CFTR function and the usage of a high concentration of Fsk (20 µM) already saturated CFTR activity [22], thus hindering further potentiation of CFTR-channel activity. Indeed, such results were also observed for our investigational compounds in the FIS assay of intestinal organoids (R334W/R334W genotype). At the lowest concentration of Fsk (0.02 µM), there was a clear difference between the negative control and single potentiators with the co-potentiator treatments. However, upon the incremental concentration of Fsk, such differences could no longer be easily detected due to high residual function. Similar behavior was also reported for D614G, another mutation with residual CFTR function [60], reinforcing thus the need for assessing the concentration of Fsk (or other CFTR activators) when investigating a residual function mutation or when testing a 'highly effective' modulator combination.
Although the R334W mutation causes neither folding nor trafficking impairment, some studies have demonstrated that R334W-CFTR-dependent chloride secretion can be enhanced by increasing the number of R334W-CFTR channels at the plasma membrane with correctors, such as VX-809 and VX-661 [39,40,46,61]. In a recent report, intestinal organoids from an individual with CF (1677delTA/R334W) were incubated with VX-770, VX-809/VX-770, VX-661/VX-770 and VX-445/VX-661/VX-770 [62]. Upon Fsk stimulation, an increase in FIS values was observed with the different modulator combinations but to a comparable extent to that of VX-770 alone [62]. Another report assessed the efficacy of VX-445/VX-661/VX-770 in F508del/R334W intestinal organoid-derived epithelial monolayers and demonstrated rescue of CFTR-mediated anion transport [63]. However, it has been suggested that the triple-combo therapy only modestly improved R334W-CFTR function and most of the gain in function was attributed to F508del-CFTR rescue [63]. While the corrector activity of VX-445 on R334W-CFTR needs to be further elucidated, our data indicate that this compound does not potentiate R334W-CFTR function when used alone or in combination with LSO molecules.

Conclusions
In summary, the present study identified four small molecules that are able to potentiate R334W-CFTR function, a mutation for which no modulator therapy is currently approved. Despite the suboptimal efficacy of LSO-24, LSO-25, LSO-38, and LSO-77, they have different scaffolds from that of previous potentiators and may thus represent a valuable starting point to design analog molecules with improved CFTR potentiator activity. These compounds also demonstrate an additive potentiation of CFTR function in combination with VX-770 in both R334W-heterologously expressing cells and intestinal organoids from a R334W homozygous CF subjects. The additive potentiation effects support the idea that they act by a mechanism distinct from that of VX-770, and their potential utility to enhance CFTR function for other mutations should be exploited in upcoming studies.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented here are available on request from the corresponding author. The data are not publicly available due to privacy and ethical issues.