Normal primary HBEC (Wildtype; WT) from two donors (passage 1) were obtained from the University of Alabama CF Research and Translation Core Center under a Material Transfer Agreement with the University of Florida (UAB A-302178; IRB 00000726). Primary HBEC with the homozygous Class II mutation F508del^+/+ were harvested from a donor lung (Pediatric tissue and data bank; IRB #201602392) under regulations of the Pediatric and Adult Pulmonary Center at the University of Florida. Primary HBEC (passage 2) with the heterozygous Class I mutations G542X/R785X and W1282X/R1162X were received from Dr. S. Randell, Marsico Lung Institute, University of North Carolina, Chapel Hill, NC, United States, and primary HBEC (passage 2) with the homozygous Class I mutation G542X^+/+ were a gift from Dr. B. Bridges, Rosa Franklin University, Chicago, IL, United States. Because the cells were obtained from deceased individuals with minor, de-identified information, their use does not constitute human subjects research as defined by CFR 46.102. The providers assured that the cells were acquired in accordance with the Common Rule, and the projects were approved by the Institutional Review Boards of the University of Alabama, the University of North Carolina, and Rosa Franklin University. A written consent was obtained before procurement of the cells. All experiments were performed by the guidelines and regulations described by the Declaration of Helsinki and the Huriet-Serusclat and Jardet law on human research ethics, and the protocols to obtain, culture, store, and study HBEC were approved by the Institutional Review Board of the University of Florida (IRB #201602392).

Primary HBEC were expanded in expansion media (PneumaCult Ex Plus; StemCell Technologies, United States), and cells of passage 3 were seeded on collagen IV-coated (Sigma, United States) permeable snapwell or transwell inserts (12 or 24 mm, 0.4 µM pore polyester membrane; Corning Costar, United States) at 5 × 10⁵ cells·cm⁻² in expansion medium containing 1% penicillin/streptomycin and kept at 37°C and 5% CO₂. After cells reached 95% confluence, cells were differentiated in PneumaCult ALI medium (StemCell Technologies, United States) containing 1% penicillin/streptomycin and 2% Ultroser G (Crescent Chemical Co., United States) at an air-liquid interface. The ALI medium was changed every 2 days until cells were fully differentiated (25–30 days). Differentiated HBEC, characterized by cilia motility, were either treated with DMSO (max. of 0.05%) or with a combination of the CFTR correctors VX661 (tezacaftor, 18 μM; MedChemExpress, United States, cat # HY-15448), and VX445 (elexacaftor, 3 μM; cat # HY-111772), and the potentiator VX770 (ivacaftor, 1 μM; cat # HY-13017) for 24 h prior to the experiments (Keating et al., 2018).

Differentiated WT- and CF-HBEC growing on snapwell inserts were mounted in pre-warmed, calibrated Ussing chambers (VCC MC8; Physiologic Instruments, United States). Each side was bathed in 5 mL Ringer’s solution containing (mM): 113.8 Na⁺, 93.6 Cl⁻, 25 HCO₃ ⁻, 5.2 K⁺, 2.4 HPO₄ ⁻, 0.4 H₂PO₄ ⁻, 1.2 Mg²⁺, 1.2 Ca²⁺, and 75 mannitol, with an osmolarity of 300 mOsm and a pH of 7.4. The basolateral bathing solution contained 5 mM of glucose. Chambers were bubbled with 95% O₂ and 5% CO₂ and maintained at 37°C. After a 30-min equilibration period, transepithelial short circuit current (I _sc, in µeq·h⁻¹·cm⁻²) and transepithelial electrical resistance (TEER) were recorded every 30 s while continuously clamping the membrane potential to zero. Voltage clamping eliminated passive ion movements across the membranes due to electrochemical potential gradients, osmotic gradients, and hydrostatic forces. Any current recorded under zero membrane potential was due to active ion movements, represented as I _sc.

To measure the anion current, apical ENaC activity was blocked using 6 µM benzamil (Tocris, United States) for 15 min, and the benzamil-insensitive I _sc was recorded. In one set of experiments, CFTR was stimulated using 10 µM forskolin (Cayman Chemical, United States) added to the apical and basolateral sides of the Ussing chambers. This was followed by sequential inhibition of CFTR using 20 µM CFTRinh-172 (Tocris, United States), and anoctamin (ANO1) inhibition using 10 µM CaCCinh-A01 (Tocris, United States) added to the apical and basolateral sides of the Ussing chambers. Any remaining anion current was inhibited using 20 µM bumetanide (Tocris, United States), a Na–K–2Cl (NKCC1) cotransporter inhibitor added to the basolateral sides. Bumetanide prevents basolateral uptake of Cl^−, which is essential for its apical exit via anion channels. In another set of experiments, the contribution of SLC26A9 to the benzamil-insensitive anion I _sc was evaluated by adding the specific SLC26A9 inhibitor, S9A13 (10 µM), to the apical side of the chamber before sequential inhibition of CFTR, ANO1, and NKCC1. The SLC26A9 inhibitor was kindly provided by Dr. Wan Namkung and Dr. Ikyon Kim from the College of Pharmacy, Yonsei Institute of Pharmaceutical Sciences, Yonsei University, Incheon, South Korea (Jo et al., 2022). An additional set of cells that did not receive forskolin or S9A13 during the initial 15-min recording of the benzamil-insensitive I _sc served as controls for the response patterns. Changes in I _sc responses (delta I _sc) were calculated by subtracting the I _sc recordings taken after 15 min of inhibitor or stimulator addition from those taken before their addition.

In the initial experiments, CF-HBEC (F508del^+/+ and G542X^+/+) were exposed to single AAs added to the apical and basolateral side of the Ussing chambers at a concentration of 8 mM. AAs were ranked based on their effect on benzamil-sensitive, benzamil-insensitive, forskolin-stimulated, CFTRinh-172-, CaCCinh-A01-, and bumetanide-sensitive I _sc. The selected AAs that showed the highest increase in benzamil-insensitive anion I _sc and a decrease or unchanged benzamil-sensitive I _sc were L-glycine, L-cysteine, L-proline, L-tyrosine, and L-lysine. These AAs were further evaluated for their transport characteristics. The response of the selected AAs on benzamil-insensitive I _sc was maximized by optimizing the AA concentrations using saturation kinetic studies in CF-HBEC. Briefly, HBEC were bathed in Ringer’s solution, ENaC was blocked with benzamil, and increasing concentrations of AAs were added to the apical side at 3-min intervals, until benzamil-insensitive I _sc saturated. Mannitol was added to the basolateral side at the same concentration to compensate for any osmotic gradients. The maximum effect of AAs on benzamil-insensitive I _sc (V _max), the Hill constant (K _M), and the Hill coefficient n were used to evaluate individual AA performance, AA transport characteristics, and optimal AA concentrations for maximum benzamil-insensitive I _sc.

After optimizing the concentrations, the selected AAs (SAA) were formulated as follows (mM): 30 L-glycine, 22.5 L-cysteine, 15 L-proline, 1.2 L-tyrosine, and 1 L-lysine (Ajinomoto, United States), diluted in Ringer’s solution containing (mM): 113.8 Na⁺, 93.6 Cl⁻, 25 HCO₃ ⁻, 5.2 K⁺, 2.4 HPO₄ ⁻, 0.4 H₂PO₄ ⁻, 1.2 Mg²⁺, 1.2 Ca²⁺ at pH 7.4, and adjusted to 300 mOsm with mannitol. This optimized SAA formulation was subsequently utilized in the experiments described in this study.

Chloride secretion was evaluated using unidirectional and net ³⁶Cl isotope flux experiments in Ussing chambers with the membrane potential continuously clamped to zero. Transepithelial net Cl⁻ flux (J _netCl) was calculated for WT-HBEC and CF-HBEC with various mutations. Unidirectional fluxes for Cl⁻ from the apical-to-basolateral (J _ms) and the basolateral-to-apical sides (J _sm) were measured, and the net ion flux across the cell culture was calculated using the formula J _msCl - J _smCl = J _netCl. The results were expressed as µeq·h⁻¹·cm⁻². After blocking ENaC activity with 6 µM benzamil, HBEC were paired based on a similar TEER. Chloride isotopes (³⁶Cl) were added to either the basolateral or apical side (hot side) of each pair, and samples were taken every 15 min from the contralateral sides (cold side) and replaced by fresh bath solution. ³⁶Cl activities were analyzed in a liquid scintillation counter (LS6500 Multipurpose Scintillation Counter, Beckman Coulter).

Differentiated WT- and CF-HBEC (F508del^+/+, and G542X/F785R) were seeded on glass coverslips in ALI medium, maintained for 24–48 h, and then treated with DMSO or VX for 24 h before electrophysiological recordings. Coverslips were placed on the stage of an inverted microscope with a bath perfusion chamber. The internal pipette solution consisted of (mM): 20 KCl, 120K-Gluconate, 5 HEPES, 0.2 EGTA, 0.5 MgCl₂. The bath Ringer’s solution consisted of (in mM): 113.8 Na⁺, 93.6 Cl⁻, 25 HCO₃ ⁻, 5.2 K⁺, 4.8 HPO₄ ⁻, 0.4 H₂PO₄, 1.2 Mg²⁺, 1.2 Ca²⁺ and 75 mannitol, which was bubbled with 95% O₂ and 5% CO₂. The electrolyte mixture for the AA formulation (SAA) consisted of (mM): 30 L-glycine, 22.5 L-cysteine, 15 L-proline, 1.2 L-tyrosine, and 1 L-lysine, diluted in Ringer’s solution containing (mM) 113.8 Na⁺, 93.6 Cl⁻, 25 HCO₃ ⁻, 5.2 K⁺, 2.4 HPO₄ ⁻, 0.4 H₂PO₄ ⁻, 1.2 Mg²⁺, 1.2 Ca²⁺ at pH 7.4, and adjusted to 300 mOsm with mannitol. The current density was calculated as the amount of current (in picoamp; pA) carried through the cell membrane after normalizing to the cell size (in picofarad; pF) and expressed as pA/pF. To evaluate the amount of CFTR current blocked by CFTRinh-172 (10 µM), the current density was set at minus 80 mV, and sets of recordings were performed using a voltage clamp ramp protocol from minus 80 mV to plus 80 mV with a holding potential of minus 10 mV.

Differentiated WT- and CF-HBEC grown on 12 mm transwell filters were treated with DMSO or VX for 24 h, followed by apical exposure to Ringer’s solution or SAA for 4 h at 37°C and 5% CO₂. After incubation, the cells were washed with ice-cold PBS and harvested from the inserts using a cell scraper. The cell suspension was incubated in 1 mM dithiothreitol diluted in PBS for 20 min, then washed three times with ice-cold PBS, with each wash followed by centrifugation at 1,200 × g for 5 min. After the final centrifugation, lysis buffer (RNeasy Mini Kit; Qiagen, United States) was added to the cell pellet and vortexed for 2 min before further processing according to manufacturer’s instructions. The mRNA concentration was determined by using the spectrophotometric nanodrop method (Agilent BioTek Epoch, United States), and cDNA synthesis was performed using the iScript cDNA Synthesis Kit (Bio-Rad, United States) following manufacturer’s instructions. A total of 1,000 ng of mRNA was reverse transcribed to cDNA using a cDNA Synthesis kit (BioRad, United States). Quantitative PCR reactions were performed in duplicates from four independent experiments using 10 ng of mRNA-equivalent cDNA, the CFX-Connect Real-Time PCR detection system (Bio-Rad, United States), SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, United States), and primers (250 nM) at their optimum melting and annealing temperatures. The following primers were used: SLC26A9 (FW: CGT​GGT​AGA​CAG​AGC​CGC​AT and Rev: GGC​TGA​GGA​ACA​TCT​GAA​GGC), ANO1 (FW: GAG​CCA​AAG​ACA​TCG​GAA​TCT​G, and Rev: TGA​AGG​AGA​TCA​CGA​AGG​CAT), CFTR (FW: GCG​AAG​ATC​TTG​CTG​CTT​GA, and Rev: CTG​CCG​CAC​TTT​GTT​CTC​TT), and RPS13 (FW: CGA​AAG​CAT​CTT​GAG​AGG​AAC​A, and REV: TCG​AGC​CAA​ACG​GTG​AAT). To identify the read-through mRNA, a CFTR primer at the 3′end of all the mutations was used. The cycle numbers at the linear range of the amplification curve were used for the calculation. Relative quantification of mRNA expression was determined using the Pfaffl method (Pfaffl, 2001), with RPS13 as the reference gene. The mRNA levels were expressed as fold change relative to WT-HBEC for all genes studied. Melt peaks were analyzed to confirm specific primer binding.

Differentiated WT- and CF-HBEC grown on 12 mm transwell filters were treated with DMSO or VX for 24 h, followed by apical exposure to 200 µL of Ringer’s solution or SAA for 1 hour at 37°C and 5% CO₂. Cells were then fixed in 4% paraformaldehyde and embedded in paraffin. Cross-sections (4 µm) were mounted on silane-coated glass slides, deparaffinized, rehydrated, and heat pre-treated in retrieval buffer at pH 6.0 (Biocare Medical, United States) as previously described (Sasidharan et al., 2024). After blocking with Dako Protein Block, sections were incubated overnight at 4°C with the following antibodies: mouse monoclonal anti-CFTR (clone 596; Dr. Martina Gentzsch, CFTR Antibody Distribution Program, cat # 596; RRID: AB_2923486), rabbit polyclonal anti-TMEM16 A (Abcam, cat # ab53212; RRID: AB_883075) and rabbit polyclonal anti-SLC26A9 (LSBio, cat # LS-C682325; RRID: AB_2923487) diluted in Dako Antibody Diluent (1:100). Secondary antibodies, goat anti-mouse superclonal recombinant antibody conjugated with AlexaFluor488 (ThermoFisher, cat # A28175; RRID: AB_2536161) or goat anti-rabbit superclonal recombinant antibody conjugated with AlexaFluor647 (ThermoFisher, cat # A27040; RRID: AB_2536101) were used at a concentration of 1 μg/mL incubated for 1 hour at room temperature. Nuclei were stained with DAPI for 15 min, and cells were mounted in an aqueous mounting medium before analysis. Signals were analyzed at ×600 magnification using the Laser Scanning Olympus Fluoview FV1000 confocal microscope.

To simulate inhalation therapy as a treatment option for SAA-mediated epithelial cell secretion, 200 µL of Ringer’s solution or SAA was nebulized onto the surface of WT-HBEC and Class I-HBEC (G542X/R785X) culture inserts using the Vitrocell Cloud exposure system (Vitrocell, Germany), as described by Chung et al. (2019). Based on the treatment volume, the nebulizer surface area (approx. 116.5 cm²), the filter surface area (1.12 cm²), and the AA concentrations in SAA (70 mM), approximately 20–30 mM of AAs were deposited onto each culture filter. Immediately after nebulization (0 h) and 4 h later, high-resolution images of the cell cultures were used to estimate ASL volumes using meniscus scanning (Harvey et al., 2011). Briefly, the HBEC culture plate was placed on an Epson scanner without the lid, at room temperature, and with a humidity of at least 50%–60% to prevent dehydration during recording. Data were analyzed using ImageJ and Dr. Myerburg’s software. ASL volume was estimated based on the meniscus light refraction, and delta ASL at 4 h was calculated, representing the difference between the 0-h and 4-h measurements, with a positive value indicating ASL secretion.

Ciliary beat frequency of nebulized WT- and Class I-HBEC (G542X/R785X) was recorded at 0 h (immediately after nebulization) and 4 h post-nebulization using a high-speed Basler acA645 camera (Basler, Ahrensburg, Germany) mounted on a Zeiss Axiovert 200 M (Carl Zeiss, Jena, Germany) running SAVA software (Sisson et al., 2003), as previously described (Salathe and Bookman, 1999). Briefly, the HBEC culture plate was mounted on the microscope stage at room temperature, and five videos were recorded for each culture at designated time points.

WT- and Class I-HBEC (G542X/R785X) grown on 12 mm transwell inserts were nebulized with 200 μL Ringer’s solution or SAA 4 h before recording MCT (Sears et al., 2015). Briefly, 10 µL of 1-µm fluorescently labeled carboxylate-modified polystyrene microbeads (Fluospheres; ThermoFisher, United States) were added to the culture surface at a 1:10,000 dilution. The velocity of the beads was recorded every 3 s for 30 s at an emission of 515 nm. The MCT speed was estimated (μm/s) using the Manual Tracking ImageJ plugin.

Results are presented as mean ± standard error of the mean (SEM), and analyses were performed using the OriginPro 2021 software package. For saturation kinetics, OriginPro software was used to calculate K _M, V _max, and the Hill coefficient n via the Hill1 equation. The Shapiro-Wilk test assessed normal distribution for each treatment group. If normally distributed, a two-sample t-test compared the overall effect of Ringer’s solution and SAA on measured parameters, while a paired t-test compared values before and after adding inhibitors, and ASL and CBF values between two time points. For multi-sample comparisons (Ringer, SAA, and VX), one-way ANOVA with post hoc Bonferroni was used for pairwise comparison. For non-normally distributed samples, the Kruskal–Wallis test followed by the Mann-Whitney U test was used. A P-value <0.05 was considered statistically significant.

Twenty AAs were screened at 8 mM concentration by adding respective AAs to the apical and basolateral side of CF-HBEC (F508del^+/+, and G542X^+/+) mounted in Ussing chambers and ranked based on benzamil-sensitive and -insensitive I _sc (Figures 1A,B). The absence of functional CFTR disrupts anion secretion, leading to hyperpolarization of the HBEC. This hyperpolarization enhances ENaC-mediated electrogenic cation I _sc as the cell attempts to counteract the lost anion function. Since hyperpolarization is essential for activating potential anion channels, it is critical to maintain a low ENaC-mediated cation I _sc to ensure effective anion secretion. CF-HBEC with the F508del^+/+mutation were used to screen AAs based on their ability to increase CFTR or other anion channel-mediated I _sc and decrease benzamil-sensitive (ENaC) I _sc. CFTR and/or anion channel activity was increased using forskolin. Because the basolateral Cl⁻ uptake mediated by the Na-K-2Cl cotransporter 1 (NKCC1) is crucial for apical anion secretion in HBEC, inhibiting NKCC1 with bumetanide (Figures 1C,D) indirectly measures apical anion channel activity. In addition, CF-HBEC with the Class I mutation G542X^+/+ were used to rank individual AAs based on their effects on both benzamil-sensitive and -insensitive I _sc. These cells were also employed to assess changes in I _sc resulting from sequential inhibition of CFTR, ANO1, and NKCC1 using CFTRinh-172, CaCCinh-A01, and bumetanide, respectively (Supplementary Figures S1A–E). Subsequently, the AAs proline, glycine, cysteine, lysine, and tyrosine were selected for their ability to decrease benzamil-sensitive I _sc and increase the anion I _sc.

To evaluate the AA concentration-dependent response on benzamil-insensitive I _sc, increasing concentrations of selected AAs (cysteine, glycine, proline, lysine and tyrosine) were added to the apical side of CF-HBECs (F508del^+/+, and G542X^+/+) at 3-min intervals. As a result, a dose-dependent increase in benzamil-insensitive I _sc was observed for cysteine, glycine, proline, and lysine in both CF mutations (Figure 2). Due to tyrosine’s limited solubility its response did not exhibit a dose-dependent saturable kinetic profile. Cysteine and glycine exhibited early signs of saturation at lower concentrations, yet I _sc continued to increase with higher AA concentration. This biphasic pattern suggests involvement of more than one apical AA transporter. In contrast, the AAs proline and lysine showed saturation consistent with a single transporter (Figures 2C,D,G,H). The early saturation therefore represents a high-affinity transporter with a low K _M and low Vmax, whereas continued increases at higher AA concentrations reflect a secondary low-affinity transporter with high K _M and high V_max. The sigmoidal nature of these saturation curves is characteristic of ion-coupled transporters and allosteric enzyme kinetics. Kinetic parameters—K _M, V_max, and the Hill coefficient n—were derived using the Hill1 equation (Lolkema and Slotboom, 2015). The hill coefficient n provides additional information on AA-transporter binding characteristics with n > 1 suggesting two or more binding sides or protein subunits, and positive cooperativity for substrate binding (binding of one substrate facilitates binding of another substrate). In contrast, n = 1 refers to a single substrate binding site or multiple binding sites that do not interact cooperatively, and n < 1 indicates negative cooperativity for substrate binding.

Based on saturation kinetics, cysteine appears to utilize at least two apical transport systems in both mutations (Figures 2A,E); Supplementary Figures S2A,B and S3A,B). In F508del^+/+-HBEC, a high-affinity transporter with a K _M of 0.12 mM, V _max of 0.2 µeq·h⁻¹·cm⁻², and n of 0.45 suggested a single substrate binding side with negative cooperativity for multiple substrate bindings. The second low-affinity transporter with a K _M of 32.3 mM and V _max of 0.56 µeq·h⁻¹·cm⁻² and n of 2.15 indicated a positive substrate binding cooperativity or ion-coupled transport mechanism with two or more substrate binding sides (Supplementary Figures S2A,B).

Glycine transport showed similar properties (Supplementary Figures S2C,D: for F508del^+/+; Supplementary Figures S3C,D: for G542X^+/+) with one high-affinity transporter at K _M of 0.07 mM, V _max of 0.2 µeq·h⁻¹ cm⁻² and n of 0.7 in F508del^+/+-HBEC, while the second transporter had low-affinity binding features with K _M of 30.5 mM, V _max of 0.6 µeq·h⁻¹ cm⁻² and n of 2.3 at higher glycine concentrations suggesting the involvement of more than one apical AA transporter for glycine. In contrast, the proline transporter required lower substrate concentrations than cysteine and glycine at similar V _max (F508del^+/+: 0.6 µeq·h⁻¹ cm⁻² with K _M of 10.8 mM and n of 2.8 and G542X^+/+: 0.4 µeq h⁻¹ cm⁻² with K _M of 13.3 mM and n of 2.1) indicating a different transport mechanism with positive substrate-binding cooperativity and potentially two or more binding sides (Figures 2C,G). The response of lysine was mediated at low concentrations through a high-affinity, positive cooperative substrate binding transporter with a K _M of 0.14 mM, V _max of 0.2 µeq·h⁻¹ cm⁻², and n of 1.5 in F508del^+/+-HBEC and K _M of 0.09 mM, V _max of 0.2 µeq h⁻¹ cm⁻², and n of 1.4 in G542X^+/+-HBEC. Thus, lysine contributed only marginally to the overall anion-secretory response of SAA (Figures 2D,H: for both mutations). Due to its low solubility, tyrosine saturation kinetics were not performed. The concentrations of AAs used in the formulation were selected based on obtaining high anion I _sc (V _max) and solubility of the respective AA. Cysteine, proline, glycine, lysine, and tyrosine were used at 22.5, 15, 30, 1, and 1.2 mM in the formulation.

To evaluate dose-response and determine the K _M and V _max of the formulation, increasing concentrations of the combination of all five AAs (SAA) were applied to the apical side of WT- and CF-HBEC with various mutations. The response followed a sigmoidal saturation curve in WT- (Figure 3A) and CF-HBEC with F508del^+/+ (Figure 3B), G542X/R785X (Figure 3C), and G542X^+/+ (Figure 3D), similar to that observed with cysteine and glycine. At low SAA concentrations, an initial saturation point was observed, consistent with a high-affinity transport mechanism seen with cysteine, glycine, and lysine. With further increases in SAA concentration, a second peak in I _sc was reached, with a V _max of 1.6 µeq h⁻¹ cm⁻² in WT-HBEC and 0.6–0.7 µeq h⁻¹ cm⁻² in CF-HBEC. The final formulation concentration was selected based on the AA concentrations that produced maximal anion secretion (V _max). Notably, the formulation altered K _M in different CF genotypes: K _M decreased in F508del^+/+-HBEC, increased in G542X^+/+-HBEC, and remained unchanged in G542X/R785X-HBEC compared to WT-HBEC. These findings were used to optimize the formulation concentration for maximal anion secretion.

Exposure of both WT- and CF-HBEC to SAA in Ussing chambers increased benzamil-insensitive I _sc and Cl⁻ secretion (Figures 4A,B), and either decreased or did not change benzamil-sensitive I _sc in CF-HBEC irrespective of the CF mutation, compared to Ringer’s solution.

Benzamil-sensitive I _sc (electrogenic Na⁺ absorption) was significantly higher in HBEC with all CF mutations (F508del^+/+: 2.39 ± 0.06 µeq h⁻¹ cm⁻²; W1282X/R1162X: 2.62 ± 0.09 µeq h⁻¹ cm⁻²; G542^+/+: 3.10 ± 0.1 µeq h⁻¹ cm⁻²) except in G542X/R785X-HBEC where electrogenic Na⁺ absorption decreased (1.82 ± 0.06 µeq h⁻¹ cm⁻²; P = 0.002) compared to WT-HBEC (2.14 ± 0.06 µeq h⁻¹ cm⁻²). When CF-HBEC were exposed to SAA, benzamil-sensitive I _sc significantly decreased in G542X/R785X-HBEC (1.66 ± 0.04 µeq h⁻¹ cm⁻², P = 0.033) but remained unchanged in F508del^+/+ (2.24 ± 0.09 µeq h⁻¹ cm⁻²), W1282X/R1162X (2.49 ± 0.08 µeq h⁻¹ cm⁻²), and G542^+/+ (2.95 ± 0.09 µeq h⁻¹ cm⁻²) compared to corresponding Ringer controls, suggesting a modest impact of SAA on electrogenic Na⁺ absorption. A 24-h treatment with VX did not change the benzamil-sensitive I _sc.

In the presence of SAA, the benzamil-insensitive I _sc increased significantly in CF-HBEC compared to their corresponding Ringer values (Figure 4A): F508del^+/+ (Ringer: 0.07 ± 0.004 µeq h⁻¹ cm⁻² vs. SAA: 0.55 ± 0.01 µeq·h⁻¹ cm⁻²), W1282X/R1162X (Ringer: 0.07 ± 0.005 µeq h⁻¹ cm⁻² vs. SAA: 0.56 ± 0.02 µeq h⁻¹ cm⁻²), G542X/R785X (Ringer: 0.07 ± 0.005 µeq h⁻¹ cm⁻² vs. SAA: 0.60 ± 0.02 µeq h⁻¹ cm⁻²) and G542X^+/+ (Ringer: 0.04 ± 0.004 µeq h⁻¹ cm⁻² vs. SAA: 0.49 ± 0.01 µeq h⁻¹ cm⁻²). Although benzamil-insensitive I _sc of CF-HBEC exposed to SAA remained lower than that of WT-HBEC exposed to Ringer’s solution (0.68 ± 0.02 µeq h⁻¹ cm⁻²), it reached 80.4% in F508del^+/+, 81.2% in W1282X/R1162X, 87.5% in G542X/R785X, and 70.8% in G542X^+/+ compared to WT-Ringer controls. In contrast, exposure to VX for 24 h increased benzamil-insensitive I _sc in Class II mutation (F508del^+/+) but not in CF-HBEC with Class I mutations compared to CF-HBEC treated with Ringer (Ringer: 0.07 ± 0.004 µeq h⁻¹ cm⁻² vs. VX: 0.52 ± 0.01 µeq h⁻¹ cm⁻²; Figure 4A), which represents 84.1% to that of WT-HBEC exposed to Ringer’s solution. The increase in benzamil-insensitive I _sc with SAA was further confirmed by unidirectional and net ³⁶Cl flux studies. J _netCl showed a significant increase with SAA in CF-HBEC with different mutations suggesting electrogenic Cl⁻ secretion (Figure 4B ; Table 1).

The elevated benzamil-insensitive I _sc induced by SAA in WT- HBEC (Figure 5A), F508del^+/+-HBEC (Figure 5B) and G542X^+/+-HBEC (Figure 5C) was sequentially inhibited by various anion channel blockers targeting SLC26A9, CFTR, and ANO1 using S9A13, CFTRinh-172, and CaCCinh-A01, respectively. The portion of the benzamil-insensitive I _sc not inhibited by the anion channel blockers was further inhibited by an NKCC1 inhibitor (bumetanide), which prevents the basolateral uptake of Cl⁻, essential for its apical exit. Examining the average current changes for WT-, F508del^+/+-, G542X/R785X, W1282X/R1162X, and G542X^+/+−HBEC showed that exposure to SAA resulted in a higher S9A13 inhibition in CF-HBEC compared to WT-HBEC (Figures 5D–H). In WT-HBEC, addition of SAA increased S9A13-sensitive I _sc to 0.25 ± 0.02 µeq·h⁻¹·cm⁻² compared to Ringer (0.05 ± 0.01 µeq·h⁻¹·cm⁻²), while CFTRinh-172-sensitive I _sc was unaltered. The highest S9A13 inhibition was observed in the G542X/R785X mutation (0.38 ± 0.03 µeq·h⁻¹·cm⁻²) (Figure 5F), followed by F508del^+/+ (0.28 ± 0.01 µeq·h⁻¹·cm⁻²) (Figure 5E), W1282X/R1162X (0.26 ± 0.01 µeq·h⁻¹·cm⁻²) (Figure 5G), and G542X^+/+ (0.24 ± 0.02 µeq·h⁻¹·cm⁻²) (Figure 5H).

To specifically determine the magnitude of CFTR-mediated I _sc in the presence of SAA, in separate experiments, CFTRinh-172 was used in the absence of any other anion channel blocker, both in the absence (Supplementary Figures S4A,C,E, and S5) or presence of forskolin Supplementary Figures S4B,D,F, and S6). CFTRinh-172-sensitive I _sc without prior activation of CFTR was significantly lower in WT-HBEC exposed to SAA when compared to Ringer’s solution but was higher in F508del^+/+-, and G542C^+/+-HBEC (Supplementary Figures S4A,C,E).

These studies demonstrated that basal SAA-mediated CFTR activity was lower in F508del^+/+ (P = 0.002), G542X^+/+ (P < 0.001), and G542X/R785X (P < 0.001) when SLC26A9 was inhibited prior to CFTR inhibition (Figures 5D–H), compared to HBEC without SLC26A9 inhibition (Supplementary Figure S5).

The CaCCinh-A01-sensitive I _sc was low in both WT-HBEC and CF-HBEC when compared to CFTR- and SLC26A9-mediated anion I _sc (Figure 5). However, CF-HBEC showed significantly higher CaCCinh-A01-sensitive I _sc in the presence of SAA when HBEC were not inhibited by S9A13 (Supplementary Figure S5). Similarly, NKCC1-dependent I _sc was lower with prior addition of S9A13 in SAA-exposed CF-HBEC.

The magnitude of the benzamil-insensitive I _sc contributed by various anion channels and the remaining NKCC1-mediated I _sc was not significant between the CF mutations.

Increasing cAMP using the adenylate cyclase activator forskolin increased benzamil-insensitive I _sc in WT-HBEC but not in CF-HBEC (Supplementary Figures S4B,D,F). The addition of 10 µM forskolin to the apical and basolateral sides in Ussing chambers did not further increase the benzamil-insensitive I _sc in CF-HBEC exposed to SAA (Supplementary Figure S6).

To investigate the magnitude of Cl⁻ secretion that is mediated by changes in I _sc, J _netCl was compared to the benzamil-insensitive I _sc in WT-HBEC and HBEC with Class I and Class II mutations (Table 1). The results showed that in HBEC with F508del^+/+, 38.5% of the benzamil-insensitive I _sc was accounted by J _netCl in the presence of SAA, compared to 17.2% in the presence of VX. In HBEC with Class I mutations, exposure to SAA resulted in a significantly higher ³⁶Cl flux compared to VX. In these cells, J _netCl accounted for 44%–57% of the benzamil-insensitive I _sc. These studies demonstrated that a significant portion of the benzamil-insensitive I _sc is because of electrogenic Cl⁻ secretion.

In a different set of experiments, we tested whether Cl⁻ secretion (J _netCl) was responsible for S9A13-sensitive I _sc, in WT-HBEC, and in HBEC with F508del^+/+ and G542X/F785X mutations (Table 2). The studies demonstrated that the increased S9A13-sensitive I _sc in the presence of SAA was associated with a significant increase in Cl⁻ secretion in both WT-HBEC and CF-HBEC. The increased Cl⁻ secretion contributed to 72%, 50%, and 39.5% of the S9A13-inhibitable I _sc in WT-, F508del ^+/+, and G542X/R785X, respectively. This suggests that a significant portion of S9A13-inhibitable I _sc is due to electrogenic Cl⁻secretion but could also be partially attributed to HCO₃ ⁻ secretion.

These findings indicate that SAA increased benzamil-insensitive I _sc by electrogenic Cl⁻ secretion in both Class I and Class II mutations, but electrogenic HCO₃ ⁻ secretion may have contributed to the higher anion I _sc. The majority of SAA-mediated I _sc and J _netCl changes were attributed to increased SLC26A9 activity. The level of anion secretion achieved was comparable to that in normal WT-HBEC. In addition, the SAA-stimulated benzamil-insensitive I _sc exhibited significantly higher electrogenic Cl-secretion when compared to treatment with VX in F508del^+/+, though the total benzamil-insensitive I _sc was higher in the presence of VX.

The study thus far has shown the effectiveness of the SAA in stimulating electrogenic Cl⁻ secretion and potential HCO₃ ⁻ secretion. Whole-cell patch clamp recordings were used to study ion channel activity, and the measurements showed that SAA or VX increased current density, which could be blocked by CFTRinh-172 (Figure 6) or S9A13 (Figure 7) suggesting inhibition of Cl⁻ secretion. Class II HBEC (F508del^+/+) were treated with VX or DMSO for 24 h and then exposed to Ringer’s solution or SAA for 15 min during the recordings. Representative current traces in CF-HBEC exposed to SAA or treated with VX before and after the addition of CFTRinh-172 are shown in Figures 6A,B. Exposure of WT-HBEC to CFTRinh-172 resulted in a significant reduction in current density (−2.37 ± 0.39 pA/pF vs. −1.16 ± 0.25 pA/pF, n = 10, P < 0.05; Figure 6C). When CF-HBEC (F508del^+/+) were exposed to SAA, current density significantly increased from −0.63 ± 0.04 pA/pF to −1.51 ± 0.29 pA/pF (n = 8; P = 0.01; Figure 6C). Total current density decreased in the presence of CFTRinh-172 in SAA- or VX-treated cells, from −1.38 ± 0.39 to −0.64 ± 0.15 pA/pF (n = 4; P = 0.06), and from −0.86 ± 0.21 to −0.55 ± 0.17 pA/pF (n = 6; P = 0.02), respectively, representing CFTR-mediated currents (Figure 6C). In a separate series of experiments, F508del^+/+ -HBEC exhibited no detectable CFTR-activity when exposed to Ringer’s solution (−1.13 ± 0.25 pA/pF vs. −1.26 ± 0.32 pA/pF, n = 4; Figure 6D). These results are consistent with an increase in CFTR function as demonstrated in SAA-exposed CF-HBEC in Ussing chamber experiments.

To test the efficacy of the S9A13 blocker and the amount of SLC26A9-carried current in WT-, F508del ^+/+- and G542X/R785X-HBEC, a series of IV protocols were run for several minutes with an interval of 3 min before and after adding the blocker (Figures 7A,B). Adding S9A13 showed a significant change in SLC26A9-carried current in WT-, F508del^+/+- and G542X/R785X-HBEC perfused with SAA compared to Ringer’s solution (Figures 7C,D). These experiments were repeated on cells perfused with SAA for 15 min and washed with Ringer’s solution for another 15 min before adding the blocker.

In addition, IV protocols performed in G542X/R785X were maintained at a holding potential of minus 20 mV for a total recording of up to 70 min. These experiments were performed in the presence of Ringer’s solution, SAA, and both SAA and the S9A13 blocker. The amount of current was compared at minus 15–20 versus 50–60 mV to highlight a percentage change in current. In G542X/R785X-HBEC, Ringer’s solution showed a 95.6% ± 6% change (n = 6; N.S.); exposure to SAA showed a 203.6% ± 37.6% change, (n = 5; P = 0.045); and treatment with SAA and S9A13 showed a 63.8% ± 9.6% change (n = 5; P = 0.006).

The mRNA expression levels of anion channels such as CFTR, SLC26A9, and ANO1 were measured in CF-HBEC with homozygous F508del^+/+ and G542X^+/+ mutations exposed to SAA or VX (Figures 8A–F). CFTR mRNA levels (Figures 8A,B) were reduced in CF-HBEC compared to WT-HBEC (WT: 1.01 ± 0.06; F508del^+/+: 0.18 ± 0.02; P < 0.001) in Ringer’s solution. However, apical exposure to SAA or treatment with VX increased CFTR mRNA levels compared to untreated CF controls (SAA: 0.28 ± 0.01, P = 0.002; VX: 0.22 ± 0.01, P = 0.04 vs. 0.18 ± 0.02). In G542X^+/+, CFTR mRNA levels were significantly lower compared to WT-HBEC (WT: 0.90 ± 0.06, G542X: 0.14 ± 0.02; P < 0.001; Figure 8B). Addition of SAA to the apical side of CF-HBEC showed a small but significant increase in CFTR mRNA levels (0.18 ± 0.01; P = 0.05). However, in VX-treated G542X^+/+-HBEC, mRNA levels remained unchanged (Figures 8A,B). These studies suggest the activation of a potential read-through mechanism by SAA.

The mRNA expression of SLC26A9 (Figure 8C) was lower in HBEC with the F508del^+/+ mutation compared to WT-HBEC (F508del^+/+: 0.65 ± 0.07; WT: 1.02 ± 0.08, P < 0.007). Treatment with VX significantly increased SLC26A9 mRNA levels and such an increase was not seen with exposure to SAA in HBEC with F508del^+/+ (VX: 1.62 ± 0.08, P < 0.001). SLC26A9 mRNA expression levels in G542X^+/+ were significantly higher compared to WT-HBEC (WT: 1.02 ± 0.07; G542X ^+/+: 2.06 ± 0.13; P < 0.001). Exposure to SAA showed a 3.16-fold increase in SLC26A9 mRNA levels when compared to WT-HBEC, and this increase was significantly higher than in G542X ^+/+-HBEC exposed to Ringer’s solution (3.16 ± 0.57, P < 0.03). VX-treated G542X ^+/+-HBEC did not show a similar increase in SLC26A9 mRNA (Figure 8D).

In contrast, ANO1 mRNA levels (Figures 8E,F) were increased in both F508del^+/+-, and G542X^+/+-HBEC when compared to WT-HBEC (WT: 1.02 ± 0.08; F508del^+/+: 2.98 ± 0.29, P < 0.001; G542X^+/+: 9.31 ± 0.80, P < 0.001). Treatment with VX and exposure to SAA further increased ANO1 mRNA levels in F508del-HBEC (VX: 3.94 ± 0.27, P = 0.03; SAA: 4.51 ± 0.22, P < 0.007), but such an increase was not observed in G542X^+/+-HBEC.

Immunofluorescence imaging for CFTR and SLC26A9 protein expression in WT-HBEC cultures showed a strong signal along the apical membranes of both ciliated and non-ciliated cells (Figure 9A; Supplementary Figure S7). Considering the low abundance of ionocytes in bronchial epithelium (<1%), we did not specifically differentiate this cell type from other non-ciliated cells by immunofluorescence labeling. However, cell morphology and localization of single, positively labeled cells indicate the presence of CFTR in this cell type. CFTR-positive signals were also present within or near nuclei, a phenomenon previously described due to CFTR protein expressed within the endoplasmic reticulum, due to non-specific staining of nuclear protein, or contaminating mouse Ig when using the monoclonal mouse antibody clone 596 (van Meegen et al., 2011; Lukasiak and Zajac, 2021; Sato et al., 2021). SLC26A9 was co-expressed with CFTR but of a lower intensity. Additionally, SLC26A9 was independently present within tight junctions, separate from CFTR (Figure 9A). In CF-HBEC, CFTR membrane expression was significantly reduced in F508del^+/+-HBEC and absent in G542X^+/+-HBEC, although some SLC26A9 signals remained visible at the apical membrane. Exposure to SAA in HBEC with Class I or II mutations resulted in notably higher CFTR and SLC26A9 expression at the apical membrane. Additionally, increased SLC26A9 signals were observed independently of CFTR at the apical membrane. VX treatment enhanced CFTR membrane expression in HBEC with F508del^+/+ but showed no CFTR signal in G542X^+/+. These findings suggest that SAA elevates CFTR and SLC26A9 protein expression at the apical membrane of CF-HBEC and may promote new synthesis of both proteins, regardless of the mutation type. In WT-HBEC, ANO1 protein was abundantly expressed along the apical membrane of both ciliated and non-ciliated cells (Figure 9B). Exposure to SAA, but not VX treatment, increased ANO1 protein levels in HBEC with Class I and II mutations. Further studies are necessary to determine the significance of the increased ANO1 protein levels in CF-HBEC, as its expression along the apical membrane was reduced in immunofluorescence images. However, the CaCCinh-A01-sensitive I _sc showed a small but significant increase with SAA treatment in all CF mutations.

Airway surface liquid, CBF, and MCT are critical for clearing inhaled particles and pathogens from the airways—functions that are compromised in individuals with CF. To simulate inhalation therapy, we nebulized either Ringer’s solution or SAA onto the apical surface of wildtype- and Class I CF-HBEC (G542X/R785X). Four hours post-exposure, ASL levels were elevated in both WT- and CF-HBEC exposed to SAA compared to those exposed to Ringer’s solution (Figures 10A–D). A similar trend was observed in CBF, which also increased in CF-HBEC following SAA exposure (Figures 10E–H). Summary data for ASL and CBF responses indicate that, while ASL was higher in SAA-treated WT-HBEC than in those exposed to Ringer (Figure 10I), CBF was reduced in WT-HBEC following SAA exposure yet increased in CF-HBEC under the same conditions (Figure 10J). Additionally, MCT was higher in CF-HBEC exposed to either Ringer or SAA compared to WT-HBEC under both treatment conditions (Figure 10K). These results contrast with the well-established mucociliary dysfunction typically seen in CF and warrant further investigation. Enhancing ASL, CBF, and MCT could offer significant therapeutic benefits and improve quality of life for patients with CF.