The SLC26A9 inhibitor S9- A13 provides no evidence for a role of SLC26A9 in airway chloride secretion but suggests a contribution to regulation of ASL pH and gastric proton secretion

The solute carrier 26 family member A9 (SLC26A9) is an epithelial anion transporter that is assumed to contribute to airway chloride secretion and surface hydra-tion. Whether SLC26A9 or CFTR is responsible for airway Cl − transport under basal conditions is still unclear, due to the lack of a specific inhibitor for SLC26A9. In the present study, we report a novel potent and specific inhibitor for SLC26A9, identi-fied by screening of a drug- like molecule library and subsequent chemical modifications. The most potent compound S9- A13 inhibited SLC26A9 with an IC 50 of 90.9 ± 13.4 nM. S9- A13 did not inhibit other members of the SLC26 family and had no effects on Cl − channels such as CFTR, TMEM16A, or VRAC. S9- A13 inhibited SLC26A9 Cl − currents in cells that lack expression


| INTRODUCTION
The epithelial anion transporter SLC26A9 (solute carrier 26 family member A9) is thought to contribute to hydration of the airway surface liquid (ASL) by operating as an Cl − transporter. [1][2][3][4][5][6][7] A recent cryo-EM structure and functional analysis revealed that SLC26A9 operates as an uncoupled chloride transporter with a high turnover rate due to a rapid alternate-access mechanism. 8 Biochemical properties, trafficking, and interaction between SLC26A9 and cystic fibrosis transmembrane conductance regulator (CFTR) are well noticed, but the true contribution of SLC26A9 to airway ion transport remains unclear. Although SLC26A9 operates as a Cl − transporter rather than a Cl − channel, it nevertheless produces substantial Cl − currents that may contribute to basal airway Cl − transport, that is, Cl − transport under resting conditions in nonstimulated airway epithelial cells. 4,9,10,11 Notably, in gastric parietal cells, SLC26A9 was shown to be relevant for acid secretion, probably by operating as a parallel Cl − secretory pathway. 6 CFTR and members of the SLC26A solute transporter family (SLC26A3, 4,6,8,9) physically and functionally interact via R (regulatory) and STAS (sulfate transporter and antisigma factor antagonist) domains. 4,12,13,14,15,16,17 . Transport of bicarbonate (HCO 3 − ) by SLC26A9 has been proposed by some scientists, 3,18,19 but was not found by other laboratories. 1,2,4,8 Loss of SLC26A9 was shown to alter intestinal HCO 3 − transport, acid secretion, and fluid absorption. 5,6 SLC26A9 is inhibited by nonselective chloride channel inhibitors, including 4,4'-diisothiocyanatostilbene-2,2′-disulfonic acid (DIDS), NS3623, flufenamic acid, niflumic acid, and GlyH-101. However, these compounds inhibit a large range of Cl − channels and showed a low potency and only partial inhibition of SLC26A9. 1,2,4 Thus the lack of potent and selective inhibitors of SLC26A9 has hampered further investigations into the physiological function of SLC26A9.
In the present paper, we describe S9-A13, a highly potent and selective inhibitor of SLC26A9. Because we found no cross-inhibition of other SLC26 transporters or chloride channels, it was possible to analyze the contribution of SLC26A9 and CFTR to airway transport. While SLC26A9 does not contribute to airway Cl − secretion, it controls ASL pH, probably by operating as a Cl − / HCO 3 − exchanger. However, in cells that do not express CFTR, SLC26A9 may operate as Cl − transporter and may facilitate proton secretion in parietal cells.

| Animals and treatments
Allergen challenge of mice has been described previously. 20  and CHO-K1 cells expressing ANO1, CFTR, or SLC26A4 were cultured using methods described previously. 21 For SLC26A9 inhibitor screening and selectivity assays,  LN215 cells expressing YFP-F46L/H148Q/I152L together  with SLC26A3, SLC26A6, or SLC26A9 were generated  by a lentiviral system containing the vectors pLenti6P.3-slc26a3, pLenti6P.3-slc26a6, and pLenti6P.3-slc26a9, respectively. Stable cell lines were selected using 1 μg/ml puromycin. Human embryonic kidney 293 (HEK293) cells were grown in DMEM media supplemented with 2 mM l-glutamine. cDNA encoding human SLC26A9 with eGFP fused to its C-terminal was transfected in pcDNA3.1. Cells were transfected using standard protocols for Lipofectamine3000 (Thermo Fisher Scientific, Darmstadt, Germany). All media were supplemented with 10% heat-inactivated fetal calf serum (Capricorn Scientific, Ebsdorfergrund, Germany). BCi-NS1 cells (kindly provided by Prof. R. Crystal, Weill Cornell Medical College, New York, USA) were cultured in supplemented bronchial epithelial cell growth medium (BEGM; Lonza, Basel, Switzerland). BCi-NS1 cells were grown on permeable supports for up to 30 days (Snapwell #3801, Corning, New York, USA) in an air/ liquid interface (ALI). Human nasal cells were obtained from one adult donor using a cytological nasal brush and all procedures were ethically approved by the Institutional Medical Research Ethics Committee of the University Medical Centre, Utrecht (protocol ID: 16/586). The cells were expanded and fully differentiated as previously described. 22 The human gastric tumor cell line HGT-1 was cultured as described earlier. 23

| Cell-based high-throughput
screening and YFP-fluorescence quenching LN215 cells expressing YFP-F46L/H148Q/I152L and SLC26A9 were plated at a density of 25 000 cells per well in 96-well microplates. Each well was washed two times with PBS (200 μl per well) and 50 μl of regular solution was added afterward. Test compounds were applied to each well at 25 μM and plates were incubated for 10 min at 37°C. The 96-well plates were transferred to a FLUOstar Omega Microplate Reader (BMG Labtech, Ortenberg, Germany) to measure SLC26A9 activity. Baseline fluorescence was recorded for 0.8 s, then 50 μl of NaI-substituted regular solution (140 mM NaI replacing 140 mM NaCl) was injected by a syringe pump. Fluorescence was recorded for 8 s. The initial slope of YFP fluorescence was used to analyze SLC26A9-mediated I − flux rate. The YFP fluorescence quenching assay for assessment of SLC26A4, SLC26A6, SLC26A9, CFTR, ANO1, and VRAC activity were performed accordingly, as described in a previous study. 21

| Cell viability assay
A Cell Titer 96® AQueous One Solution Assay kit (MTS) (Promega, WI, USA) was used to determine the effect of compounds on cell viability. Calu-3 cells were cultured in 96-well plates in growth medium supplemented with 10% FBS for 24 h until cell density reached ~40%. Cells were treated with S9-A13 and cisplatin for 24 h followed by MTS analysis according to the supplier's protocol. The absorbance was measured by an Infinite M200 microplate reader (Tecan, Männedorf, Switzerland) at a wavelength of 490 nm.

| Immunocytochemistry
Mouse lung sections were fixed using 4% paraformaldehyde (PFA) and 3.4% sucrose in PBS. Lung sections were deparaffinized with xylene and rehydrated through a series of ethanol. Cells grown on permeable inserts were fixed with 4% PFA in PBS and embedded in paraffin. Sections or cells were incubated with rabbit anti-SCL26A9 antibody (1:100, raised against mouse SCL26A9 aa 11-29, DRAAYSLSLFDDEFEKKDR, Davids Biotechnologie, Regensburg, Germany) overnight at 4°C. Nonspecific binding of the antibody to the airway epithelium was excluded in previous publications showing lack of binding in airways from patients carrying the F508del-CFTR/F508del-CFTR mutation, which causes a lack of apical expression of SLC26A9. 24 Antigen retrieval was performed in preheated Tris-EDTA buffer (pH 9.0) for 15 min, using a microwave before blocking. Cells were incubated with secondary anti-rabbit antibody conjugated with Alexa Fluor 488 or Alexa Fluor 546 (1:400) for 1 h at room temperature. Nuclei were counterstained with 5 μM Hoe33342 (Thermo Fisher Scientific, Darmstadt, Germany). Section or cells were mounted with a fluorescence mounting medium (DAKO Cytomation, Hamburg, Germany). Immunofluorescence was examined with an Axio Observer microscope equipped with Axiocam 503 mono, ApoTome.2, and ZEN 3.0 (blue edition) software (Zeiss, Oberkochen, Germany).

| Airway surface liquid pH and assessment of proton secretory activity
Airway surface liquid pH (ASL pH) was measured using a temperature and CO 2 -controlled plate reader (TECAN SPARK 10 M), as previously described. 28 Briefly, the ASL was stained with 3 μl of a mixture of dextran-coupled pH-sensitive pHrodo Red (0.5 mg/ml, λex: 565 nm, λem: Proton secretory activity and corresponding data analysis were conducted by means of the pH-sensitive fluorescence dye 1,5 carboxy-seminaphto-rhodafluor acetoxymethylester (SNARF-1-AM; Life Technologies), as described previously. 23

| Materials and statistical analysis
All compounds used were of highest available grade of purity. Data are reported as mean ± SEM. Student's t-test (for paired or unpaired samples as appropriate) or ANOVA were used for statistical analysis. A p-value <.05 was accepted as significant difference. For each experimental series, the number of animals used and the number of measurements/assays/reactions is provided.

| S9-A13 inhibits SLC26A9 but shows little effects on CFTR currents in HEK293 cells
HEK293 cells were transfected with SLC26A9-cDNA, which induced strong expression of SLC26A9 protein ( Figure 5A,B). Partially membrane-localized SLC26A9 caused robust constitutive whole cell currents as detected by whole cell patch clamping, which were inhibited by S9-A13 in a dose-dependent manner ( Figure 5C-F). Currents in mock-transfected cells were not affected by S9-A13. Moreover, CFTR currents were only marginally inhibited at the highest concentration of S9-A13 (5 µM), and only at strongly depolarized clamp voltages ( Figure 5G,H). Notably, SLC26A9 currents were not further stimulated by increase in intracellular cAMP using IBMX/forskolin ( Figure 5I).

| S9-A13 has a minor effect on ion transport in human airway epithelial cells
We analyzed the effects of S9-A13 in BCi-NS1 human airway epithelial cells, which strongly express CFTR. 25 Figure 6A, Figure S1). The same pattern for SLC26A9 variants was detected in other human airway epithelial cell lines, such as Calu3 or CFBE14o-(not shown). The larger band size for SLC26A9 in airways of about 120 kDa is most likely due to glycosylation, 4 while overexpression of SLC26A9 in HEK293 cells produced a band at only around 90 kDa ( Figure 5A). Immunostaining of SLC26A9 in plastic grown BCi-NS1 cells demonstrated intracellular and plasma membrane localization ( Figure 6B). Surprisingly, S9-A13-inhibitable Cl − currents were not detected, suggesting no Cl − transport by SLC26A9 in normal airways.
( Figure 6C). BCi-NS1 cells were grown on permeable supports in PneumaCult™ differentiation media under ALI. BCi-NS1 formed a well differentiated epithelium with mostly ciliated epithelial cells ( Figure 7A). The effects of S9-A13 on short circuit currents were examined in Ussing chamber recordings. We found a small albeit significant inhibition of basal Isc by increasing concentrations of S9-A13 ( Figure 7B,C). However, when compared to the pronounced inhibition of IF-activated Isc by CFTR inh -172 (CFinh), inhibition of Isc by S9-A13 was negligible. In BCi-NS1 airway cells, CFTR inh -172 inhibited also basal Isc present under non-stimulated conditions ( Figure S4C-E ). Essentially identical results were obtained in primary human airway epithelial cells ( Figure S4A,B). The results indicat that CFTR is in charge of both basal and stimulated Cl − secretion in human airways.

| Basal transport in mouse trachea is due to CFTR but not Slc26a9
We also examined expression of Slc26a9, constitutive transport (I'sc; measured under open circuit conditions) and inhibition by S9-A13 in freshly excised mouse tracheas. Immunohistochemistry demonstrated apical expression of Slc26a9 in ciliated cells of mouse airway epithelium, while non-ciliated cells such as club cells and goblet cells did not express SLC26A9 ( Figure 8A). S9-A13 did not inhibit constitutive (basal) ion transport in mouse trachea ( Figure 8B,C). In contrast, CFTR inh -172 dose dependently blocked basal Isc, again demonstrating CFTR In contrast to Ca 2+ -dependent (Tmem16a) Cl − secretion (ATP), cAMP activated transport (CFTR; IF) was small, due to low expression of CFTR in mouse trachea. 29 The small activation of I'sc by IF in the presence of CFTR inh -172, is explained by activation of KCNQ1/ KCNE3 K + channels. 30

| SLC26A9 supports alkalization of ASL pH and H + secretion by gastric cells
Because we found little contribution of SLC26A9 to Cl − transport, we examined if SLC26A9 has a different function in airways. In ASL pH measurements, fully differentiated nasal epithelial cells were exposed to S9-A13 or control solution. Online recordings of ASL pH under thin film conditions indeed showed a sustained decrease in ASL pH by S9-A13. (Figure 9A-D). Two hours after application of S9-A13, ASL pH dropped by 0.23 ± 0.11 (n = 6) units. Subsequent stimulation of CFTR by forskolin caused only a small but not significant increase in ASL pH in airways exposed to S9-A13. These results strongly suggest a role of SLC26A9 for alkalization of ASL pH. As SLC26A9 was proposed as apical Cl − transporter relevant for HCl secretion by gastric parietal cells, 6 we also examined the effect of S9-A13 on histamine-induced H + secretion by HGT-1 human gastric cells. Remarkably, S9-A13 significantly inhibited H + secretion by HGT-1 cells, supporting the concept of SLC26A9 being a Cl − release channel in parietal cells 5,6,19 (Figure 9E).

| DISCUSSION
In the present study, we demonstrate S9-A13 as the first specific and highly potent inhibitor of SLC26A9. S9-A13 inhibits SLC26A9 in the low nanomolar range. It does not show cytotoxic effects, even at micromolar concentrations. The inhibitor was used to examine the potential role of SLC26A9 for ion transport in the airway epithelium. This is of particular interest in cystic fibrosis, as genome-wide association studies demonstrated SLC26A9 as an important modifier of CF lung and gastrointestinal disease: (i) SLC26A9 is expressed in the intestinal and airway epithelium as well as other CF-relevant tissues, (ii) single nucleotide polymorphisms (cSNPs) of SLC26A9 were found to be associated with variable anion transport properties, and (iii) SLC26A9 was shown to modulate CFrelated intestinal disease, diabetes and airway response to CFTR-therapeutics. [31][32][33][34][35] A number of studies argue in favor of CFTR being responsible for constitutive (basal) anion secretion that is absent in CF airways. 26,36,37,38 Other reports provided evidence for the role of SLC26A9 in basal airway Cl − transport. 4,7,11,39,40 So far it was difficult to discriminate between both conductances due to the lack of specific inhibitors for SLC26A9. SLC26A9 inhibitors used in earlier studies were not specific and inhibited also other Cl − channels. 1,2,4 Moreover, expression and activity of both CFTR and SLC26A9 are functionally connected. Intracellular trafficking and constitutive activity of SLC26A9 are compromised in cystic fibrosis (CF) airway epithelia expressing the trafficking mutant F508del-CFTR. [40][41][42] In the presence of F508del-CFTR, binding of the PDZ-domain protein CAL inhibits the function of SLC26A9 due to enhanced endoplasmic reticulum-associated proteasomal degradation. 43 Vice versa, SLC26A9 may support biogenesis and/or stabilization of CFTR. 44 Here we present evidence that SLC26A9 is located in the apical membrane of ciliated airway epithelial cells, but not in non-ciliated club or goblet cells. Overexpression of SLC26A9 induced a basal Cl − -conductance in HEK293 cells, which was well inhibited by S9-A13. Unlike reported recently, 39 SLC26A9 in HEK293 cells could not be further activated by increase in intracellular cAMP ( Figure 5I). SLC26A9 is expressed endogenously in human airway epithelial cells such as CFBE, Calu3, and BCI-NS1, but Cl − transport by SLC26A9 could not be detected in these cells using S9-A13. As airways express CFTR, we speculate regulation of SLC26A9 mediated HCO 3 − transport by CFTR.
In CFTR-expressing BCi-NS1 cells, primary human airway epithelial cells and mouse trachea a substantial SLC26A9-mediated Cl − transport could not be detected (Figures 6-8, Figure S4). In contrast, airway surface liquid pH was significantly reduced by S9-A13, suggesting a role of SLC26A9 for HCO 3 − secretion ( Figure 9A-D).
In cells that only express SLC26A9 but not CFTR such as HEK293 cells or LN215 cells, Cl − /HCO 3 − exchange by SLC26A9 could not be detected, while Cl − transport by SLC26A9 is clearly present (Figures 1-3, and 5, Figure S3). Stomach mucosa expresses only low levels of CFTR, 47 Here SLC26A9-mediated Cl − transport may parallel H + secretion by the proton pump. This is supported by S9-A13-dependent inhibition of H + secretion in HGT-1 human gastric cells 5,6,19 (Figure 9). We speculate that depending on expression of CFTR, SLC26A9 may transport Cl − or HCO 3 − 3 . While Cl − secretion is due to CFTR, basal HCO 3 − secretion might be due to SLC26A9 ( Figure 8, Figure S4C,D). Immunofluorescence suggested pronounced expression in ciliated cells, while transcriptome analysis indicated SLC26A9 enriched in pulmonary neuroendocrine cells, where it could have sensory, neuroendocrine or other functions to protect airways from mucus obstruction . Interestingly, even in the presence of the inhibitor of adenylate cyclase type 1, ST034307, or in the presence of the adenosine 2AB inhibitor PSB603, basal Cl − secretion by CFTR is still detectable ( Figure S4E). This suggests that CFTR does not need to be activated by cAMP/PKA to produce a basal Cl − conductance, but may be activated by SLC26A9 44 The novel SLC26A9 inhibitor