The K+ Channel Opener 1-EBIO Potentiates Residual Function of Mutant CFTR in Rectal Biopsies from Cystic Fibrosis Patients

Background The identification of strategies to improve mutant CFTR function remains a key priority in the development of new treatments for cystic fibrosis (CF). Previous studies demonstrated that the K+ channel opener 1-ethyl-2-benzimidazolone (1-EBIO) potentiates CFTR-mediated Cl− secretion in cultured cells and mouse colon. However, the effects of 1-EBIO on wild-type and mutant CFTR function in native human colonic tissues remain unknown. Methods We studied the effects of 1-EBIO on CFTR-mediated Cl− secretion in rectal biopsies from 47 CF patients carrying a wide spectrum of CFTR mutations and 57 age-matched controls. Rectal tissues were mounted in perfused micro-Ussing chambers and the effects of 1-EBIO were compared in control tissues, CF tissues expressing residual CFTR function and CF tissues with no detectable Cl− secretion. Results Studies in control tissues demonstrate that 1-EBIO activated CFTR-mediated Cl− secretion in the absence of cAMP-mediated stimulation and potentiated cAMP-induced Cl− secretion by 39.2±6.7% (P<0.001) via activation of basolateral Ca2+-activated and clotrimazole-sensitive KCNN4 K+ channels. In CF specimens, 1-EBIO potentiated cAMP-induced Cl− secretion in tissues with residual CFTR function by 44.4±11.5% (P<0.001), but had no effect on tissues lacking CFTR-mediated Cl−conductance. Conclusions We conclude that 1-EBIO potentiates Cl−secretion in native CF tissues expressing CFTR mutants with residual Cl− channel function by activation of basolateral KCNN4 K+ channels that increase the driving force for luminal Cl− exit. This mechanism may augment effects of CFTR correctors and potentiators that increase the number and/or activity of mutant CFTR channels at the cell surface and suggests KCNN4 as a therapeutic target for CF.


Introduction
The early onset multiorgan disease cystic fibrosis (CF) is caused by more than 1,800 mutations in the CFTR gene and remains the most common fatal monogenetic disease in Caucasian populations [1,2]. Therefore, the identification of strategies that improve impaired function of mutant CFTR remains a high priority in the development of causal therapies for patients with CF. The CFTR gene encodes a cAMP-regulated Cl 2 channel that localizes to the luminal membrane of epithelial cells, where it plays an important role in transepithelial ion and fluid transport. CFTR mutations, via different molecular mechanisms, can reduce the number, or impair the regulation of CFTR Cl 2 channels inserted into the apical cell membrane [3,4,5,6,7]. As a result, defective or reduced CFTR-mediated Cl 2 secretion causes dehydration of epithelial surfaces and dysfunction of many epithelial organs including the small and large intestine, the pancreas, the hepato-biliary system, and the airways [2,8,9].
Current pharmacological strategies to rescue and/or improve CFTR function in CF focus on i) the development of CFTR correctors that increase the number of apical membrane CFTR Cl 2 channels by improving the processing or synthesis of mutant CFTR; and ii) the development of potentiators that improve the open probability (P O ) of mutant CFTR Cl 2 channels that are delivered to the cell surface, but exhibit impaired gating [3,10]. In a recent phase 2 clinical trial in CF patients carrying the gating mutation G551D, the CFTR potentiator VX-770 [11] induced CFTR-mediated Cl 2 secretion in nasal epithelia in vivo to levels of ,20% of normal [12,13]. A similar level of functional correction was obtained in CF patients with CFTR nonsense mutations after treatment with PTC124, a small molecule compound that improves read through at premature stop codons to produce full-length CFTR [14,15]. Previous studies on the correlation between CFTR function and CF disease severity demonstrated that detection of residual CFTR function in this range (,20% of normal) in native nasal or intestinal epithelia was associated with a less severe disease phenotype, but was not sufficient to prevent the onset and progression of CF organ disease [16,17]. Therefore, additional pharmacological strategies that further improve Cl 2 secretion through mutant CFTR channels may improve therapeutic effects of current CFTR potentiator and corrector compounds.
Besides activation of luminal CFTR Cl 2 channels, CFTRmediated Cl 2 secretion relies on parallel activation of basolateral Ca 2+ -and cAMP-dependent K + channels that hyperpolarize the membrane potential and generate the electrical driving force for luminal Cl 2 exit [8,18,19]. A series of functional and genetic studies in cultured cells and mice demonstrated that the Ca 2+activated intermediate conductance K + channel in the basolateral membrane of colonocytes is encoded by KCNN4 and inhibited by clotrimazole [20,21,22,23,24], whereas the cAMP-dependent K + channel is formed by the KCNQ1/KCNE3 complex and inhibited by the chromanol 293B [25,26,27,28,29,30]. In this context, previous studies in colonic cells (T84) and mouse colon identified 1-ethyl-2-benzimidazolone (1-EBIO) as a compound that potentiates CFTR-mediated Cl 2 secretion by coordinate activation of the basolateral Ca 2+ -activated K + channels KCNN4 (SK4, IK1) and luminal CFTR Cl 2 channels [31,32,33,34,35]. However, the effects of 1-EBIO on Cl 2 secretion mediated by wild-type and mutant CFTR, and the role of KCNN4 and KCNQ1/KCNE3 K + channels in native human colon have not been studied.
We previously demonstrated that bioelectric measurements in rectal biopsies mounted in perfused micro-Ussing chambers enable sensitive assessment of CFTR function in native human tissues [17,36]. In this study, we used this approach to determine the effects of 1-EBIO on CFTR-mediated Cl 2 secretion in rectal biopsies from non-CF individuals and CF patients carrying a wide spectrum of CFTR mutations. First, we used protocols to study the effect of 1-EBIO on the activity of luminal CFTR Cl 2 channels. Second, we studied effects of 1-EBIO on basolateral Ca 2+activated and cAMP-dependent K + channels and determined expression of KCNN4 and KCNQ1 K + channel transcripts in human rectal biopsies. Finally, we determined the capacity of 1-EBIO to potentiate Cl 2 secretion mediated by wild-type and mutant CFTR. The results of our studies provide novel insights into pharmacological modulation of mutant CFTR function in native human CF epithelia and identified co-activation of the basolateral K + channel KCNN4 as a potential strategy to augment therapeutic effects of CFTR potentiator and corrector drugs.

Subjects
The study was approved by the Ethical Committees at the University Hospitals of Heidelberg, Freiburg, Lisbon and Vienna, and all subjects gave their written informed consent. For children under 18, parents obtained detailed information and gave their signed informed consent. The effects of 1-EBIO were studied in rectal biopsies from 47 CF patients (31 PI, 16 PS; mean age 10.461.4 yr, range 3 mo to 47 yr) and 57 age-matched non-CF controls (11.461.5 yr, range 6 mo to 45 yr) who were enrolled in the study between 1999-2011 at the CF Centers at the University Hospitals of Heidelberg, Freiburg, Lisbon and Vienna. The diagnosis of CF was established by clinical symptoms characteristic of CF, increased sweat Cl 2 concentrations ($ 60 mmol/L) and/ or detection of disease-causing mutations in both CFTR alleles [37,38]. Exocrine pancreatic insufficiency (PI) was defined by a history of malabsorption and fecal elastase E1 levels of ,200 mg/g stool [39]. CFTR mutations were identified by sequential screening by allele specific PCR (Elucigene CF20 kit, Cellmark, Abingdon, UK) and denaturing gradient gel electrophoresis, followed by automated DNA sequencing of the 27 exons of the CFTR gene using the ABI Prism TM BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, Weiterstadt, Germany) as previously described [17]. Genotypes of CF patients are summarized in table 1. In all control subjects, sweat Cl 2 concentrations were normal (,30 mmol/L) and genetic screening for 20 common CFTR mutations was negative (Elucigene CF20 kit, Cellmark, Abingdon, UK).

Experimental Protocols and Analysis of Ussing Chamber Experiments
Rectal tissues were equilibrated for 40 min in the presence of amiloride (10 mmol/L, luminal) to block electrogenic Na + absorption and indomethacin (10 mmol/L, basolateral) to inhibit prostaglandin E2 synthesis and endogenous cAMP formation. Previous studies showed that under these experimental conditions, CFTR activity is inhibited and Ca 2+ -dependent activation with carbachol (CCH, 100 mmol/L, basolateral) results in a transient lumen-positive I sc ' response reflecting K + secretion that is unmasked in the absence of anion secretion [8,19,40]. When CFTR Cl 2 channels are activated under these conditions, Ca 2+dependent co-stimulation produces lumen-negative (Cl 2 secretory) responses by activation of basolateral Ca 2+ -activated K + channels that increase the electrical driving force for luminal Cl 2 exit [19,40,41]. To determine the effects of 1-EBIO on CFTRmediated Cl 2 secretion in the absence of cAMP-mediated activation (indomethacin), we therefore measured the effects of 1-EBIO (500 mmol/L, basolateral) on basal I sc ' and CCH-induced I sc ', and determined the effects of the CFTR inhibitor CFTR inh -172 (20 mmol/L, basolateral) on 1-EBIO-induced responses.
To study the effect of 1-EBIO on basolateral Ca 2+ -activated and cAMP-dependent K + channels that generate the driving force for CFTR-mediated Cl 2 secretion, control rectal tissues were stimulated with 3-isobutyl-1-methylxanthine (IBMX; 100 mmol/ L, basolateral) and forskolin (1 mmol/L, basolateral) to obtain maximal cAMP-mediated activation of CFTR, and the effects of 1-EBIO on cAMP-induced I sc ' and CCH-induced I sc ' were measured before and after adding either clotrimazole (30 mmol/ L, basolateral), an inhibitor of the basolateral Ca 2+ -activated K + channel KCNN4 [20,22,23], or 293B (10 mmol/L, basolateral), an inhibitor of the cAMP-dependent heteromeric KCNQ1/ KCNE3 K + channel [25,26,28]. Dose-response curves for the effects of 1-EBIO in the presence of cAMP-mediated activation (IBMX/forskolin) were obtained by measuring the change in I sc ' induced by exposing tissues to increasing concentrations of 1-EBIO (10 28 to 10 23 mol/L) and plotted as 1-EBIO-induced I sc ' normalized to the maximal 1-EBIO-induced I sc ' (I sc ' max), and EC 50 values were determined by fitting dose-response data to the Hill equation using Origin version 6.1 (OriginLab Corp., Northampton, MA).
To determine potentiator effects of 1-EBIO on CFTR-mediated Cl 2 secretion in native control and CF rectal epithelia, tissues were activated with IBMX and forskolin and effects of 1-EBIO on cAMP-induced I sc ' and CCH-induced I sc ' were measured in control tissues, CF rectal tissues with no detectable Cl 2 secretion (CF absent ) and CF tissues with residual cAMP-mediated Cl 2 secretion (CF residual ). In the presence of cAMP stimulation, CCH induced i) a transient monophasic lumen-negative (Cl 2 secretory) response (control tissues), ii) a transient monophasic lumen-positive (K + secretory) response (CF absent tissues), or iii) a biphasic response characterized by a lumen-positive peak followed by a lumen-negative deflection (CF residual ) [17], and effects of 1-EBIO on CCH-induced Cl 2 secretion were determined from lumennegative deflections (control and CF residual tissues) or the lumenpositive plateau (CF absent tissues). In a subset of experiments, we used CFTR inh -172 (20 mmol/L, luminal), and bumetanide (100 mmol/L, basolateral), an inhibitor of the Na + -K + -2Cl 2 cotransporter (NKCC1), to block transepithelial Cl 2 secretion [40]. For different experimental protocols, bioelectric measurements were generally performed in one rectal biopsy per individual. Occasionally, more than one sample from one individual was studied by the same experimental protocol, and data were averaged to obtain a single value for each individual subject.

Statistical Analysis
Data were analyzed with SigmaStat version 3.1 (Systat Software, Erkrath, Germany) and are reported as mean 6 SEM (n = number of individuals studied). Statistical analyses were performed using paired Student's t test and Wilcoxon signed rank test as appropriate, and P,0.05 was accepted to indicate statistical significance.

1-EBIO activates CFTR-dependent Clsecretion in native rectal epithelia
To study the effects of 1-EBIO on luminal CFTR Cl 2 channels in native human rectal epithelia, tissues were pretreated with indomethacin to block endogenous cAMP-mediated CFTR activation. Under these conditions, V te and I sc ' approached ,zero and cholinergic activation with CCH induced a monophasic lumen-positive response reflecting luminal K + secretion ( Fig. 1 A,C,D) indicating that luminal CFTR Cl 2 channels were inactive [19,40]. In control tissues, addition of 1-EBIO resulted in a sustained lumen-negative Cl 2 secretory response (DI sc ' = 225.066.6 mA/cm 2 , n = 12, P = 0.001) ( Fig. 1 A,C). In the presence of 1-EBIO, CCH induced a biphasic response with the lumen-positive K + secretory peak followed by a lumennegative peak reflecting a transient increase in 1-EBIO-induced Cl 2 secretion ( Fig. 1 A,D) [41]. Effects of 1-EBIO on basal and CCH-induced I sc ' were completely reversible upon washout ( Fig. 1  C,D). In rectal tissues from CF patients with two severe CFTR mutations and no detectable cAMP-mediated Cl 2 secretion (CF absent ) ( Table 1), 1-EBIO failed to induce Cl 2 secretion (DI sc ' = 20.760.8 mA/cm 2 , n = 7, P = 0.3) and had no effect on CCH-induced I sc ' (Fig. 1 B,C,E). In rectal biopsies from the control group, 1-EBIO-induced Cl 2 secretion was significantly inhibited when tissues were pretreated with the CFTR inhibitor CFTR inh -172 (Fig. 1 F). Similar, CCH-induced Cl 2 secretory (lumen-negative) responses detected in the presence of 1-EBIO were inhibited by CFTR inh -172 (Fig. 1 G). Taken together, these data provide pharmacologic and genetic evidence that 1-EBIO activated luminal CFTR Cl 2 channel activity and transepithelial Cl 2 secretion in native human rectal epithelia in the absence of cAMP-mediated stimulation.
1-EBIO activates basolateral Ca 2+ -dependent but not cAMP-dependent K + channels in native rectal epithelia Next, we determined the effect of 1-EBIO on basolateral K + channels in human rectal tissues. In these experiments, CFTR was first activated by maximal cAMP stimulation (IBMX/forskolin) to generate conditions under which basolateral K + channels became limiting for CFTR-mediated Cl 2 secretion [8,30]. Under these conditions, we then tested the effects of clotrimazole, an inhibitor of the Ca 2+ -activated KCNN4 K + channel (Fig. 2), and 293B, an inhibitor of the cAMP-dependent KCNQ1/KCNE3 K + channel complex (Fig. 3), on 1-EBIO-induced responses in control rectal tissues.
Stimulation of tissues with IBMX and forskolin produced a large and sustained Cl 2 secretory response that was significantly increased by 1-EBIO (DI sc ' = 264.6613.5 mA/cm 2 , n = 17, P = 0.001) in a reversible and dose-dependent fashion (EC 50 = 583.8621.2 mmol/l; n = 6) ( Fig. 2 A,B,C). As expected from previous studies, cholinergic co-activation induced a transient increase in Cl 2 secretion reflecting an increase in the driving force for CFTR-mediated Cl 2 secretion generated by activation of basolateral K + channels [8,19]. CCH-induced Cl 2 secretory responses were significantly increased after tissues were pretreated with 1-EBIO (Fig. 2 A,D). The effects of 1-EBIO on cAMP-induced (IBMX/forskolin) and CCH-induced Cl 2 secretion were completely inhibited when tissues were perfused with clotrimazole ( Fig. 2 A,B,D). RT-PCR analyses demonstrated that transcripts encoding for the clotrimazole-sensitive Ca 2+ -activated K + channel KCNN4, which is pharmacologically activated by 1-EBIO and responsible for driving Ca 2+ -activated Cl 2 in mouse colon [22,23,31], are expressed in native rectal tissues from control subjects and CF patients (Fig. 2 E).
Perfusion of rectal tissues with 293B abolished cAMP-induced Cl 2 secretion as expected [19], and RT-PCR demonstrated expression of KCNQ1 transcripts in native rectal epithelia (Fig. 3A,B,D). However, in contrast to clotrimazole (Fig. 2), 293B did neither inhibit 1-EBIO-induced Cl 2 secretion, nor the augmentation of cholinergic Cl 2 secretion by 1-EBIO in the presence of cAMP activation (Fig. 3A,B,C). Collectively, these results indicated that 1-EBIO augments CFTR-mediated Cl 2 secretion in human colon in the presence of cAMP stimulation by activation of the basolateral Ca 2+ -activated KCNN4 K + channel, but not the cAMP-dependent KCNQ1/KCNE3 K + channel complex, and suggested that 1-EBIO may potentiate residual CFTR-mediated Cl 2 secretion in CF tissues via this mechanism.

1-EBIO potentiates residual CFTR-mediated Cl 2 secretion in CF rectal biopsies
To test if pharmacological co-activation of the Ca 2+ -regulated KCNN4 K + channel promotes CFTR-mediated Cl 2 secretion in native CF tissues, we next determined the effects of 1-EBIO in rectal biopsies from CF patients carrying a large spectrum of CFTR mutations (Table 1). Following the equilibration in the presence of amiloride and indomethacin, CF tissues were stimulated with IBMX and forskolin to achieve maximal activation of luminal CFTR Cl 2 channels and basolateral KCNQ1/KCNE3 K + channels, and co-stimulated with CCH to activate Ca 2+ -regulated KCNN4 K + channels, thus maximizing the driving force for CFTR-mediated Clsecretion [8,19]. Based on Cl 2 secretory responses observed with this protocol, tissues were stratified into two groups, namely one group including tissues expressing residual CFTR Cl 2 channel function (CF residual ) and another group with no detectable Cl 2 secretion (CF absent ) [17]. The CF genotypes of patients assigned to each group are summarized in table 1. Tissues from age-matched control subjects were included in these studies as a reference of wild-type CFTR function.
In the control group, cAMP-mediated stimulation induced a large Cl 2 secretory response that was significantly increased by cholinergic co-activation as expected [19] (Fig. 4 and 5). In the CF absent group, cAMP-mediated stimulation induced small lumenpositive responses and co-activation with CCH produced a monophasic lumen-positive I sc ' response reflecting K + secretion [40] (Fig. 4 and 5). In the CF residual group, cAMP-mediated activation induced an attenuated but sustained lumen-negative Cl 2 secretory response ( Fig. 4 and 5). Consistent with previous studies [17], co-activation of CF residual tissues with CCH produced biphasic responses with an initial lumen-positive (K + secretory) deflection followed by a transient lumen-negative (Cl 2 secretory) response ( Fig. 4 and 5). Mean Cl 2 secretory responses in CF residual tissues accounted for ,34% of cAMP-mediated Cl 2 secretion and ,17% of CCH-induced Cl 2 secretion observed in control rectal tissues (Fig. 5E,F). Similar to control tissues, Cl 2 secretory responses in CF residual tissues were inhibited by bumetanide and CFTR inh -172 demonstrating that residual cAMP-and CCHinduced Cl 2 secretion was mediated by CFTR (Fig. 5A-D).

Discussion
This study provides important information on the pharmacological modulation of CFTR-mediated Cl 2 secretion in native human normal and CF rectal epithelia. First, by using experimental conditions under which CFTR Cl 2 channels were inactive, we demonstrate that the benzimidazolone compound 1-EBIO [31,32,33,34,35] activated CFTR-mediated Cl 2 secretion in normal rectal tissues in the absence of cAMP-mediated activation (Fig. 1). As previously shown for the CFTR opener genistein [41], 1-EBIO-induced Cl 2 secretory responses were augmented by cholinergic stimulation, consistent with parallel activation of basolateral Ca 2+ -activated K + channels that increase the driving force for CFTR-mediated Cl 2 secretion [8,19]. The notion that 1-EBIO acted as a CFTR Cl 2 channel opener in the absence of cAMP-mediated activation was supported by the observations i) that 1-EBIO-induced Cl 2 secretory responses were inhibited by the CFTR inhibitor CFTR inh -172, and ii) that 1-EBIO failed to induce Cl 2 secretion in rectal epithelia from CF patients carrying two severe CFTR mutations that caused a lack of functional CFTR Cl 2 channels in the apical cell membrane (Table 1) [7,17]. Second, our results show that 1-EBIO activated coltrimazolesensitive basolateral K + channels in rectal epithelia (Fig. 2,3). This effect of 1-EBIO augmented CFTR-mediated Cl 2 secretion in normal tissues when luminal CFTR Clchannels were fully activated, as evidenced by a substantial increase in steady-state and CCH-induced Cl 2 secretory responses in the presence of cAMP stimulation. As shown in figure 2, this augmentation of Cl 2 secretion by 1-EBIO was completely abrogated when tissues were pretreated with clotrimazole. In contrast, pretreatment of tissues with 293B had no effect on 1-EBIO-induced Cl 2 secretion (Fig. 3). This pharmacological profile, together with mRNA transcripts analyses showing that both clotrimazole-sensitive KCNN4 K + channels, as well as 293B-sensitive KCNQ1 K + channels were expressed in rectal tissues, support the concept that 1-EBIO activates basolateral KCNN4 K + channels and thereby increases the electrical driving force for CFTR-mediated Cl 2 secretion in native human colon. The observation that 1-EBIO promoted transepithelial Cl 2 secretion even in the absence of Ca 2+dependent stimulation ( Fig. 2A-C) is consistent with previous patch-clamp studies in cultured colonocytes (T84) and heterologous cells, which demonstrated that 1-EBIO activates KCNN4 K + channels at low levels of free Ca 2+ [31,44]. These results suggest that pharmacological activation of basolateral KCNN4 K + channels can potentiate CFTR-mediated Cl 2 secretion in the colon under basal conditions, as well as in the presence of secretagogues activating cAMP-and Ca 2+ signaling pathways in vivo. Taken together, our data showing a coordinate activation of luminal CFTR Cl 2 channels and basolateral KCNN4 K + channels by 1-EBIO in human rectal epithelia are consistent with previous results in colonic cells [22,31,32] and mouse colon [23,33,34,35]. These results thus confirmed that this dual mode of action of 1-EBIO is also operative in native human colon, and formed the rationale for the hypothesis that 1-EBIO may potentiate Cl 2 secretion in native CF tissues expressing CFTR mutants with residual Cl 2 channel function.
We tested this hypothesis by studying the effects of 1-EBIO in rectal tissues from CF patients carrying a wide spectrum of CF genotypes (Table 1). These measurements focused on potentiator effects of 1-EBIO in cAMP pre-stimulated tissues and demonstrated that pharmacological activation of basolateral KCNN4 K + channels by 1-EBIO induced an ,44% increase in cAMPmediated Cl 2 secretion in CF residual tissues, which expressed at least one mild CFTR mutation producing Cl 2 channels with residual function (Table 1, Fig. 4, 5). In contrast, consistent with findings in the colon of Cftr-deficient (Cftr -/-) mice [33,34], 1-EBIO failed to induce Cl 2 secretion in CF absent tissues expressing two severe CFTR mutations that resulted in a lack of functional CFTR Cl 2 channels in the luminal membrane of colonocytes (Table 1, Fig. 4, 5), including rectal tissues from patients homozygous for the common processing mutation F508del-CFTR. However, it is noteworthy that 1-EBIO was able to increase residual function of F508del-CFTR after low temperature PLoS ONE | www.plosone.org correction in F508del-expressing human bronchial epithelial cells [45]. When viewed in combination, these results demonstrate that 1-EBIO, via activation of basolateral KCNN4 K + channels, improves cAMP-mediated Cl 2 secretion in human CF tissues expressing several CFTR mutants with residual Cl 2 channel activity including F508del-CFTR when inserted into cell membrane.
The relevance of these findings is underscored by recent results driven from phase 2 clinical trials testing the CFTR potentiator VX-770 [12] and the CFTR corrector PTC124 [15] in CF patients carrying CFTR mutations responsive to these drugs. In these studies, measurements of nasal potential difference indicated that on average, both therapeutic strategies produced similar levels of functional correction of CFTR-mediated Cl 2 secretion in the range of ,20% of normal [12,13,15]. The results from previous studies correlating CF disease severity with CFTR-mediated Cl 2 secretion in nasal and rectal epithelia predict that functional correction to this level (i.e. ,20%) may ameliorate disease severity, but will be insufficient to prevent CF organ disease [16,17]. Our data provide a proof-of-concept that pharmacological co-activation of basolateral KCNN4 K + channels substantially potentiates residual CFTR function in native CF tissues (Fig. 4,5), and may thus provide an opportunity to improve therapeutic effects of current CFTR potentiator and corrector compounds developed either to improve the open probability (P O ) or to increase the number of mutant CFTR Cl 2 channels in the apical cell membrane [10,46,47]. However, no potent KCNN4 K + channel openers have so far become available for clinical testing. While previous structure-activity studies were able to identify benzimidazolone derivatives that were up to 100-fold more potent in activating KCNN4 than the reference compound 1-EBIO (EC 50 ,600 mmol/l), these improvements were not sufficient for clinical development [48]. We expect that recent developments in functional high-throughput screening including cell-based fluorescent screens for K + channel activity will enable the identification of new KCNN4 K + channel activators that may be useful as lead compounds for CF drug development [46,47,49].
We previously demonstrated that transepithelial bioelectric measurements in rectal biopsies provide a sensitive tool to detect residual CFTR function in native tissues from individual CF patients carrying mild CFTR mutations, and that the magnitude of residual residual CFTR function correlated with clinical outcomes including age at diagnosis, exocrine pancreatic function, and lung function [17]. In these prior studies, the magnitute of normal cAMP-induced CFTR-mediated Cl 2 secretion detected with our modified micro-Ussing chambers in rectal tissues from healthy individuals was ,5 to 40-fold greater than values reported with other experimental setups [17,36,50,51] indicating that our protocols detected CFTR function at high fidelity. However, the sensitivity of this assay to detect effects of CFTR potentiators and correctors was not known. Our present results indicate that this technique provides a sensitive tool to detect potentiator effects on low levels of residual CFTR function in individual CF patients (Fig. 4,5). These data suggest that assessment of CFTR function in rectal biopsies may provide a sensitive biomarker to evaluate effects of CFTR potentiator and corrector drugs in future clinical trials aimed at correcting the CFTR-associated defects [52]. We expect that this new outcome measure, together with bioelectric measurments of CFTR function in sweat glands and nasal epithelia [13,53], will help to determine the level of functional correction that is required in different organ systems to achieve therapeutic benefits in CF patients. In the future, assessment of CFTR function in rectal biopsies and pre-testing of drugs ex vivo may also aid individualized CF therapy. In this context, this measurement could help to stratify CF patients carrying CFTR mutations with unknown functional consequences for treatment with CFTR potentiator or corrector drugs, and also pre-assess which patients will be 'responders' to a given drug by determining its therapeutic effects on CFTR function directly in native tissues from each individual CF patient.
In summary, we demonstrate that 1-EBIO activates CFTRmediated Cl 2 secretion in native human colon by coordinate activation of luminal CFTR Cl 2 channels and basolateral Ca 2+ -activated KCNN4 K + channels, and that this mode of action potentiates residual cAMP-mediated Cl 2 secretion in native CF rectal tissues expressing CFTR mutants that retain residual Cl 2 channel function. Our results suggest that pharmacological co-activation of KCNN4 K + channels may augment therapeutic effects of CFTR potentiator and corrector compounds that are currently in active preclinical and clinical development, and suggest KCNN4 as a therapeutic target for CF.