Bioactivity-Guided Fractionation of an Antidiarrheal Chinese Herb Rhodiola kirilowii (Regel) Maxim Reveals (-)–Epicatechin-3-Gallate and (-)–Epigallocatechin-3-Gallate as Inhibitors of Cystic Fibrosis Transmembrane Conductance Regulator

Cystic fibrosis transmembrane conductance regulator (CFTR) is the principal apical route for transepithelial fluid transport induced by enterotoxin. Inhibition of CFTR has been confirmed as a pharmaceutical approach for the treatment of secretory diarrhea. Many traditional Chinese herbal medicines, like Rhodiola kirilowii (Regel) Maxim, have long been used for the treatment of secretory diarrhea. However, the active ingredients responsible for their therapeutic effectiveness remain unknown. The purpose of this study is to identify CFTR inhibitors from Rhodiola kirilowii (Regel) Maxim via bioactivity-directed isolation strategy. We first identified fractions of Rhodiola kirilowii (Regel) Maxim that inhibited CFTR Cl- channel activity. Further bioactivity-directed fractionation led to the identification of (-)–epigallocatechin-3-gallate (EGCG) as CFTR Cl- channel inhibitor. Analysis of 5 commercially available EGCG analogs including (+)–catechins (C), (-)–epicatechin (EC), (-)–epigallocatechin (EGC), (-)–epicatechin-3-gallate (ECG) and EGCG revealed that ECG also had CFTR inhibitory activity. EGCG dose-dependently and reversibly inhibited CFTR Cl- channel activity in transfected FRT cells with an IC50 value around 100 μM. In ex vivo studies, EGCG and ECG inhibited CFTR-mediated short-circuit currents in isolated rat colonic mucosa in a dose-dependent manner. In an intestinal closed-loop model in mice, intraluminal application of EGCG (10 μg) and ECG (10 μg) significantly reduced cholera toxin-induced intestinal fluid secretion. CFTR Cl- channel is a molecular target of natural compounds EGCG and ECG. CFTR inhibition may account, at least in part, for the antidiarrheal activity of Rhodiola kirilowii (Regel) Maxim. EGCG and ECG could be new lead compounds for development of CFTR-related diseases such as secretory diarrhea.


Introduction
Maintenance of an appropriate amount of intestinal fluid is vital for digestion and clearance of the luminal contents. It is a passive process driven by the active anion, predominantly Cl -, transport from blood to the intestinal lumen [1,2]. The major components in fluid secretion involve Clintake via Na + /K + /2Clcotransporter (NKCC1) through the basolateral membrane and Clexit to the lumen via cystic fibrosis transmembrane conductance regulator (CFTR) and Ca 2+ -activated Clchannels (CaCCs) in apical membrane of secretory epithelial cells [1,3,4].
CFTR belongs to the superfamily of ATP-binding cassette (ABC) proteins, whose core units contain two membrane-spanning domains (MSDs) and two nucleotide-binding domains (NBDs). CFTR contains a regulatory (R) region, which is unique to this superfamily. Activity of CFTR is regulated by binding and hydrolysis of ATP at NBDs and by phosphorylation of the R region [5,6]. Though CFTR is not the sole pathway for apical Clexit, it is the predominant pathway for Cltransport in active fluid secretion evoked by cholera toxin and E. coli heatstable enterotoxin [7][8][9]. CFTR is a well-validated target for development of inhibitors for therapy of secretory diarrheas [10][11][12].
Small-molecule blockers of CFTR have been proven valuable for the development of drugs to treat cholera and traveler's diarrhea [13,14]. So far, several CFTR inhibitors have been identified and characterized [10,[15][16][17][18][19], among which the most prominent one is the thiazolidinone CFTR inh -172, a CFTR selective blocker identified from a combinatorial small molecule library. Though CFTR inh -172 is highly specific to CFTR protein and could potently reduce cholera toxin-induced intestinal fluid secretion in rodents, poor water solubility (<5 μM) of the compound greatly limits its potential use in the treatment of diarrhea [20].
Natural products have long been the major resources for new drugs, and many successful drugs originated from natural compounds [21][22][23]. Natural compounds are highly diverse in structure and often provide highly specific biological activities [24][25][26]. Traditional Chinese herbal medicine contains large numbers of therapeutic compounds for a broad spectrum of human diseases including secretory diarrhea. Systematic investigation on the pharmacology of active ingredients and mechanisms are crucial for transforming traditional herbal practices into evidence-based medicine.
We report here the identification of CFTR Clchannel inhibitors from a traditional Chinese herbal antidiarrheal medicine. We found two galloyl-containing catechins (EGCG and ECG) as CFTR inhibitors. Galloyl-containing catechins are major components of Rhodiola kirilowii (Regel) Maxim and green tea that have been reported to have many biological (mainly anticancer and cancer-preventive) activities. Here, we report a new activity for EGCG and ECG, providing a molecular mechanism for the antidiarrheal efficacy of Rhodiola kirilowii (Regel) Maxim.

CFTR inhibition by fractions of Rhodiola kirilowii (Regel) Maxim
Rhodiola kirilowii (Regel) Maxim was extracted using 95% ethanol on Soxhlet reflux apparatus, and then the extract was fractionated into 80 fractions by preparative HPLC with a linear gradient of 0-90% methanol (MeOH). The fractions were dried and dissolved in DMSO to generate 5 mg/ml solutions in a 96-well plate. To identify CFTR inhibitors, we used a cell-based fluorescence assay (Fig. 1A). The analysis revealed inhibition of CFTR-mediated Iinflux by a cluster of fractions (Fig. 1B). Fig. 1C shows that fractions 21-27 significantly inhibited CFTR-mediated Iinflux; whereas the other fractions had little effect. Fig. 2A shows preparative HPLC chromatogram of the Rhodiola kirilowii (Regel) Maxim 95% ethanol extract. The active fractions 21-27 (appeared as a clusters) were eluted at 11-14 min. We next optimized extraction and purification methods for CFTR inhibitor isolation under bioactivity-directed strategy. Rhodiola kirilowii (Regel) Maxim was first extracted with 95% ethanol by using microwave extraction equipment, and then the crude extract was extracted sequentially using a series of solvents including petroleum ether, dichloromethane (DCM), ethyl acetate, and n-butyl alcohol. The plate reader assay indicated that the ethyl acetate fraction was the most effective one, so, the fraction was subjected to further isolation. After two cycles of silica gel chromatography, two active fractions (Fr.7 and Fr.7-2) were obtained, respectively. The active fraction Fr.7-2 was further purified by HPLC to get active fraction C2. Detailed activitydirected fractionation protocol was seen in the Material and Method section and summarized in Fig. 2B. Analytical HPLC chromatograms of ethyl acetate extract, Fr.7, Fr.7-2, and C2 were shown in Fig. 2C.

Bioactivity-directed fractionation and structure determination
Because fraction C2 shows a single peak with a purity >99%, the compound was applied to mass spectrometry (MS) analysis, which gave the molecular size (458.4 Da) and known fragmentation pattern of EGCG (Fig. 3). Nuclear magnetic resonance (NMR) spectrometry gave the 1 H and 13C pattern identical to EGCG, which further confirmed the molecule identity of the active compounds as EGCG.

Structure-activity analysis
Next, we analyzed the effects of commercially available EGCG and 4 structural analogs (C, EC, EGC, and ECG) on CFTR Clchannel activities by fluorescent quenching assay. The results confirmed the activity of EGCG and found that ECG also exhibited significant CFTR inhibitory activity. Fig. 4 summarized the structures and activities of EGCG and 4 analogs. Further timecourse results indicated that rapid inhibition of CFTR Clchannel activities by EGCG and ECG, with half-maximal activation in 5 min; and the inhibitions could be fully reversed at 30 min after EGCG and ECG washout (data not shown).

Biological studies
CFTR is expressed in the crypt cells in the distal small intestine and colon, and it is the major pathway for Clsecretion into intestinal lumen [4,27]. The efficacies of EGCG and ECG were tested ex vivo in isolated rat colonic mucosa by Ussing chamber short-circuit assay with CFTR inh -172 as a positive control. Amiloride (10 μM) and indomethacin (10 μM) were present in the chamber solutions for entire experimental periods to prevent Na + current and prostaglandin generation. As shown in Fig. 5, EGCG and ECG dose-dependently inhibited shortcircuit currents in intact rat colonic mucosa. Next, we investigated the in vivo efficacies of EGCG and ECG in inhibiting cholera toxininduced intestinal fluid secretion in live mice. As shown in the images in Fig. 6A and summarized data in Fig 6B. Massive fluid was accumulated in cholera toxin-treated loops as compared to the saline control group. Intraluminal injection of EGCG (10 μg) and ECG (10 μg) significantly reduced cholera toxin-induced intestinal fluid secretion as quantified from loop weightto-length ratio, supporting that EGCG and ECG are active antidiarrheal ingredients of Rhodiola kirilowii (Regel) Maxim through inhibition of CFTR chloride channel.

Discussion
The system of Chinese herbal medicine is built on the experience of herb users over thousands of years. The efficacies of the herbs in many diseases are already well documented [28,29]. However, because of insufficient studies on material basis and the pharmacology of traditional Chinese medicine (TCM), the molecular mechanisms responsible for their therapeutic effectiveness are still largely unclear. Before TCM has been accepted globally as complementary and alternative medicine for disease treatment and prevention, it is crucial to understand the molecular basis for their effects [30]. To understand the therapeutic mechanisms of TCM, it is critical to identify the active compounds against molecular targets that participate in a particular disease.
The purpose of this study is to identify CFTR inhibitors from traditional Chinese herbal antidiarrheal medicine. Two natural small molecules, EGCG and ECG, were identified as CFTR inhibitors. EGCG and ECG inhibited CFTR Clcurrents in CFTR-transfected FRT cells and isolated rat colonic mucosa. In vivo, both compounds effectively reduced cholera toxin-induced intestinal fluid secretion in mouse closed-loop model. Their inhibitory effects on CFTR Clcurrents and intestinal fluid secretion may provide a molecular basis for the antidiarrheal properties of Rhodiola kirilowii (Regel) Maxim [31,32].
One of the fundamental advantages in TCM is the multicomponent therapy, in which several active compounds interact with one molecular target and/or one active compound interacts with several different molecular targets simultaneously. EGCG and ECG, belonging to galloylcontaining catechins, are found in a wide range of natural products and extracts. These compounds have been reported to have various biological activities, including reduced incidence of cardiovascular disease, diabetes, cancer, stroke, cataracts, inflammation, and obesity [33,34].  Previous studies have shown that catechins had potent antibacterial effects against a number of bacterial pathogens including Vibrio cholerae and Escherichia coli [32,35]. Here, we found that EGCG and ECG significantly reduced cholera toxin-induced fluid secretion, suggesting that these compounds may protect patients against Vibrio cholerae infection disease via multiple pathways. The major pathways for Cltransports in intestine include CFTR and CaCC [4,27,36]. Namkung et al [36] reported that EGCG and ECG inhibited TMEM16A/CaCC channel activity. EGCG and ECG had inhibitory effect on both CFTR and CaCC Clchannel activities suggested that mild inhibition of CFTR and CaCC may result in a significant inhibition of intestinal salt and water secretion. These results further support the proposition that TCM are multicomponent and multi-target agents. Elucidation of multicomponent activity of TCM is critical not only for developing new therapeutic strategies to conquer complex diseases, but also for reducing side effects of drugs. EGCG is more soluble in water compared to CFTR inh -172, with a maximum solubility~23.6 mg/mL [37]. Recent study indicated that cocrystallization technique could significantly improve the bioavailability of EGCG which may greatly increase the chance to develop EGCG as a therapeutic agent [37].
Core structure of the five tested catechins contains three types of functional groups, hydroxyl, pyrogallol and gallic groups, on a phenyl benzopyrane ring. Structure-activity relationship analysis of the catechins suggested that gallic group on the phenyl benzopyrane ring is needed for CFTR inhibition, as evidenced by the lack of activity of C, EC and EGC. Further studies of medicinal chemistry are required to optimize the CFTR-inhibitory activity of catechin compounds.
In summary, we have identified EGCG and ECG from antidiarrheal herb Rhodiola kirilowii (Regel) Maxim as inhibitors of CFTR Clchannel. As a new molecular target of natural compounds EGCG and ECG, CFTR inhibition may account, at least in part, for the antidiarrheal activity of Rhodiola kirilowii (Regel) Maxim. EGCG and ECG could be new lead compounds for developing therapeutics of CFTR-related diseases such as secretory diarrhea and polycystic kidney disease.

Ethics statement
All animal usage procedures in this study were carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and were approved by the Liaoning Normal University Committee on Animal Research. All surgery was performed under sodium pentobarbital anesthesia, and all possible efforts were made to minimize suffering.

General HPLC experimental procedures
The HPLC fractionation was performed on a Waters autopurification system equipped with a constant flow pump (Waters 2525), a compensation pump (Waters 515), a dual channel detector (Waters 2487), an automatic collector (Waters 2767), using a C18 reversed-phase column (XTerra Pre, 19 Ã 150 mm, 5 mm particle size, USA). The purity of active fractions was analyzed on a Waters Alliance HPLC system (Waters 2695) with a diode assay detector (Waters 2996), using C18 column reversed-phase column (XTerra OBD, 2.1 Ã 150 mm, 5 mm particle size, USA).

CFTR-mediated iodide influx assay
The iodide influx fluorescence measurements in CFTR expressing FRT cells were done on a BMG plate reader (Fluostar Optima, BMG Lab Technologies) equipped with HQ500/20X (500 ± 10 nm) excitation and HQ535/30M (535 ± 15 nm) emission filters (Chroma Technology Corp., Brattleboro, VT) as references [10,35]. Briefly, the FRT cells were seeded in 96-well blackwalled-clear-bottomed tissue culture plates (Corning Inc., Corning, NY) at high density (20,000 per well) and incubated in CO 2 incubator for 24-36 h until the cells were confluent. The cells were washed 3 times with PBS, prestimulated with a cocktail (5 μM FSK, 100 μM IBMX and 25 μM GEN in PBS) for 10 min, and then incubated with test samples for another 10 min. Iodide influx rate (d[I -]/dt) of each well was assayed by continuously recording fluorescence for 2 s (baseline) and 12 s after addition of 120 μL I --containing PBS (in which 137 mM Clwas replaced by I -). The d [I -]/dt values were computed from fluorescence data by nonlinear regression. The detailed assay and computation methods were described in the references [10,41,42].

Intestinal fluid secretion measurements
In vivo efficacy of EGCG and ECG was evaluated in a closed-loop mice model of cholera toxininduced intestinal fluid secretion. Briefly, KM mice (age 8-10 weeks) were deprived of food and free access to water for 24 hours. The mice were anaesthetized with intraperitoneal sodium pentobarbital (40 mg/kg). Three closed ileal loops (length 10-15 mm) proximal to the cecum were produced by sutures. Loops in each mouse were injected with saline alone, saline containing cholera toxin (0.5 μg) without or with 10 μg EGCG (or ECG). At 6 hours, the mice were sacrificed with one overdose intraperitoneal sodium pentobarbital (100 mg/kg), and intestinal fluid secretion was quantified from loop weight-to-length ratio. Details were described in reference [10].

Statistical analysis
Data are expressed as the mean±SE or as representative traces. Student's t test was used to compare test and control values, P values < 0.05 were considered to be statistically significant.