ION CHANNELS, RECEPTORS AND TRANSPORTERS

Glucose enhances rotavirus enterotoxin-induced intestinalchloride secretion

Liangjie Yin1& Rejeesh Menon1

& Reshu Gupta1 & Lauren Vaught1 & Paul Okunieff1 &

Sadasivan Vidyasagar1

Received: 8 February 2017 /Revised: 20 April 2017 /Accepted: 24 April 2017 /Published online: 10 May 2017# Springer-Verlag Berlin Heidelberg 2017

Abstract Rotavirus causes severe diarrhea in small childrenand is typically treated using glucose-containing oral rehydra-tion solutions; however, glucose may have a detrimental im-pact on these patients, because it increases chloride secretionand presumably water loss. The rotavirus enterotoxin non-structural protein 4 (NSP4) directly inhibits glucose-mediated sodium absorption. We examined the effects ofNSP4 and glucose on sodium and chloride transport in mousesmall intestines and Caco-2 cells. Mouse small intestines andCaco-2 cells were incubated with NSP4114–135 in the presence/absence of glucose. Absorption and secretion of sodium andchloride, fluid movement, peak amplitude of intracellularcalcium fluorescence, and expression of Ano1 and sodium-glucose cotransporter 1 were assessed. NHE3 activityincreased, and chloride secretory activity decreased withage. Net chloride secretion increased, and net sodium absorp-tion decreased in the intestines of 3-week-old mice comparedto 8-week-old mice with NSP4. Glucose increased NSP4-stimulated chloride secretion. Glucose increased NSP4-stimulated increase in short-circuit current measurements(Isc) and net chloride secretion. Ano1 cells with siRNA knock-down showed a significant difference in Isc in the presence ofNSP4 and glucose without a significant difference in peakcalcium fluorescence intracellular when compared to non-silencing (N.S.) cells. The failure of glucose to stimulate

significant sodium absorption was likely due to the inhibitionof sodium-hydrogen exchange and sodium-glucose cotrans-port by NSP4. Since glucose enhances intestinal chloridesecretion and fails to increase sodium absorption in the pres-ence of NSP4, glucose-based oral rehydration solutions maynot be ideal for the management of rotaviral diarrhea.

Keywords Rotavirus . NSP4 . SGLT1 . NHE3 . Ano1

Introduction

Rotavirus is a major cause of severe diarrhea in children<5 years of age worldwide and causes diarrhea andvomiting, often with fever and abdominal pain, for 3 to8 days [7, 12, 38]. Since the introduction of the rotavirusvaccination in 2006, hospitalizations of children in devel-oped countries due to rotavirus-induced gastroenteritishave declined by an estimated 60 to 89% [9, 23, 35],but developing and underdeveloped countries have notexperienced such promising results [19, 25, 28, 49].

Rotavirus consists of 11 double strands of RNA that code forsix structural and five nonstructural proteins. One of the non-structural proteins, NSP4, is primarily responsible for theetiopathogenesis of rotavirus-induced diarrhea [1]. Exposureto NSP4 activates a protein kinase C (PKC)/inositol triphos-phate (IP3) pathway, resulting in increased intracellular calciumthat induces electrogenic secretion of chloride and fluid [10].This intracellular calcium-activated chloride secretion is the pri-mary factor leading to NSP4-induced secretory diarrhea [31],although additional mechanisms may be involved [2, 6, 20].Rotavirus primarily affects the mature and differentiated smallintestinal villus epithelial cells [21]. These mature cells at the tipof the villus possess the specific transporters necessary for theabsorption of electrolytes and nutrients. In contrast to villous

Electronic supplementary material The online version of this article(doi:10.1007/s00424-017-1987-x) contains supplementary material,which is available to authorized users.

* Sadasivan Vidyasagarvidyasagar@ufl.edu

1 Department of Radiation Oncology, University of Florida HealthCancer Center, Cancer and Genomic Research Complex, 2033Mowry Rd., Box 103633, Gainesville, FL 32610, USA

Pflugers Arch - Eur J Physiol (2017) 469:1093–1105DOI 10.1007/s00424-017-1987-x

epithelial cells, crypt cells lack fully developed micro-villi or brush border membranes and the associatedtransport proteins necessary for the absorptive functionsof the enterocytes. Crypt cells actively secrete chloridewhen stimulated with secretagogues, such as cyclicadenosine monophosphate (cAMP), cyclic guanosinemonophosphate (cGMP), or intracellular calcium, whichpromotes passive movement of water into the lumen. Ina normal, uninfected small intestine, net absorption ofelectrolytes and fluid stems from the dynamic bidirec-tional flux of electrolytes and water movement betweenthe gut lumen and the vascular compartment. Thus,there is spatial segregation of absorptive and secretoryprocesses, with active secretion thought to occur pre-dominantly in the crypts.

In addition to electroneutral absorption of sodium andchloride (Na+-H+ exchange coupled with Cl−-HCO3

− ex-change), the villous epithelial cell is responsible fornutrient-coupled electrolyte absorption [e.g., absorptionof sodium and glucose by sodium-glucose cotransporter1 (SGLT1)]. Our recent study shows that absorption ofglucose at concentrations of less than 0.6 mM results incalcium-activated active chloride secretion in addition towell-characterized sodium-coupled glucose absorption[53]. Others have also noted that glucose might stimu-late active chloride secretion [4, 32, 40, 43].

Current management of rotavirus-induced diarrheaprimarily involves prevention of dehydration throughthe use of oral rehydration solutions (ORS) that, likethe standard World Health Organization hypoosmolarORS (WHO-HO-ORS), utilize glucose stimulation ofsodium transport to improve hydration. However, glu-cose stimulation of chloride secretion accompanied byfluid secretion could explain the well-documented fail-ure of WHO-HO-ORS, a glucose-containing rehydrationsolution, to dramatically decrease stool volume (i.e., di-arrhea) in some children with diarrheal diseases [47].

Since glucose and NSP4 both stimulate calcium-activated chloride secretion, we hypothesized that oralrehydration solutions containing glucose further en-hance active chloride secretion during rotavirus infec-tions. Sodium absorption could also be disrupted, asNSP4 directly inhibits activity of SGLT1 [13]. In ad-dition, rotavirus-induced damage to epithelial cells atthe villous tip can severely compromise electroneutralabsorption of sodium (Na+-H+ exchange) and chloride(Cl−-HCO3

− exchange), further shifting the electrolyteand water transport balance from absorption to secre-tion [13, 53]. Given the prevalence of ORS as a treat-ment for diarrheal illness, we chose to identify theabsorptive and secretory changes associated with theaddition of glucose in a model of rotavirus-induceddiarrhea.

Materials and methods

NSP4114–135 peptide

A high-performance liquid chromatography-purified toxigenicregion (NSP4114–135) of rotavirus enterotoxin that contains themembrane interaction domain and increases intracellular calci-umwas used [16, 44]. NH2-DKLTTREIEQVELLKRIYDKLT-CONH2, a peptide with a molecular weight of 2705 that causesdose-dependent diarrhea in rodents, was obtained from thePeptide Synthesis Facility at the University of Pittsburgh(Pittsburgh, PA) [1].

Animal preparation

Three-, six-, and eight-week-old, male NIH Swiss mice pur-chased from the NCIMouse Repository (Frederick,MD) werefed Teklad LM-485 Mouse/Rat Sterilizable Diet #7912/7012(Envigo, Huntington, Cambridgeshire, UK) and housed atfour mice per cage. Bedding was also purchased fromEnvigo (#7092, 1/8″ corn cobs, 40 lbs). Experiments did notutilize fasting or any extra enrichment or interventions.Following CO2 inhalation, mice were humanely euthanizedby cervical dislocation. After exsanguination, the ileal mucosawas obtained, as previously described [54, 55]. All experi-ments were approved by the University of FloridaInstitutional Animal Care and Use Committee and adheredto the Animal Research: Reporting of In Vivo Experiments(ARRIVE) guidelines.

Electrophysiology

Stripped ileal sheets from mice or Caco-2 cells grown for21 days post-confluence on Costar Snapwell culture inserts(Corning, NY, USA) with 0.4-mm pores were mounted be-tween two halves of an Ussing chamber (P2304; PhysiologicInstruments, San Diego, CA, USA) with 0.3 or 1.13 cm2 ofexposed surface area, respectively. The Ringer solution(140 mM of Na+, 119.8 mM of Cl−, 5.2 mM of K+, 2.4 mMof HPO4

−, 0.4 mM of H2PO4−, 1.2 mM of Mg2+, 1.2 mM of

Ca2+, and 25 mM of HCO3−) was bubbled bilaterally with

95% O2 and 5% CO2 and maintained at 37 °C. After allowingthe tissues to stabilize for 45 min, the short-circuit current (Isc,expressed as μeq h−1 cm−2) and conductance (G, expressed asmS/cm2) were recorded using a computer-controlled voltage/current clamp device (VCC MC-8; Physiologic Instruments),as previously described [54, 55].

For isotope flux studies, 0.5 μCi/chamber of sodium (22Na)and 0.3 μCi/chamber of chloride (36Cl) were used to obtain anactivity of 20,000 counts per minute (CPM) for sodium and4500 CPM for chloride to determine the unidirectional fluxesacross the ileal mucosa in the presence and absence ofNSP4114–135 and glucose, as described previously [48, 55].

1094 Pflugers Arch - Eur J Physiol (2017) 469:1093–1105

Samples (500 μl) were collected at regular intervals of 15 minfrom the side opposite to the side where isotopes were added;an average of three such flux periods was used. The activity of22Na was measured using a gamma counter (Wizard 2, 2480Automatic Gamma Counter, Perkin Elmer, USA), and 36Clwas measured using a liquid scintillation counter (LS 6500Multipurpose Scintillation Counter, Beckman Coulter, Inc.,Brea, CA, USA). After the basal electrical measurements weremade, 7 μM of NSP4114–135 was added to the mucosal sideand equilibrated for 45 min [34] before changes in current andconductance were recorded. Net flux (Jnet) was calculatedfrom mucosa to serosa and serosa to mucosa unidirectionalfluxes determined from separate tissues, which were pairedbased upon conductance similarity (<5% difference) and cal-culated using the formula Jnet = Jms − Jsm and expressed asμeq h−1 cm−2.

Glucose saturation kinetics

To separate the contribution of sodium-coupled glucose trans-port to Isc from that of glucose-stimulated chloride secretion,experiments were conducted in the presence of the calcium-activated chloride channel blocker niflumic acid (20 μM;Thermo Scientific) [26, 46, 50]. These experiments were alsoused to determine the saturation kinetics for sodium-coupledglucose transport. To determine the effect of NSP4114–135 onthe rate of glucose-coupled sodium transport across the bio-logical membrane, saturation kinetics studies were repeated inthe presence of glucose and NSP4114–135 on ileal tissuesmounted in Ussing chambers. Increasing concentrations ofglucose were added to the lumenal side to elicit an increasein Isc in the presence and absence of NSP4114–135 and niflumicacid. The Vmax and KM were determined by fitting data to theHill equation. To eliminate the glucose and NSP4-stimulatedchloride secretion and determine only the glucose-stimulatedsodium absorption, we first attempted to block the NSP4-stimulated chloride secretion using niflumic acid orCaCCinh-A01 before determining the saturation kinetics ofglucose-stimulated sodium absorption. Addition of the inhib-itors prior to performing the glucose-stimulated sodium satu-ration kinetics resulted in poor tissue viability and ambiguousresults. Therefore, all further experiments were performedwithout the inhibitors.

The presence of basal electroneutral Na+-H+ exchange ac-tivity was assessed using 10 μM of 5-(N,N-hexamethylene)amiloride (HMA; Sigma), a selective inhibitor of Na+-H+ ex-change protein 3 (NHE3) [29]. Basal Cl−-HCO3

− exchangeactivity was inhibited using 100 μMof 4′-diisothiocyano-2,2′-stilbenedisulfonic acid (DIDS; Thermo Scientific, Waltham,MA), an anion exchange inhibitor that was dissolved inDMSO, so that the total concentration of the DMSO did notexceed 0.1% in the final working Ringer solution [8, 17, 18].

Ileal sac assays

An ex vivo method using 10-cm-long ileal segments ligated atone end and then filled with 200 μl of Ringer solution con-taining glucose (8 mM) or glucose and a calcium-activatedchloride channel blocker (e.g., niflumic acid) both in the pres-ence and absence of NSP4114–135 were used to measure watermovement. The loops were incubated for 1 h in Ringer solu-tion bubbled with 95% O2 and 5% CO2. The beaker wasslowly agitated in a water bath at 37 °C. The percentage ofnet fluid absorption was calculated using the formula (I −O)/I,where I = volume of fluid added to the lumen andO = volumeof fluid in the lumen at the end of the incubation time.

Immunohistochemistry and western blot analysis

Immunohistochemistry was performed using methodsdescribed previously [54, 55]. Ileal tissues collected fol-lowing ileal sacs assays were fixed in formalin (10%)for 20 h at room temperature and then embedded inparaffin. Tissue sections (4 μm) obtained using a micro-tome (Leica RM2245, Leica Microsystems, Inc., IL,USA) were used. Primary antibodies against Ano1(Cat# AB53212 from Abcam, USA) and fluorophore-conjugated secondary antibody (AlexaFluor 647 GoatAnti-Rabbit IgG, A-11034, Life Technologies Corp.,Carlsbad, CA, USA) was used before being imaged un-der a confocal laser-scanning microscope (Fluoview1000 IX81, Olympus) using a 488-nm laser.

Western blot analysis was performed using brushborder membrane vesicles from intestinal sacs incubat-ed with Ringer solution or with glucose-containingRinger solution both in the presence and absence ofNSP4114–135. The brush border membrane vesicleswere isolated using methods modified from Stiegeret al. [41] and Binder et al. [5].

Protein concentrations were determined using theBradford assay, and protein (30 μg) was resolved byelectrophoresis through SDS-7.5% polyacrylamide gels,as previously described [36, 53]. Blots were subsequent-ly reacted with rabbit anti-Ano1 antibody (1:500–1:1000dilution) followed by peroxidase-coupled secondary an-tibody (Bio-Rad, 1:3000). For ANO1 western analysis,tissues incubated with forskolin, an activator of cAMP-stimulated chloride secretion via the cystic fibrosistransmembrane conductance regulator (CFTR) channel,was used as a negative control. Immunoreactive bandswere visualized by enhanced chemiluminescence (PierceECL Western blotting substrate, Thermo Scientific) andautoradiography. Blots were stained with Ponceau S so-lution (Sigma) to confirm comparable sample loadingand transfer.

Pflugers Arch - Eur J Physiol (2017) 469:1093–1105 1095

Lentiviral-mediated siRNA targeted against Ano1in Caco-2 cells

Post-confluent Caco-2 cells were used, because smallintestinal cells failed to adhere to the coverslips. Ano1was silenced using a short interfering RNA sequence(5′-AAAAGCACGATTGTCTATGAGAT-3′) that flanksexons 5 and 6 to determine if Ano1 is responsible forthe increased chloride secretion in the presence of NSP4and glucose. Intracellular calcium measurements and Iscwere recorded in the Caco-2 cells transfected with aninducible siRNA plasmid encoding for a sequence di-rected at silencing Ano1 (siRNA) or a non-silencing(N.S.) scrambled sequence as a control. For negativesiRNA control, scrambled RNA sequences with noknown targets in Caco-2 cells were used (5′-AAAACATGGAGTGGCACGTAGGT-3′). The sequences weredesigned using the National Center for BiotechnologyInformation (NCBI) Basic Local Alignment SearchTool (BLAST). A BLAST search using the selectedsequence was conducted to rule out off-targeting. Thesequences were selected based upon the transfectionefficiency and viability of the cells from four siRNAsequences. These siRNAs were cloned into a LentisiRNA GFP expression system (Applied BiologicalMaterials, Richmond BC; catalog: LV300). After trans-fection, ∼95% of the cells were GFP positive. Caco-2(1 × 105) cells were grown on a Costar 12-mmSnapwell insert in growth medium containing 20%FBS. The transepithelial electrical resistance (TEER)was monitored, as described in our previous publication[53]. The cells selected for transfection had a TEERbetween 400 and 500 Ω/cm2.

Confocal Ca2+ fluorescence microscopy

Caco-2 cells grown for 22 days post-confluence devel-oped characteristics of small intestinal epithelium [39].Caco-2 cells grown on coverslips were used for confo-cal calcium fluorescence microscopy with a RC-21BRbath chamber (25 mm in diameter) attached to aSeries 20 stage adapter (Warner Instruments, LLC,Hamden, CT, USA). A single-channel tabletop tempera-ture controller (TC-324B, Warner Instruments) main-tained cells at 37 °C. Cells were loaded with 0.5 μMof Rhod-2 AM (Cat #R1245MP, Thermo Fisher, USA)on a fluorescent calcium indicator and incubated for45 min at room temperature. Confocal laser-scanningmicroscopy was performed using an inverted Fluoview1000 IX81 microscope (Olympus, Tokyo, Japan) with aU Plan S-Apo 20× objective with an excitation wave-length of 552 nm and emission wavelength of 581 nm.Ringer solution with NSP4, glucose or with 1,2-Bis(2-

aminophenoxy)ethane-N ,N ,N ′ ,N ′- te t raacet ic acidtetrakis(acetoxymethyl ester) (BAPTA-AM; Sigma), acell-permeable, selective calcium chelator was added tothe bath chamber using a multivalve perfusion system(VC-8, Warner Instruments). Fluorescence intensity wasdetermined from the mean peak amplitude and normal-ized to baseline to obtain F/F0. This experiment wasrepeated 10 times; one representative experiment isshown (Supplementary Fig. 1). At the end of the exper-iment, the Ringer solution (RR) was used to wash thesecretagogues from the cells.

Statistical analysis

Results are presented as means ± SEM. Statistical anal-yses were performed with paired and unpaired t tests.Analysis of variance (ANOVA) was used to comparedifferences between several experimental control andnonparametric Mann-Whitney U tests was used tocompare between groups. p < 0.05 was consideredsignificant.

Results

Mice at 3 weeks of age showed a greater NSP4-stimulatedchloride secretion and minimal electroneutral sodiumand chloride absorption

Since rotavirus infection is particularly severe in chil-dren, chloride secretory activity and sodium absorptionwith age were studied in 3-, 6-, and 8-week-old mice inthe presence of NSP4114–135 and/or glucose. Net chlo-ride absorption increased with age of the mice in intes-tines (Fig. 1a, black bars), but in the presence ofNSP4114–135 maximum secretion was observed in 3-week-old mice when compared to 6- and 8-week-oldmice (Fig. 1a, hatched bars). Increased JnetCl in 3-week-old mice resulted mostly from a decrease inJmsCl; however, the difference was not statistically sig-nificant (Supplementary Table 1). The addition of DIDS(100 μM), an anion exchange blocker used to inhibitCl−-HCO3

− exchange, did not cause significant inhibi-tion of NSP4-induced anion secretion in 3-week-oldmice (−2.1 ± 0.2 vs −2.2 ± 0.2 μeq h−1 cm−2), 6-week-old (−1.3 ± 0.5 vs −1.8 ± 0.5 μeq h−1 cm−2), or8 - w e e k - w e e k - o l d m i c e ( − 0 . 7 ± 0 . 2 v s−1.0 ± 0.6 μeq h−1 cm−2). As such, most of our subse-quent studies were performed using 3-week-old mice.

Since electroneutral sodium transport is a major mecha-nism of sodium absorption in the small intestine, we assessedthe function of NHE3 in the presence and absence of HMA, arelatively specific blocker of NHE3. As with chloride

1096 Pflugers Arch - Eur J Physiol (2017) 469:1093–1105

absorption, sodium absorption was maximal in 8-week-oldmice (Fig. 1b, Supplementary Table 1). The addition of10μMofHMA-reduced baseline levels of net sodium absorp-tion (Fig. 1b) in the intestinal tissues of 3-, 6-, and 8-week-oldmice. Unidirectional flux studies showed that the decrease inJnetNa was because of a decreased JmsNa following the addi-tion of HMA. This decrease was maximal in 8-week-old mice16.3 ± 0.6· vs 13.4 ± 0.7 μeq h−1 cm−2) when compared to 3-and 6-week-old mice (Supplementary Table 1), emphasizingthe importance of electroneutral sodium absorption thatincreases with age.

Tissues from 3-week-old mice incubated with NSP4 result-ed in a significant decrease in JNetNa (0.7 ± 0.2 vs−0.1 ± 0.4 μeq h−1 cm−2). The addition of HMA in the pres-ence of NSP4 did not decrease JNetNa beyond the value ob-served with NSP4 alone, suggesting a lack of NHE3 activityin the presence of NSP4 (Fig. 1c). The addition of glucose totissues from 3-week-old mice substantially increased Isc(1.1 ± 0.2 vs 3.5 ± 1.1 μeq h−1 cm−2, p < 0.05, n = 15), butthe Jnet sodium in the presence of glucose was not significant-ly different when compared to tissue bathed in Ringer solution(0.7 ± 0.2 vs 1.2 ± 0.4 μeq cm−2 h−1). These studies suggestthat in 3-week-old mice, the major portion of the glucose-stimulated increase in Isc is not due to glucose-coupled sodiumabsorption. Tissues incubated with glucose and NSP4 did notresult in a significant increase in JNetNa when compared totissues incubated with NSP4 alone (0.3 ± 0.2 vs−0.1 ± 0.4 μeq h−1 cm−2). The addition of HMA to the mu-cosal side of tissues incubated with NSP4 and glucose did notsignificantly inhibit JnetNa. These studies suggest that NSP4-mediated attenuation of NHE3-mediated sodium absorption isnot corrected in the presence of glucose. The addition of NSP4to small intestinal tissues from 3-week-old (39.2 ± 1.2 vs

47 ± 1.9 mS, p < 0.001, n = 20) and 8-week-old (43.3 ± 1.2vs 52.7 ± 2.1 mS, p < 0.001, n = 20) mice significantly in-creased conductance. The addition of glucose to the tissuestreated with NSP4 did not significantly decrease conductancein 3-week-old (47 ± 1.9 vs 44.6 ± 2.2 mS, n = 20) or 8-week-old mice (52.7 ± 2.1 vs 58.2 ± 5.4 mS, n = 20).

NSP4114–135 increased short-circuit current and netchloride secretion in ileal segments via a calcium-activatedchloride channel

The addition of 7 μM of NSP4114–135 to the lumenal sideof ileal tissues from 3-week-old mice led to a significantincrease in Isc (2.3 ± 0.1 vs 1.6 ± 0.1 μeq h−1 cm−2,p < 0.05, Fig. 2a). Niflumic acid (20 μM), a calcium-activated chloride channel blocker, completely abolishedthe NSP4114–135-stimulated increase in Isc (Fig. 2a).Similar results were observed for the calcium-activatedchloride channel blocker CaCC(inh)-A01 (data notshown). Since there was no significant difference in thedegree of inhibition of NSP4114–135-stimulated change inIsc in the presence of niflumic acid compared toCaCC(inh)-A01, we continued to use niflumic acid insubsequent experiments.

Flux studies to specifically examine net chloride flux(JnetCl

−) showed minimal basal absorption of chloride(0.2 ± 0.3 μeq h−1 cm−2, Fig. 2b). The addition of 7 μM ofNSP4114–135 to the lumenal side resulted in a significant de-crease in JnetCl

− and a shift from net absorption to net secre-tion (−2.1 ± 0.4 μeq h−1 cm−2, p < 0.001, Fig. 2b). The addi-tion of 10 μM of niflumic acid in the presence of NSP4114–135fully restored net chloride absorption (0.6 ± 0.2 μeq h−1 cm−2,Fig. 2b). Unidirectional flux studies demonstrated that the net

Fig. 1 Three-week-old mice display a greater NSP4-stimulated chloridesecretion and reduced sodium absorption than do older mice. Ilealsegments were obtained from 3-, 6-, and 8-week-old NIH Swiss mice. aNet chloride flux (JNetCl) was assessed by measuring 36Cl flux in Ussingchamber studies.Hatched bars show JNetCl in the first 45min under basalconditions. Solid black bars show JNetCl in the presence of 7 μMNSP4114–135. b Net sodium flux (JNetNa) was determined by 22Na fluxstudies in tissues mounted in Ussing chambers. The addition of NSP4114–135 abolished sodium absorption (cross-hatched bar); the addition of10 μM of HMA, a selective inhibitor of NHE3, did not further decrease

sodium absorption (solid bar). In separate experiments, the addition ofHMA under basal conditions decreased basal JNetNa (hatched bars), andthe addition of HMA to tissues incubated with NSP4114–135 and glucosein an Ussing chamber abolished JNetNa (gray cross-hatched bars). c Ilealsegments obtained from 3-week-old NIH Swiss mice were mounted in anUssing chamber. Basal JNetNa (hatched bars) and 10 μM of HMA-sensitive JNetNa (solid bars) showed a maximal inhibition in 3-week-old mice. Studies are from n = 10 intestinal segments per group. NS notsignificant

Pflugers Arch - Eur J Physiol (2017) 469:1093–1105 1097

chloride secretion in the presence of NSP4114–135 was due to asignificant increase in JsmCl, without a significant change inJmsCl (Supplementary Table 1). Taken together, these findingssuggest that NSP4114–135 activates niflumic acid-sensitive,electrogenic chloride secretion.

Glucose enhanced the increases in short-circuit currentand chloride secretion observed in ileal segments exposedto NSP4144–135

The combination of glucose and NSP4114–135 significantlyincreased Isc compared to NSP4114–135 alone [5.4 ± 0.3 vs2.3 ± 0.1 μeq h−1 cm−2, p < 0.001, from n = 62 (NSP4 andglucose) and n = 12 (NSP4), Fig. 2c]. The addition of niflumicacid again decreased Isc (1.4 ± 0.2 vs 5.4 ± 0.3 μeq h−1 cm−2,p < 0.001, Fig. 2c).

Studies of JNetCl confirmed that the glucose-induced changein Isc represented, at least in part, changes in chloride transport.Glucose and NSP4114–135 in combination increased chloride se-cretion beyond the increase observed with NSP4 alone(−4.9 ± 0.7 μeq h−1 cm−2, Fig. 2d vs. −2.1 ± 0.2 μeq h−1 cm−2,

p < 0.001, Fig. 2b). The niflumic acid-inhibitable current wassignificantly higher in the presence of NSP4 and glucose com-pared to NSP4 alone (4.0 vs 1.2 μeq h−1 cm−2) (SupplementaryTable 1). Niflumic acid restored baseline chloride absorption(0.2 ± 0.4 μeq h−1 cm−2, Fig. 2d), suggesting inhibition ofcalcium-activated chloride secretion.

NSP4 reduced glucose-coupled sodium absorption

Since glucose failed to stimulate robust sodium absorp-tion in NSP4-treated tissues, we assessed saturation ki-netics for sodium-coupled glucose transport (SGLT1). Inthe presence or absence of NSP4114–135, the Isc forintestinal segments exposed to glucose at low concen-trations increased in a near-linear fashion with an in-creasing substrate concentration. The Isc values includecontributions of both glucose-stimulated sodium currentand NSP4-stimulated electrogenic chloride current. Thiswould explain why the Isc values were higher in thepresence of NSP4 than in its absence. The addition ofinhibitors of the calcium-activated chloride channel

Fig. 2 NSP4114–135 increases short-circuit current and net chloridesecretion in ileal segments via a niflumic acid-sensitive chloridechannel, and glucose further increases short-circuit current and chloridesecretion. Ileal segments of 3-week-old mice were mounted in Ussingchambers. a Short-circuit current (Isc) and b net chloride flux (JNetCl)were measured after 45 min under basal conditions (diamonds) in thepresence of 7 μM of NSP4114–135 added to the lumenal side of theintestine (circles) or both 7 μM of NSP4114–135 and 10 μM of niflumicacid (NFA), an inhibitor of calcium-activated chloride channels (squares).

c Isc was measured in the presence of 7 μMofNSP4114–135 (open circles);7 μM of NSP4 and 8 mM of glucose (closed triangle); and NSP4114–135,glucose, and niflumic acid (open inverted triangle). d Net chloride flux(JNetCl) was measured under basal conditions (closed diamonds), in thepresence of 7 μMofNSP4114–135 and 8 mMof glucose (closed triangles),or in the presence of 7 μM of NSP4114–135, 8 mM of glucose, and 10 μMniflumic acid (NFA) (open inverted triangles). n = 21 intestinal segmentsper group for a, 10 tissue pairs for b, and 9 intestinal tissue pairs per group(c and d)

1098 Pflugers Arch - Eur J Physiol (2017) 469:1093–1105

resulted in poor tissue viability for the duration of theexperiment. Therefore, further experiments were per-formed in the absence of ANO1 inhibitors. At higherconcentrations of glucose, the Isc no longer increasedin proportion to the increase in glucose concentration,suggesting saturation of the SGLT1 transporter. Thissaturation occurred at a higher concentration in the pres-ence of NSP4 (Fig. 3a). With glucose Vmax, the maxi-mum increase in Isc was 2.6 ± 0.2 μeq h−1 cm−2 in theabsence of NSP4 and 5.2 ± 0.5 μeq h−1 cm−2 in thepresence of NSP4. KM, the Michaelis constant or theglucose concentration that gave rise to half-maximal in-crease in Isc, was 0.7 ± 0.1 mM in the absence of NSP4and 3.2 ± 0.1 mM glucose in the presence of NSP4(Fig. 3b).

Glucose-mediated improvements in NSP4-induced fluidsecretion were limited by calcium-activated chloridesecretion

Studies on ex vivo ileal loops confirmed that there was nosignificant change in fluid absorption with Ringer solutionand niflumic acid. The addition of glucose to the Ringer solu-tion increased net fluid absorption (18.64 ± 3.83 vs.30.64 ± 2.81, p < 0.01), and the addition of niflumic acidfurther increased absorption (to 47.09 ± 3.55, p < 0.01). Ilealloops incubated with Ringer solution containing NSP4114–135displayed a net secretion that was converted to net absorptionin the presence of niflumic acid (−9.72 ± 5.46 vs. 3.27 ± 1.2,p < 0.001, Fig. 4), which is consistent with rotavirus causingniflumic acid-sensitive, electrogenic chloride secretion.Glucose also converted NSP4-induced fluid secretion to fluidabsorption (7.8 ± 4.2 vs. −9.7 ± 5.5, p < 0.001, Fig. 4), andniflumic acid further increased net fluid absorption (7.8 ± 4.2vs. 38.7 ± 5.0, p < 0.001, Fig. 4). This observation suggeststhat glucose-mediated rehydration is limited by concomitantcalcium-activated chloride secretion and fluid secretion.

Glucose enhanced the NSP4-induced increasesin intracellular calcium and Isc in post-confluent Caco-2cells

In both the silencing and N.S. siRNA transfected cells, theamplitude of calcium fluorescence was sustained for atleast 200 s (Supplementary Fig. 1) in the baseline condi-tions. The NSP4114–135-induced increase in calcium fluo-rescence was abolished by the addition of BAPTA-AM tothe bath solutions (F/Fo of 1.6 ± 0.2 vs 1.1 ± 0.1; p < 0.05)in N.S. cells and siRNA cells (F/Fo of 1.5 ± 0.1 vs1.1 ± 0.1; p < 0.05). Glucose added to the bathing mediumof Caco-2 cells of the N.S. and siRNA groups increased

Fig. 3 Glucose saturation kinetics show increased KM in the presence ofNSP4. Ileal segments obtained from 3-week-old mice were mounted inUssing chambers. a The addition of increasing concentrations of glucose(0–13 mM) to the lumen side under basal conditions or in the presence of

7 μM of NSP4114–135 yielded a dose-dependent increase in Isc withsaturable kinetics. b The Michaelis constant (KM) for glucose wascalculated by fitting the Isc data from a to the Hill equation. n = 12intestinal segments per glucose concentration

Fig. 4 Glucose-mediated improvements in NSP4-induced fluid secretionare limited by calcium-activated chloride secretion. Ileal segmentsapproximately 10 cm in length were obtained from 3-week-old miceand ligated to form sacs. The sacs were filled with 200 μL of Ringersolution with or without 7 μM of NSP4114–135, 8 mM of glucose, and/or 10 μM of niflumic acid (NFA) and incubated for 1 h before the finalvolume of fluid in the sac was measured. n = 6 intestinal sacs percondition

Pflugers Arch - Eur J Physiol (2017) 469:1093–1105 1099

amplitude of calcium fluorescence (Fig. 5a); this wasinhibited by BAPTA-AM with a mean peak amplitude(F/Fo) of 1.5 ± 0.1 vs 1.0 ± 0.1 and 1.3 ± 0.1 vs0.9 ± 0.1, p < 0.05, respectively (Fig. 5a). The additionof BAPTA-AM to the bathing medium significantly de-creased calcium amplitude to calcium fluorescence(1.0 ± 0.1 vs 0.9 ± 0.1), whereas the addition of glucosein the continued presence of NSP4114–135 resulted in sig-nificantly higher amplitude of calcium fluorescence insiRNA [F/Fo of 1.3 ± 0.1 vs 1.9 ± 0.2 (p < 0.01)] andN.S. transfected cells (1.6 ± 0.1 vs 2.1 ± 0.3 p < 0.001,Fig. 5b). These results confirm that both NSP4 and glucoseincrease intracellular calcium and that the intracellular cal-cium levels in the presence of NSP4114–135 and/or glucoseare not affected by Ano1 silencing. Western blot analysesusing the cell lysates from siRNA and N.S. cells showedthat Ano1 protein levels were significantly lower in siRNAcells, suggesting successful transfection (Fig. 5d).

Human intestinal Caco-2 cells were transfected withsiRNA targeted atAno1. Cells transfectedwithAno1 siRNA

had a TEER of 550 ± 13 Ω/cm2, and cells transfected withnonspecific siRNA had a TEER of 558 ± 8Ω/cm2. Similarly,basalIscwasnotsignificantlydifferentbetweenthesiRNAandN.S. groups (0.1±0.01vs0.1±0.02μeqh−1 cm−2, Fig. 5c).A15-min incubation with NSP4114–135 stimulated a negativecurrent in N.S. cells when compared to siRNA cells(−0.04 ± 0.01 vs 0.1 ± 0.01μeq h−1 cm−2) but did not lead to asignificant increase in Isc (Fig. 5c); these results contrastwiththose observed for incubations ofNSP4114–135with ileal seg-ments (Fig.2a).Theadditionofglucose toNSP4-treatedcellsresulted in a negative Isc only in N.S. cells (−0.1 ± 0.03 vs0.1±0.01μeqh−1cm−2,Fig.6c),similartotheresultswithilealsegments incubated with glucose and NSP4114–135 (Fig. 1c).The negative Isc was partially decreased by the addition ofCaCC(inh)-A01 to the apical sideof the chamber inN.S. cellsand not in siRNAcells (p<0.05, Fig. 5c). Changes in Iscwereassociatedwithchanges inTEER.TheTEERforcells inbasalRinger solutionwas 540± 13/8Ω cm−2,whereas cells treatedwith NSP4 and glucose had resistances of 244.4 ± 28.3 and240.4±37.9Ω cm−2, respectively (p<0.01).

Fig. 5 Glucose enhances NSP4-induced increases in intracellularcalcium and Isc in post-confluent Caco-2 cells. a NSP4114–135 (7 μM) orglucose (8 mM) increased intracellular calcium in Caco-2 cells withsilencing Ano1 (siRNA) or a non-silencing (N.S.) scrambled sequence.b Glucose increased the NSP4114–135-stimulated increase in intracellularcalcium in siRNA and N.S. Caco-2 cells. The addition of BAPTA-AM(calcium chelator; 50 μM) to the bath solution decreased intracellularcalcium fluorescent intensity. The values are from n = 10 coverslips percondition. c Isc measurements in siRNA and N.S. Caco-2 cells. NSP4114–135, glucose, and CaCC(inh)-A01 was added in sequence to control

tissues under basal conditions. d Western blot analysis showing Ano1protein levels in whole-cell lysate from siRNA and N.S. Caco-2 cells. eWestern blot analysis showing Ano1 protein levels in brush bordermembrane vesicles (BBMV) prepared from intestinal tissues from miceincubated with Ringer solution in the presence or absence of 8 mM ofglucose, 7 μM of NSP4114–135, 8 mM of glucose plus 7 μM of NSP4114–135, forskolin (negative control), or carbachol (positive control). Thisexperiment was repeated five times; one representative experiment isshown

1100 Pflugers Arch - Eur J Physiol (2017) 469:1093–1105

NSP4 and glucose in combination increased protein levelsand expression of Ano1

We previously observed that Ano1 expression is increased inileal tissues exposed to glucose [53], and Ano1 function hasbeen linked to rotaviral pathogenesis [34]. The molecularweight of Ano1 was found to be different in human cells andmouse intestinal epithelial cells because of the complete dele-tion of exon 1 and exon 2 in mouse ANO1 when compared tohuman ANO1. The possibility of such deletions has been pre-viously reported [30]. However, ileal tissues treated with glu-cose and NSP4 in combination showed a substantial increase inAno1 protein levels (Fig. 5e). Carbachol, a cholinergic drug,was used as a positive control. Ano1 protein levels were in-creased in brush border membrane vesicles (BBMV) isolatedfrom intestinal tissues incubated with carbachol. Intestinal tis-sues incubated with forskolin had a small decrease in Ano1protein levels and were used as a negative control (Fig. 5e).Ileal tissues incubated with Ringer solution showed a minimalexpression of Ano1, while tissues incubated with NSP4114–135showed an expression along the basal region of the crypt andvillus (Fig. 6, white arrowheads in middle panel), with no in-crease in expression in the crypt region. However, tissues fromileal loops incubated with NSP4114–135 and glucose in combi-nation showed an increased expression of Ano1 along the brushborder of the villus (arrowheads) and markedly increased im-munofluorescence in the basal one-third of the crypt (Fig. 6,white arrows in right panel).

Discussion

Although an effective vaccine against rotavirus is now beingused in developed countries, rotavirus infection remains a sig-nificant clinical issue in developing and underdevelopedcountries, especially in children <5 years of age. Theontogenic nature of the NHE3 transporter, the predominanttransporter of electroneutral electrolyte absorption, has beendemonstrated in mouse small intestines. Following entry intosmall intestinal enterocytes, rotavirus replicates in the nondi-viding mature enterocytes, damaging the villus tip and secret-ing the enterotoxin NSP4. Destruction of the mature epithelialcells in the villus that are responsible for electrolyte absorptioncauses electrolyte and nutrient absorption defects; the NSP4protein increases intracellular calcium, resulting in electrogen-ic chloride secretion and diarrhea [16, 44]. Secretion of chlo-ride via an electrogenic process shifts the potential of the lu-men of the intestine to more negative values, and the cellbecome more negative with every chloride secreted.Accordingly, the most important counteraction followingchloride secretion is the secretion of potassium, which is awell-documented process that must be corrected during rehy-dration. In addition to chloride, bicarbonate is also lost fromstool; oftentimes, diarrheal stool is referred to as Bplasma-like,^ as both have similar electrolyte content. Since the bodyhas ∼25 mM of bicarbonate, a patient with acute severe diar-rhea is likely to develop metabolic acidosis. Lack of electro-lyte and fluid absorption and active chloride secretion

Fig. 6 NSP4 and glucose in combination increase villous and cryptexpression of the calcium-activated chloride channel anoctamin 1(Ano1Paraffin sections (4 μm) were obtained from ileal sacs incubatedfor 1 h with Ringer solution in the presence or absence of 7 μM ofNSP4114–135 or 7 μM of NSP4114–135 plus 8 mM of glucose.

Arrowheads indicate Ano1 expression along the villous brush border,and arrows indicate Ano1 expression in the intestinal crypts. Nuclei arestained with DAPI in blue, and corresponding white-field images areshown in the lower panels. Staining was repeated at least five times;one representative section per condition is shown

Pflugers Arch - Eur J Physiol (2017) 469:1093–1105 1101

stimulated by NSP4 can cause a variety of symptoms, includ-ing watery diarrhea, nausea, and vomiting. The loss of fluidand electrolytes associated with gastroenteritis can lead to de-hydration, electrolyte imbalances, and even death. Chloridesecretion has been implicated in the pathogenesis of manyenteric infections, including those caused by Vibrio cholerae,rotavirus, and enterotoxigenic Escherichia coli. For all ofthese enteric infections, glucose-containing ORS remain thestandard of care [14, 51].

Na+ absorption mediated by SGLT1 is the driving force forpassive water absorption from the lumen. Although WHO-HO-ORS has dramatically reduced mortality for children, ithas been less efficacious in developing countries. This couldbe attributed to the failure of WHO-HO-ORS to reduce stoolvolume and frequency in some children with diarrheal dis-eases. We previously showed that glucose increases intracel-lular calcium [53], which then stimulates calcium-activatedchloride secretion. The present study illustrates the impact ofglucose on electrolyte transport in the context of rotavirusinfection, as NSP4 activates calcium-activated chloride secre-tion that is further enhanced in the presence of glucose.

Our studies on 3-, 6-, and 8-week-old mice showed that thelowest baseline NHE3 function and reduced electrolyteabsorption occurred at 3 weeks of age, which progressivelyincreased with age (Fig. 1). NSP4 also inhibited already pre-carious sodium and chloride absorption in these young miceand activated electrogenic calcium-activated chloride secre-tion, shifting the balance to net electrolyte and fluid secretion(Fig. 1a, c). Glucose-containing WHO-HO-ORS further en-hanced net chloride secretion by increasing intracellular calci-um and the expression and localization of Ano1. In addition,sodium-coupled glucose absorption was reduced in the pres-ence of NSP4, suggesting decreased SGLT1 activity. The KM

for glucose-mediated sodium absorption was increased in thepresence of NSP4 (Fig. 3). Others have achieved similar re-sults and found that NSP4 inhibited SGLT1 in the small

intestines of young rabbits [13]. These studies demonstratethat glucose-mediated hydration mechanisms are impaired/reduced in the presence of NSP4.

Glucose can also stimulate sodium absorption via NHE3 inthe small intestines of normal mice and mice exposed to chol-era toxin [15, 27, 45]. In the presence of NSP4, glucose didnot compensate for the loss of SGLT1 activity. Moreover, inthe presence of NSP4 and/or glucose, no significant sodiumabsorption occurred (Fig. 1c). Although, in 3-week-old mice,addition of glucose to tissues substantially increased Isc, it didnot translate into an increase in Jnet sodium when compared totissue bathed in Ringer solution. These studies suggest that in3-week-old mice, the major portion of the glucose-stimulatedincrease in Isc is not due to glucose-coupled sodium absorp-tion. Similarly, addition of glucose to NSP4 treated tissues didnot result in a significant increase in Jnet sodium. The reducedelectrolyte absorption and reduced ability to activate sodiumabsorption in the very young mice might help explain theseverity of rotavirus-induced diarrhea as well as the failureof glucose to exert its effect on SGLT1-mediated and NHE3-mediated sodium absorption. Since NHE3-mediated sodiumabsorption and SGLT1-mediated sodium absorption is knownto occur in the mature and differentiated villous epithelialcells, reduced or minimal sodium absorption suggest de-creased presence of differentiated epithelial cells. This togeth-er with calcium-activated chloride secretion in the presence ofglucose could explain why WHO-HO-ORS does not substan-tially decrease stool volume and frequency in some children,especially in children living in underdeveloped or developingcountries. It has been shown in environmental enteropathythat villous height is substantially reduced and that theglucose-stimulated sodium absorption is also reduced [24].Indeed, we found that niflumic acid caused a substantial de-crease in fluid output/input ratio in blind loop experiments inwhich intestinal tissues were incubated with NSP4 and glu-cose in combination (Fig. 4). These observations support the

Fig. 7 Schematic representationof intestinal villi showing a basalNHE3 activity (electroneutralNa + /H+) and calcium-activatedchloride channel activity (Ano1);b NSP4 increased intracellularcalcium and Ano1 activity anddecreased NHE3 function; and cNSP4 and glucose furtherincreased intracellular calciumand Ano1 activity. Previousstudies by various groups haveshown that NSP4 directlyinterferes with ANO1 [34]

1102 Pflugers Arch - Eur J Physiol (2017) 469:1093–1105

notion that glucose-stimulated chloride secretion may contrib-ute to the continued diarrhea observed in patients treated withWHO-HO-ORS.

Glucose stimulates increases in intracellular calcium,resulting in calcium-activated chloride secretion in normal in-testines [4, 32, 40, 43, 53] and NSP4-exposed intestines(Fig. 2c). Notably, treatment with an inhibitor of calcium-activated chloride secretion reduced fluid output in intestinesexposed to both glucose and NSP4 (Fig. 4). These observationsmay explain why WHO-ORS often fails to reduce stool vol-ume, a phenomenon that may partially account for the persis-tent and alarming underutilization of ORS to treat infectiousgastroenteritis [3, 11]. Efforts to modify and improve ORS todecrease stool output have been only modestly successful [3].This is perhaps not surprising given the evidence showing thatglucose can stimulate active Cl− secretion, thereby contributingto watery diarrhea [4, 32, 40, 43, 53]. This raises the possibilitythat a non-glucose-based ORS might more effectively promotehydration in some individuals with infectious diarrhea (Fig. 7).

The mature and differentiated epithelial cells of the intestinalvilli are thought to be primarily responsible for absorption ofelectrolytes and nutrients, whereas crypt epithelial cells are gen-erally responsible for active electrolyte secretion.Major transportproteins involved in chloride secretion include CFTR andcalcium-activated chloride channels. CFTR, which is mainly re-stricted to the crypt cells, has been well-characterized for its rolein cAMP-activated chloride secretion [42]. Calcium-activatedchloride channels were identified through bioinformatic analysis,with Ano1/TMEM16A the first to be cloned [52]. Ano1 andCFTR play important roles in intestinal fluid secretion in normaland diseased states [33, 37]. Studies have shown that rotavirus-infectedmice treatedwith theAno1-selective inhibitor CaCCinh-A01 have reduced stool volume compared to control mice [22].Furthermore, NSP4 has been shown to induce diarrhea by acti-vating Ano1 via increases in intracellular calcium and activationof the phospholipase-C/inositol triphosphate pathway [10]. Ourobservation of increased Ano1 expression in both the villus andcrypt of intestines exposed to NSP4 and glucose suggests that, inpathogenic states, secretory activity may be present at both re-gions of the intestinal lining. The presence of ANO1 in the cryptand villus suggests that NSP4-stimulated anion secretion couldoccur from both the crypt and the villus. The secretory responseof certain chloride channels to signaling cascades triggered bytoxins has evolved over time to flush out pathogens.Accordingly, the therapeutic goal is to restore electrolyte andwater homeostasis but not necessary to reach a Bnormal^ stoolvolume. Our finding that glucose enhances chloride secretionand fails to enhance sodium absorption should be consideredwhen using glucose-based oral rehydration therapy in the man-agement of rotaviral diarrhea and may also have advantages indiarrheal diseases that are associated with villous atrophy and/orother diarrheal disease conditions associated with calcium-activated chloride secretion.

Compliance with ethical standards

Funding The work was partially funded through grants from (1) TheNational Space Biomedical Research Institute (NSBRI) through NASANCC 9-58, RE02901, and a research agreement with the University ofFlorida; (2) Florida High Tech Corridor Council (FHTCC) matchingfunds research program at the University of Florida; and (3) matchingfunds from Entrinsic Health Solutions, Project number P0001295 andAGR00003327.

Conflict of interest Sadasivan Vidyasagar and Paul Okunieff havesignificant shares and royalties in Entrinsic Health Systems, LLC, whichpartially funded the work presented in this article. All other authors haveno conflicts of interest to disclose.

Research involving human participants and/or animals There wereno human participants in this study. All procedures performed in studiesinvolving animals were in accordance with the ethical standards of theinstitution or practice at which the studies were conducted.

Informed consent Not applicable.

References

1. Ball JM, Tian P, Zeng CQ, Morris AP, Estes MK (1996) Age-dependent diarrhea induced by a rotaviral nonstructural glycopro-tein. Science 272:101–104. doi:10.1126/science.272.5258.101

2. Beau I, Berger A, Servin AL (2007) Rotavirus impairs thebiosynthesis of brush-border-associated dipeptidyl peptidaseIV in human enterocyte-l ike Caco-2/TC7 cells. CellMicrobiol 9:779–789. doi:10.1111/j.1462-5822.2006.00827.x

3. Binder HJ, Brown I, Ramakrishna BS, Young GP (2014) Oral re-hydration therapy in the second decade of the twenty-first century.Curr Gastroenterol Rep 16:376. doi:10.1007/s11894-014-0376-2

4. Binder HJ, Powell DW, Curran PF (1972) Effect of hexoses on iontransport in guinea pig ileum. Am J Phys 223:538–544

5. Binder HJ, Stange G, Murer H, Stieger B, Hauri HP (1986)Sodium-proton exchange in colon brush-border membranes. AmJ Phys 251:G382–G390

6. Brunet JP, Cotte-Laffitte J, Linxe C, Quero AM, Geniteau-Legendre M, Servin A (2000) Rotavirus infection induces an in-crease in intracellular calcium concentration in human intestinalepithelial cells: role in microvillar actin alteration. J Virol 74:2323–2332. doi:10.1128/JVI.74.5.2323-2332.2000

7. Carlson JA, Middleton PJ, Szymanski MT, Huber J, Petric M(1978) Fatal rotavirus gastroenteritis: an analysis of 21 cases. AmJ Dis Child 132:477–479

8. Clarke LL, Stien X, Walker NM (2001) Intestinal bicarbonate se-cretion in cystic fibrosis mice. JOP 2:263–267

9. Cortes JE, Curns AT, Tate JE, Cortese MM, Patel MM, ZhouF, Parashar UD (2011) Rotavirus vaccine and health care uti-lization for diarrhea in U.S. children. N Engl J Med 365:1108–1117. doi:10.1056/NEJMoa1000446

10. Dong Y, Zeng CQ, Ball JM, Estes MK, Morris AP (1997) Therotavirus enterotoxin NSP4 mobilizes intracellular calcium in hu-man intestinal cells by stimulating phospholipase C-mediated ino-sitol 1,4,5-trisphosphate production. Proc Natl Acad Sci U S A 94:3960–3965

11. Greenough WB 3rd (1993) A simple solution: curing diarrhea. JDiarrhoeal Dis Res 11:1–5

Pflugers Arch - Eur J Physiol (2017) 469:1093–1105 1103

12. Gurwith M, Wenman W, Hinde D, Feltham S, Greenberg H (1981)A prospective study of rotavirus infection in infants and youngchildren. J Infect Dis 144:218–224

13. Halaihel N, Lievin V, Alvarado F, Vasseur M (2000) Rotavirusinfection impairs intestinal brush-border membrane Na(+)-solutecotransport activities in young rabbits. Am J Physiol GastrointestLiver Physiol 279:G587–G596

14. Hediger MA, Rhoads DB (1994) Molecular physiology of sodium-glucose cotransporters. Physiol Rev 74:993–1026

15. Hu Z, Wang Y, Graham WV, Su L, Musch MW, Turner JR (2006)MAPKAPK-2 is a critical signaling intermediate in NHE3 activa-tion following Na+−glucose cotransport. J Biol Chem 281:24247–24253. doi:10.1074/jbc.M602898200

16. Huang H, Schroeder F, Zeng C, Estes MK, Schoer JK, Ball JM(2001) Membrane interactions of a novel viral enterotoxin: rotavi-rus nonstructural glycoprotein NSP4. Biochemistry 40:4169–4180.doi:10.1021/bi002346s

17. Ikuma M, Geibel J, Binder HJ, Rajendran VM (2003)Characterization of cl-HCO3 exchange in basolateral membraneof rat distal colon. Am J Physiol Cell Physiol 285:C912–C921.doi:10.1152/ajpcell.00396.2002

18. Jessen F, Sjoholm C, Hoffmann EK (1986) Identification of theanion exchange protein of Ehrlich cells: a kinetic analysis of theinhibitory effects of 4,4′-diisothiocyano-2,2′-stilbene-disulfonic ac-id (DIDS) and labeling of membrane proteins with 3H-DIDS. JMembr Biol 92:195–205

19. John J, Sarkar R, Muliyil J, Bhandari N, Bhan MK, Kang G (2014)Rotavirus gastroenteritis in India, 2011-2013: revised estimates ofdisease burden and potential impact of vaccines. Vaccine 32(Suppl1):A5–A9. doi:10.1016/j.vaccine.2014.03.004

20. Jourdan N, Brunet JP, Sapin C, Blais A, Cotte-Laffitte J, Forestier F,Quero AM, Trugnan G, Servin AL (1998) Rotavirus infection re-duces sucrase-isomaltase expression in human intestinal epithelialcells by perturbing protein targeting and organization of microvillarcytoskeleton. J Virol 72:7228–7236

21. Jourdan N, Maurice M, Delautier D, Quero AM, Servin AL,Trugnan G (1997) Rotavirus is released from the apical surface ofcultured human intestinal cells through nonconventional vesiculartransport that bypasses the Golgi apparatus. J Virol 71:8268–8278

22. Ko EA, Jin BJ, Namkung W, Ma T, Thiagarajah JR,Verkman AS (2014) Chloride channel inhibition by a redwine extract and a synthetic small molecule preventsrotaviral secretory diarrhoea in neonatal mice. Gut 63:1120–1129. doi:10.1136/gutjnl-2013-305663

23. Koo HL, Neill FH, Estes MK, Munoz FM, Cameron A, DuPontHL, Atmar RL (2013) Noroviruses: the most common pediatricviral enteric pathogen at a large University hospital after introduc-tion of rotavirus vaccination. J Pediatric Infect Dis Soc 2:57–60.doi:10.1093/jpids/pis070

24. Korpe PS, Petri WA Jr (2012) Environmental enteropathy: criticalimplications of a poorly understood condition. TrendsMolMed 18:328–336. doi:10.1016/j.molmed.2012.04.007

25. Leshem E, Tate JE, Steiner CA, Curns AT, Lopman BA, ParasharUD (2015) Acute gastroenteritis hospitalizations among US chil-dren following implementation of the rotavirus vaccine. JAMA313:2282–2284. doi:10.1001/jama.2015.5571

26. Liantonio A, Giannuzzi V, Picollo A, Babini E, Pusch M,Conte Camerino D (2007) Niflumic acid inhibits chloride con-ductance of rat skeletal muscle by directly inhibiting the CLC-1 channel and by increasing intracellular calcium. Br JPharmacol 150:235–247. doi:10.1038/sj.bjp.0706954

27. Lin R, Murtazina R, Cha B, Chakraborty M, Sarker R, ChenTE, Lin Z, Hogema BM, de Jonge HR, Seidler U, Turner JR,Li X, Kovbasnjuk O, Donowitz M (2011) D-glucose acts viasodium/glucose cotransporter 1 to increase NHE3 in mousejejunal brush border by a Na+/H+ exchange regulatory factor

2-dependent process. Gastroenterology 140:560–571. doi:10.1053/j.gastro.2010.10.042

28. Liu L, Johnson HL, Cousens S, Perin J, Scott S, Lawn JE, Rudan I,Campbell H, Cibulskis R, LiM,Mathers C, Black RE, Child HealthEpidemiology Reference Group of WHO, Unicef (2012) Global,regional, and national causes of child mortality: an updated system-atic analysis for 2010 with time trends since 2000. Lancet 379:2151–2161. doi:10.1016/S0140-6736(12)60560-1

29. Masereel B, Pochet L, Laeckmann D (2003) An overview of inhib-itors of Na(+)/H(+) exchanger. Eur J Med Chem 38:547–554

30. Mazzone A, Bernard CE, Strege PR, Beyder A, Galietta LJ,Pasricha PJ, Rae JL, Parkman HP, Linden DR, Szurszewski JH,Ordog T, Gibbons SJ, Farrugia G (2011) Altered expression ofAno1 variants in human diabetic gastroparesis. J Biol Chem 286:13393–13403. doi:10.1074/jbc.M110.196089

31. Morris AP, Scott JK, Ball JM, Zeng CQ, O’Neal WK, Estes MK(1999) NSP4 elicits age-dependent diarrhea and ca(2+)mediatedI(−) influx into intestinal crypts of CF mice. Am J Phys 277:G431–G444

32. Munck BG (1970) Interactions between lysine, Na+ and cl- trans-port in rat jejunum. Biochim Biophys Acta 203:424–433

33. Ousingsawat J, Martins JR, Schreiber R, Rock JR, Harfe BD,Kunzelmann K (2009) Loss of TMEM16A causes a defect in epi-thelial Ca2+−dependent chloride transport. J Biol Chem 284:28698–28703. doi:10.1074/jbc.M109.012120

34. Ousingsawat J, Mirza M, Tian Y, Roussa E, Schreiber R, Cook DI,Kunzelmann K (2011) Rotavirus toxin NSP4 induces diarrhea byactivation of TMEM16A and inhibition of Na+ absorption. PflugersArch 461:579–589. doi:10.1007/s00424-011-0947-0

35. Payne DC, Staat MA, Edwards KM, Szilagyi PG, WeinbergGA, Hall CB, Chappell J, Curns AT, Wikswo M, Tate JE,Lopman BA, Parashar UD, New Vaccine Surveillance N(2011) Direct and indirect effects of rotavirus vaccination uponchildhood hospitalizations in 3 US counties, 2006-2009. ClinInfect Dis 53:245–253. doi:10.1093/cid/cir307

36. Ranganathan PN, Lu Y, Jiang L, Kim C, Collins JF (2011) Serumceruloplasmin protein expression and activity increases in iron-deficient rats and is further enhanced by higher dietary copper in-take. Blood 118:3146–3153. doi:10.1182/blood-2011-05-352112

37. Rock JR, Futtner CR, Harfe BD (2008) The transmembrane proteinTMEM16A is required for normal development of the murine tra-chea. Dev Biol 321:141–149. doi:10.1016/j.ydbio.2008.06.009

38. Rodriguez WJ, Kim HW, Brandt CD, Schwartz RH, GardnerMK, Jeffries B, Parrott RH, Kaslow RA, Smith JI, KapikianAZ (1987) Longitudinal study of rotavirus infection and gas-troenteritis in families served by a pediatric medical practice:clinical and epidemiologic observations. Pediatr Infect Dis J6:170–176

39. Rousset M (1986) The human colon carcinoma cell lines HT-29 andCaco-2: two in vitro models for the study of intestinal differentia-tion. Biochimie 68:1035–1040

40. Schultz SG, Zalusky R, Gass AE Jr (1964) Ion transport in isolatedrabbit ileum. 3. Chloride fluxes. J Gen Physiol 48:375–378

41. Stieger B, Marxer A, Hauri HP (1986) Isolation of brush-bordermembranes from rat and rabbit colonocytes: is alkaline phosphatasea marker enzyme? J Membr Biol 91:19–31

42. Strong TV, Boehm K, Collins FS (1994) Localization of cysticfibrosis transmembrane conductance regulatormRNA in the humangastrointestinal tract by in situ hybridization. J Clin Invest 93:347–354. doi:10.1172/JCI116966

43. Taylor AE, Wright EM, Schultz SG, Curran PF (1968) Effect ofsugars on ion fluxes in intest-ine. Am J Phys 214:836–842

44. Tian P, Ball JM, Zeng CQ, Estes MK (1996) The rotavirus non-structural glycoprotein NSP4 possesses membrane destabilizationactivity. J Virol 70:6973–6981

1104 Pflugers Arch - Eur J Physiol (2017) 469:1093–1105

45. Turner JR, Black ED (2001) NHE3-dependent cytoplasmic alkalin-ization is triggered by Na(+)-glucose cotransport in intestinal epi-thelia. Am J Physiol Cell Physiol 281:C1533–C1541

46. Verkman AS, Galietta LJ (2009) Chloride channels as drug targets.Nat Rev Drug Discov 8:153–171. doi:10.1038/nrd2780

47. Victora CG, Bryce J, Fontaine O, Monasch R (2000) Reducingdeaths from diarrhoea through oral rehydration therapy. BullWorld Health Organ 78:1246–1255

48. Vidyasagar S, Barmeyer C, Geibel J, Binder HJ, Rajendran VM(2005) Role of short-chain fatty acids in colonic HCO3 secretion.Am J Physiol Gastrointest Liver Physiol 288:G1217–G1226. doi:10.1152/ajpgi.00415.2004

49. Walker CL, Rudan I, Liu L, Nair H, Theodoratou E, BhuttaZA, O’Brien KL, Campbell H, Black RE (2013) Global bur-den of childhood pneumonia and diarrhoea. Lancet 381:1405–1416. doi:10.1016/S0140-6736(13)60222-6

50. White MM, Aylwin M (1990) Niflumic and flufenamic acids arepotent reversible blockers of Ca2(+)-activated cl- channels inXenopus oocytes. Mol Pharmacol 37:720–724

51. Wright EM, Loo DD, Hirayama BA, Turk E (2004) Surprisingversatility of Na+−glucose cotransporters: SLC5. Physiology(Bethesda) 19:370–376. doi:10.1152/physiol.00026.2004

52. Yang YD, Cho H, Koo JY, TakMH, Cho Y, ShimWS, Park SP, LeeJ, Lee B, Kim BM, Raouf R, Shin YK, Oh U (2008) TMEM16Aconfers receptor-activated calcium-dependent chloride conduc-tance. Nature 455:1210–1215. doi:10.1038/nature07313

53. Yin L, Vijaygopal P, MacGregor GG, Menon R, Ranganathan P,Prabhakaran S, Zhang L, Zhang M, Binder HJ, Okunieff P,Vidyasagar S (2014) Glucose stimulates calcium-activated chloridesecretion in small intestinal cells. Am J Physiol Cell Physiol 306:C687–C696. doi:10.1152/ajpcell.00174.2013

54. Zhang H, Ameen N,Melvin JE, Vidyasagar S (2007) Acute inflam-mation alters bicarbonate transport in mouse ileum. J Physiol 581:1221–1233. doi:10.1113/jphysiol.2007.129262

55. Zhang K, Yin L, Zhang M, Parker MD, Binder HJ, Salzman P,Zhang L, Okunieff P, Vidyasagar S (2011) Radiation decreasesmurine small intestinal HCO3- secretion. Int J Radiat Biol 87:878–888. doi:10.3109/09553002.2011.583314

Pflugers Arch - Eur J Physiol (2017) 469:1093–1105 1105