Sodium and chloride absorptive defects in the small intestine in Slc26a6 null mice

TRANSPORTERS

Sodium and chloride absorptive defects in the small intestinein Slc26a6 null mice

Ursula Seidler & Ingrid Rottinghaus & Jutta Hillesheim &

Mingmin Chen & Brigitte Riederer & Anja Krabbenhöft &Regina Engelhardt & Martin Wiemann &

Zhaouhui Wang & Sharon Barone & Michael P. Manns &

Manoocher Soleimani

Received: 16 April 2007 /Revised: 4 June 2007 /Accepted: 3 July 2007 / Published online: 1 September 2007# Springer-Verlag 2007

Abstract PAT1 (Slc26a6) is located on the apical mem-brane of the small intestinal villi, but its role for saltabsorption has not been studied. To ascertain the role ofSlc26a6 in jejunal sodium and chloride absorption, and itsinterplay with NHE3, muscle-stripped jejuna from Slc26a6+/+ and −/− and NHE3 +/+ and −/− mice were mounted inUssing chambers and electrical parameters, and 36Cl− and22Na+ fluxes were measured. In parallel studies, expressionof the apical Na+/H+ exchanger (NHE3) was examined byimmunofluorescence labeling and immunoblot analysis inbrush border membrane (BBM). In the basal state, net Cl−

and Na+ fluxes were absorptive in Slc26a6−/− and +/+jejuni, but significantly decreased in −/− animals. Uponforskolin addition, net Na+ absorption decreased, Iscstrongly increased, and net Cl− flux became secretory inSlc26a6−/− and +/+ jejuni. When luminal glucose wasadded to activate Na+/glucose cotransport, concomitant Cl−

absorption was significantly reduced in Slc26a6 −/− jejuni,while Na+ absorption increased to the same degree inSlc26a6 −/− and +/+ jejuni. Identical experiments in NHE3-deficient jejuni also showed reduced Na+ and Cl− absorp-tion. Results further demonstrated that the lack of NHE3rendered Na+ and Cl− absorption unresponsive to inhibitionby cAMP, but did not affect glucose-driven Na+ and Cl−

absorption. Immunoblotting revealed comparable NHE3abundance and distribution in apical membranes inSlc26a6−/− and +/+ mice. The data strongly suggests thatSlc26a6 acts in concert with NHE3 in electroneutral saltabsorption in the small intestine. Slc26a6 also serves toabsorb Cl− during glucose-driven salt absorption.

Keywords Electrolyte absorption . Intestine .

Anion exchange . Sodium absorption . Chloride secretion .

Glucose absorption

Introduction

Slc26a6 (PAT1, CFEX) is a member of a large, conservedfamily of anion exchangers (Slc26) that encompasses atleast ten distinct genes [2, 8, 11, 12, 18, 24–27, 30, 35, 36,40, 47]. All, with the exception of Slc26a5 (prestin),function as versatile anion exchangers and display limitedand selective tissue expression [2, 8, 11, 12, 18, 24–27, 30,35, 36, 40, 47]. The chloride/base exchanger Slc26a6, alsoknown as PAT1 or CFEX, is located on the apicalmembranes of small intestine, pancreatic duct, and kidneyproximal tubule [14, 18, 24, 36, 41]. The dominant modesof transport mediated by Slc26a6 include Cl�

�HCO�

3

exchange and Cl−/oxalate exchange [6, 16, 21, 32, 44].Studies in Slc26a6 null mice demonstrate significant

Pflugers Arch - Eur J Physiol (2008) 455:757–766DOI 10.1007/s00424-007-0318-z

U. Seidler (*) : I. Rottinghaus : J. Hillesheim :M. Chen :B. Riederer :A. Krabbenhöft :R. Engelhardt :M. P. MannsDepartment of Gastroenterology, Hepatology and Endocrinology,Hannover Medical School,Hannover, Germanye-mail: seidler.ursula@mh-hannover.de

M. WiemannDepartment of Physiology, University of Essen,Essen, Germany

Z. Wang : S. Barone :M. SoleimaniDepartment of Medicine, University of Cincinnati,Cincinnati, OH, USA

M. SoleimaniResearch Services, Veterans Affairs Medical Center,Cincinnati, OH, USA

reductions in baseline bicarbonate secretion and apicalCl�

�HCO�

3 exchange activity in the duodenum [42]. Inaddition, Slc26a6 was shown to mediate the majority ofPGE-2-mediated bicarbonate secretion in the duodenum[37]. Functional studies in perfused duodena of wild-typeand Slc26a6 null mice demonstrated that Slc26a6 mediatesalmost all of the apical Cl�

�HCO�

3 exchange at the tip ofthe villi [34], mirroring the expression pattern of thisexchanger in the duodenum as determined by immunocy-tochemistry [28, 41, 42].

Both Slc26a6 and NHE3, which function as Cl��HCO�

3

and Na+/H+ exchangers, respectively, are coexpressed in theapical membrane of the small intestinal villi (this paper, and[42]), strongly suggesting that the absorption of chloride andsodium in the small intestine may in part be mediated viathese two exchangers working in concert [13, 29, 36, 45]. Inthe present study, we attempted to ascertain the role ofSlc26a6 in chloride and sodium absorption in the jejunum,and its interplay with NHE3, using Slc26a6 and NHE3 komice and their wild-type littermate controls. The resultsshowed that Slc26a6 is involved in electroneutral NaClabsorption (mediated by Na+/H+ and Cl�

�HCO�

3 ex-change), and, surprisingly, in Cl− absorption driven byglucose. The absence of NHE3, on the other hand, rendersCl− absorption unresponsive to cAMP inhibition, but doesnot affect glucose-stimulated Cl− absorption.

Materials and methods

Animal models Slc26a6 ko mice were generated byimplanting targeted 129 SVJ stem cells in a C57/BL6blastocyst and the chimerical mice bred to C57/BL6 and theprogeny studied [42]. For studies at Hannover MedicalSchool, Slc26a6−/− and +/+ mice were bred on a C57/B6background in the animal facility of the Hannover MedicalSchool. NHE3-deficient mice were originally generated atthe University of Cincinnati as described [29], bred forseveral years on the original congenic background at theUniversity of Essen, and then further bred on the original 129background at the Hannover Medical School. Age and sex-matched littermates were used at approximately 4 months ofage. Animal experiments followed approved protocols at theHannover Medical School and local authorities for theregulation of animal welfare (Regierungspräsidium) andthe University of Cincinnati Medical Center. Animals hadaccess to water and chow ad libitum.

Preparation of isolated intestinal mucosa After briefnarcosis with 100%CO2, the mice were killed by cervicaldislocation. The abdomen was opened by a midline incisionand the proximal to mid jejunum, was removed andimmediately placed in ice-cold, oxygenized ringer’s solu-

tion containing 1 μmol/l indomethacin to suppress trauma-induced prostaglandin release. The intestine was openedalong the mesenteric border and stripped of external serosaland muscle layers with fine forceps.

Ussing chamber experiments The jejunal mucosa wasmounted between two chambers with an exposed area of0.625 cm2 and placed in an Ussing chamber. Parafilm “O”rings were used to minimize edge damage to the tissuewhere it was secured between the chamber halves. Tissueswere bathed with HCO�

3 containing solutions on both sidesgassed with 95% O2/5% CO2. The composition (in mil-limolar) was 108 NaCl, 22 NaHCO3, 3 KCl, 1.3 MgSO4,2 CaCl2, 1.5 KH2PO4, and pH 7.4. The serosal bath contain-ed (in millimolar) 8.9 glucose, 10 sodium pyruvate, 10−3

indomethacin, and 10−3 TTX, the luminal bath 8.9 mannitoland 10−2 amiloride (to block potential amiloride-sensitiveNa+ channels). Isc, PD, and tissue conductivity (Gt) wererecorded using the Mussler 6-channel voltage clamp system(Mussler, Aachen, Germany). 22Na+ and 36Cl− flux studieswere performed during voltage clamp to zero PD. Exactly74 kBq/ml 22Na+ or 36Cl− was added either to the serosal ormucosal solutions after reaching stable electrical parame-ters. After stabilization (approximately 20–30 min aftermounting), a 45-min period of equilibration followed, thenaliquots were taken in 15-min intervals (two intervals forbasal flux, two after forskolin, and two after luminalglucose). For the presented results, we used the valuesfrom the second basal flux period, the first period afterforskolin, and the first period after glucose. There were nostatistically significant differences between the valuesobtained in the first and second flux period after forskolinand after glucose. Radioactivity was determined in a liquidscintillation counter, and bidirectional flux rates for therespective substance calculated. The values for Isc representthe average value of the 15-min period, respectively.

Preparation of brush border membranes and Westernanalysis Enterocytes were isolated according to the methodby Van den Berghe et al. [39], which sequentially separatesvillus and crypt cells of the lamina propria by Ca2+

chelation and rapid mechanical vibration. Isolated cells orbrush border membranes were immediately taken up inpreparation buffer (see below) and lysed by sonication.Mouse BBM vesicles from scrapings of small intestinalepithelium were prepared by the divalent cation precipita-tion technique as previously described [31] except for minormodifications as follows. The preparation buffer consistedof 100 mM mannitol, 3 mM EDTA, 1 mM DTT, 4 mMbenzamidin, and 10 mm Hepes at pH 7.5 which in additioncontained 40 μg/ml PMSF, 20 μg/ml leupeptin, 20 μg/mlpepstatin A, and 20 μg/ml antipain. The BBM pellets weresignificantly enriched for alkaline phosphatase (21±2.2-

758 Pflugers Arch - Eur J Physiol (2008) 455:757–766

fold in the Slc26a6+/+ and 24.1±1.8-fold in the Slc26a6−/−)but not for Na+/K+ ATPase, a basolateral marker enzyme(0.7±0.5 in WT vs 1.0±0.7 in KO).

For Western blotting, 30 μg of purified brush bordermembranes, or 100 μg of whole enterocyte lysate weresize-fractionated on 10% SDS polyacrylamide minigels(Amersham Biosciences) under denaturing conditions,transferred to PVDF membranes (Hybond-P, AmershamBiosciences), and blocked with 5% nonfat dry milk in TBS-Tween. Blots were probed with anti-NHE3 antibody,stripped, and probed again for villin, which is an integralBBM membrane protein. Affinity purified anti-rat-NHE3-IgG (alpha-diagnostics international) was used at a con-centration of 0.1 μg/ml, in TBS-Tween overnight at 4°C.Anti-villin antibody (HCT-8, BD biosciences, Pharmingen,Germany) was used in TBS-Tween at a concentration of25 ng/ml. The secondary antibody (goat anti-rabbit IgGconjugated to horseradish peroxidase, KPL) was diluted1:10,000 in TBS-Tween and incubated for 1 h at roomtemperature. The antigen–antibody complexes on the PVDFmembranes were visualized by chemiluminescence (ECLWestern blotting detection reagents, Amersham PharmaciaBiotech) and the image captured on light sensitive imagingfilm (Hyperfilm ECL, Amersham Biosciences). Bands weredetected and digitized by the KODAKDigital Science ImageStation 440CF and the optical density was measured byTotalLab software (Nonlinear Dynamics, Durham, UK). Theratio NHE3/villin was calculated from each individual lane,each representing protein from one wild-type or homozygoteknockout littermate.

Immunofluorescence staining

Mouse jejunum was rinsed with ice-cold phosphate-buffered saline (PBS) and fixed for 2 h at 4°C with 2%paraformaldehyd in PBS. Fixed tissue was rinsed with PBSand transferred to 30% sucrose in PBS overnight. The tissuewas embedded with tissue-freezing medium (TissueTecO.C.T., Sakura). Cryosectioning was done with a microtomecryostat at −20°C and 10-μm-thick sections were collectedon microscope slides (SuperFrost Plus, Menzel-Gläser,Germany). Sections were treated sequentially with PBS for5 min, washing buffer of PBS with 50 mMNH4Cl two timesfor 10 min each, background reducing buffer (DAKO) for20 min for blocking and the first antibody, affinity purifiedanti-rat-NHE3-IgG (alpha-diagnostics international) at aconcentration of 1.7 μg/ml in blocking buffer. Washingfour times for 5 min in washing buffer was followed byblocking for 20 min in 3% goat serum in PBS and twotimes washing for 5 min each. Secondary antibody (AlexaFluor 488-labeled goat anti-rabbit IgG, Molecular Probes)was incubated for 1 h at room temperature in a concentra-

tion of 2 μg/ml in background reducing buffer. After twowashes for 5 min each, a nuclear staining with 2 μg/mlpropidium iodide in PBS for 5 min and another two washesfollowed. Cover slides were imaged on the confocalmicroscope (Leica TCS SP2). Samples were excited with488 nm and emission was collected at 500–550 nm (AlexaFluor) and 600–680 mn (propidium iodide). Sections fromSlc26a6−/− and +/+ jejunum were imaged with identicalconfocal settings.

Materials Nitrocellulose filters and other chemicals werepurchased from Sigma Chemical (St. Louis, MO, USA).22Na and 36Cl were purchased from GE Healthcare LifeSciences, (Freiburg, Germany).

Statistical analyses Values are expressed as mean±SEM.Statistical analysis was examined using Student t test,Wilcoxon rank test, or ANOVA. P<0.05 was consideredstatistically significant.

Results

Cl− transport in Slc26a6 and +/+ small intestine

Bilateral 36Cl− fluxes were measured in isolated Slc26a6+/+and −/− jejunal epithelium at 15-min intervals in the basalstate, or after the addition of forskolin [10−5 M] or luminalglucose [25 mM]. Net fluxes were calculated by subtractingthe serosal to mucosal flux from the mucosal to serosal fluxfor each individual experiment. As shown in Fig. 1a (leftpanel), mucosal to serosal Cl− absorption was decreased inSlc26a6−/− mice. Forskolin-mediated inhibition of Cl−

uptake was similar in the small intestinal epithelium ofSlc26a6−/− and +/+ littermates (Fig. 1a, middle panel).Addition of luminal glucose 30 min after forskolinstimulated chloride (Cl−) absorption in Slc26a6+/+ butnot Slc26a6−/− mice (Fig. 1a, right panel vs middle panel).Serosal to mucosal Cl− flux was not significantly differentin the basal state and increased to a similar degree inresponse to the addition of forskolin in Slc26a6−/− and +/+tissue (Fig. 1b). The resultant net Cl− flux is depicted inFig. 1c and shows a significant reduction in net chlorideabsorption in Slc26−/− jejuna under basal condition and areversal to net chloride secretion in response to forskolin.The addition of glucose reversed the net chloride flux toabsorption only in Slc26a6+/+ but not in Slc26a6−/− jejuna(Fig. 1c). The forskolin-induced ΔIsc was not significantlydifferent in Slc26a6−/− and +/+ tissue, but it is interestingto note that the glucose-induced ΔIsc was significantlylarger in Slc26a6−/−mice (185±16 in +/+ vs 284±26 in −/−tissue, p<0.05, no figure, because of similar data shown in

Pflugers Arch - Eur J Physiol (2008) 455:757–766 759

Fig. 2d). Tissue conductance was 48±12 and 60±14 inSlc26a6+/+ and Slc26a6−/− jejuna, respectively (p>0.05),and did not change significantly after the addition offorskolin. Tissue conductance increased mildly after theaddition of glucose by 8±12 in Slc26a6+/+ jejuna and by18±24 in Slc26a6−/− jejuna. Thus, these results show asignificant disturbance in small intestinal Cl− absorption inSlc26a6−/− small intestine.

22Na+ transport in Slc26a6−/− and +/+ small intestine

We next determined whether the decrease in Cl− absorptionwas paralleled by a decrease in Na+ absorption. Toward thisend, we performed identical experiments as in Fig. 1, using22Na+ as a tracer. The mice were of the same gender andgenetic background as those used for the Cl− flux experi-ments. The absorption, secretion, and net flux of radio-labeled sodium are shown in Fig. 2a–c. As noted, thebilateral Na+ flux rates were significantly higher than theCl− flux rates, reflecting the fact that small intestinal tightjunctions are cation-selective. However, basal net Na+ andCl− absorptive rates were similar. Net Na+ absorption wassignificantly reduced in Slc26a6−/− compared to +/+jejunum, and this was because of a reduction in mucosalto serosal Na+ flux rates (Fig. 2a,c). Significant inhibitionof Na+ flux by forskolin was observed in Slc26a6−/− and

+/+ tissue (Fig. 2a,c). Glucose-stimulated Na+ absorptiverates were not statistically different Slc26a6−/− and +/+tissue (Fig. 2a,c). Taken together, these results indicate thatelectroneutral Na+ and Cl− absorption is reduced in Slc26a6−/− compared to +/+ jejunum. It is interesting to note thatglucose-stimulated Cl− absorption is reduced in Slc26a6−/−mice (Fig. 1) whereas glucose-stimulated Na+ absorption isnot affected (Fig. 2a). The forskolin-mediated ΔIsc was170±17 in Slc26a6+/+ and 147±13 in −/− tissue; glucose-mediated Isc was larger in the Slc26a6−/− tissue (230±33in WT vs 331±39 in KO, p<0.05) (Fig. 2d).

Transepithelial Cl− fluxes in the absence of NHE3

The previous results suggest that Slc26a6 function is coupledto an electroneutral sodium transporter in the absence ofglucose. To ascertain the role of NHE3 as the possible couplingpartner with Slc26a6, the impact of NHE3 deletion on jejunalCl− transport was examined. Toward this end, bilateral Cl−

and Na+ flux experiments were performed as above inisolated jejuni of NHE3+/+ and −/− sex-matched littermates.

In NHE3−/− jejunum, basal Cl− mucosal to serosal fluxwas insignificantly lower than in the NHE3+/+ jejunum.However, whereas forskolin caused a significant inhibitionof mucosal to serosal Cl− transport in the NHE3+/+jejunum, this inhibition was completely absent in NHE3−/−

Fig. 1 Bilateral Cl− flux experi-ments in Slc26a6+/+ and −/−jejuna. a Mucosal to serosal36Cl− flux in the basal state (leftpanel), after forskolin (10−5 M)addition (middle panel), andafter subsequent addition of 25-mM glucose to the luminal bath(right panel). b Serosal to mu-cosal 36Cl− flux from differenttissue sections of the same miceunder identical experimentalconditions. c Net 36Cl− absorp-tion, calculated by subtractingeach individual s–m flux fromthe matched m–s flux value. n=7 individual mice, *p<0.05,**p<0.01. Cl− absorption inSlc26a6-deficient jejuna is sig-nificantly reduced in the basalstate and after luminal glucoseaddition. However, the inhibi-tion of Cl− absorption as well asthe Cl− secretory response toforskolin was normal

760 Pflugers Arch - Eur J Physiol (2008) 455:757–766

jejunal tissue (Fig. 3a). Addition of luminal glucose 30 minafter forskolin stimulated Cl− uptake to the same degree inNHE3+/+ and −/− jejuna (Fig. 3a).

The serosal to mucosal Cl− flux was significantly higherin NHE3−/− jejunum and was stimulated by forskolin to thesame degree in both NHE3+/+ and NHE3−/− jejuna(Fig. 3b). The serosal to mucosal Cl− flux was notsignificantly influenced by luminal glucose (Fig. 3b). Thisresulted in a significant reduction in net Cl− absorption inNHE3−/− mice (Fig. 3c). Although the mucosal to serosalCl− flux was not affected by forskolin in the NHE3−/−tissue, the identical increase in serosal to mucosal Cl− fluxin NHE3−/− and +/+ mice resulted in inhibition of net Cl−

absorption in the jejuna of NHE3−/− and +/+ mice,respectively. This is exclusively because of the effects offorskolin on Cl− secretion. The Isc response to forskolinwas slightly but significantly lower in NHE3−/− jejuna,whereas the response to the luminal glucose was notaffected (Fig. 4d). Tissue conductance was significantlylower in NHE3−/− jejuna (55±16 in NHE3+/+ and 39±13in NHE3−/−, p<0.05).

Transepithelial Na+ fluxes in NHE3+/+ and −/− jejunum

We next evaluated the impact of NHE3 deletion on 22Na+

flux in the jejunum by examining tissues from sex- and age-

matched NHE3−/− and +/+ mice. Mucosal to serosal Na+

flux was significantly reduced in NHE3−/− compared to +/+tissues (Fig. 4a). Forskolin had a significant inhibitory effecton mucosal to serosal Na+ flux in NHE3+/+ tissue, whereasluminal glucose stimulated mucosal to serosal Na+ uptake tothe same degree in +/+ and −/− tissues (Fig. 4a). Serosal tomucosal Na+ fluxes were not significantly different, ascompared to the Cl− fluxes. Neither forskolin nor glucosehad a significant effect on serosal to mucosal Na+ fluxes ineither NHE3+/+ or −/− jejunum (Fig. 4b). The resulting netNa+ absorption was significantly lower in NHE3−/−jejunum, and, unlike NHE3+/+ jejunum, was not signifi-cantly inhibited by forskolin. However, glucose-driven Na+

uptake was not significantly different in +/+ and −/− tissue(Fig. 4c). The values for Isc and tissue conductance did notdiffer from those obtained in the Cl− flux experiments (datanot shown). The above results demonstrate that the inhibitionof mucosal to serosal Na+ and Cl− absorption by cAMP inwild-type animals is absent in NHE3 null mice. Thissuggests a tight coupling of the absorptive action of apicalCl�

�HCO�

3 exchangers to the regulation of NHE3. On theother hand, the loss of Slc26a6, one of the two major apicalSlc26 Cl�

�HCO�

3 exchangers expressed in the smallintestine, reduces electroneutral Cl− as well as Na+ absorp-tion, but leaves the inhibition of the remaining NaClabsorption by cAMP intact. The results further indicate that

Fig. 2 Bilateral Na+ flux experi-ments in Slc26a6+/+ and −/−mice. a Mucosal to serosal 22Na+

flux in the basal state (left panel),after forskolin (10−5 M) addition(middle panel), and after subse-quent addition of 25-mM glucoseto the luminal bath (right panel).b Serosal to mucosal 22Na+ fluxfrom different tissue sections ofthe same mice under identicalexperimental conditions. c Net22Na+ absorption, calculated bysubtracting each individual s–mflux from the matched m–s fluxvalue. d Short circuit currentresponse ΔISC to forskolin andto luminal glucose. Because theresponse values were not signif-icantly different from those seenin the experiments performed inFig. 1, they are only shown once.n=9, *p<0.05, **p<0.01. Asignificant reduction in basal Na+

absorption accompanies the re-duction in Cl− absorption in thejejuna of Slc26a6-deficient mice.However, the stimulation of Na+

absorption by luminal glucosewas not affected

Pflugers Arch - Eur J Physiol (2008) 455:757–766 761

Fig. 4 Bilateral Na+ fluxexperiments in NHE3+/+ and−/− jejuna. a Mucosal to serosal22Na+ flux in the basal state (leftpanel), after forskolin (10−5 M)addition (middle panel), andafter subsequent addition of 25-mM glucose to the luminal bath(right panel). b Serosal to mu-cosal 22Na+ flux from differenttissue sections of the same miceunder identical experimentalconditions. c Net 22Na+ absorp-tion, calculated by subtractingeach individual s–m flux fromthe matched m–s flux value. n=8, *p<0.05, **p<0.01. A sig-nificant decrease in basal Na+

absorption and a lack of inhibi-tion by forskolin was observed,whereas glucose-mediated sodi-um absorption was normal

Fig. 3 Bilateral Cl− flux experi-ments in NHE3+/+ and −/−jejuna. a Mucosal to serosal36Cl− flux in the basal state (leftpanel), after forskolin (10−5 M)addition (middle panel), andafter subsequent addition of25-mM glucose to the luminalbath (right panel). b Serosal tomucosal 36Cl− flux from differ-ent tissue sections of the samemice under identical experimen-tal conditions. c Net 36Cl− ab-sorption, calculated bysubtracting each individual s–mflux from the matched m–s fluxvalue. d Short circuit currentresponse ΔISC to forskolin andto luminal glucose. n=6, *p<0.05, **p<0.01. A significantreduction in basal net Cl− ab-sorption and a complete lack ofinhibition of mucosal to serosalCl−1 flux by forskolin are theprominent abnormalities inNHE3-deficient jejuna

762 Pflugers Arch - Eur J Physiol (2008) 455:757–766

the electrogenic, glucose-driven Na+ uptake stimulatesSlc26a6, as evidenced by the strongly reduced glucose-mediated stimulation of Cl− absorption in Slc26a6−/−jejunum, which is preserved in NHE3 null mice.

NHE3 abundance in Slc26a6−/− and +/+ enterocytehomogenate and brush border membrane

To determine whether deletion of Slc26a6 decreased NHE3abundance, immunoblot analysis of NHE3 was performedon enterocytes or brush border membrane proteins isolatedfrom jejunum of WT and Slc26a6 ko mice. As shown inFig. 5, NHE3 abundance was comparable in both entero-cytes (Fig. 5a) and brush border membrane proteins(Fig. 5b) in wild-type and Slc26a6 null mice. Immunohis-tochemical staining revealed an identical NHE3 distributionin the brush border membrane of Slc26a6+/+ and −/−jejuna (Fig. 6).

Discussion

The role of Slc26a6 in jejunal sodium and chloride absorp-tion, as well as its interplay with NHE3, was examined inSlc26a6 and NHE3+/+ and −/− mice. Deletion of Slc26a6decreased net Cl− and Na+ fluxes in the basal state (Figs. 1and 2). Similar experiments in NHE3-deficient jejuna alsoshowed decreased Na+ and Cl− absorption (Figs. 3 and 4).While the reductions in Na+ and Cl− absorption observed inNHE3 and Slc26a6 deficient jejunum were almost identical,demonstrating the important and previously unrecognizedrole of Slc26a6 in small intestinal electrolyte absorption,there were two surprising differences. The lack of NHE3rendered mucosal Na+ and Cl− absorption unresponsive toinhibition by cAMP, whereas Slc26a6 deletion did not.Furthermore, NHE3 deletion did not affect the glucose-driven Cl− absorption whereas Slc26a6 deletion bluntedthis response. Immunoblotting and immunohistochemistryrevealed comparable NHE3 abundance and distribution inapical membranes of jejunum in Slc26a6−/− and +/+ mice.

The results demonstrate that Slc26a6 is an importantpartner for NHE3 in the electroneutral absorption of salt inthe small intestine. Based on comparable abundance ofNHE3 in brush border membrane vesicles of Slc26a6+/+and −/− mice (Fig. 5), we suggest that decreased NHE3transport rate in Slc26a6 ko mice is likely at posttransla-tional levels and is likely because of functional down-

Fig. 6 NHE3 immunohistochemistry in Slc26a6+/+ and −/− jejunum.NHE3 immunostaining (green) is detected in the brush bordermembrane of the enterocytes along the jejunal villi. The nuclei arecounterstained in red (propidium iodide). No difference in the stainingpattern was observed between Slc26a6+/+ and −/− jejunum. Multiplesections were stained from several different areas of each jejunum,from three individual matched pairs of Slc26a6+/+ and −/− mice

Fig. 5 Immunoblot analysis of NHE3 in enterocytes or brush bordermembrane proteins isolated from jejunum of WT and Slc26a6 komice. a NHE3 abundance was measured in relation to the integralbrush border membrane protein villin and was not different inenterocytes of wild-type and Slc26a6 null mice. b Likewise, NHE3abundance was not different in brush border membrane proteins inwild-type and Slc26a6 null mice. The graph shows a representativeexperiment with four samples on the gel, but the graph is derived fromthe results of seven individual experiments with one pair of +/+ and −/−mice each, for enterocyte homogenate, and for BBM preparation.NHE3 runs at the estimated size of approx 88 kd, and villin-antibodystained a band of slightly smaller size than NHE3 (approx 81 kd)

Pflugers Arch - Eur J Physiol (2008) 455:757–766 763

regulation of the exchanger. Knickelbein et al. [19, 20]studied Na+/H+ and Cl�

�HCO�

3 exchanges in isolatedrabbit ileal BBM vesicles and brought forth the hypothesisthat the coupling of the two exchange mechanisms for saltuptake may occur via their effect on the intracellular pH.Recent studies suggest that a narrow region of subapicalpH-climate change generated by the activation of oneexchanger may be sufficient to result in the concomitantactivation of the other [3, 4, 5, 25]. Thus, the lack of eitherexchange mechanism would result in a shift of this pH-microclimate towards less optimal conditions and thereforeblunt the coordinated absorption of Na+ and Cl−.

An additional mode of activation of electroneutral saltabsorption could be a structural interaction of NHE3 withSlc26a6. Both exchangers contain PDZ-domain bindingmotifs and bind to a common set of adaptor proteins fromthe NHERF family [23]. A structural interaction of Slc26a6and CFTR has been shown to result in a reciprocal activationof both transport proteins [22]. If a similar situation existsfor NHE3 and Slc26a6, the lack of the latter could alsoresult in a functional downregulation of the former.

The fact that Cl− absorption is still responsive to cAMPinhibition in the absence of Slc26a6, but not in the absenceof NHE3, suggests that NHE3 is functionally coupled tomore than one anion exchanger in the small intestine. Thelikely candidate is Slc26a3, which is expressed in murine[34] and human [15] small intestine. However, othermembers of the Slc26 gene family are also expressed inthe small intestine and their contribution to salt absorptionis unclear. Similar experiments in the Slc26a3 knockoutjejunum will help clarify the relative contribution of thistransporter to small intestinal salt absorption.

The complete loss of response in mucosal Cl− absorptionto cAMP in NHE3-deficient small intestine stronglysuggests that this effect occurs via functional coupling ofanion exchange to NHE3. It further demonstrates thatNHE3 is the only Na+/H+ exchanger that can carry out theinhibition of Cl− absorption by cAMP in this intestinalsegment. The stimulation of Cl− secretion by cAMP, whichoccurs via CFTR, was only mildly blunted in the absence ofNHE3, as evidenced by the slight decrease in Isc-responseto forskolin in NHE3-deficient jejunum. This stimulation ofCl− secretion results in an inhibition of salt (and fluid)absorption even in the absence of NHE3 (although lesspronounced), but the reason for this appears to be solely thepreserved function of cAMP on stimulation of Cl− secretion.These data further show that in native murine intestine, thereciprocal PKA regulatory interactions between CFTR andNHE3, as seen in transfected renal cell lines [1], do notrelate to PKA-mediated CFTR activation. This may bebecause of the fact that the majority of CFTR is expressedin the villous base and cryptal region, where little or noNHE3 expression occurs.

Thus, the results demonstrate that the presence of NHE3is essential for the inhibitory effect of cAMP on chlorideuptake from the luminal bath. Such a regulatory responsesuggests that NHE3, with its associated proteins, such asmembers of the NHERF adapter protein family, ezrin, andothers, is likely the main target of cAMP inhibitory effect(Fig. 4). How the inhibition of NHE3 confers inhibition ofanion absorption via the Slc26a6 and/or Slc26a3 (in thedistal intestinal tract) is still unclear. Murine Slc26a6, ex-pressed in Xenopus oocytes, is slightly stimulated, notinhibited, by cAMP and this stimulation is increased bycoexpression with CFTR [6]. In the small intestine, the lackof Slc26a6 does not significantly inhibit cAMP-stimulatedCl− secretion, suggesting that Slc26a6 is not a major partnerfor CFTR in this intestinal segment.

Similar Na+ and Cl− flux experiments have been per-formed previously in NHE3-deficient jejunum by Gaweniset al. [10]. While those studies and the present experimentsarrive at similar conclusions in several aspects, notabledifferences exist, including a higher basal net Na+ than Cl−

absorption in that study, a stronger inhibitory effect offorskolin and IBMX on unidirectional and net Na+ absorp-tion compared to forskolin alone in our studies, and aninhibitory effect of forskolin and IBMX on serosal tomucosal Na+ and Cl− fluxes and tissue conductance in thatstudy, rather than a stimulation of serosal to mucosal Cl−

flux (indicative of CFTR activation) which we observed inboth NHE3+/+ and −/− jejunum [10]. In addition, there wasa residual effect of forskolin and IBMX on mucosal toserosal Na+ and Cl− absorption which we did not observe[10]. There could be several reasons for the apparentlyconflicting results in the two studies. One major reasoncould be because of the differences in tissue preparation. Inour studies, the mucosa is stripped from the submucosaltissue, which omits influences of muscular thickness andtone and, together with TTX and indomethacin, of neuraltransmitter release and prostaglandin-mediated secretorytone. This may explain our equal net reabsorptive rates forNa+ and Cl− and the lack of effect of forskolin on tissueconductivity, the absence of the strong decrease in s–m fluxfor both Na+ and Cl− in the absence of NHE3 (which maybe because of hypertrophy of some submucosal structuredue to the diarrhea present in NHE3 ko mice), and markedlyhigher Isc response to forskolin in our study. In addition, weadd 10-μM amiloride to the luminal bath to inhibitelectrogenic Na+ absorption. Lastly, we apply a long initialequilibration period, because we noticed that it can take upto 1 h before the initially higher flux rates have stabilized.

A strikingly conspicuous aspect of the present studies isthe stimulatory effect of glucose on chloride absorption inthe jejunum of wild-type mice which was blunted inSlc26a6 ko mice (Fig. 1). These results suggest that Slc26a6is involved in glucose-stimulated chloride absorption in the

764 Pflugers Arch - Eur J Physiol (2008) 455:757–766

small intestine. The molecular basis of enhanced Slc26a6-mediated chloride absorption by glucose in the jejunumremains speculative. Luminal glucose activates luminalSGLT1, Glut5, and possibly Glut2, and, in addition toglucose uptake, results in a strong increase in electrogenicNa+ uptake [7, 17, 43]. The increase in sodium absorption byglucose is significantly blunted in the absence of luminalchloride [9]. The concomitant glucose-mediated Cl− absorp-tion has been thought to occur via paracellular pathways, butthe current data show that a transcellular route is alsoinvolved. The mechanism of Slc26a6 activation by glucoseis unknown. Na+ glucose cotransport activation has beenshown to stimulate the apical Na+/H+ exchanger NHE3 inintestinal cells [33, 38, 46]. It is plausible that the stimulatoryeffect of glucose on chloride absorption by PAT1 issecondary to the stimulation of NHE3, which can alter thecell volume and/or increase intracellular bicarbonate, therebyincreasing the driving force for Cl�

�HCO�

3 exchange viaSlc26a6. Against this possibility is the fact that deletion ofNHE3 does not significantly affect the stimulation of chlorideabsorption by glucose (Figs. 3 and 4). An alternative possi-bility is that alteration in Na+, glucose, cellular solute load, orthe change in membrane potential stimulates a regulatoryprotein, which can then activate Slc26a6-mediated chloridetransport. In this respect, it will be important to study theeffect of PI3K/MAPK signaling on Slc26a6 activity.

In conclusion, our data strongly suggest that Slc26a6(PAT1) acts in concert with NHE3 in mediating electro-neutral salt absorption in the small intestine, and in concertwith SGLT1 in mediating glucose-dependent salt absorp-tion. Because cAMP inhibition of Na+ as well as Cl−

absorption is still present in the absence of Slc26a6, butabsent in the absence of NHE3, it is likely that NHE3 actsin concert with more than one anion exchanger, but is itselfan essential component for the inhibition of Cl− reabsorp-tion via apical anion exchange. Future work will need todelineate the molecular details of NHE3–Slc26a6 andSGLT1–Slc26a6 interaction.

Acknowledgments These studies were supported by grants fromDeutsche Forschungsgemeinschaft Se 460/9-5 and Se 460/17-1 andSonderforschungsbereich 621/C9 (to US) and NIH grants DK 62809(MS) and DK50596 (GES) and Merit Review Award (to MS). Thismanuscript contains work performed in fulfillment of the requirementsfor the doctoral theses of Ingrid Rottinghaus, Jutta Hillesheim, andMingmin Chen.

References

1. Bagorda A, Guerra L, Di Sole F, Hemle-Kolb C, Cardone RA,Fanelli T, Reshkin SJ, Gisler SM, Murer H, Casavola V (2002)Reciprocal protein kinase A regulatory interactions between cysticfibrosis transmembrane conductance regulator and Na+/H+

exchanger isoform 3 in a renal polarized epithelial cell model. JBiol Chem 277:21480–21488

2. Bissig M, Hagenbuch B, Stieger B, Koller T, Meier PJ (1994)Functional expression cloning of the canalicular sulfate transportsystem of rat hepatocytes. J Biol Chem 269:3017–3021

3. Charney AN, Egnor RW, Alexander-Chacko J, Cassai N, SidhuGS (2002) Acid-base effects on intestinal Na(+) absorption andvesicular trafficking. Am J Physiol Cell Physiol 283:C971–C979

4. Charney AN, Egnor RW, Cassai N, Sidhu GS (2002) Carbondioxide affects rat colonic Na+ absorption by modulatingvesicular traffic. Gastroenterology 122:318–330

5. Charney AN, Egnor RW, Henner D, Rashid H, Cassai N, Sidhu GS(2004) Acid-base effects on intestinal Cl- absorption and vesiculartrafficking. Am J Physiol Cell Physiol 286:C1062–C1070

6. Chernova MN, Jiang L, Friedman DJ, Darman RB, Lohi H, KereJ, Vandorpe DH, Alper SL (2005) Functional comparison ofmouse Slc26a6 anion exchanger with human SLC26A6 polypep-tide variants: differences in anion selectivity, regulation, andelectrogenicity. J Biol Chem 280:8564–8580

7. Drozdowski LA, Thomson AB (2006) Intestinal sugar transport.World J Gastroenterol 12:1657–1670

8. Everett LA, Glaser B, Beck JC, Idol JR, Buchs A, Heyman M,Adawi F, Hazani E, Nassir E, Baxevanis AD, Sheffield VC, GreenED (1997) Pendred syndrome is caused by mutations in a putativesulphate transporter gene (PDS). Nat Genet 17:411–422

9. Fordtran JS (1975) Stimulation of active and passive sodium ab-sorption by sugars in the human jejunum. J Clin Invest 55:728–737

10. Gawenis LR, Stien X, Shull GE, Schultheis PJ, Woo AL, WalkerNM, Clarke LL (2002) Intestinal NaCl transport in NHE2 andNHE3 knockout mice. Am J Physiol Gastrointest Liver Physiol282:G776–G784

11. Hastbacka J, de la Chapelle A, Mahtani MM, Clines G, Reeve-DalyMP, Daly M, Hamilton BA, Kusumi K, Trivedi B, Weaver A,Coloma A, Lovett M, Buckler A, Kaitila I, Lander ES (1994) Thediastrophic dysplasia gene encodes a novel sulfate transporter:positional cloning by fine-structure linkage disequilibrium mapping.Cell 78:1073–1087

12. Hoglund P, Haila S, Socha J, Tomaszewski L, Saarialho-Kere U,Karjalainen-Lindsberg ML, Airola K, Holmberg C, Chapelle A,Kere J (1996) Mutations of the Down-regulated in adenoma (DRA)gene cause congenital chloride diarrhoea. Nat Genet 14:316–319

13. Hoogerwerf WA, Tsao SC, Devuyst O, Levine SA, Yun CH, Yip JW,Cohen ME, Wilson PD, Lazenby AJ, Tse CM, Donowitz M (1996)NHE2 and NHE3 are human and rabbit intestinal brush-borderproteins. Am J Physiol 270:G29–G41

14. Ishiguro H, Namkung W, Yamamoto A, Wang Z, Worrell RT, Xu J,Lee MG, Soleimani M (2006) Effect of Slc26a6 deletion on apicalCl-/HCO3- exchanger activity and cAMP-stimulated bicarbonatesecretion in pancreatic duct. Am J Physiol Gastrointest LiverPhysiol 292:G447–G455

15. Jacob P, Rossmann H, Lamprecht G, Kretz A, Neff C, Lin-Wu E,Gregor M, Groneberg DA, Kere J, Seidler U (2002) Down-regulated in adenoma mediates apical Cl-/HCO3- exchange inrabbit, rat, and human duodenum. Gastroenterology 122:709–724

16. Jiang Z, Grichtchenko II, Boron WF, Aronson PS (2002)Specificity of anion exchange mediated by mouse Slc26a6. JBiol Chem 277:33963–33967

17. Kellett GL, Brot-Laroche E (2005) Apical GLUT2: a majorpathway of intestinal sugar absorption. Diabetes 54:3056–3062

18. Knauf F, Yang CL, Thomson RB, Mentone SA, Giebisch G,Aronson PS (2001) Identification of a chloride-formate exchangerexpressed on the brush border membrane of renal proximal tubulecells. Proc Natl Acad Sci USA 98:9425–9430

19. Knickelbein R, Aronson PS, Atherton W, Dobbins JW (1983)Sodium and chloride transport across rabbit ileal brush border. I.Evidence for Na-H exchange. Am J Physiol 245:G504–G510

Pflugers Arch - Eur J Physiol (2008) 455:757–766 765

20. Knickelbein R, Aronson PS, Schron CM, Seifter J, Dobbins JW(1985) Sodium and chloride transport across rabbit ileal brushborder. II. Evidence for Cl-HCO3 exchange and mechanism ofcoupling. Am J Physiol 249:G236–G245

21. Ko SB, Zeng W, Dorwart MR, Luo X, Kim KH, Millen L, Goto H,Naruse S, Soyombo A, Thomas PJ, Muallem S (2004) Gating ofCFTR by the STAS domain of SLC26 transporters. Nat Cell Biol6:343–350

22. Ko SB, Zeng W, Dorwart MR, Luo X, Kim KH, Millen L, GotoH, Naruse S, Soyombo A, Thomas PJ, Muallem S (2004) Gatingof CFTR by the STAS domain of SLC26 transporters. Nat CellBiol 6:343–350

23. Lamprecht G, Schaefer J, Dietz K, Gregor M (2006) Chloride andbicarbonate have similar affinities to the intestinal anion exchangerDRA (down regulated in adenoma). Pflugers Arch 452:307–315

24. Lohi H, Kujala M, Kerkela E, Saarialho-Kere U, Kestila M, Kere J(2000) Mapping of five new putative anion transporter genes inhuman and characterization of SLC26A6, a candidate gene forpancreatic anion exchanger. Genomics 70:102–112

25. Lohi H, Kujala M, Makela S, Lehtonen E, Kestila M, Saarialho-Kere U, Markovich D, Kere J (2002) Functional characterizationof three novel tissue-specific anion exchangers SLC26A7, -A8,and -A9. J Biol Chem 277:14246–14254

26. Melvin JE, Park K, Richardson L, Schultheis PJ, Shull GE (1999)Mouse down-regulated in adenoma (DRA) is an intestinal Cl(−)/HCO(3)(−) exchanger and is up-regulated in colon of micelacking the NHE3 Na(+)/H(+) exchanger. J Biol Chem274:22855–22861

27. Mount DB, Romero MF (2004) The SLC26 gene family ofmultifunctional anion exchangers. Pflugers Arch 447:710–721

28. Petrovic S, Wang Z, Ma L, Seidler U, Forte JG, Shull GE,Soleimani M (2002) Colocalization of the apical Cl-/HCO3-exchanger PAT1 and gastric H-K-ATPase in stomach parietalcells. J Physiol Gastrointest Liver Physiol 283:G1207–G1216

29. Schultheis PJ, Clarke LL, Meneton P, Miller ML, Soleimani M,Gawenis LR, Riddle TM, Duffy JJ, Doetschman T, Wang T,Giebisch G, Aronson PS, Lorenz JN, Shull GE (1998) Renal andintestinal absorptive defects in mice lacking the NHE3 Na+/H+exchanger. Nat Genet 19:282–285

30. Schweinfest CW, Henderson KW, Suster S, Kondoh N, Papas TS(1993) Identification of a colon mucosa gene that is down-regulated in colon adenomas and adenocarcinomas. Proc NatlAcad Sci USA 90:4166–4170

31. Hillesheim JB, Riederer B, Tuo B, Chen M, Biber C, Yun C,Kocher O, Seidler U (2007) Downregulation of small intestinalion transport in PDZK1 (CAP70NHERF3) deficient mice.Pflügers Arch 454(4):575–586

32. Shcheynikov N, Wang Y, Park M, Ko SB, Dorwart M, Naruse S,Thomas PJ, Muallem S (2006) Coupling modes and stoichiometryof CI-/HCO3- exchange by slc26a3 and slc26a6. J Gen Physiol127:511–524

33. Shiue H, Musch MW, Wang Y, Chang EB, Turner JR (2005) Akt2phosphorylates ezrin to trigger NHE3 translocation and activation.J Biol Chem 280:1688–1695

34. Simpson JE, Schweinfest CW, Shull GE, Gawenis LR, WalkerNM, Boyle KT, Soleimani M, Clarke LL (2007) PAT-1 (Slc26a6)is the predominant apical membrane CI-/HCO3- exchanger in theupper villous epithelium of the murine duodenum. Am J PhysiolGastrointest Liver Physiol 292(4):G1079–1088

35. Soleimani M, Greeley T, Petrovic S, Wang Z, Amlal H, Kopp P,Burnham CE (2001) Pendrin: an apical Cl-/OH-/HCO3- exchangerin the kidney cortex. Am J Physiol Renal Physiol 280:F356–F364

36. Soleimani M (2006) Expression, regulation and the role of SLC26CI-/HCO3- exchangers in kidney and gastrointestinal tract.Novartis Found Symp 273:91–102

37. Tuo B, Riederer B, Wang Z, Colledge WH, Soleimani M, Seidler U(2006) Involvement of the anion exchanger SLC26A6 in prostaglan-din E2- but not forskolin-stimulated duodenal. Gastroenterology130:349–358

38. Turner JR, Black ED (2001) NHE3-dependent cytoplasmicalkalinization is triggered by Na(+)-glucose cotransport inintestinal epithelia. Am J Physiol Cell Physiol 281:C1533–C1541

39. van den Berghe N, Nieuwkoop NJ, Vaandrager AB, de Jonge HR(1991) Asymmetrical distribution of G-proteins among the apicaland basolateral membranes of rat enterocytes. Biochem J 278(Pt2):565–571

40. Vincourt JB, Jullien D, Kossida S, Amalric F, Girard JP (2002)Molecular cloning of SLC26A7, a novel member of the SLC26sulfate/anion transporter family, from high endothelial venules andkidney. Genomics 79:249–256

41. Wang Z, Petrovic S, Mann E, Soleimani M (2002) Identificationof an apical Cl(−)/HCO3(−) exchanger in the small intestine. AmJ Physiol Gastrointest Liver Physiol 282:G573–G579

42. Wang Z, Wang T, Petrovic S, Tuo B, Riederer B, Barone S,Lorenz JN, Seidler U, Aronson PS, Soleimani M (2005) Renaland intestinal transport defects in Slc26a6-null mice. Am J PhysiolCell Physiol 288:C957–C965

43. Wright EM, Loo DD (2000) Coupling between Na+, sugar, andwater transport across the intestine. Ann N YAcad Sci 915:54–66

44. Xie Q, Welch R, Mercado A, Romero MF, Mount DB (2002)Molecular characterization of the murine Slc26a6 anion exchang-er: functional comparison with Slc26a1. Am J Physiol RenalPhysiol 283:F826–F838

45. Zachos NC, Tse M, Donowitz M (2005) Molecular physiology ofintestinal Na+/H+ exchange. Annu Rev Physiol 67:411–443

46. Zhao H, Shiue H, Palkon S, Wang Y, Cullinan P, Burkhardt JK,Musch MW, Chang EB, Turner JR (2004) Ezrin regulates NHE3translocation and activation after Na+-glucose cotransport. ProcNatl Acad Sci USA 101:9485–9490

47. Zheng J, Shen W, He DZ, Long KB, Madison LD, Dallos P(2000) Prestin is the motor protein of cochlear outer hair cells.Nature 405:149–155

766 Pflugers Arch - Eur J Physiol (2008) 455:757–766