Acid-base transport in isolated rabbit duodenal villus and crypt cells

Human Enteroids as a Model of Upper Small Intestinal Ion Transport Physiology and Pathophysiology

BACKGROUND & AIMS

Human intestinal crypt-derived enteroids are a model of intestinal ion transport that require validation by comparison with cell culture and animal models. We used human small intestinal enteroids to study neutral Na(+) absorption and stimulated fluid and anion secretion under basal and regulated conditions in undifferentiated and differentiated cultures to show their functional relevance to ion transport physiology and pathophysiology.

METHODS

Human intestinal tissue specimens were obtained from an endoscopic biopsy or surgical resections performed at Johns Hopkins Hospital. Crypts were isolated, enteroids were propagated in culture, induced to undergo differentiation, and transduced with lentiviral vectors. Crypt markers, surface cell enzymes, and membrane ion transporters were characterized using quantitative reverse-transcription polymerase chain reaction, immunoblot, or immunofluorescence analyses. We used multiphoton and time-lapse confocal microscopy to monitor intracellular pH and luminal dilatation in enteroids under basal and regulated conditions.

RESULTS

Enteroids differentiated upon withdrawal of WNT3A, yielding decreased crypt markers and increased villus-like characteristics. Na(+)/H(+) exchanger 3 activity was similar in undifferentiated and differentiated enteroids, and was affected by known inhibitors, second messengers, and bacterial enterotoxins. Forskolin-induced swelling was completely dependent on cystic fibrosis transmembrane conductance regulator and partially dependent on Na(+)/H(+) exchanger 3 and Na(+)/K(+)/2Cl(-) cotransporter 1 inhibition in undifferentiated and differentiated enteroids. Increases in cyclic adenosine monophosphate with forskolin caused enteroid intracellular acidification in HCO3(-)-free buffer. Cyclic adenosine monophosphate-induced enteroid intracellular pH acidification as part of duodenal HCO3(-) secretion appears to require cystic fibrosis transmembrane conductance regulator and electrogenic Na(+)/HCO3(-) cotransporter 1.

CONCLUSIONS

Undifferentiated or crypt-like, and differentiated or villus-like, human enteroids represent distinct points along the crypt-villus axis; they can be used to characterize electrolyte transport processes along the vertical axis of the small intestine. The duodenal enteroid model showed that electrogenic Na(+)/HCO3(-) cotransporter 1 might be a target in the intestinal mucosa for treatment of secretory diarrheas.

5-HT induces duodenal mucosal bicarbonate secretion via cAMP- and Ca2+-dependent signaling pathways and 5-HT4 receptors in mice

In previous studies, we have found that 5-hydroxytryptamine (5-HT) is a potent stimulant of duodenal mucosal bicarbonate secretion (DMBS) in mice. The aim of the present study was to determine the intracellular signaling pathways and 5-HT receptor subtypes involved in 5-HT-induced DMBS. Bicarbonate secretion by murine duodenal mucosa was examined in vitro in Ussing chambers. 5-HT receptor involvement in DMBS was inferred from pharmacological studies by using selective 5-HT receptor antagonists and agonists. The expression of 5-HT(4) receptor mRNA in duodenal mucosa and epithelial cells was analyzed by RT-PCR. cAMP-dependent signaling pathway inhibitors MDL-12330A, Rp-cAMP, and H-89 and Ca(2+)-dependent signaling pathway inhibitors verapamil and W-13 markedly reduced 5-HT-stimulated duodenal bicarbonate secretion and short-circuit current (I(sc)), whereas cGMP-dependent signaling pathway inhibitors NS-2028 and KT-5823 failed to alter these responses. Both SB-204070 and high-dose ICS-205930 (selective 5-HT(4) receptor antagonists) markedly inhibited 5-HT-stimulated bicarbonate secretion and I(sc), whereas methiothepine (5-HT(1) receptor antagonist), ketanserin (5-HT(2) receptor antagonist), and a low concentration of ICS-205930 (5-HT(3) receptor antagonist) had no effect. RS-67506 (partial 5-HT(4) receptor agonist) concentration-dependently increased bicarbonate secretion and I(sc), whereas 5-carboxamidotryptamine (5-HT(1) receptor agonist), alpha-methyl-5-HT (5-HT(2) receptor agonist), and phenylbiguanide (5-HT(3) receptor agonist) did not significantly increase bicarbonate secretion or I(sc). RT-PCR analysis confirmed the expression of 5-HT(4) receptor mRNA in murine duodenal mucosa and epithelial cells. These results demonstrate that 5-HT regulates DMBS via both cAMP- and Ca(2+)-dependent signaling pathways and 5-HT(4) receptors located in the duodenal mucosa and/or epithelial cells.

Human duodenal mucosal brush border Na(+)/H(+) exchangers NHE2 and NHE3 alter net bicarbonate movement

The proximal duodenal mucosa secretes HCO that serves to protect the epithelium from injury. In isolated human duodenal enterocytes in vitro, multiple luminal membrane proteins are involved in acid/base transport. We postulated that one or more isoforms of the Na(+)/H(+) exchanger (NHE) family is located on the apical surface of human duodenal mucosal epithelial cells and thereby contributes to duodenal mucosal HCO transport. Duodenal biopsies were obtained from human volunteers, and the presence of NHE2 and NHE3 was determined by using previously characterized polyclonal antibodies (Ab 597 for NHE2 and Ab 1381 for NHE3). In addition, proximal duodenal mucosal HCO(3)(-) transport was measured in humans in vivo in response to luminal perfusion of graded doses of amiloride; 10(-5)--10(-4) M amiloride was used to inhibit NHE2 and 10(-3) M amiloride to inhibit NHE3. Both NHE2 and NHE3 were localized principally to the brush border of duodenal villus cells. Sequential doses of amiloride resulted in significant, step-wise increases in net duodenal HCO(3)(-) output. Inhibition of NHE2 with 10(-5) M and 10(-4) M amiloride significantly increased net HCO(3)(-) output. Moreover, there was an additional, equivalent increase (P < 0.05) in duodenal HCO(3)(-) output with 10(-3) M amiloride, which inhibited NHE3. We conclude that 1) NHE2 and NHE3 are localized principally to the brush border of human duodenal villus epithelial cells; 2) sequential inhibition of NHE2 and NHE3 isoforms resulted in step-wise increases in net HCO(3)(-) output; 3) NHE2 and NHE3 participate in human duodenal villus cell HCO(3)(-) transport; and 4) the contribution of NHE-related transport events should be considered when studying duodenal HCO(3)(-) transport processes.

Molecular and functional evidence for electrogenic and electroneutral Na(+)-HCO(3)(-) cotransporters in murine duodenum

Inward Na(+)-HCO(3)(-) cotransport has previously been demonstrated in acidified duodenal epithelial cells, but the identity and localization of the mRNAs and proteins involved have not been determined. The molecular expression and localization of Na(+)-HCO(3)(-) cotransporters (NBCs) were studied by RT-PCR, sequence analysis, and immunohistochemistry. By fluorescence spectroscopy, the intracellular pH (pH(i)) was recorded in suspensions of isolated murine duodenal epithelial cells loaded with 2',7'-bis(2-carboxyethyl)-5(6)-carboxyfluorescein. Proximal duodenal epithelial cells expressed mRNA encoding two electrogenic NBC1 isoforms and the electroneutral NBCn1. Both NBC1 and NBCn1 were localized to the basolateral membrane of proximal duodenal villus cells, whereas the crypt cells did not label with the anti-NBC antibodies. DIDS or removal of extracellular Cl(-) increased pH(i), whereas an acidification was observed on removal of Na(+) or both Na(+) and Cl(-). The effects of inhibitors and ionic dependence of acid/base transporters were consistent with both inward and outward Na(+)-HCO(3)(-) cotransport. Hence, we propose that NBCs are involved in both basolateral electroneutral HCO(3)(-) transport as well as basolateral electrogenic HCO(3)(-) transport in proximal duodenal villus cells.

Role of Na(+)HCO(3)(-) cotransporter NBC1, Na(+)/H(+) exchanger NHE1, and carbonic anhydrase in rabbit duodenal bicarbonate secretion

BACKGROUND & AIMS

HCO(3)(-) supply to the enterocyte is rate limiting for duodenal HCO(3)(-) secretion (J(HCO3-)). This study defines the molecular nature of the major HCO(3)(-) uptake pathways in rabbit duodenocytes and investigates their physiologic significance and regulation during basal and stimulated J(HCO3-).

METHODS & RESULTS

pH gradient-driven (22)Na(+) uptake into duodenal basolateral membrane vesicles was partly HCO(3)(-) dependent, stilbene sensitive, and therefore mediated by Na(+)HCO(3)(-) cotransport, and partly HCO(3)(-) independent, Hoechst 642 sensitive, and therefore mediated by the Na(+)/H(+) exchanger isoform NHE1. Semiquantitative polymerase chain reaction (PCR) revealed high duodenal expression levels for the NBC1 isoform of the Na(+)HCO(3)(-) cotransporter gene family and NHE1. Cloning and comparison of full-length rabbit with human gastrointestinal and kidney NBC1 subtype revealed a conserved protein kinase A consensus sequence in the cytoplasmic N-terminus of the gastrointestinal NBC1. Inhibition of either Na(+)HCO(3)(-) cotransport or carbonic anhydrase reduced ouabain-sensitive J(HCO3-) in in vitro rabbit duodenal mucosae by approximately 50%, but did not affect 8-Br-cAMP-induced DeltaJ(HCO3-), suggesting cAMP-mediated up-regulation of the alternative pathway. However, inhibition of both Na(+)HCO(3)(-) cotransport and either carbonic anhydrase or NHE1 strongly reduced DeltaJ(HCO3-).

CONCLUSIONS

NBC1 and NHE1 are the major base importers in rabbit duodenocytes. Na(+)HCO(3)(-) cotransport and CO(2) hydration/Na(+)/H(+) exchange are equally important pathways for duodenal HCO(3)(-) supply and are up-regulated during cAMP-mediated stimulation.

Identification of transport abnormalities in duodenal mucosa and duodenal enterocytes from patients with cystic fibrosis

BACKGROUND & AIMS

The duodenum is a cystic fibrosis transmembrane conductance regulator (CFTR)-expressing epithelium with high bicarbonate secretory capacity. We aimed to define the role of CFTR in human duodenal epithelial bicarbonate secretion in normal (NL) subjects and patients with cystic fibrosis (CF).

METHODS

Endoscopic biopsy specimens of the duodenal bulb were obtained from 9 CF patients and 16 volunteers. Tissues were mounted in modified Ussing chambers. Bicarbonate secretion and short-circuit current (Isc) were quantitated under basal conditions and in response to dibutyryl adenosine 3',5'-cyclic monophosphate (db-cAMP), carbachol, and the heat-stable toxin of Escherichia coli (STa). Duodenocytes were also isolated and loaded with the pH-sensitive fluoroprobe BCECF/AM, and intracellular pH (pH(i)) was measured at rest and after intracellular acidification and alkalinization.

RESULTS

Basal HCO(3)(-) secretion and Isc were significantly lower in the CF vs. NL duodenal mucosa. In contrast to NL, db-cAMP failed to alter either HCO(3)(-) or Isc in CF tissues. However, in CF, carbachol resulted in an electroneutral HCO(3)(-) secretion, whereas STa induced electrogenic HCO(3)(-) secretion that was similar to NL. In CF and NL duodenocytes, basal pH(i) and recovery from an acid load were comparable, but pH(i) recovery after an alkaline load in CF duodenocytes was Cl(-) dependent, whereas in NL duodenocytes it was Cl(-) independent.

CONCLUSIONS

These findings implicate CFTR in NL duodenal alkaline transport and its absence in CF. Although duodenal bicarbonate secretion is impaired in CF tissues, alternate pathway(s) likely exist that can be activated by carbachol and STa.

Activation of transepithelial ion transport by secretin in human intestinal Caco-2 cells

Secretin stimulates bicarbonate secretion from pancreatic duct cells, but what influence secretin exerts on intestinal tissues remains to be clarified. The aim of this study is to examine effects of secretin on ion transport in intestinal epithelial Caco-2 cells. We mounted monolayers of Caco-2 cells grown on permeable supports for 21-28 d in a Ussing chamber and measured short-circuit currents (I(sc)). Addition of secretin (5-100 nM) to the basolateral solution dose-dependently induced biphasic increases of I(sc) (transient and sustained phase). Dibutyryl cyclic AMP (200 microM), forskolin (10 microM), and 3-isobutyl-1-methylxanthine (IBMX, 1 mM) also induced I(sc) responses similar to the administration of secretin. Addition of 5-nitro-2-(3-phenylpropylamino) benzoic acid (NPPB, 100 microM) or benzamil (100 microM) to the apical solution markedly reduced the secretin-induced I(sc) increase in the transient phase. A selective antagonist of cAMP-dependent protein kinase (PKA), N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide (H-89, 1 microM), and a membrane permeable Ca(2+) chelator, 1, 2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid tetrakis(acetoxymethyl ester) (BAPTA/AM, 10 microM) reduced the secretin-induced I(sc). Basolateral addition of 4, 4'-diisothiocyanostilbene-2,2'-disulfonic acid (DIDS, 1 mM) suppressed the sustained phase I(sc) increase. Secretin also induced alkalinization of the apical solution (DeltapH, 0.053 +/- 0.013). The alkalinization did not occur when DIDS (1 mM) was added to the basolateral solution or Na(+) was removed from the solutions. Taken together, our observations suggest: (1) secretin stimulates a benzamil-sensitive Na(+) influx and an NPPB-sensitive Cl(-) efflux across the apical membrane through PKA-dependent and Ca(2+)-sensitive pathways; and (2) secretin also induces alkalinization of the apical solution through the activation of a DIDS-sensitive Na(+)-HCO(3)(-) cotransport in the basolateral membrane of Caco-2 cells.

NHE1, NHE2, and NHE3 contribute to regulation of intracellular pH in murine duodenal epithelial cells

Na(+)/H(+)-exchangers (NHE) mediate acid extrusion from duodenal epithelial cells, but the isoforms involved have not previously been determined. Thus we investigated 1) the contribution of Na(+)-dependent processes to acid extrusion, 2) sensitivity to Na(+)/H(+) exchange inhibitors, and 3) molecular expression of NHE isoforms. By fluorescence spectroscopy the recovery of intracellular pH (pH(i)) was measured on suspensions of isolated acidified murine duodenal epithelial cells loaded with 2', 7'-bis(2-carboxyethyl)-5(6)-carboxyfluorescein. Expression of NHE isoforms was studied by RT-PCR and Western blot analysis. Reduction of extracellular Na(+) concentration ([Na(+)](o)) during pH(i) recovery decreased H(+) efflux to minimally 12.5% of control with a relatively high apparent Michaelis constant for extracellular Na(+). The Na(+)/H(+) exchange inhibitors ethylisopropylamiloride and amiloride inhibited H(+) efflux maximally by 57 and 80%, respectively. NHE1, NHE2, and NHE3 were expressed at the mRNA level (RT-PCR) as well as at the protein level (Western blot analysis). On the basis of the effects of low [Na(+)](o) and inhibitors we propose that acid extrusion in duodenal epithelial cells involves Na(+)/H(+) exchange by isoforms NHE1, NHE2, and NHE3.

Regulation of intracellular pH and blood flow in rat duodenal epithelium in vivo

Duodenal mucosal defense was assessed by measuring blood flow and epithelial intracellular pH (pHi) of rat proximal duodenum in vivo. Fluorescence microscopy was used to measure epithelial pHi using the trapped, pHi-indicating dye 2', 7'-bis(2-carboxyethyl)-5(6)-carboxyfluorescein-AM. Blood flow was measured with laser-Doppler flowmetry. The mucosa was briefly superfused with NH4Cl, pH 2.2 buffer, the potent Na+/H+ exchange inhibitor 5-(N,N-dimethyl)-amiloride (DMA), or the anion exchange and Na+-HCO-3 cotransport inhibitor DIDS. Cryostat sections localized dye fluorescence to the villus tip. Steady-state pHi was 7. 02 +/- 0.01, which remained stable for 60 min. Interventions that load the cells with protons without affecting superfusate pH (NH4Cl prepulse, nigericin with low superfusate K+ concentration, DMA, and DIDS) all decreased pHi, supporting our contention that the dye was faithfully measuring pHi. An acid pulse decreased pHi, followed by a DIDS-inhibitable overshoot over baseline. Intracellular acidification increased duodenal blood flow independent of superfusate pH, which was inhibited by DMA, but not by DIDS. We conclude that we have established a novel in vivo microscopy system enabling simultaneous measurements of pHi and blood flow of duodenal epithelium. Na+/H+ exchange and Na+-HCO-3 cotransport regulate baseline duodenal epithelial pHi. Intracellular acidification enhances duodenal blood flow by a unique, amiloride-inhibitable, superfusate pH-independent mechanism.

Calcium signaling in cultured human and rat duodenal enterocytes

Vagal stimuli increase duodenal mucosal HCO-3 secretion and may provide anticipatory protection against acid injury, but duodenal enterocyte (duodenocyte) responses and cholinoceptor selectivity have not been defined. We therefore developed a stable primary culture model of duodenocytes from rats and humans. Brief digestion of scraped rat duodenal mucosa or human biopsies with collagenase/dispase yielded cells that attached to the extracellular matrix Matrigel within a few hours of plating. Columnar cells with villus enterocyte morphology that exhibited spontaneous active movement were evident between 1 and 3 days of culture. Rat duodenocytes loaded with fura 2 responded to carbachol with a transient increase in intracellular calcium concentration ([Ca2+]i), with an apparent EC50 of approximately 3 microM. In a first type of signaling pattern, [Ca2+]i returned to basal or near basal values within 3-5 min. In a second type, observed in cells with enlarged vacuoles characteristic of crypt cell morphology, the initial transient increase was followed by rhythmic oscillations. Human duodenocytes responded with a more sustained increase in [Ca2+]i, and oscillations were not observed. Rat as well as human duodenocytes also responded to CCK-octapeptide but not to vasoactive intestinal polypeptide. Equimolar concentrations (100 nM) of the subtype-independent muscarinic antagonist atropine and the M3 antagonist 4-diphenylacetoxy-N-methylpiperidine methiodide prevented the response to 10 microM carbachol, whereas the M1 antagonist pirenzepine and the M2 antagonists methoctramine and AF-DX 116BS had no effect at similar concentrations. Responses in rat and human duodenocytes were similar. A new agonist-sensitive primary culture model for rat and human duodenocytes has thus been established and the presence of enterocyte CCK and muscarinic M3 receptors demonstrated.