Characterization of an electrogenic sodium/glucose cotransporter in a human colon epithelial cell line

Review: A species comparison of the kinetic homogeneous and heterogeneous organization of sodium-dependent glucose transport systems along the intestine

The targeted use of carbohydrates by feed and food industries to create balanced and cost-effective diets has generated a tremendous amount of research in carbohydrate digestion and absorption in different species. Specifically, this research has led us to a larger observation that identified different organizations of intestinal sodium-dependent glucose absorption across species, which has not been previously collated and reviewed. Thus, this review will compare the kinetic segregation of sodium-dependent glucose transport across the intestine of different species, which we have termed either homogeneous or heterogeneous systems. For instance, the pig follows a heterogeneous system of sodium-dependent glucose transport with a high-affinity, super-low-capacity (Ha/sLc) in the jejunum, and a high-affinity, super-high-capacity (Ha/sHc) in the ileum. This is achieved by multiple sodium-dependent glucose transporters contributing to each segment. In contrast, tilapia have a homogenous system characterized by high-affinity, high-capacity (Ha/Hc) throughout the intestine. Additionally, we are the first to report glucose transporter patterns across species presented from vertebrates to invertebrates. Finally, other kinetic transport systems are briefly covered to illustrate possible contributions/modulations to sodium-dependent glucose transporter organization. Overall, we present a new perspective on the organization of glucose absorption along the intestinal tract.

Regulation of nutrient and electrolyte absorption in human organoid-derived intestinal epithelial cell monolayers

Recently developed human intestinal epithelial 3D organoid cultures are a useful cell culture model to study intestinal transport physiology. From these, 2D monolayer cultures can be generated in which apical transporters are exposed to the medium, thereby better facilitating in vitro investigation of intestinal absorption processes. However, whether nutrient and electrolyte absorption can be physiologically regulated in human organoid-derived monolayers has not been determined. Constitutive nitric oxide (cNO) is known to regulate multiple gastrointestinal physiological functions. Previous studies using in vivo and in vitro mammalian animal models indicate that enhanced intracellular cNO differentially regulates the two primary apical Na transporters in small intestinal epithelial cells. Here, we generated human jejunal organoid-derived monolayers to determine whether apical nutrient and electrolyte transporter function is regulated by cNO in human enterocytes. Western blot analysis and immunocytochemical staining showed that organoid-derived 2D cultures express markers of enterocyte differentiation and form intact monolayers of apical-basal polarized epithelial cells. Uptake studies demonstrated that jejunal monolayers exhibit functional activity of Na-glucose cotransporter 1 (SGLT1; SLC5A1) and Na-H exchanger 3 (NHE3; SLC9A3). In response to physiological increases in cNO, the two primary apical Na transporters were differentially regulated in human intestinal organoid-derived monolayers, across multiple human specimens. An increase in cNO stimulated SGLT1, while NHE3 was inhibited. These results are similar to what is seen in vivo and in vitro in different animal intestinal models. Thus, human jejunal organoid-derived monolayers are an ideal in vitro model to better understand how intestinal nutrient absorption is regulated.

Sodium-glucose cotransporters: new targets of cancer therapy?

Glucose, the key source of metabolic energy, is imported into cells by two categories of transporters: 1) facilitative glucose transporters (GLUTs) and 2) secondary active sodium-glucose cotransporters (SGLTs). Cancer cells have an increased demand for glucose uptake and utilisation compared to normal cells. Previous studies have demonstrated the overexpression of GLUTs, mainly GLUT1, in many cancer types. As the current standard positron emission tomography (PET) tracer 2-deoxy-2-(18F)fluoro-D-glucose (2-FDG) for imaging tumour cells via GLUT1 lacks in sensitivity and specificity, it may soon be replaced by the newly designed, highly sensitive and specific SGLT tracer α-methyl-4-(F-18)fluoro-4-deoxy-Dglucopyranoside (Me-4FDG) in clinical detection and tumour staging. This tracer has recently demonstrated the functional activity of SGLT in pancreatic, prostate, and brain cancers. The mRNA and protein expression of SGLTs have also been reported in colon/colorectal, lung, ovarian, head, neck, and oral squamous carcinomas. So far, SGLTs have been poorly investigated in cancer, and their protein expression and localisation are often controversial due to a lack of specific SGLT antibodies. In this review, we describe current knowledge concerning SGLT1 and SGLT2 (over)expression in various cancer types. The findings of SGLTs in malignant cells may help in developing novel cancer therapies with SGLT2 or SGLT1/SGLT2 inhibitors already used in diabetes mellitus treatment.

Sigmoidal kinetics define porcine intestinal segregation of electrogenic monosaccharide transport systems as having multiple transporter population involvement

Kinetic characterization of electrogenic sodium-dependent transport in Ussing chambers of d-glucose and d-galactose demonstrated sigmoidal/Hill kinetics in the porcine jejunum and ileum, with the absence of transport in the distal colon. In the jejunum, a high-affinity, super-low-capacity (Ha/sLc) kinetic system accounted for glucose transport, and a low-affinity, low-capacity (La/Lc) kinetic system accounted for galactose transport. In contrast, the ileum demonstrated a high-affinity, super-high-capacity (Ha/sHc) glucose transport and a low-affinity, high-capacity (La/Hc) galactose transport systems. Jejunal glucose transport was not inhibited by dapagliflozin, but galactose transport was inhibited. Comparatively, ileal glucose and galactose transport were both sensitive to dapagliflozin. Genomic and gene expression analyses identified 10 of the 12 known SLC5A family members in the porcine jejunum, ileum, and distal colon. Dominant SGLT1 (SLC5A1) and SGLT3 (SLC5A4) expression was associated with the sigmoidal Ha/sLc glucose and La/Lc galactose transport systems in the jejunum. Comparatively, the dominant expression of SGLT1 (SLC5A1) in the ileum was only associated with Ha glucose and La galactose kinetic systems. However, the sigmoidal kinetics and overall high capacity (Hc) of transport is unlikely accounted for by SGLT1 (SLC5A1) alone. Finally, the absence of transport and lack of pharmacological inhibition in the colon was associated with the poor expression of SLC5A genes. Altogether, the results demonstrated intestinal segregation of monosaccharide transport fit different sigmoidal kinetic systems. This reveals multiple transporter populations in each system, supported by gene expression profiles and pharmacological inhibition. Overall, this work demonstrates a complexity to transporter involvement in intestinal electrogenic monosaccharide absorption systems not previously defined.

Intestinal electrogenic sodium-dependent glucose absorption in tilapia and trout reveal species differences in SLC5A-associated kinetic segmental segregation

Electrogenic sodium-dependent glucose transport along the length of the intestine was compared between the omnivorous Nile tilapia ( Oreochromis niloticus) and the carnivorous rainbow trout ( Oncorhynchus mykiss) in Ussing chambers. In tilapia, a high-affinity, high-capacity kinetic system accounted for the transport throughout the proximal intestine, midintestine, and hindgut segments. Similar dapagliflozin and phloridzin dihydrate inhibition across all segments support this homogenous high-affinity, high-capacity system throughout the tilapia intestine. Genomic and gene expression analysis supported findings by identifying 10 of the known 12 SLC5A family members, with homogeneous expression throughout the segments with dominant expression of sodium-glucose cotransporter 1 (SGLT1; SLC5A1) and sodium-myoinositol cotransporter 2 (SMIT2; SLC5A11). In contrast, trout's electrogenic sodium-dependent glucose absorption was 20-35 times lower and segregated into three significantly different kinetic systems found in different anatomical segments: a high-affinity, low-capacity system in the pyloric ceca; a super-high-affinity, low-capacity system in the midgut; and a low-affinity, low-capacity system in the hindgut. Genomic and gene expression analysis found 5 of the known 12 SLC5A family members with dominant expression of SGLT1 ( SLC5A1), sodium-glucose cotransporter 2 (SGLT2; SLC5A2), and SMIT2 ( SLC5A11) in the pyloric ceca, and only SGLT1 ( SLC5A1) in the midgut, accounting for differences in kinetics between the two. The hindgut presented a low-affinity, low-capacity system partially attributed to a decrease in SGLT1 ( SLC5A1). Overall, the omnivorous tilapia had a higher electrogenic glucose absorption than the carnivorous trout, represented with different kinetic systems and a greater expression and number of SLC5A orthologs. Fish differ from mammals, having hindgut electrogenic glucose absorption and segment specific transport kinetics.

Inhibition of the intestinal glucose transporter GLUT2 by flavonoids

We tested whether the dominant intestinal sugar transporter GLUT2 was inhibited by intestinal luminal compounds that are inefficiently absorbed and naturally present in foods. Because of their abundance in fruits and vegetables, flavonoids were selected as model compounds. Robust inhibition of glucose and fructose transport by GLUT2 expressed in Xenopus laevis oocytes was produced by the flavonols myricetin, fisetin, the widely consumed flavonoid quercetin, and its glucoside precursor isoquercitrin [corrected]. IC50s for quercetin, myricetin, and isoquercitirin [corrected]were approximately 200- to 1000-fold less than glucose or fructose concentrations, and noncompetitive inhibition was observed. The two other major intestinal sugar transporters, GLUT5 and SGLT1, were unaffected by flavonoids. Sugar transport by GLUT2 overexpressed in pituitary cells and naturally present in Caco-2E intestinal cells was similarly inhibited by quercetin. GLUT2 was detected on the apical side of Caco-2E cells, indicating that GLUT2 was in the correct orientation to be inhibited by luminal compounds. Quercetin itself was not transported by the three major intestinal glucose transporters. Because the flavonoid quercetin, a food component with an excellent pharmacology safety profile, might act as a potent luminal inhibitor of sugar absorption independent of its own transport, flavonols show promise as new pharmacologic agents in the obesity epidemic.

Potent diarrheagenic mechanism mediated by the cooperative action of three enteropathogenic Escherichia coli-injected effector proteins

Enteropathogenic Escherichia coli (EPEC) induces a severe watery diarrhea responsible for several hundred thousand infant deaths each year by a process correlated with the loss (effacement) of absorptive microvilli. Effacement is linked to the locus of enterocyte effacement pathogenicity island that encodes an "injection system," "effector" proteins, and the Intimin outer membrane protein. Here, we reveal that effacement (i) is a two-step process, (ii) requires the cooperative action of three injected effectors (Map, EspF, and Tir) as well as Intimin, and (iii) leads to the retention, not release (into the extracellular milieu), of the detached microvillar material. We also discover that EPEC rapidly inactivates the sodium-d-glucose cotransporter (SGLT-1) by multiple mechanisms. Indeed, the finding that one mechanism occurs more rapidly than microvilli effacement provides a plausible explanation for the rapid onset of severe watery diarrhea, given the crucial role of SGLT-1 in the daily uptake of approximately 6 liters of fluids from the normal intestine. The importance of SGLT-1 in the disease process is supported by severe EPEC diarrheal cases being refractory to oral rehydration therapy (dependent on SGLT-1 function). Moreover, the identification of effector activities that alter microvilli structure and SGLT-1 function provides new tools for studying the underlying regulatory processes.

Apical uptake and transepithelial transport of sphingosine monomers through intact human intestinal epithelial cells: physicochemical and molecular modeling studies

The mechanism of absorption of sphingosine was studied in human intestinal epithelial cells Caco-2 and HT-29-D4. The experiments were performed below the critical micellar concentration of sphingosine which was evaluated to 6 microM by surface tension measurements. [3H]Sphingosine uptake was not inhibited by Na+-free conditions, ATP depletion, L-cycloserine or methyl-beta-cyclodextrin, consistent with a passive diffusion mechanism independent of lipid raft integrity. Molecular modeling studies suggested that sphingosine can adopt two distinct conformations: a high-energy "snake-like" conformer in water and an extended low-energy conformer in lipid phases. We propose that the energy stored in the compressed snake-like conformer is transformed into kinetic energy, allowing: (i) the motion of sphingosine through the unstirred water layer bathing the mucosal enterocyte surface, and (ii) its insertion into the enterocyte brush border membrane. Dietary lipids that stabilized the extended sphingosine conformer in mixed micelles (e.g., cholesterol and sphingomyelin) induced a marked inhibition of sphingosine absorption.

Cultured monolayers of the dog jejunum with the structural and functional properties resembling the normal epithelium

The development of a culture of the normal mammalian jejunum motivated this work. Isolated crypt cells of the dog jejunum were induced to form primary cultures on Snapwell filters. Up to seven subcultures were studied under the electron microscope and in Ussing chambers. Epithelial markers were identified by RT-PCR, Western blot, and immunofluorescent staining. Confluent monolayers exhibit a dense apical brush border, basolateral membrane infoldings, desmosomes, and tight junctions expressing zonula occludens-1, occludin-1, and claudin-3 and -4. In OptiMEM medium fortified with epidermal growth factor, hydrocortisone, and insulin, monolayer transepithelial voltage was -6.8 mV (apical side), transepithelial resistance was 1,050 Omega.cm(2), and short-circuit current (I(sc)) was 8.1 microA/cm(2). Transcellular and paracellular resistances were estimated as 14.8 and 1.1 kOmega.cm(2), respectively. Serosal ouabain reduced voltage and current toward zero, as did apical amiloride. The presence of mRNA of alpha-epithelial Na(+) channel (ENaC) was confirmed. Na-d-glucose cotransport was identified with an antibody to Na(+)-glucose cotransporter (SGLT) 1. The unidirectional mucosa-to-serosa Na(+) flux (19 nmol.min(-1).cm(-2)) was two times as large as the reverse flux, and net transepithelial Na(+) flux was nearly double the amiloride-sensitive I(sc). In plain Ringer solution, the amiloride-sensitive I(sc) went toward zero. Under these conditions plus mucosal amiloride, serosal dibutyryl-cAMP elicited a Cl(-)-dependent I(sc) consistent with the stimulation of transepithelial Cl(-) secretion. In conclusion, primary cultures and subcultures of the normal mammalian jejunum form polarized epithelial monolayers with 1) the properties of a leaky epithelium, 2) claudins specific to the jejunal tight junction, 3) transepithelial Na(+) absorption mediated in part by SGLT1 and ENaC, and 4) electrogenic Cl(-) secretion activated by cAMP.

Interaction of cholesterol with sphingosine

Molecular associations between sphingomyelin and cholesterol provide a molecular basis for the colocalization of these lipids in plasma membrane microdomains (lipid rafts) and for the inhibitory effect of sphingomyelin on the intestinal absorption of cholesterol. Using surface pressure measurements at the air-water interface, we showed that sphingosine, the common sphingoid backbone of most sphingolipids, formed condensed lipid complexes with cholesterol. Structure-activity relationship studies with long-chain analogs of sphingosine, together with molecular mechanics simulations, were consistent with a specific interaction between sphingosine and the alpha face of cholesterol. The uptake of micellar cholesterol and the effect of sphingosine on cholesterol absorption were studied with two human model intestinal epithelial cell lines, Caco-2 and HT-29-D4. Real-time PCR quantifications of the putative cholesterol transporter Niemann-Pick C1 like 1 (NPC1L1) mRNA revealed that, in these cell lines, the activity of cholesterol transport correlated with the level of NPC1L1 expression. In both cell lines, sphingosine induced a dose-dependent decrease of cholesterol absorption. Yet the effect of sphingosine was more dramatic in Caco-2 cells, which also displayed the highest expression of NPC1L1 mRNA. Altogether, these data suggested that sphingosine interacts specifically with cholesterol and inhibits the intestinal NPC1L1-dependent transport of micellar cholesterol.

Expression of SGLT-1 in preneoplastic and neoplastic lesions of the head and neck

Tumors show an increased glucose uptake that is mediated by glucose transport proteins. We have analyzed the expression of the sodium-dependent glucose co-transporters SGLT-1 and-2 in short-term cultures of squamous cell carcinomas of the head and neck (HNSCC) by RT-PCR. Distribution of the SGLT-1 protein in HNSCC tissues was investigated by immunohistochemistry. While we observed in 17/36 HNSCC short-term cultures the SGLT-1 mRNA, we found no SGLT-2 expression. SGLT-1 mRNA expression occurred preferentially in cultures originating from moderately and well differentiated HNSCC. In tumor tissues a heterogeneous SGLT-1 staining restricted to differentiated tumor cells was seen in 27/30 HNSCC analyzed. In normal mucosa SGLT-1 staining was also confined to differentiated compartments and lacked in dysplastic areas. Our data indicate a differentiation-dependent expression of SGLT-1 in HNSCC. This is important knowledge for the planning of glucose-targeting therapies and suggest SGLT-1 as a differentiation marker in head and neck tissues.

Glycosphingolipid (GSL) microdomains as attachment platforms for host pathogens and their toxins on intestinal epithelial cells: activation of signal transduction pathways and perturbations of intestinal absorption and secretion

Glycosphingolipid (GSL)-enriched microdomains are used as cellular binding sites for various pathogens including viruses and bacteria. These attachment platforms are specifically associated with transducer molecules, so that the binding of host pathogens (or their toxins) to the cell surface may result in the activation of signal transduction pathways. In the intestinal epithelium, such pathogen-induced dysregulations of signal transduction can elicit a severe impairment of enterocytic functions. In this study, we demonstrate that the interaction of a bacterial toxin (cholera toxin) and a viral envelope glycoprotein (HIV-1 gp120) with the apical plasma membrane of intestinal cells is mediated by GSL-enriched microdomains that are associated with G regulatory proteins. These microbial proteins induce a GSL-dependent increase of intestinal fluid secretion by two mechanisms: activation of chloride secretion and inhibition of Na+ -dependent glucose absorption. Taken together, these data support the view that GSL-enriched microdomains in the apical plasma membrane of enterocytes are involved in the regulation of intestinal functions.

Differences in the electrophysiological response to I- and the inhibitory anions SCN- and ClO-4, studied in FRTL-5 cells

The electrophysiological properties of the Na+/I- symporter (NIS) were examined in a cloned rat thyroid cell line (FRTL-5) using the whole-cell patch-clamp technique. When the holding potential was between -40 mV and -80 mV, 1 mM NaI and NaSCN induced an immediate inward current which was greater with SCN- than with I-. The reversal potential for I- and SCN- induced membrane currents was +50 mV. This is close to the value of +55 mV calculated by the Nernst equation for Na+. These results are consistent with I- and SCN- translocation via the NIS that is energized by the electrochemical gradient of Na+ and coupled to the transport of two or more Na+. There was no change in the membrane current recording with ClO-4 indicating that ClO-4 was either not transported into the cell, or the translocation was electroneutral. ClO-4 addition, however, did reverse the inward currents induced by I- or SCN-. These effects of I-, SCN- and ClO-4 on membrane currents reflect endogenous NIS activity since the responses duplicated those seen in CHO cells transfected with NIS. There were additional currents elicited by SCN- in FRTL-5 cells under certain conditions. For example at holding potentials of 0 and +30 mV, 1 mM SCN- produced an increasingly greater outward current. This outward current was transient. In addition, when SCN- was washed off the cells a transient inward current was detected. Unlike SCN-, 1-10 mM I- had no observable effect on the membrane current at holding potentials of 0 and +30 mV. The results indicate FRTL-5 cells may have a specific SCN- translocation system in addition to the SCN- translocation by the I- porter. Differences demonstrated in current response may explain some of the complicated influx and efflux properties of I-, SCN- and ClO-4 in thyroid cells.

Thyroid hormone regulation of the Na+/glucose cotransporter SGLT1 in Caco-2 cells

The expression of the Na+/glucose cotransporter (SGLT1) in response to thyroid hormone [3,5,3'-tri-iodo-l-thyronine (T3)] was investigated in the enterocytic model cell line Caco-2/TC7. In differentiated cells, T3 treatment induces an average 10-fold increase in glucose consumption as well as a T3 dose-dependent increase in SGLT1 mRNA abundance. Only cells grown on glucose-containing media, but not on the non-metabolizable glucose analogue alpha-methylglucose (AMG), could respond to T3-treatment. The Vmax parameter of AMG transport was enhanced 6-fold by T3 treatment, whereas the protein abundance of SGLT1 was unchanged. The role of Na+ recycling in the T3-related activation of SGLT1 activity was suggested by both the large increase in Na+/K+ATPase protein abundance and the inhibition, down to control levels, of AMG uptake in ouabain-treated cells. Further investigations aimed at identifying the presence of a second cotransporter that could be expressed erroneously in the colon cancer cell line were unsuccessful: T3-treatment did not modify the sugar-specificity profile of AMG transport and did not induce the expression of SGLT2 as assessed by reverse transcription-PCR. Our results show that T3 can stimulate the SGLT1 cotransport activity in Caco-2 cells. Both transcriptional and translational levels of regulation are involved. Finally, glucose metabolism is required for SGLT1 expression, a result that contrasts with the in vivo situation and may be related to the fetal phenotype of the cells.

Interrelationship between the Na+/glucose cotransporter and CFTR in Caco-2 cells: relevance to cystic fibrosis

Both the Na+-dependent glucose cotransporter (SGLT1) and the cystic fibrosis transmembrane conductance regulator (CFTR) modulate Na+ and fluid movement, although in opposite directions. Yet few studies have investigated a possible interrelationship between these two transporters. By using the Caco-2 human colon carcinoma cell line, we confirmed that the activities of these transporters increased with spontaneous differentiation to the enterocytic phenotype. We showed that SGLT1 was positively regulated by Cl- and that optimal activity of CFTR was dependent on the presence of glucose. We also demonstrated that inhibition of CFTR by glibenclamide or diphenylamine-2-carboxylate did not modify the activity of SGLT1 and inhibition of SGLT1 by phlorizin did not modify the activity of CFTR, although it resulted in inhibition of glycoconjugate synthesis. These results point to positive substrate-cross regulation of SGLT1 and CFTR and suggest that NaCl and glucose are important for not only Na+ absorption and fluid movement, but also for cAMP-dependent Cl- efflux, and glycoconjugate synthesis, functions that are known to be anomalous in cystic fibrosis.

Effect of atrial natriuretic peptide on alpha-methyl-D-glucoside intestinal active uptake in rats

In vivo, atrial natriuretic peptide (ANP) inhibits water and sodium absorption by the intestine. In addition, ANP inhibits glucose (re)absorption at the level of both the intestine and kidney. ANP also decreases sodium absorption in the rat small intestine in vitro, but only if glucose is present on the luminal side of the tissue. These findings suggest that ANP inhibits the sodium-glucose cotransporter (SGLT) of enterocytes. In the present study the inhibitory effect of 1 microM ANP on SGLT1 in rat small intestine and colon was tested. For this purpose, the apparent kinetic constants of SGLT1 were determined using radioactive alpha-methyl-D-glucoside (alpha-MG), a non-metabolizable glucose analogue that selectively serves the luminal Na+-dependent intestinal uptake, but not the serosal-facilitated diffusion sugar carrier. In both tissues, incubation with ANP increased Km without modifying the Vmax. In addition, Vmax in the small intestine was found to be higher than in the colon. The evidence presented here suggests that ANP, through its second messenger, may be a competitive inhibitor of SGLT1. Since SGLT1 is also expressed in the brush-border membrane of the renal proximal tubule, we suggest that this peptide might regulate the hydro-saline balance at intestinal and proximal tubular nephron levels.

Direct effect of type 1 human immunodeficiency virus (HIV-1) on intestinal epithelial cell differentiation: relationship to HIV-1 enteropathy

Human immunodeficiency virus (HIV)-infected patients display severe impairments of gastrointestinal functions, including diarrhea and malabsorption, even in the absence of opportunistic infections. Since HIV-1 proteins and nucleic acids have been detected in several cell types of the intestinal mucosa, it has been postulated that HIV-1 itself could alter enterocytic functions. In the present study, we analyzed the effect of HIV-1 on the differentiation process of the epithelial intestinal cell clone HT-29-D4, which mimics the maturation of enterocytes along the crypt-villus axis of the small intestine. We found that HIV-1 infection impairs cellular differentiation (i) by affecting the barrier function of the epithelium, as evidenced by a decrease in the transepithelial electrical resistance, and (ii) by inhibiting the activity of one major glucose absorption function, i.e., sodium/glucose cotransport. At the morphological level, HIV-1 infection of HT-29-D4 cells was associated with the formation of lumina, which are representative of a defect in cellular organization. These morphofunctional perturbations induced by HIV-1 could be mimicked by nocodazole, a microtubule-disrupting agent. Correspondingly, HIV-1 exposure of HT-29-D4 cells evoked a massive disruption of microtubules, as revealed by alpha-tubulin indirect immunofluorescence staining. A similar effect was observed after incubation of the cells with either recombinant gp120 or a monoclonal antibody against galactosylceramide (GalCer), the intestinal receptor for HIV-1 gp120, suggesting that the effect of HIV-1 was mediated by the binding of gp120 to GalCer. Based on these data, we propose that HIV-1 may selectively alter enterocytic functions through a direct effect on the intracellular architecture of the cells. In contrast with previous theories for HIV-1 enteropathy, our data support the concept that HIV-1 may perturb intestinal functions without necessarily infecting intestinal epithelial cells.