| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TRANSLATIONAL PHYSIOLOGY
Section of Digestive Diseases, Department of Medicine, West Virginia University, School of Medicine
Submitted 28 December 2007 ; accepted in final form 17 April 2008
ABSTRACT
Glutamine, the primary metabolic fuel for the mammalian small intestinal enterocytes, is primarily assimilated by Na-amino acid cotransporters. Although Na-solute cotransport has been shown to exist in the brush border membrane (BBM) of the absorptive villus cells, the identity of Na-glutamine cotransport in rabbit small intestinal villus cells was unknown. Na-dependent glutamine uptake is present in villus BBM vesicles. An intravesicular proton gradient did not stimulate this Na-dependent glutamine uptake, whereas Li+ did not significantly suppress this uptake. These observations in concert with amino acid substitution studies suggested that Na-glutamine cotransporter in the villus cell BBM was the newly identified cotransporter B0AT1 (SLC6A19). Quantitative real-time PCR identified the message for this cotransporter in villus cells. Thus a full-length cDNA of B0AT1 was cloned and expressed in MDA-MB-231 cells. This expressed cotransporter exhibited characteristics similar to those observed in villus cells from the rabbit small intestine. Antibody was generated for B0AT1 that demonstrated the presence of this cotransporter protein in the villus cell BBM. Kinetic studies defined the kinetic parameters of this cotransporter. Thus this study describes the identification, cloning, and characterization of the Na-amino acid cotransporter responsible for the assimilation of a critical amino acid by the absorptive villus cells in the mammalian small intestine.
regulation of intestinal absorption; Na-amino acid cotransport; epithelial transport; villus cells
In the intestine, heterogeneity of expression of amino acid cotransporters may be secondary to entry of different kinds of food that interact for cotransporter expression. In the mammalian small intestine, larger neutral amino acids have been shown to be transported by three different transport systems, namely system B0, system B0,+ (Na-dependent transporter for neutral and cationic amino acids), and system ASC (Na-dependent transporter for midsize neutral amino acids; Refs. 11, 12). In addition, proton-amino acid cotransporter 1 (PAT1) transports small neutral amino acids (1). The activity of the intestinal system B0 has also been observed in the renal cortex (9, 14). Recently, successful identification of a mouse B0-type transporter was reported by Broer and colleagues (4). This transporter was named B0AT1 (SLC6A19) and belongs to a cluster of orphan transporters within the family of Na-dependent neurotransmitters and amino acid transporters (SLC6). When expressed in Xenopus oocytes, the B0AT1 cDNA induces a Na-dependent uptake of neutral amino acids.
It is noteworthy, however, that several related gene products, two in human (products of SLC6A18 and 20) and three in mice (SLC6A18 and two genes corresponding to human SLC6A20), are expressed in the kidney brush border and, to some extent, in the small intestine as well. Interestingly, whereas the gene product of SLC6A19 (B0AT1) is expressed mainly in the early part of the proximal tubule, the product of Slc6a18 is most abundant in the late proximal segments. In contrast, the gene products derived from the SLC6A20 homologues appear to be expressed all along the proximal tubule (20). It was hypothesized that these gene products perform B0-type Na-amino acid cotransport similar to that of B0AT1, but with somewhat different selectivity, apparent affinity, and axial distribution. However, it is not known which Na-dependent amino acid cotransporter is responsible for the assimilation of glutamine in the mammalian small intestine.
Given the reported heterogeneity of Na-dependent amino acid cotransporters, the aims of the present study were to identify the predominant Na-dependent glutamine cotransporter in the mammalian intestinal absorptive villus cells, then clone, express, and characterize it.
MATERIALS AND METHODS
Reagents. L-Glutamine (Glu), L-alanine (Ala), L-serine (Ser), L-methyl-aminoisobutyric acid (MeAIB), L-arginine (Arg), L-tryptophan (Trp), L-histidine (His), L-asparagine (Asn), L-aspartate (Asp), L-glutamate (Glu), and β-hydroxybutyric acid from Sigma Chemical (St. Louis, MO) and insulin from Novo Nordisk (Clayman, NC) were purchased. HEPES buffer, Leibovitz's L-15 medium powder (L-15), bovine fetal serum, and all supplements were supplied from Invitrogen (Grand Island, NY) and [3H]glutamine from Amersham Biosciences (Little Chalfont, UK).
Cell isolation. Pathogen-free New Zealand White male rabbits weighing 2–2.5 kg were euthanized according to the Guiding Principles for the Care and Use of Laboratory Animals according to a protocol approved by the West Virginia University IACUC. Villus cells were isolated from normal rabbit small intestine by a Ca2+ chelation technique as previously described (16). Briefly, a 3-ft section of ileum was filled with buffer containing (in mM) 0.15 EDTA, 112 NaCl, 25 NaHCO3, 2.4 K2HPO4, 0.4 KH2PO4, 2.5 L-glutamine, 0.5 β-hydroxybutyrate, 0.5 dithiothreitol, and gassed with 95% O2-5% CO2, pH 7.4, at 37°C. The intestine was incubated in this solution for 3 min and gently palpitated for another 3 min to facilitate cell dispersion. The fluid was then drained from the ileal loop, phenylmethylsulfonyl fluoride was added, and the suspension was centrifuged at 100 g for 3 min. Cells used for BBM vesicle (BBMV) preparation were frozen immediately in liquid nitrogen and stored at –80°C until required. Previously established criteria were used to validate good separation of villus cells. These criteria included 1) marker enzymes (e.g., thymidine kinase, alkaline phosphatase), 2) transporter specificity (e.g., Na-glucose cotransport and Na+/H+ exchange are present on the BBM of villus cells), and 3) morphological differences (e.g., villus cells are larger, with better developed BBM compared with crypt cells).
BBMV preparation. BBMV from rabbit intestinal villus cells were prepared by CaCl2 precipitation and differential centrifugation as previously described (17). Briefly, frozen villus cells were thawed and suspended in 2 mM Tris·HCl buffer (pH 7.0) containing 50 mM mannitol. The suspension was homogenized and 10 mM CaCl2 was added. The suspension was then centrifuged at 8,000 g for 15 min, and the resulting supernatant was centrifuged at 20,000 g for 30 min. The pellet was then resuspended in 10 mM Tris·HCl buffer (pH 7.5) containing 100 mM mannitol and homogenized. Vesicles were formed by adding MgCl2 to 10 mM. The homogenate was centrifuged at 2,000 g for 15 min to remove debris, and the BBMV were precipitated by centrifugation at 27,000 g for 15 min. BBMV were resuspended in a medium appropriate to each experiment. BBMV purity was assured by marker enzyme (e.g., alkaline phosphatase) enrichment.
Uptake studies in BBMV. Vesicle uptake studies were performed by the rapid-filtration technique as previously described (17). In brief, 5 µl of BBMV resuspended in 100 mM choline chloride, 0.10 mM MgSO4, 50 mM HEPES-Tris (pH 7.4), 50 mM mannitol, and 75 mM KCl were incubated in 95 µl of reaction medium that contained 50 mM HEPES-Tris buffer (pH 7.4), 2 nmol glutamine, 10 µCi [3H]glutamine, 0.10 mM MgSO4, 75 mM KCl, 50 mM mannitol, 100 mM of either NaCl or choline chloride, 10 µM valinomycin, and 100 µM carbonyl cyanide p-(tri-fluoromethoxy) phenyl-hydrazone. At the desired time, uptake was arrested by mixing with ice-cold stop solution (50 mM HEPES-Tris buffer, 0.10 mM MgSO4, 75 mM KCl, and 100 mM choline chloride, 2 mM glutamine, pH 7.4). The mixture was filtered on a 0.45-µm Millipore (HAWP) filter and washed two times with 5 ml of ice-cold stop solution. Filters with BBMV were dissolved in 5 ml of Ecoscint A and kept in the dark for 12 h, and radioactivity was determined in a Beckman Coulter LS 6500 scintillation counter.
RTQ-PCR studies. Total RNA was isolated from villus cells by use of TRIzol reagent from Invitrogen Life Technologies. Real-time quantitative PCR (RTQ-PCR) was performed using total RNA isolated by a two-step method. First-strand cDNA synthesis from total RNA was performed by using SuperScript III from Invitrogen Life Technologies using an equal mixture of oligo(dT) primer and random hexamers. The cDNA generated was used as template for real-time PCR using TaqMan Universal PCR master mix from Applied Biosystems (Foster City, CA) according to the manufacturer's protocol.
B0AT1 specific primer and probe sequences were as follows: forward primer 5'-GGATCCTGCTGTGCCTCA-3'; reverse primer 5'-CGAGGTAGGGTAGCGTAGAAGT-3'; TaqMan probe 5'-6FAM- CACCATCCGCGGAATCGAGACAAC-TAMRA-3'.
RTQ-PCR for rabbit β-actin was run along with the B0AT1 RTQ-PCR as an endogenous control using rabbit β-actin-specific primers and probes. The expression of β-actin was used to normalize the expression levels of B0AT1 between the individual samples. The sequences of the rabbit β-actin primers and probes were as follows: forward primer 5'-GCTATTTGGCGCTGGACTT-3'; reverse primer 5'-GCGGCTCGTAGCTCTTCTC-3'; TaqMan probe 5'-FAM- AAGAGATGGCCACGGCCGCGAAC-TAMRA-3'.
Final concentrations of primers and probes for B0AT1 and β-actin were 500 and 100 nM, respectively. The parameters for RTQ-PCR were: 95°C for 15 s, and 58°C for 1 min. All experiments were performed in triplicate and repeated at least thrice with RNA obtained from separate animals. Serial dilution experiments of cDNA were performed to establish that the efficiency of PCR was the same between β-actin and B0AT1 transporters (data not shown).
Rabbit B0AT1 cloning. Full-length B0AT1 was obtained by PCR using Taq polymerase PCR kit from Promega (Madison, WI). For the purpose of transient transfection, full-length B0AT1 was cloned into pcDNA3.1/Zeo(+), a mammalian expression vector from Invitrogen Life Technologies (Carlsbad, CA). Rabbit B0AT1 sequence corresponding to the NH2 and COOH termini of the protein were obtained by aligning the mouse B0AT1 protein sequence against the rabbit genomic sequence available from the public genomic sequence source and identifying the genomic sequence of maximum homology to the mouse B0AT1 protein sequence. Thus the nucleotide sequence was obtained and used to design PCR primers specific to rabbit B0AT1 (GenBank accession number EU095941). The upstream and downstream primer sequences are as follows: upstream primer 5'-GCC GCC ACC ATG GTG AGG CTG GTG CTG-3'; downstream primer 5'-TCT AGA TCA GTA CTT GAG GTC CCC ATT-3'. Forward primer contained Kozak consensus sequence upstream of the start codon (GCC GCC ACC) to aid in efficient transcription and the downstream primer contained the stop codon followed by XbaI restriction site to aid in unidirectional cloning into the expression vector pcDNA3.1/Zeo(+). The underlined sequence in the upstream primer is the Kozak consensus sequence followed by the start codon. The resultant PCR product was cloned into pGEM-T Easy vector from Promega (Madison, WI) and the B0AT1 insert was released by digesting with EcoRI and was inserted into EcoRI-digested pcDNA3.1/Zeo(+). Both sense and antisense strands of the cDNAs were sequenced by primer walking by use of the custom sequencing service from Agencourt Bioscience (Beverly, MA).
Expression and characterization of B0AT1.
The normal, diploid, human breast adenocarcinoma (MDA-MB-231) cell line (HTB-26, American Type Culture Collection, Manassas, VA/ Rockville, MD) was used between passages 5 and 20. Density of cells were 
0.4 x 106 cells/well, grown on 12-well plastic plate (Corning, NY) in L-15 containing 2 mM L-glutamine, 10% vol/vol fetal bovine serum, and 0.02% insulin, without any antibiotics in a humidified atmosphere at 37°C. To determine the function of the novel sequence of B0AT1, MDA-MB-231 cells were transiently transfected with B0AT1 cDNA. When the cells attained 70 to 90% confluency, generally at 24 h postseeding, 1.6 µg B0AT1 plasmid was transiently transfected into the cells by use of Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol. pcDNA 3.1 vector without the B0AT1 cDNA was used as control. The transfected cells were used for transporter assay 72 h posttransfection.
Amino acid transport studies were performed in plastic plates using triplicate wells for each time point. Cells were rinsed once with oxygenated Leibovitz's (L-15, 100% oxygen) medium (pH 7.4; pH was adjusted with tetramethyl ammoniun (TMA)-OH) supplemented with 10% fetal calf serum, 20 mM HEPES, 200 U/l insulin (0.02%) and incubated with L-15 at room temperature for 1 h. Cells were washed once with TMA-HEPES buffer (4.7 mM KCl, 1 mM MgSO4, 1.2 mM KH2PO4, 20 mM HEPES, 125 mM CaCl2, 130 mM TMA-Cl, pH 7.4; pH was adjusted with TMA-OH) and incubated with TMA-HEPES (Na-free) buffer for 10 min in room temperature. Cells were treated with reaction mixture containing 2 nmol of cold glutamine and [3H]glutamine, in TMA-HEPES buffer (Na-free) and with 130 mM NaCl (Na-HEPES) for a specified time. The reaction was stopped and washed twice with ice-cold TMA-HEPES (Na-free) buffer. NaOH (1 M, 800 µl) was added in each well and incubated for 30 min at 70°C to digest the cells. Cell extracts were transferred into a scintillation vial and 5 ml of scintillation fluid (Ecoscint A) was added and counted [3H]glutamine in a scintillation counter.
Western blot studies. Polyclonal antibody against B0AT1 was raised in goat by using the custom antibody services provided by Invitrogen. Antigenic peptide with the sequence-CDRFNKDIEFMIGHKPN-coupled to Keyhole limpet hemocyanin was used as the immunogen. Western blotting for B0AT1 was performed according to the standard protocols (2). Briefly, plasma membranes from villus cells were solubilized in RIPA buffer (50 mM Tris·HCl pH 7.4, 1% Igepal, 150 mM NaCl, 1 mM EDTA, 1 mM PMSF, 1 mM Na3VO4, 1 mM NaF) containing protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO). Equal volume of 2x SDS/sample buffer (100 mM Tris, 25% glycerol, 2% SDS, 0.01% bromphenol blue, 10% 2-ME, pH 6.8) was added and the proteins were separated on a 4–20% Ready Gel (Bio-Rad Laboratories, Hercules, CA). The separated proteins were transferred on to polyvinylidene difluoride membrane and probed with the primary antibody against rabbit B0AT1. Secondary antibody coupled to horseradish peroxidase was used to monitor the binding of the primary antibody. ECL Western blotting detection reagent (GE Healthcare Bio-Sciences, Piscataway, NJ) was used to detect the B0AT1 specific signal. The resultant chemiluminescent signals were quantitated by using a Molecular Dynamics (Sunnyvale, CA) densitometric scanner. All experiments were performed in triplicate.
IHC studies. Localization of B0AT1 protein was determined by immunohistochemistry (IHC). Normal rabbit small intestine was fixed in 10% (vol/vol) neutral-buffered formalin and embedded in paraffin blocks according to the standard protocols. Sections 4 µm thick were made by using a microtome and were mounted on glass slides. The sectioned tissues were deparaffinized in xylene and were hydrated gradually by graded ethanol. The sectioned tissues were incubated in 10 mM sodium citrate buffer, pH 6.0, at 95°C for 5 min to retrieve the antigens. Nonspecific binding sites in tissue sections were blocked by incubation with bovine normal serum (10%, 20 min, room temperature). The tissue sections were incubated overnight (4°C) with 1:200 diluted B0AT1 primary antibody. Excess antibody was removed by washing with PBS (5 min, 3 times) and incubated with (1:1,000) bovine anti-goat IgG secondary antibody coupled to tetramethyl rhodamine isothiocyanate (TRITC; Santa Cruz Biotechnology, Santa Cruz, CA) at room temperature for 1 h. Excess secondary antibody was removed by washing with PBS (5 min, 3 times) and the tissue sections were mounted with ProLong Gold Antifade Reagent (Invitrogen Life Technologies). Finally, the signals generated by TRITC were observed under Zeiss LSM510 confocal microscope and photographed.
Protein determination. Total protein was measured by using the Bio-Rad protein assay kit with BSA as standard.
Statistical analysis. Data were analyzed by Student's t distribution using mean values and the associated standard errors. A comparative P value of less than 0.05 was considered significant.
RESULTS
Na-dependent glutamine uptake in villus cell BBMV. To determine the presence of Na-glutamine cotransport in the BBM of villus cells, glutamine uptake was performed in the presence or absence of Na at given time points (Fig. 1). Na-dependent glutamine uptake, defined as 3[H]glutamine uptake in the presence of Na minus uptake in the absence of Na, was present in villus cells. These data indicate that an active Na-dependent glutamine cotransporter is present in villus cell BBM from the rabbit small intestine.
|
|
|
|
|
|
Western blotting studies. Since mRNA levels do not necessarily correlate with immunoreactive B0AT1 protein, Western blot studies were performed in crypt and villus cell BBMV using anti-B0AT1 polyclonal antibody. A representative Western blot is shown in Fig. 6A where the position of the 75-kDa protein corresponding to B0AT1 is marked with an arrow. Figure 6B shows the quantitative data of B0AT1 protein expression in villus and crypt cells. These data indicated that B0AT1 protein expression is four times higher in the villus cells compared with the crypt cells (P < 0.05, n = 3). Taken together, the functional and molecular data indicate that B0AT1 is the major Na-glutamine cotransporter expressed in the villus but not crypt cells in the rabbit intestine.
|
|
|
|
In the mammalian small intestine amino acids are primarily absorbed by Na-dependent amino acid cotransporters. These cotransporters are secondarily active depending on Na+-K+-ATPase to provide the favorable transcellular Na+ gradient (7, 8). Furthermore, to date they have been shown to be present on the BBM of absorptive villus cells but not secretory crypt cells in the mammalian small intestine (1). Glutamine is an important amino acid because it is the preferred nutrient for the intestinal enterocytes (7, 21). A number of amino acid transport systems have been reported in the intestine of various animals and human and some of these are capable of transporting glutamine (10, 11, 15). However, specifically which Na amino acid cotransporter is primarily responsible for the assimilation of glutamine in the normal mammalian small intestine was previously unknown.
This study for the first time has identified, cloned, and characterized the Na-dependent glutamine cotransporter in the absorptive villus cells of the rabbit small intestine. Na-stimulated glutamine uptake was present in villus cell BBM (Fig. 1). Both system N and system A have been reported to be present in different organs of various species and are capable of transporting glutamine (3, 4, 9, 10). However, both system N and system A are stimulated by an intravesicular proton load (5). Our studies demonstrated that the Na-glutamine cotransporter in the rabbit intestine villus cell BBM is not stimulated by an intravesicular proton load (Fig. 2). Thus the rabbit intestinal villus cell Na-glutamine cotransporter is not system N or system A. Furthermore, our studies demonstrated that the villus cell BBM Na-glutamine cotransporter is not lithium tolerant (Fig. 3). These data also indicate that this Na-glutamine cotransporter also does not belong to system A. Subsequently the substrate inhibition studies of the Na-dependent glutamine cotransporter in the villus cell BBM demonstrated that the system that is most likely responsible for Na-glutamine cotransport in the villus cell BBM is B0AT1 (Table 1). Kinetic studies demonstrated that the Na-glutamine cotransporter in the villus cell BBM follows Michaelis-Menten kinetics. The Vmax of this cotransporter was 4.66 pmol·mg protein–1 at 6 s. The Km for this cotransporter was 34.3 mM (Fig. 4).
In view of the data to this point, which suggested that the rabbit intestinal villus cell BBM Na-glutamine cotransporter is B0AT1, primers and probes were designed for quantitative PCR of this cotransport process. RTQ-PCR demonstrated that B0AT1 mRNA is abundantly present in villus cells compared with the crypt cells from the normal rabbit small intestine (Fig. 5B). This cotransporter was then cloned. To characterize this cloned cotransporter it was transiently expressed in MDA-MB-231 cells. The expressed cotransporter also had substrate specificity that is consistent with B0AT1 properties. Specifically, it is known that Gln, Trp, Ser, and Asn are substrates for B0AT1, and indeed all four of these amino acids significantly inhibited the transport of radiolabeled glutamine by the expressed cotransporter (Table 2). However, it is known that Glu and Asp are not substrates for B0AT1, and once again these two amino acids did not inhibit the uptake of glutamine by the expressed cotransporter. Given that the data indicated that B0AT1 is the predominant Na-amino acid cotransporter responsible for glutamine assimilation in the villus cells of the rabbit intestine, polyclonal antibody for B0AT1 was raised in goat. Western blot and immunohistochemical studies using this antibody demonstrated that B0AT1 is indeed present in the BBM of the rabbit intestinal villus, but not crypt cells (Figs. 6A and 7).
Thus this study has identified, cloned, and characterized the Na glutamine cotransporter responsible for glutamine absorption in the villus cell of the rabbit small intestine. It is well established that glutamine is the preferred substrate for the mammalian small intestinal enterocytes. Thus it is important for the normal growth and maturation of the intestine. Perhaps it is even more vitally important in pathophysiological states of the small intestine where there is substantial loss of the absorptive surface such as conditions characterized by chronic inflammation [e.g., inflammatory bowel disease (IBD)]. In these conditions, return to health of the mucosa will undoubtedly require the preferred nutrient glutamine. Identification and characterization of the primary transporter responsible for glutamine assimilation in the normal small intestine will allow us to explore the alterations in this transporter during chronic intestinal inflammation, which would potentially allow for more efficacious treatment modalities for this very important intestinal disorder.
It is likely that B0AT1 is affected in chronic intestinal inflammation because other amino acid assimilation pathways have previously been demonstrated to be affected. For example, Na-neutral amino acid cotransport, specifically Na-alanine cotransport, mediated by ASCT1 in the rabbit small intestine is inhibited during chronic intestinal inflammation (18). The inhibition in ASCT1 is not secondary to an alteration in the Na+ extruding capacity of the villus cell but rather a direct effect on the cotransporter at the BBM level of the villus cells. It was demonstrated that the mechanism of inhibition of ASCT1 was likely secondary to a decrease in the affinity of the cotransporter for alanine without an alteration in the number of cotransporters on the BBM (18). Likewise it was also previously demonstrated that the proton dipeptide cotransporter PEPT1, yet another pathway of assimilation of amino acids in the intestine, was also significantly altered during chronic intestinal inflammation (18). PEPT1 was present in villus but not crypt cells of the rabbit small intestine. PEPT1 was significantly inhibited during chronic intestinal inflammation. Further kinetic studies demonstrated that the mechanism of this inhibition was secondary to a reduction in the affinity of the cotransporter rather than an alteration in the number of cotransporters on the BBM of villus cells (18).
Fortunately as it relates to the treatment of IBD, these alterations in the chronically inflamed intestine lend themselves to modulation. For example, it was previously demonstrated that a very common modality of treatment of IBD, specifically corticosteroids, had significant effects on the altered ASCT1 during chronic intestinal inflammation (19). It was demonstrated that corticosteroids reversed the inhibition of ASCT1 during chronic intestinal inflammation. Furthermore, it was of note that the mechanism of reversal was specifically secondary to a restoration in the affinity of the cotransporter without altering the number of cotransporter in the chronically inflamed rabbit intestine. This suggested that corticosteroids, being a broad-spectrum immune modulator, more than likely inhibited the immune-inflammatory mediator that was responsible for the inhibition of this Na-amino acid cotransporter. Although these studies demonstrated that amino acid and dipeptide transport are significantly affected during chronic intestinal inflammation, how the amino acid that is the primary metabolite for intestinal epithelial cells might be affected in this condition was unknown. The reason for this was that it was previously unknown specifically which Na-amino acid cotransporter was responsible for the assimilation of glutamine in the mammalian small intestine. Having identified this cotransporter as B0AT1, this study now provides the means to further delineating its role in chronic intestinal inflammation.
In conclusion, this study has for the first time identified the Na-amino acid cotransporter in the mammalian small intestinal absorptive villus cells responsible for the assimilation of the most important nutrient for the enterocytes as B0AT1. B0AT1 is functionally present in villus cells; it was cloned and characterized from the rabbit intestinal villus cells. Finally, the functional protein and its message were also shown to be present in the rabbit small intestinal villus, but not crypt cells.
GRANTS
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Research Grants DK-45062 and DK-58034 to U. Sundaram.
Address for reprint requests and other correspondence: U. Sundaram, Section of Digestive Diseases, Dept. of Medicine, PO Box 9161, One Medical Center Dr., West Virginia Univ. School of Medicine, Morgantown, WV 26506 (e-mail: usundaram{at}hsc.wvu.edu)
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
