Proc. Natl. Acad. Sci. USA
`Vol. 94, pp. 2705–2710, March 1997
`Pharmacology
`
`Regulation of intestinal uroguanylin兾guanylin receptor-mediated
`responses by mucosal acidity
`F. KENT HAMRA*†‡, SAMMY L. EBER*†, DAVID T. CHIN†, MARK G. CURRIE†§, AND LEONARD R. FORTE*†¶
`*Truman Veterans Affairs Medical Center and †Departments of Pharmacology and Biochemistry and Molecular Biology Program, Missouri University, Columbia,
`MO 65212; and §Searle Research and Development, St. Louis, MO 63167
`
`Communicated by Philip Needleman, Monsanto Company, St. Louis, MO, January 2, 1997 (received for review March 27, 1996)
`
`Guanylin and uroguanylin are intestinal
`ABSTRACT
`peptides that stimulate chloride secretion by activating a
`common set of receptor–guanylate cyclase signaling molecules
`located on the mucosal surface of enterocytes. High mucosal
`acidity, similar to the pH occurring within the fluid micro-
`climate domain at the mucosal surface of the intestine,
`markedly enhances the cGMP accumulation responses of T84
`human intestinal cells to uroguanylin. In contrast, a mucosal
`acidity of pH 5.0 renders guanylin essentially inactive. T84
`cells were used as a model epithelium to further explore the
`concept that mucosal acidity imposes agonist selectivity for
`activation of the intestinal receptors for uroguanylin and
`guanylin, thus providing a rationale for the evolution of these
`related peptides. At an acidic mucosal pH of 5.0, uroguanylin
`is 100-fold more potent than guanylin, but at an alkaline pH
`of 8.0 guanylin is more potent than uroguanylin in stimulating
`intracellular cGMP accumulation and transepithelial chlo-
`ride secretion. The relative affinities of uroguanylin and
`guanylin for binding to receptors on the mucosal surface of
`T84 cells is influenced dramatically by mucosal acidity, which
`explains the strong pH dependency of the cGMP and chloride
`secretion responses to these peptides. The guanylin-binding
`affinities for peptide–receptor interaction were reduced by
`100-fold at pH 5 versus pH 8, whereas the affinities of
`uroguanylin for these receptors were increased 10-fold by
`acidic pH conditions. Deletion of the N-terminal acidic amino
`acids in uroguanylin demonstrated that these residues are
`responsible for the increase in binding affinities that are
`observed for uroguanylin at acidic pH. We conclude that
`guanylin and uroguanylin evolved distinctly different struc-
`tures, which enables both peptides to regulate, in a pH-
`dependent fashion, the activity of receptors that control
`intestinal salt and water transport via cGMP.
`
`Guanylin and uroguanylin are structurally related peptides
`that were isolated from intestinal mucosa and urine (1–5). A
`receptor for guanylin and uroguanylin that has been identified
`at the molecular level is a transmembrane form of guanylate
`cyclase, termed GC-C (6). This membrane protein was origi-
`nally discovered as an intestinal receptor for the heat-stable
`toxin (ST) peptides, which are secreted intraluminally by
`enteric bacteria that cause traveler’s diarrhea (7). Bacterial ST
`peptides are related in primary structure to uroguanylin and
`guanylin, thus acting as molecular mimics of the enteric
`peptide hormones (reviewed in refs. 8 and 9). Membrane
`receptor–guanylate cyclases are found on the luminal surface
`of enterocytes throughout the small and large intestine and in
`other epithelia (10–13). Binding of peptide agonists to an
`extracellular domain of the receptor activates the intracellular
`
`catalytic domain producing the second messenger cGMP
`within target enterocytes (1–6). Intracellular cGMP stimulates
`transepithelial chloride secretion by regulating the phosphor-
`ylation state and chloride channel activity of the cystic fibrosis
`transmembrane conductance regulator, an apical protein that
`is located with the receptors for uroguanylin, guanylin, and ST
`peptides (14–16).
`Isolation of uroguanylin from opossum urine (2) followed by
`the cloning of a colon cDNA that encodes opossum pre-
`prouroguanylin (17) revealed that the uroguanylin and gua-
`nylin genes are evolutionarily related (18–20). Furthermore,
`the mRNAs and precursor proteins for both uroguanylin and
`guanylin are expressed together throughout the mucosa of
`small and large intestine along with their receptors (5, 11,
`17–20). This raised a question of whether the differences in
`primary structure between guanylin and uroguanylin evolved
`to regulate intestinal salt and water transport through a
`cooperative mechanism using common receptor–guanylate
`cyclase signaling molecules located on the mucosal surface of
`the intestine.
`During the isolation of uroguanylin, guanylin, and their
`prohormone precursors, we observed that acidic column re-
`agents markedly attenuated the cGMP responses of T84 cells
`to guanylin, but enhanced the responses to uroguanylin (4, 5).
`This pH dependency for activation of guanylate cyclase was
`successfully used to detect guanylin and uroguanylin during
`their separation and purification from intestinal mucosa. The
`possibility was then considered that the primary structures of
`guanylin and uroguanylin could have evolved to regulate the
`enzymatic activity of a common set of receptors over the wide
`range of mucosal acidity that occurs within the intestinal lumen
`during digestion (21–24). We report here that high mucosal
`acidity rendered guanylin ineffective as a cGMP agonist and
`chloride secretogogue, whereas an acid pH markedly enhanced
`the potency of uroguanylin. A mucosal pH of 8.0 substantially
`increased the potency of guanylin but decreased the potency of
`uroguanylin. These changes in agonist potencies were ex-
`plained by corresponding directional shifts in the affinities of
`guanylin and uroguanylin for binding to receptors at pH 5.0
`versus 8.0. Uroguanylin and guanylin cooperatively regulate
`the guanylate cyclase activity of a common set of mucosal
`receptors in a pH-dependent fashion, thus providing an enteric
`signaling pathway for the intrinsic, paracrine regulation of
`intestinal salt and water transport.
`
`MATERIALS AND METHODS
`cGMP Accumulation Assay in T84 Cells. T84 cells were
`cultured in 24-well plastic dishes, and the cGMP levels were
`
`The publication costs of this article were defrayed in part by page charge
`payment. This article must therefore be hereby marked ‘‘advertisement’’ in
`accordance with 18 U.S.C. §1734 solely to indicate this fact.
`Copyright 䉷 1997 by THE NATIONAL ACADEMY OF SCIENCES OF THE USA
`0027-8424兾97兾942705-6$2.00兾0
`PNAS is available online at http:兾兾www.pnas.org.
`
`Abbreviation: ST, heat-stable toxin.
`‡Present address: Howard Hughes Medical Institute and Department
`of Pharmacology, University of Texas Southwest Medical Center,
`5323 Harry Hines Boulevard, Dallas, TX 75235-9050.
`¶To whom reprint requests should be addressed at: Department of
`Pharmacology, School of Medicine, University of Missouri, Colum-
`bia, MO 65212. e-mail: Leonard_R._Forte@muccmail.missouri.edu.
`
`2705
`
`MYLAN EXHIBIT - 1021
`Mylan Pharmaceuticals, Inc. v. Bausch Health Ireland, Ltd. - IPR2022-00722
`
`

`

`2706
`
`Pharmacology: Hamra et al.
`
`Proc. Natl. Acad. Sci. USA 94 (1997)
`
`measured in control and agonist-stimulated cells by radioim-
`munoassay (2). Synthetic peptides were suspended in 200 ␮l of
`each of two assay buffers; pH 8.0 buffer [Dulbecco’s modified
`Eagle’s medium (DMEM)兾20 mM N-(2-hydroxyethyl)pipera-
`zine-N⬘-(2-ethanesulfonic acid) (Hepes)兾50 mM sodium bi-
`carbonate, pH 8.0兾1 mM isobutylmethylxanthine (IBMX)] and
`pH 5.0 buffer [DMEM兾20 mM 2-(N-morpholino)ethanesul-
`fonic acid (Mes), pH 5.0兾1 mM IBMX]. T84 cells were washed
`twice with 200 ␮l of the respective pH 8.0 and pH 5.0 buffers
`before addition of the peptides. These solutions containing the
`bioactive peptides were then added to T84 cells and incubated
`at 37⬚C for 40 min. Following incubation, the reaction medium
`was aspirated, and 200 ␮l of 3.3% perchloric acid was added
`per well to stop the reaction and extract cGMP. The extract was
`adjusted to pH 7.0 with KOH, centrifuged, and 50 ␮l of the
`supernatant was used to measure cGMP.
`Measurement of Short Circuit Current in T84 Cell Mono-
`layers. T84 cells were raised on collagen-coated permeable
`filters and mounted in custom-made Ussing chambers for
`measurement of transepithelial chloride secretion as described
`(15, 25). The buffer in the basolateral reservoir was a Krebs–
`Ringer solution (pH 7.4) (25 mM sodium bicarbonate), con-
`taining 10 mM glucose. Buffers in the apical reservoir were
`Krebs–Ringer solutions containing 10 mM glucose and ad-
`justed to either pH 5.5 (25 mM Mes, minus sodium bicarbon-
`ate) or pH 7.8 (60 mM sodium bicarbonate). The pH of
`basolateral and apical reservoir buffer solutions was main-
`tained by bubbling 95% O2兾5% CO2 through the medium
`(except for the apical reservoir at pH 5.5, which did not use a
`bicarbonate兾CO2 buffering system). The signals for short
`circuit current (Isc) and potential difference across the epi-
`thelium were measured every 20 sec, digitized with an ADC-1
`data recording system (Remote Measurement Systems, Seat-
`tle) and stored for later analysis. The Isc observed with T84
`cells cultured on permeable filters has been shown to be caused
`by the net secretion of chloride across T84 cell monolayers
`when cells were treated with various chloride secretogogues,
`including guanylin, uroguanylin, and Escherichia coli ST (1, 2,
`15, 26).
`Competitive Radioligand-Binding Assay in T84 Cells. Com-
`petitive radioligand-binding experiments were performed with
`intact, T84 cells cultured in 24-well plastic dishes using meth-
`ods that were previously described (15), but at medium pH
`values of 5.0 and 8.0. Identical buffer conditions at pH 5.0 and
`pH 8.0 were used in the competitive radioligand-binding assays
`as those used in the cGMP accumulation bioassays. 125I-ST-
`(1–19) was used as the radioligand (2, 15). Concentration-
`response curves for cGMP accumulation and competitive
`radioligand-binding curves performed with each agonist were
`analyzed with the computer program PRISM (Graphpad, San
`Diego). A better fit of the binding data was consistently
`obtained with a two-site model as compared with a single-site
`model for all agonists at either pH 5 or pH 8 (15). The
`concentrations at which specific binding of the radioligand at
`each binding site was inhibited by 50% IC50, were obtained by
`nonlinear regression of the untransformed competition bind-
`ing data. The apparent equilibrium dissociation constants, Ki,
`for the competing ligands were calculated from the computed
`IC50 values using the previously reported estimates of the
`affinity of the radioligand in these cells, Kd ⬇15 nM (27): Ki ⫽
`IC50兾1 ⫹ (L兾Kd), where L equals the radioligand concentra-
`tion. It should be noted that the calculated IC50 and Ki values
`are essentially identical because the concentration of the
`radioligand used in these studies (⬇120 pM) was a small
`fraction of the reported binding affinity of the radioligand.
`Synthesis of Uroguanylin, Guanylin, and ST Peptides. Hu-
`man uroguanylin (NDDCELCVNVACTGCL) and human
`guanylin (PGTCEICAYAACTGC), and the opossum forms of
`uroguanylin95–109 (QEDCELCINVACTGC), uroguany-
`lin98–109 (CELCINVACTGC), and E. coli ST-(5–17) (CCEL-
`
`CCNPACAGC) were synthesized by the solid-phase method
`on an Applied Biosystems model 431A peptide synthesizer and
`purified by reverse-phase C18 chromatography as previously
`(1–3). The structure and mass of synthetic peptides were
`verified by electrospray mass spectrometry, gas-phase se-
`quence analysis, and amino acid composition analysis.
`Cell Culture. T84 cells (passage 21 obtained from Jim
`McRoberts, Harbor–University of California Los Angeles
`Medical Center, Torrance, CA) were cultured in DMEM and
`Ham’s F-12 medium (1:1) containing 5% fetal bovine serum
`and 60 ␮g of penicillin plus 100 ␮g of streptomycin per ml as
`described (2, 15).
`
`RESULTS
`The relative potencies of the synthetic forms of human urogua-
`nylin and guanylin for stimulation of cGMP accumulation in
`intact T84 intestinal cells were assessed at medium pH values
`of 5.0 and 8.0, which represent the extremes of microclimate
`pH found at the mucosal surface of the intestine. Experiments
`using opossum uroguanylin and guanylin provided additional
`insights into the optimal medium pH values used in this study
`(4). The potency of guanylin for eliciting cGMP accumulation
`responses in T84 cells was 10-fold greater when tested at pH
`8.0 compared with its potency at pH 5.0 (Fig. 1A). In contrast,
`the potency of uroguanylin was reduced by 10-fold at pH 8.0
`compared with its potency at pH 5.0 (Fig. 1B). We previously
`reported that E. coli ST-(5–17) was 2- to 3-fold more potent at
`acidic pH compared with alkaline pH (4). Under acidic
`conditions, uroguanylin was 100-fold more potent than gua-
`nylin. At pH 5.0, 3000 nM guanylin was required to stimulate
`⬇200-fold increases in cellular cGMP, whereas 30 nM urogua-
`nylin elicited this magnitude of cGMP response. However, this
`rank order of potency was reversed at pH 8.0 with guanylin
`becoming 3-fold more potent than uroguanylin. For example,
`30 nM guanylin caused ⬇50-fold increases in cGMP levels,
`whereas 100 nM uroguanylin was required at pH 8.0. Thresh-
`old stimulation of cGMP levels was observed with ⬇0.1 nM
`uroguanylin and ⬇10 nM guanylin at pH 5.0, whereas at pH 8,
`⬇0.3 nM guanylin and ⬇3 nM uroguanylin were required to
`stimulate cGMP increases by at least 2-fold over the basal
`cGMP levels. Thus, variations in mucosal pH similar to those
`that may occur within the intestinal lumen during digestion
`markedly and differentially influenced the cGMP responses of
`T84 cells to guanylin and uroguanylin.
`We further tested the effects of mucosal acidity on the
`relative potencies of guanylin, uroguanylin, and ST-(5–17) for
`stimulation of chloride secretion across monolayers of T84
`cells cultured on permeable filters and mounted in Ussing
`chambers. T84 cells secrete chloride in the serosal to mucosal
`direction and the magnitude of chloride transport can be
`measured as the input current required to maintain a trans-
`epithelial potential difference equal to zero (short circuit
`current, Isc). Guanylin, uroguanylin, and E. coli ST stimulate
`the Isc of T84 cells when added to the mucosal bath of Ussing
`chambers (1–3, 15, 26). In these experiments, the basolateral
`(serosal) surfaces of T84 cells were maintained at pH 7.4,
`whereas the apical (mucosal) surface was maintained at either
`pH 5.5 or pH 7.8. The potency of guanylin for stimulation of
`chloride secretion was markedly increased at an apical pH of
`7.8 compared with pH 5.5 (Fig. 2 Upper). In contrast, urogua-
`nylin was considerably more potent in the stimulation of
`chloride secretion when the apical pH was 5.5 compared with
`its potency at pH 7.8 (Fig. 2 Lower). Fig. 3 compares the
`relative potencies of guanylin, uroguanylin, and ST-(5–17) for
`stimulating chloride secretion at a mucosal pH of 5.5 compared
`with pH 7.8 over a wide range of agonist concentrations.
`ST-(5–17) and uroguanylin act similarly by stimulating greater
`increases in chloride secretion at a mucosal pH of 5.5 com-
`pared with the stimulation observed at pH 7.8 (Fig. 3). The
`
`

`

`Pharmacology: Hamra et al.
`
`Proc. Natl. Acad. Sci. USA 94 (1997)
`
`2707
`
`250 A.
`
`200
`
`• pH 8.0
`150 o pH 5.0
`
`100
`
`50
`
`0
`-10
`
`-9
`
`-8
`
`-7
`
`[Guanylin], Log (M)
`
`E
`
`• pH 7.8
`
`ST (5-17)
`
`2
`
`o pH 5.5
`
`Guanylin
`
`2
`a
`.
`
`E
`c
`o
`
`/
`
`/
`
`2
`c
`a
`c.,
`
`1,
`
`2
`o
`o
`
`ST (5-17)
`2
`
`Uroguanylin
`
`35.
`
`30.
`
`25-
`
`20-
`
`15-
`
`10-
`
`5-
`
`0-
`50
`
`40
`
`30
`
`20
`
`10
`
`10
`
`20
`
`30
`
`40
`
`50
`
`60
`
`0
`
`80
`
`Time, (min)
`
`FIG. 2. Effects of mucosal pH on the stimulation of Isc in T84 cells
`by guanylin and uroguanylin. T84 cells were cultured on collagen-
`coated membranes and mounted in Ussing chambers as described. At
`the arrows, the indicated concentrations of guanylin (Upper), urogua-
`nylin (Lower), and E. coli ST-(5–17) were added to the apical reservoirs
`containing medium at either pH 5.5 (E) or pH 7.8 (F). For all
`experiments, the basolateral reservoir contained medium at pH 7.4.
`Data are representative experiments of at least five separate experi-
`ments performed with each peptide at the mucosal pH values of 5.5
`and pH 7.8. Concentrations of guanylin and uroguanylin greater than
`100 nM were not used in these experiments because they consumed
`excessive amounts of the peptides.
`
`affinity for binding to the low affinity site when tested at pH
`8.0 compared with pH 5.0 (Fig. 4 Top). In contrast, parallel 9-
`to 10-fold increases in affinities for the high and low affinity
`sites were observed when uroguanylin was tested at pH 5.0
`versus pH 8.0, indicating that mucosal acidity increased the
`affinity of uroguanylin for binding to receptors (Fig. 4 Middle).
`In agreement with a previous report (27), mucosal acidity had
`little influence on the binding affinities for ST-(5–17) inter-
`action with the receptors on T84 cells (Fig. 4 Bottom). The
`remarkable effects of mucosal pH on cGMP accumulation and
`chloride secretion responses to these peptides may be ex-
`plained by pH-dependent shifts in the affinities of uroguanylin
`and guanylin for binding to guanylate cyclase effector mole-
`cules on the apical surface of this model intestinal epithelium.
`The unique acidic residues at the N terminus of uroguanylin
`were postulated to be involved in the increased potencies and
`binding affinities for uroguanylin in the interaction of this
`peptide with T84 cell receptors at acidic versus alkaline pH. A
`truncated form of opossum uroguanylin95–109 was synthesized
`without the N-terminal Gln95–Glu96–Asp97 amino acids to test
`the hypothesis that conformational changes in these residues
`within the uroguanylin95–109 molecule contribute to the in-
`
`pmoles cGMP per well
`
`pmoles cGMP per well
`
`250 B.
`
`200
`• pH 8.0
`150 o pH 5.0
`
`100
`
`50
`
`0
`-10
`
`-9
`
`-8
`
`-7
`
`-6
`
`[Uroguanylin], Log (M)
`
`FIG. 1. Comparison of cGMP accumulation responses to the
`synthetic forms of human guanylin and uroguanylin in T84 cells under
`acidic and alkaline conditions. T84 cells were treated with the indi-
`cated concentrations of guanylin (A) or uroguanylin (B) for 40 min at
`a mucosal pH of 5.0 (E) or pH 8.0 (F). Data are the mean of duplicate
`experiments performed with each agonist at each pH value. Concen-
`trations of guanylin and uroguanylin greater than 10 ␮M and 1 ␮M,
`respectively, were not used because it consumed excessive amounts of
`the peptides.
`
`rank order of potencies for agonist-mediated stimulation of
`chloride secretion was ST ⬎ uroguanylin ⬎ guanylin at acidic
`pH and ST ⬎ guanylin ⬎ uroguanylin at an alkaline pH (Fig.
`3). The relative potencies of uroguanylin, guanylin, and ST-
`(5–17) in the stimulation of transepithelial chloride secretion
`across monolayers of T84 cells at acidic versus alkaline pH
`matched their relative potencies for stimulation of cGMP
`levels under these conditions.
`Modulation by mucosal acidity of the relative affinities of
`uroguanylin and guanylin for binding to a common set of
`receptors (1, 2, 15) could account for the effects of medium pH
`on the cGMP and chloride secretion responses elicited by these
`peptides in T84 cells. This hypothesis was tested using com-
`petitive radioligand-binding assays in cultured T84 cells with
`125I-ST-(1–19) as the radioligand (2, 15). Uroguanylin, guany-
`lin, and ST-(5–17) fully inhibited the binding of 125I-ST-(1–19)
`to apical receptors on T84 cells when tested at medium pH
`values of 5.0 and 8.0. Examination of the radioligand-binding
`data using computer-assisted fitting of curves to a two-site
`model (15) confirms that T84 cells have both high- and
`low-affinity binding sites for each ligand. The Ki values for each
`peptide at medium pH values of 5.0 versus 8.0 are found in the
`legend to Fig. 4. Guanylin had a 100-fold increase in affinity
`for binding to the high affinity site and a 30-fold increase in
`
`

`

`2708
`
`Pharmacology: Hamra et al.
`
`Proc. Natl. Acad. Sci. USA 94 (1997)
`
`O pH 5.0
`100
`
`pH 8.0
`
`75
`
`50
`
`25
`
`Guanylin
`
`►
`
`0
`-12 -11 - 0 -9
`
`-8
`
`-7
`
`-6
`
`-5
`
`100
`
`75
`
`50
`
`25
`
`Uroguanylin
`
`0
`-12 -11 -10 -9
`
`100
`
`75
`
`50
`
`25
`
`ST (5-17)
`
`0
`-12 -11 -10 -9
`
`-8
`
`-7 • -6
`
`-5
`
`-8
`
`-7
`
`-6
`
`-5
`
`o
`m
`
`•
`
`O
`
`-
`
`(N1
`
`pH 5.5
`0 ST (5-17)
`Uroguanylin
`A Guanylin
`
`pH 7.8
`• ST (5-17)
`• Uroguanylin
`Guanylin
`
`60-
`
`50-
`
`40-
`
`30-
`
`20-
`
`10.
`
`0-
`
`30
`
`25
`
`20
`
`15
`
`10
`
`5
`
`A Isc, µA / cm2
`
`20
`
`40
`
`1 0 1 0 1 0
`60
`80
`[Peptide], nM
`
`FIG. 3. Comparison of the relative potencies of uroguanylin,
`guanylin, and E. coli ST-(5–17) in the stimulation of Isc in T84 cells at
`the mucosal pH values of 5.5 (Upper) and pH 7.8 (Lower). Peptides
`were added to the apical reservior at 10–15 min intervals starting with
`the lowest concentrations shown, followed by successive additions of
`the peptide. The basolateral reservior was maintained at pH 7.4. Data
`are the mean of five experiments performed with each agonist.
`Horizontal bars indicate the SEM for each point.
`
`creased potency of uroguanylin at acidic versus alkaline pH (4,
`17). The truncated peptide, uroguanylin98–109, stimulated
`cGMP accumulation in T84 cells similarly at pH 5.0 compared
`with pH 8.0 (Fig. 5). The potency of this peptide was actually
`increased at alkaline pH, thus demonstrating that the trun-
`cated uroguanylin98–109 peptide possessed this characteristic
`pharmacological property found with the guanylin peptides
`(Fig. 1, ref. 4). The 15 amino acid form of opossum urogua-
`nylin95–109, containing the N-terminal gln95 and the two acidic
`residues, glu96 and asp97, was substantially more potent at pH
`5.0 compared with pH 8.0. The opossum uroguanylin95–109
`peptide and human uroguanylin share this pH dependency for
`agonist potency (Fig. 1). We previously reported similar effects
`of acid pH on the potency of opossum uroguanylin96–109 that
`did not have the N-terminal gln95 residue, but retained the two
`acidic amino acids (4). The presence or absence of the
`N-terminal glutamine in the opossum form of uroguanylin did
`not influence the characteristic enhancement of agonist po-
`tency elicited by the mucosal pH of 5.0.
`In competitive radioligand-binding experiments, uroguany-
`lin98–109 bound to cell–surface receptors on T84 cells with
`similar affinities at pH 5.0 and pH 8.0 (Fig. 5). In this
`radioligand-binding assay, the Ki values for uroguanylin98–109
`binding to the high affinity site were 0.14 nM at pH 8.0
`compared with 0.19 nM at pH 5.0 and the Ki values for
`uroguanylin98–109 interaction with the low affinity site were 345
`nM at pH 8.0 versus 404 nM at pH 5.0. Thus, uroguanylin98–109
`did not exhibit an increase in the affinity of this peptide for
`
`[Peptide], Log (M)
`
`FIG. 4. Effects of medium pH on the relative affinities of guanylin,
`uroguanylin, and E. coli ST-(5–17) for binding to receptors on T84
`cells. Binding of 125I-ST-(1–19) to intact T84 cells was determined in
`the presence of the indicated concentrations of guanylin (Top),
`uroguanylin (Middle), and ST-(5–17) (Bottom) as described. The values
`shown are the composite data (mean ⫾SEM) from three experiments
`performed in duplicate with each peptide at pH 5 and pH 8 and are
`expressed as the total binding of 125I-ST-(1–19) in the absence of a
`competing ligand. Nonspecific binding was measured using 1 ␮M
`ST-(5–17). Competitive radioligand binding curves are computer-
`derived best fits of the binding data to a two-site model (15). Ki values
`obtained for the high and low affinity sites were: guanylin, pH 5 ⬇102
`nM and 2.3 ␮M, pH 8 ⬇1 nM and 77 nM; uroguanylin, pH 5 ⬇1 nM
`and 70 nM, pH 8 ⬇10 nM and 615 nM; ST-(5–17), pH 5 ⬇94 pM and
`7 nM, pH 8 ⬇440 pM and 17 nM.
`
`binding to receptors on T84 cells at acidic versus alkaline pH.
`This observation is consistent with the similar potencies mea-
`sured at pH 5.0 compared with pH 8.0 for the cGMP accu-
`mulation response to uroguanylin98–109. We conclude that the
`unique acidic amino acids at the N terminus of uroguanylin are
`required for the increased binding affinities, and accordingly,
`the enhanced potencies of uroguanylin in the stimulation of
`target cell responses under the acidic conditions of pH 5.0–5.5
`maintained at the mucosal surface of T84 cells in this model
`epithelium.
`
`DISCUSSION
`At the surface of the intestinal mucosa, between the apical
`plasma membranes of enterocytes and a protective layer of
`hydrated mucin, the ligand-binding domains of a common set
`of receptors for uroguanylin and guanylin extend into an
`aqueous (microclimate) zone that has a variable pH (6, 10–13).
`It is in this microdomain of changing mucosal acidity where the
`luminally secreted agonists uroguanylin and guanylin (28) bind
`to and activate the intestinal receptor–guanylate cyclase sig-
`
`

`

`Pharmacology: Hamra et al.
`
`Proc. Natl. Acad. Sci. USA 94 (1997)
`
`2709
`
`receptors undergo pH-dependent shifts by as much as 100-fold.
`Thus, variation in mucosal acidity within the physiological
`limits observed at the surface of the intestinal mucosa (21–24)
`influences the activation of receptors by uroguanylin and
`guanylin.
`Modulation of receptor–guanylate cyclase activity and chlo-
`ride secretion by uroguanylin would be most effective in
`regions of the intestine where the luminal microclimate do-
`main is acidic, whereas the actions of guanylin would be
`augmented when alkaline pH occurs at the mucosal surface.
`Acidic conditions occur intraluminally in the proximal small
`intestine and proximal colon during digestion (21–24). Gastric
`emptying introduces into the lumen of the duodenum a highly
`acidic chyme (22), thus increasing mucosal acidity and poten-
`tially enhancing the cGMP accumulation and chloride secre-
`tion responses of the intestine to uroguanylin, while rendering
`guanylin ineffective. The observation that uroguanylin mRNA
`is relatively abundant in the opossum duodenum compared
`with guanylin mRNA is consistent with a physiological role for
`uroguanylin in the proximal small intestine (17). Guanylin
`mRNA levels are also lower in the duodenum of other
`mammals compared with the mRNA levels in the ileum and
`colon (18). Luminal pH within the colon can become acidic
`due to the production of short chain fatty acids by enteric
`microorganisms (23, 24). High levels of uroguanylin and
`guanylin and their mRNAs are expressed in the mucosa lining
`the cecum and colon (4, 5, 11, 17, 18). Increased acidity in the
`lumen of the large intestine due to microbial metabolism could
`lower the microclimate pH, thus increasing the affinity of
`uroguanylin and reducing the affinity of guanylin for binding
`to and activation of receptors. In addition, alkalinization of the
`microclimate domain at the mucosal surface is achieved
`through bicarbonate secretion from the pancreas into the
`duodenum, and兾or from epithelial cells lining the small and
`large intestine (22, 24). Alkalinization would enhance the
`potency of guanylin, while attenuating responses to urogua-
`nylin. The intraluminal secretion of guanylin and uroguanylin
`provides an intrinsic mechanism for control of salt and water
`transport under the variable acidity conditions occurring in the
`microclimate domain that bathes the mucosal surface of
`enterocytes lining the intestinal tract (5, 11, 17–20).
`A striking difference in the primary structure of uroguanylin
`compared with guanylin is the appearance of two acidic amino
`acids at the N terminus of uroguanylin (Fig. 6). All uroguanylin
`peptides have aspartate or glutamate residues at these posi-
`tions (8, 9). Deletion of the N-terminal residues (Gln95–Glu96–
`Asp97) of opossum uroguanylin95–109 converted the truncated
`uroguanylin98–109 into a uroguanylin analogue that possessed
`the pharmacological property that is characteristically ob-
`served in the guanylin subfamily of peptide agonists. The
`truncated uroguanylin98–109 was actually somewhat more po-
`tent at pH 8.0 than at pH 5.0. We conclude that the N-terminal
`acidic residues of uroguanylin are required for the increased
`binding affinities, and therefore, the enhanced potency of
`uroguanylin for activation of receptors under acidic conditions.
`It is likely that acidic conditions influence the ionization
`and兾or conformational state of the uroguanylin molecule as a
`molecular mechanism for the increased biological activity of
`uroguanylin in this circumstance. Presently, we have no infor-
`
`FIG. 5. Effects of uroguanylin95–109 and uroguanylin98–109 on
`cGMP accumulation and the affinities of uroguanylin98–109 for binding
`to receptors on T84 cells at pH 5.0 versus pH 8.0. The data are
`representative experiments with duplicate assays. Each experiment
`was performed at least three times with similar results. The conditions
`are the same as those given in the Materials and Methods and the
`legends to Figs. 1 and 4. (Top) Stimulation of T84 cell cGMP
`accumulation by opossum uroguanylin95–109. (Middle) Stimulation of
`cGMP accumulation by opossum uroguanylin98–109. (Bottom) Inhibi-
`tion of 125I-ST binding to receptors on T84 cells by opossum urogua-
`nylin98–109.
`
`naling molecules. Using a model intestinal epithelium, we
`demonstrated that potential changes in mucosal acidity can
`differentially influence the relative potencies of uroguanylin
`and guanylin for activation of these receptors located on the
`apical surface of intestinal cells (1, 2, 15, 26). Mucosal acidity
`markedly increases the potency of uroguanylin, while render-
`ing guanylin ineffective in the stimulation of cGMP accumu-
`lation and transepithelial chloride secretion. In sharp contrast,
`a mucosal pH of 8.0 substantially increases the potency of
`guanylin, while diminishing the potency of uroguanylin. This
`striking effect of mucosal pH on agonist potencies was ex-
`plained by the corresponding shifts in affinities of guanylin and
`uroguanylin for binding to receptors on T84 cells at the
`mucosal pH values of 5.0 versus pH 8.0. As a result, the
`affinities of guanylin and uroguanylin for binding to these
`
`Uroguanylin
`
`N D D
`
`C E
`
`C
`
`V N V
`
`A C
`
`T G C
`
`L
`
`Guanylin
`
`G T
`
`C E
`
`C
`
`A Y A
`
`A C
`
`T G C
`
`E. coil ST(5-17)
`
`C
`
`C
`
`E L C
`
`C N
`
`P A
`
`C
`
`A G C
`
`FIG. 6. Comparison of the primary structures of uroguanylin, guanylin, and E. coli ST. Underlined amino acids indicate the identical residues
`shared between uroguanylin, guanylin, and ST-(5–17). Shaded boxes highlight the differences in primary structure between uroguanylin and
`guanylin.
`
`400
`
`300
`
`200
`
`100
`
`• pH 8.0
`O pH 5.0
`
`-10
`
`-9
`-8
`-7
`-6
`-5
`RIroguanylinss-los Log (M)
`
`-4
`
`• pH 8.0
`O pH 5.0
`
`-10
`-9
`-8
`-7
`-6
`-5
`[Uroguanylinsa-los], Log (M)
`
`0
`
`•
`
`•
`
`0
`•
`
`• pH 8.0
`O pH 5.0
`
`-10
`
`-9
`
`-8
`
`-7
`
`-6
`
`-5
`
`-
`
`[Uroguanylin98-109] , Log (M)
`
`pmoles cGMP per well
`
`pmoles cGMP per well
`
`300
`
`200
`
`100
`
`0
`-11
`
`100
`
`7
`
`50
`
`25
`
`O
`CO
`"5
`
`•
`OO
`03
`-
`(1)
`
`0
`-11
`
`

`

`2710
`
`Pharmacology: Hamra et al.
`
`Proc. Natl. Acad. Sci. USA 94 (1997)
`
`mation concerning the residues in guanylin that contribute to
`the remarkable increase in affinities for interaction with
`receptors at alkaline compared with acidic pH. Although these
`properties of uroguanylin and guanylin were discovered using
`the T84 cell line as a model epithelium, current studies have
`demonstrated that the short circuit current responses of mouse
`duodenum and cecum to uroguanylin in vitro are markedly
`increased when the luminal pH is acidic compared with the
`responses observed at a mucosal pH of 7.4 (29). This prelim-
`inary observation with uroguanylin using a native epithelium
`mounted into Ussing chambers is consistent with the increased
`affinity of uroguanylin for binding to and subsequent activa-
`tion of the apical membrane receptor–guanylate cyclases of
`T84 cells revealed by experiments presented in this commu-
`nication.
`E. coli ST-(5–17) binds with extraordinarily high affinities to
`the uroguanylin兾guanylin receptors on the apical surface of
`T84 cells and potently stimulates cGMP production and chlo-
`ride secretion at both alkaline and acidic pH. The interactions
`of ST peptides with these receptors is little affected by mucosal
`pH in this model epithelium. Enteric bacteria have evolved a
`single peptide toxin that serves as a molecular mimic for both
`of the intestinal hormones, uroguanylin and guanylin. The
`remarkable potencies of ST peptides compared with the
`potencies of the enteric hormones is caused by higher affinities
`for ST binding to the intestinal receptors for uroguanylin and
`guanylin. Bacteria have created superagonist peptide toxins
`and this pharmacological property contributes to the remark-
`able toxicities of ST peptides in the molecular and cellular
`mechanism underlying Travelers diarrhea (8, 9, 30, 31).
`We conclude that uroguanylin and guanylin cooperatively
`regulate a signaling pathway that modulates intestinal salt and
`water transport via an intrinsic, paracrine mechanism involving
`cGMP as a second messenger. Uroguanylin is a highly pot