Opossum colonic mucosa contains uroguanylin
`and guanylin peptides
`
`F. KENT HAMRA, WILLIAM J. KRAUSE, SAMMY L. EBER, RONALD H. FREEMAN,
`CHRISTINE E. SMITH, MARK G. CURRIE, AND LEONARD R. FORTE
`The Truman Veterans Affairs Medical Center and Departments of Pharmacology, Anatomy,
`and Physiology, School of Medicine, Missouri University, Columbia 65212;
`and Searle Research and Development, St. Louis, Missouri 63167
`
`Hamra, F. Kent, William J. Krause, Sammy L. Eber,
`Ronald H. Freeman, Christine E. Smith, Mark G. Cur-
`rie, and Leonard R. Forte. Opossum colonic mucosa con-
`tains uroguanylin and guanylin peptides. Am. J. Physiol. 270
`(Gastrointest. Liver Physiol. 33): G708—G716, 1996.—Urogua-
`nylin and guanylin are structurally related peptides that
`activate an intestinal form of membrane guanylate cyclase
`(GC-C). Guanylin was isolated from the intestine, but urogua-
`nylin was isolated from urine, thus a tissue source for
`uroguanylin was sought. In these experiments, uroguanylin
`and guanylin were separated and purified independently
`from colonic mucosa and urine of opossums. Colonic, urinary,
`and synthetic forms of uroguanylin had an isoelectric point of
`--3.0, eluted from C18 reverse-phase high-performance liquid
`chromatography (RP-HPLC) columns at 8-9% acetonitrile,
`elicited greater guanosine 3',5'-cyclic monophosphate (cGMP)
`responses in T84 cells at pH 5.5 than pH 8, and were not
`cleaved and inactivated by pretreatment with chymotrypsin.
`In contrast, colonic, urinary, and synthetic guanylin had an
`isoelectric point of -6.0, eluted at 15-16% acetonitrile on C18
`RP-HPLC columns, stimulated greater cGMP responses in
`T84 cells at pH 8 than pH 5.5, and were inactivated by
`chymotrypsin, which hydrolyzed the Phe-Ala or Tyr-Ala
`bonds within guanylin. Uroguanylin joins guanylin as an
`intestinal peptide that may participate in an intrinsic path-
`way for cGMP-mediated regulation of intestinal salt and
`water transport. Moreover, uroguanylin and guanylin in
`urine may be derived from the intestinal mucosa, thus
`implicating these peptides in an endocrine mechanism link-
`ing the intestine with the kidney.
`guanylate cyclase; TS4 cells; guanosine 3',5'-cyclic monophos-
`phate; urine
`
`UROGUANYLIN AND GUANYLIN are members of an emerg-
`ing family of peptides that function as the physiological
`ligands for an intestinal isoform of membrane guanyl-
`ate cyclase (GC-C) (7, 18; see Ref. 13 for review).
`Uroguanylin and guanylin stimulate GC-C activity,
`resulting in elevations of cellular guanosine 3',5'-cyclic
`monophosphate (cGMP) (7, 18). All species of mammals
`and birds examined express GC-C-like receptor activity
`on the apical surface of enterocytes throughout the
`intestine (21, 22). The opossum kidney also expresses
`high levels of GC-C-like receptors located in the apical
`membranes of proximal tubular cells (14). Guanylin
`was first isolated from rat jejunum as a heat-stable,
`15-amino acid peptide that activated GC-C in human
`intestinal T84 cells (7). Guanylin cDNAs encoding 115-
`to 116-amino acid precursors have been isolated from
`rat, human, and mouse intestine (19). Uroguanylin was
`initially purified as 13- to 15-amino acid peptides from
`
`opossum urine (18) and was named on the basis of its
`structural and functional similarities to guanylin and
`its isolation from urine. Uroguanylin was confirmed as a
`second member in the guanylin peptide family by the
`subsequent isolation of opossum guanylin (18) and the
`human and rat forms of uroguanylin (10, 20).
`Before the discovery of guanylin and uroguanylin,
`the only peptide agonists for GC-C were the diarrhea-
`producing, heat-stable enterotoxins (STs) (11, 34). STs
`are produced by different strains of pathogenic, enteric
`microorganisms, including Escherichia coli (25). STs
`act as molecular mimics of guanylin and uroguanylin
`by binding to GC-C and stimulating dramatic increases
`in cellular cGMP accumulation (11, 13, 31). ST-
`stimulated cGMP production decreases sodium absorp-
`tion and increases chloride secretion by enterocytes,
`which results in secretory, or "traveler's," diarrhea (11,
`25, 31). Similar to STs, uroguanylin and guanylin
`stimulate transepithelial chloride secretion from T84
`cells and intestinal tissues mounted in Ussing cham-
`bers by increasing cellular cGMP production (7, 18, 20).
`Thus previous studies of the mechanisms by which STs
`exert their pathological effects may provide insights
`into the physiological roles of uroguanylin and gua-
`nylin.
`Guanylin mRNA is found throughout the intestine,
`with the highest levels of expression in the ileum and
`colon (20). Guanylin and/or its mRNA has been re-
`ported to occur in a heterogeneous population of intesti-
`nal cell types, including absorptive enterocytes (27),
`Paneth cells (9), enterochromaffin cells (4), and goblet
`cells (6). Northern analysis has also demonstrated
`lower levels of guanylin mRNA in the kidney, adrenal
`gland, and the uterus/oviduct (33). Moreover, progua-
`nylin has been shown to circulate in the plasma of
`humans, demonstrating that guanylin may potentially
`regulate GC-C or other target receptors via an endo-
`crine mechanism (23, 24, 30). Presently, uroguanylin
`has only been isolated from urine, and little is known
`regarding the tissues that may produce this peptide
`(10, 18, 20). The high levels of uroguanylin found in the
`urine of opossums, humans, and rats could be derived
`from the kidney and/or from other tissues via filtration
`of uroguanylin from the circulation. In the current
`study, we have isolated uroguanylin and guanylin
`peptides from the colonic mucosa of opossums. Several
`independent analytical techniques were used to iden-
`tify the bioactive peptides. Thus intestinal mucosa is a
`potential tissue source for uroguanylin and guanylin
`found in urine (10, 18, 20).
`
`G708
`
`0193-1857/96 $5.00 Copyright © 1996 the American Physiological Society
`
`MYLAN EXHIBIT - 1019 (Corrected)
`Mylan Pharmaceuticals, Inc. v. Bausch Health Ireland, Ltd.
`
`

`

`UROGUANYLIN: AN INTESTINAL PEPTIDE
`
`G709
`
`MATERIALS AND METHODS
`
`Purification of colon peptides. Full-length colons, including
`cecums, were removed from adult opossums (Didelphis virgin-
`iana), and the mucosae (150 g wet wt) were scraped from
`colonic muscle with the use of a glass microscope slide. Only
`healthy opossums with hard stools were used in these stud-
`ies. The mucosae were divided into two batches, which were
`processed separately (75 g wet wt/batch). Each batch was
`suspended in 10 vol of 1 M acetic acid, heated at 100°C for 10
`min with constant stirring, homogenized, and stored at
`—20°C. The homogenate was thawed and centrifuged at
`10,000 g for 20 min, and the supernatant was made to 0.1%
`trifluoroacetic acid (TFA). The supernatants were processed
`with Waters Sep-Pak cartridges [octadecylsilane cartridges
`(C18)] as described previously (7, 18). Eluted fractions of the
`colon extract from C18 cartridges were dried in a Speed-Vac,
`resuspended in 10 ml of 25 mM ammonium acetate, pH 5.0,
`and centrifuged at 500 g for 10 min. From each batch, 8.0 ml
`of supernatant were applied to a 2.5 X 90 cm Sephadex G-25
`column, and 10-ml fractions were collected for two successive
`runs. Fractions from this step and subsequent purification
`steps were assayed for bioactivity as described in cGMP
`accumulation assay in T84 cells. Bioactive fractions from
`each batch were combined after the Sephadex G-25 step.
`Final purification by isoelectric focusing on a Rotofor appara-
`tus (Bio-Rad) and reverse-phase high-performance liquid
`chromatography (RP-HPLC) of individual uroguanylin-like
`or guanylin-like bioactive peptides was achieved using the
`same purification scheme as described previously for isolation
`of intestinal peptides from the opossum and rat (7, 18). An
`ampholyte range of pH 3.0 to 10.0 (Bio-Rad) was used for
`isoelectric focusing.
`Purification of urine peptides. Uroguanylin and guanylin
`peptides were extracted from 4.5 liters of opossum urine and
`purified as previously described, except that an isoelectric
`focusing step was added after the semipreparative HPLC step
`(18). Active fractions were identified using the bioassay as
`described below. Separate columns and Rotofor apparatus
`materials were used during purification of urine peptides
`than were used for purification of the colon peptides.
`Cell culture. T84 cells (passage 21 obtained from Jim
`McRoberts, Harbor-University of California Los Angeles Medi-
`cal Center, Torrance, CA) were cultured in Dulbecco's modi-
`fied Eagle's medium (DMEM)-Ham's F-12 medium (1:1) con-
`taining 5% fetal bovine serum, 60 pg penicillin, and 100 pg
`streptomycin per milliliter as previously described (18).
`cGMP accumulation assay in T84 cells. T84 cells were
`cultured in 24-well plastic dishes, and the cGMP levels were
`measured in control and agonist-stimulated cells by radioim-
`munoassay (18). Aliquots of column fractions and vehicle
`were suspended in 200 pl of each of two assay buffers: pH 8.0
`buffer, consisting of DMEM, 20 mM N-2-hydroxyethylpipera-
`zine-N'-2-ethanesulfonic acid (HEPES), 50 mM sodium bicar-
`bonate, pH 8.0, and 1 mM 3-isobutyl-1-methylxanthine
`(IBMX); and pH 5.5 buffer, consisting of DMEM, 20 mM
`2-(N-morpholino)ethanesulfonic acid (MES), pH 5.5, and 1
`mM IBMX. Ammonium acetate and TFA were removed from
`test samples by drying fraction aliquots in a Speed-Vac before
`suspension in assay buffers. This was done to avoid changes
`in pH caused by the column reagents. T84 cells were washed
`twice with 200 pl of the respective pH 8.0 and pH 5.5 buffers
`before addition of reagents. These solutions of medium plus
`bioactive peptides were then added to T84 cells and incubated
`at 37°C for 40 min. After incubation, the reaction medium was
`aspirated, and 200 pl 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 potassium hydroxide and centrifuged,
`and 50 pl of the extract was used to measure cGMP. For pH
`titration studies, DMEM, 20 mM HEPES, 1 mM IBMX was
`adjusted to pH 7.0, 7.5, 8.0, and 8.5 with NaOH, and DMEM,
`20 mM MES, 1 mM IBMX was adjusted to pH 5.0, 5.5, 6.0,
`and 6.5 with NaOH.
`Peptide-agonist concentration-response curves were ana-
`lyzed with the computer program Prism (GraphPad Software,
`San Diego, CA). The concentrations at which peptide agonists
`elicited 50% of the maximal cGMP accumulation response
`(EC50) were obtained by nonlinear regression of agonist-
`stimulated cGMP accumulation data.
`Synthesis of uroguanylin, guanylin, and ST peptides. Syn-
`thetic uroguanylin-(2-15), EDCELCINVACTGC, and syn-
`thetic guanylin-(1-15), SHTCEICAFAACAGC, were synthe-
`sized by the solid-phase method with an Applied Biosystems
`431A peptide synthesizer. N-(9-fluorenylmethoxycarbonyl)
`(FMOC)-protected amino acids activated with 2-(1H-benzo-
`triazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate
`were added to FMOC-Cys-(trityl)-Wang resin (Nova Bio-
`chem). Coupling efficiencies were monitored by the ultravio-
`let absorbance of the released FMOC groups. The peptides
`were cleaved from the resin, and the side chains were
`deprotected, except for the acetamidomethyl groups on Cys3
`and Cysil, by incubation in TFA, ethanedithiol, and water
`(95:2.5:2.5, vol/vol) for 2 h at room temperature. The peptides
`were cyclized with the use of air oxidation. The acetamido-
`methyl groups on Cys3 and Cysll were removed with iodine.
`The peptides were desalted with a 12-ml Whatman ODS-3
`solid-phase extraction device and purified to a single peak by
`C18 RP-HPLC (acetonitrile-ammonium acetate). The se-
`quences of uroguanylin and guanylin were confirmed by
`protein sequencing on an Applied Biosystems 470A gas-phase
`protein sequencer. Amino acid composition analysis was
`performed to estimate peptide mass, after RP-HPLC purifica-
`tion.
`E. coli ST-(5-17), CCELCCNPACAGC, was prepared as
`previously described by the solid-phase method with an
`Applied Biosystems 430A peptide synthesizer on Cys-(4-
`CH3Bz1)-OCH2-Pam resin, using double coupling cycles to
`ensure complete coupling at each step (6). The peptide was
`cyclized with the use of dimethyl sulfoxide, and its structure
`was verified by electrospray mass spectrometry, gas-phase
`sequence analysis, and amino acid composition analysis (7).
`Synthesis, purification, and composition analysis of rat
`guanylin-(101-115), rat guanylin-(93-115), and human uro-
`guanylin were performed by our previously described meth-
`ods (7, 20).
`Chymotrypsin digestion of peptides. Vehicle, synthetic pep-
`tides, or purified peptides were separated into aliquots in
`Microfuge tubes and dried in a Speed-Vac. Digestion reac-
`tions were started by resuspending the dried aliquots in 100
`pl of 10 mM HEPES, pH 8.0, containing 0.15 U (3 pg solid) of
`bovine pancreatic a-chymotrypsin (Sigma, St. Louis, MO),
`either with or without 100 pM chymostatin (Sigma). Reac-
`tions were incubated at 34°C in a water bath for 1 h. After 1 h,
`the reaction tubes were placed on ice, 100 pM chymostatin
`was added to the reactions that had been incubated in the
`absence of chymostatin, and all the tubes were frozen at
`—80°C. After freezing, the reaction mixtures were dried in a
`Speed-Vac and then resuspended in 200 pl of DMEM contain-
`ing 50 mM sodium bicarbonate (pH 8.0) and 1 mM IBMX for
`analysis in the T84 cell cGMP accumulation bioassay. The
`same experiments were repeated and confirmed with sequenc-
`ing-grade bovine pancreatic chymotrypsin from a second
`source (Boehringer-Mannheim, Indianapolis, IN).
`
`

`

`G710
`
`UROGUANYLIN: AN INTESTINAL PEPTIDE
`
`pmol cGMP per well
`
`For NH2-terminal sequence analysis of digested peptides,
`400 pmol of synthetic opossum uroguanylin, synthetic opos-
`sum guanylin, and synthetic rat guanylin-(101-115) were
`incubated for 10 h at 34°C in a Microfuge tube containing 50
`µl of 10 mM HEPES, pH 8.0, with either 20 or 200 pmol of
`bovine pancreatic a-chymotrypsin (sequencing grade, Boeh-
`ringer-Mannheim).
`
`RESULTS
`
`Selective bioassay for uroguanylin and guanylin pep-
`tides. During the initial isolation of uroguanylin and
`guanylin peptides, we observed that a reduction in
`medium pH reduced the cGMP response elicited by
`guanylin in T84 cells. In contrast, uroguanylin was less
`sensitive to changes in pH. To further investigate these
`observations, the effects of medium pH on the cGMP
`responses elicited by 30 nM of the synthetic forms of
`opossum uroguanylin and guanylin in T84 cells were
`examined (Fig. 1). Uroguanylin caused a greater in-
`crease in cellular cGMP levels when assayed at pH 5
`compared with pH 8. In contrast, guanylin caused only
`a doubling in cGMP accumulation above basal levels at
`pH 5, with the cGMP response increasing to 13-fold at
`pH 8 (Fig. 1). Results obtained with 3 nM E. coli
`ST-(5-17) were similar to those obtained with 30 nM
`uroguanylin (data not shown). Synthetic rat guanylin,
`like opossum guanylin, elicited greater cGMP re-
`sponses in T84 cells at pH 8 than at pH 5.5 (Fig. 2).
`Moreover, the synthetic form of human uroguanylin
`was like opossum uroguanylin and E. coli ST-(5-17) in
`that each peptide stimulated greater levels of cGMP
`accumulation in T84 cells at acidic relative to alkaline
`pH (Fig. 2). These experiments defined pH conditions
`that were used to estimate the potencies of these
`
`120 —
`
`a 100 -
`'a
`co c
`.5 a
`cr
`vTo
`< E 74
`a.
`•
`
`80 -
`
`60
`
`7,,
`
`40
`
`20
`
`0
`
`5.0
`
`5.5 6.0
`
`7.0
`
`7.5 8.0
`
`6.5
`pH
`Fig. 1. Effects of medium pH on uroguanylin (O) and guanylin
`(1)-stimulated guanosine 3' ,5' -cyclic monophosphate (cGMP) accumu-
`lation in T84 cells. Vehicle, 30 nM synthetic opossum uroguanylin,
`and 30 nM synthetic opossum guanylin were suspended in buffered
`assay medium previously adjusted to pH values indicated, as de-
`scribed in MATERIALS AND METHODS. Levels of T84 cell cGMP accumula-
`tion (pmol/well, average of 3 wells) elicited by vehicle and peptides in
`this experiment when tested at pH 5.0 and pH 8.0, respectively, were
`as follows: basal (vehicle control) = 0.45 and 0.78, uroguanylin = 43.9
`and 17.5, and guanylin = 0.85 and 10.0. Data are representative of 4
`experiments with similar results.
`
`80
`
`70 -
`
`60 -
`
`50 -
`
`40 -
`
`30 -
`
`T
`
`20 -
`
`10 - 1
`
`Basal
`
`ST
`
`I-
`opUGn huUGn
`
`rGn
`
`opGn
`
`Fig. 2. Agonist-stimulated cGMP accumulation in T84 cells at pH 5.5
`(open bars) and pH 8.0 (solid bars). Peptides and vehicle were
`suspended in HEPES and Dulbecco's modified Eagle's medium
`(DMEM) containing 50 mM sodium bicarbonate (pH 8.0), or 2-(N-
`morpholino)ethanesulfonic acid (MES) and DMEM at pH 5.5 (pH 5.5)
`for analysis in the T84 cell cGMP accumulation bioassay. Basal,
`vehicle control; ST, synthetic E. coli ST-(5-17); opGn, synthetic
`opossum guanylin; opUGn, synthetic opossum uroguanylin; huUGn,
`synthetic human uroguanylin; rGn, synthetic rat guanylin-(101-
`115). All peptides were tested at 30 nM except for E. coli ST-(5-17),
`which was tested at 3 nM. Error bars indicate standard error of the
`mean for 3 experiments.
`
`peptide agonists. Uroguanylin was more potent at pH
`5.5 (EC50, 200 i 50 nM, n = 5) than at pH 8 (EC50,
`1,900 ± 100 nM, n = 5) (Fig. 3). In contrast, guanylin
`was less potent at pH 5.5 (EC50, 10.6 ± 4.2 pM, n = 3)
`than at pH 8 (EC50, 0.68 ± 0.1 pM, n = 3). ST-(5-17)
`appeared similar in potency at both pH 5.5 (EC50, 40 ±
`10 nM, n = 3) and pH 8 (EC50, 130 ± 60 nM, n = 3).
`Interestingly, the rank order of potency of ST > urogua-
`nylin > guanylin at pH 5.5 changed to ST > guanylin >
`uroguanylin at pH 8 in the T84 cell bioassay. These
`results demonstrate that guanylin-selective and urogua-
`nylin-selective pH conditions can be used to identify
`these peptides.
`Comparison of cGMP responses to endogenous and
`synthetic peptides. To demonstrate that the selective
`bioassay conditions using synthetic uroguanylin and
`guanylin could be extended to endogenous uroguanylin
`and guanylin, we isolated bioactive peptides from opos-
`sum urine, which contains greater levels of urogua-
`nylin than guanylin (10, 18, 20). At each stage of
`purification before the isoelectric focusing step, the
`urine extracts elicited much larger cGMP responses in
`T84 cells at pH 5.5 compared with the cGMP responses
`at pH 8 (data not shown). The bioactivity profile
`obtained using preparative isoelectric focusing of the
`partially purified urine extract revealed a dominant
`peak of bioactivity migrating with a pl of about 3.0
`(peak 1, Fig. 4). This pI is similar to that of the
`uroguanylin peptides previously isolated from urine,
`which contained two or three acidic amino acids and no
`basic residues (18). The bioactive components in peak 1
`also elicited greater cGMP responses in T84 cells when
`assayed at pH 5.5 than at pH 8. A small amount of
`
`

`

`UROGUANYLIN: AN INTESTINAL PEPTIDE
`
`G711
`
`assayed at pH 5.5 revealed two distinct peaks of
`bioactivity that eluted within the internal volume of
`the Sephadex G-25 gel column (Fig. 5). When the
`column fractions were assayed at pH 8, a single peak of
`bioactivity was observed that was coincident with the
`peak 2 that was found when the assay was conducted at
`pH 5.5. The uroguanylin-like peptides in peak 1 elicited
`27-fold increases in cellular cGMP accumulation when
`Qcnyprl at pT-T Fl Fl and nnly fivPfnld inrroaQPQ in etalliiinr
`cGMP at pH 8. In contrast, the guanylin-like peptides
`in peak 2 stimulated greater levels of cGMP at pH 8
`(100-fold) than when assayed at pH 5.5 (28-fold).
`To further characterize the uroguanylin-like pep-
`tides, the components of peak 1 were purified as
`previously described (7, 18). After each purification
`step, a dominant activity peak was observed that
`elicited greater cGMP responses at pH 5.5 than at pH 8.
`In addition, a pI of 3.0 was observed when the bioactive
`components from peak 1 were fractionated by prepara-
`tive isoelectric focusing (data not shown), consistent
`with peak I containing an acidic, uroguanylin-like
`peptide (10, 18, 20). The chromatographic properties of
`peak 1 were characterized further with the use of
`RP-HPLC. The peak fraction of bioactive peptides
`eluting from a C18 RP-HPLC column stimulated a
`23-fold increase in cGMP when assayed at pH 5.5,
`compared with only an eightfold increase in cGMP at
`pH 8 (Fig. 6). Moreover, a characteristic uroguanylin
`elution profile (18) was observed, with the uroguanylin-
`like peptide(s) eluting at 8% acetonitrile from this C18
`RP-HPLC column (Fig. 6). This elution pattern is
`
`12
`
`• •
`
`- 10
`
`-
`
`6
`
`- 4
`
`2
`
`Peak #2
`
`30
`
`25 -
`
`Peak #1
`0
`
`20
`
`15
`
`10 -
`
`5 -
`
`0
`
`pmol cGMP per well
`
`15
`
`18
`
`3
`
`6
`
`12
`9
`Fraction
`Fig. 4. Isoelectric focusing of peptides extracted from opossum urine.
`Peptides were extracted from 4.5 liters of opossum urine using C18
`cartridges and purified as described in MATERIALS AND METHODS.
`Purified peptides were suspended in 50 ml water containing 0.8%
`ampholytes (pH range 3-10, Bio-Rad) and then fractionated on a
`preparative isoelectric focusing cell (Rotofor, Bio-Rad). Fractions
`(0.2% of fraction volume) were next assayed for the ability to elicit a
`cGMP response in T84 cells after suspension in MES, DMEM at pH
`5.5 (0), and in HEPES, DMEM adjusted to pH 8.0 with 50 mM
`sodium bicarbonate (I). Right y-axis represents pH (•) of Rotofor
`fractions obtained after isoelectric focusing of partially purified
`peptides. pH of the Rotofor fraction containing peak of bioactivity, as
`determined by T84 cell cGMP accumulation bioassay, was estimated
`as the isoelectric point (pI).
`
`1000 g A
`
`100 r
`
`10r
`
`1
`
`1000 B
`
`a)
`
`100
`
`10 r
`
`o_
`a.
`2
`
`1000 r c I
`
`100 r
`
`10 r
`
`0
`
`0
`
`ee 2
`
`4
`
`1
`
`0
`
`6 5
`10 9 8 7
`-log Peptide [M]
`Fig. 3. Concentration-response curves for stimulation of cGMP accu-
`mulation in T84 cells by synthetic opossum uroguanylin (A), syn-
`thetic opossum guanylin (B), and synthetic E. coli ST-(5-17) (C).
`Peptides were suspended in MES, DMEM at pH 5.5 (pH 5.5) and in
`HEPES, DMEM adjusted to pH 8.0 with 50 mM sodium bicarbonate
`(pH 8.0). Data are representative of 3-5 experiments with each
`agonist at both pH 5.5 (0) and pH 8.0 (•).
`
`guanylin-like activity was also observed migrating
`with a pI of 5.2 (peak 2, Fig. 4), similar to the pI of
`opossum guanylin, which contains one histidine (18).
`The guanylin-like peptide stimulated a fourfold in-
`crease in cellular cGMP when assayed at pH 8, with no
`detectable increase in cGMP at pH 5.5. These experi-
`ments confirmed that endogenous uroguanylin and
`guanylin peptides respond similarly to the synthetic
`forms of uroguanylin and guanylin when assayed for
`the ability to elicit cGMP responses at acidic vs. alka-
`line pH in T84 cells.
`Isolation of uroguanylin and guanylin from colonic
`mucosa. Preliminary examination of extracts from co-
`lonic mucosa, full-length small intestinal mucosa, kid-
`ney, and plasma of the opossum revealed the presence
`of both uroguanylin-like and guanylin-like peptides in
`each of the extracts (data not shown). Because the
`amounts of extracted uroguanylin-like bioactivity in
`colonic mucosa appeared to be the greatest per gram
`wet weight of tissue, we further characterized the
`peptides from this tissue. Bioactive peptides were
`extracted from 150 g of opossum colonic mucosa and
`fractionated by gel-filtration chromatography. The pro-
`file of bioactivity obtained when column fractions were
`
`.................
`

`

`Peak #2
`
`Peak #1
`
`LI
`•O
`
`•
`
`G712
`
`25
`
`20 -
`
`15 -
`
`1 0
`
`5
`
`pmol cGMP per well
`
`UROGUANYLIN: AN INTESTINAL PEPTIDE
`
`sequence analysis, no signal for amino acids was ob-
`served, indicating that this peptide may be blocked at
`the NH2-terminal end, a phenomenon previously en-
`countered when uroguanylin was purified from urine.
`To evaluate the bioactive components contained in
`peak 2 (Fig. 5), the same purification methods were
`utilized. When the components of peak 2 were further
`purified by isoelectric focusing, both guanylin-like and
`rmrsi-;r1 no AxIcr•cl moo nlxnarl (r,on,bo 9A onr1
`2B, Fig. 7). Peak 2A migrated with a pI of 3.0, similar to
`the pI of uroguanylin. Peak 2A also caused a greater
`cGMP response in T84 cells at pH 5.5 than it did at pH
`8, which is consistent with this peptide being urogua-
`nylin. It is likely that uroguanylin was not completely
`separated from guanylin in peak 2 (Fig. 5) during
`Sephadex G-25 chromatography. This uroguanylin-like
`activity of peak 2A exhibited the same chromatographic
`properties as authentic uroguanylin when subjected to
`further analysis by RP-HPLC using C18 columns (data
`not shown). An insufficient quantity of this purified
`peptide prevented its identification by NH2-terminal
`sequence analysis. The guanylin-like bioactivity peak
`that was resolved by isoelectric focusing migrated with
`a pI of 6.0, elicited 27-fold increases in T84 cell cGMP
`levels when assayed at pH 8, and caused only twofold
`increases in cGMP at pH 5.5 (peak 2B, Fig. 7). The pI of
`6.0 is similar to the pl estimated for opossum guanylin
`(~5.2), which contains a histidine residue (18). Peak 2B
`was further purified by a series of C18 RP-HPLC steps
`(7, 18). The guanylin-like activity eluted at about 15%
`acetonitrile (Fig. 8), which is a characteristic property
`of guanylin (18). One percent of this peak elicited a
`15-fold increase in cGMP accumulation when assayed
`at pH 8, but caused no detectable increase in cGMP
`levels in T84 cells at pH 5.5. Further purification by
`RP-HPLC followed by NH2-terminal sequence analysis
`
`12
`
`- 10
`
`-8
`
`-6
`
`-4
`
`2
`
`15
`
`_ 12
`
`Peak #2A
`0
`
`Peak #2B
`
`a. 9 -
`a.
`2
`
`6 -
`
`3 -
`
`0
`
`0 a
`
`.
`
`15
`
`18
`
`3
`
`6
`
`12
`9
`Fraction
`Fig. 7. Isoelectric focusing of peptides extracted from colonic mucosa.
`Fractions displaying predominately guanylin-like activity (including
`fractions comprising peak 2 in Fig. 5) were pooled from 2 separate
`Sephadex G-25 column runs and fractionated by isoelectric focusing
`on the Rotofor apparatus, as described in Fig. 4 legend. Fractions
`(1.0% of fraction volume) were bioassayed for cGMP accumulation in
`T84 cells using pH 5.5 (0) and pH 8.0 (U) medium, as described in
`Fig. 4 legend. •, pH of Rotofor fractions.
`
`35
`
`39
`
`19
`
`23
`
`31
`27
`Fraction
`Fig. 5. Sephadex G-25, gel-filtration chromatography elution profile
`of uroguanylin-like and guanylin-like peptides extracted from colonic
`mucosa. Peptides were extracted from colonic mucosa using Cis
`cartridges and processed as described in MATERIALS AND METHODS for
`application to a 2.5 x 90 cm Sephadex G-25 gel column. Fractions (10
`ml) were collected, and 1.0% of each fraction was assayed in MES,
`DMEM at pH 5.5 (0), and in HEPES, DMEM adjusted to pH 8.0 with
`50 mM sodium bicarbonate (■) for stimulation of cGMP accumulation
`in T84 cells. Illustrated bioactivity profile is a representative experi-
`ment from 1 of 2 Sephadex G-25 column runs.
`
`consistent with that previously observed for the 14- and
`15-amino acid forms of opossum uroguanylin that were
`isolated from urine (18). When the putative urogua-
`nylin peptide from colon was subjected to NH2-terminal
`
`8.5
`
`% 01IJHUOIO3V
`
`8
`
`7.5
`
`0
`
`6
`
`4 -
`
`2 -
`
`0
`
`pmol cGMP per well
`
`29
`
`35
`
`33
`31
`Fraction
`Fig. 6. T84 cell cGMP accumulation responses elicited by urogua-
`nylin purified from opossum colonic mucosa. Fractions were bioas-
`sayed using MES, DMEM at pH 5.5 (O), and HEPES, DMEM, 50 mM
`sodium bicarbonate at pH 8.0 (M). Uroguanylin-like peptides (includ-
`ing the fractions of peak 1 shown in Fig. 5) from 2 successive
`Sephadex G-25 column runs of colonic mucosal extracts were further
`purified as previously described (18) by a preparative isoelectric
`focusing step and 3 reverse-phase high-performance liquid chromatog-
`raphy (RP-HPLC) steps. Purified uroguanylin-like bioactivity shown
`is from fractions obtained after the third RP-HPLC step (thin solid
`line), which used acetonitrile and 10 mM ammonium acetate, pH 6.2,
`to purify peptides, using a 4.9 x 250 mm C15 column, as previously
`described (18).
`
`

`

`UROGUANYLIN: AN INTESTINAL PEPTIDE
`
`G713
`
`but not opossum uroguanylin. Highly purified, sequenc-
`ing-grade chymotrypsin was used to digest the syn-
`thetic peptides for these studies, which was followed by
`five cycles of NH2-terminal sequence analysis. Two
`amino acid sequences were obtained from chymotrypsin-
`digested opossum guanylin, SHTCE and AACAG, which
`were identical to the first five amino acids of the NH2
`
`q Control
`CS
`iii
`CT
`•CT + CS
`
`Basal
`
`Opossum
`ST-(5-17) Opossum
`Uroguanylin Guanylin
`
`nal
`Rat Guanylin
`(93-115)
`
`Rat Guanylin
`(101-115)
`
`CT
`• CT + CS
`
`A
`
`40-
`
`77;
`
`30-
`
`o.
`o.
`O 20-
`o
`0
`E C.
`
`10
`
`B
`
`100
`E
`E x
`to
`2
`
`75
`
`0: 5
`2
`
`25
`
`Synthetic
`Co on
`Uroguanylin
`
`Colon
`Synthetic
`Guanylin
`
`Fig. 9. cGMP responses elicited by chymotrypsin-treated synthetic
`peptides (A) and purified colonic peptides (B). Vehicle or peptides
`were incubated for 1 h at 34°C in 100 p1 of 10 mM HEPES, pH 8.0,
`under the following conditions: minus both chymotrypsin and chy-
`mostatin (control), minus chymotrypsin and plus chymostatin (CS),
`plus chymotrypsin and minus chymostatin (CT), and plus both
`chymotrypsin and chymostatin (CT + CS). Chymotrypsin (0.15 units)
`and chymostatin (100 pM) were used in reactions where indicated.
`After 1-h incubation, 100 pM chymostatin was added to CT reactions.
`Reaction mixtures were processed and analyzed in the T84 cell
`bioassay as described in MATERIALS AND METHODS. A: ST-(5-17) = 3
`nM; opossum guanylin, rat guanylin-(101-115), and opossum urogua-
`nylin = 30 nM; and rat guanylin-(93-115) = 100 nM. Error bars
`indicate standard error of the mean for 3 experiments. B: %maximal
`cGMP responses elicited by purified colonic forms of opossum urogua-
`nylin (Fig. 6) and guanylin (Fig. 8) are shown in comparison to cGMP
`responses elicited by 10 nM of synthetic opossum uroguanylin and
`guanylin after treatments as described for reactions CT and CT + CS.
`cGMP values (pmol/well), representing average of 2 experiments,
`minus basal values of cGMP (-0.2 pmol/well), are as follows: colonic
`uroguanylin, CT — 4.2, CT + CS = 5.7; synthetic uroguanylin, CT =
`2.16, CT + CS = 2.91; colonic guanylin, CT = 0.86, CT + CS = 45.3;
`synthetic guanylin, CT = 0.25, CT + CS = 18.7.
`
`16
`
`-15.5 D
`CDO
`
`CD
`-15 e
`
`14.5
`
`8
`
`7 -
`
`6 -
`
`5 -
`
`4 -
`
`3
`
`2
`
`1
`
`0
`
`pmol cGINAP per well
`
`95
`
`101
`
`99
`97
`Fraction
`Fig. 8. Purification of guanylin from opossum colonic mucosa by
`RP-HPLC. Guanylin-like peptides were purified from colonic mucosal
`extracts and bioassayed at pH 5.5 (O) and pH 8.0 (M) for stimulation
`of cGMP accumulation in T84 cells as described in Fig. 6 legend.
`Purified guanylin shown here was obtained after the third RP-HPLC
`step, which used acetonitrile (thin solid line) and 10 mM ammonium
`acetate, pH 6.2, to purify peptides using a 4.9 X 250 mm C15 column,
`as previously described (18).
`
`of the guanylin-like peptide confirmed that this bioac-
`tive substance was guanylin (SHTCEICAFAACAGC) (18).
`Inactivation of guanylin by chymotrypsin. As an
`additional test that the uroguanylin-like peptide identi-
`fied in opossum colon was not guanylin, we examined
`the guanylin-like and uroguanylin-like peptides that
`were isolated from colonic mucosa for sensitivity to
`inactivation by chymotrypsin in vitro. It has recently
`been reported that treatment of guanylin with chymo-
`trypsin resulted in the inactivation of rat guanylin-
`(101-115), but not a rat guanylin analogue containing
`an asparagine in place of Tyr109 (3). Because urogua-
`nylin has an asparagine instead of Tyr109, we postu-
`lated that uroguanylin would be resistant to chymotryp-
`sin. Figure 9A demonstrates that chymotrypsin
`treatment did not reduce the bioactivity of synthetic E.
`coli ST-(5-17) and synthetic opossum uroguanylin
`under conditions that completely inactivated the syn-
`thetic forms of opossum guanylin, rat guanylin-(101-
`115), and rat guanylin-(93-115). The chymotrypsin
`inhibitor, chymostatin, blocked the inactivation of syn-
`thetic guanylin peptides by chymotrypsin, as deter-
`mined by the T84 cell cGMP bioassay. The T84 cell
`cGMP responses elicited by purified colonic urogua-
`nylin and synthetic opossum uroguanylin were reduced
`by only 25% after treatment with chymotrypsin (Fig.
`9B). In contrast, the cGMP responses elicited by puri-
`fied colonic guanylin and synthetic opossum guanylin
`were reduced by >98% when pretreated with chymo-
`trypsin (Fig. 9B). Thus colonic mucosa contains chymo-
`trypsin-sensitive guanylin peptides, as well as chymo-
`trypsin-resistant uroguanylin peptides.
`NH2-terminal sequence analysis of syn