PEPTIDES
`
` seees
`ELSEVIER
`Regulatory Peptides 68 (1997) 45-56
`
`
`
`REGULATORY
`
`Purification, cDNA sequence, andtissue distribution of rat uroguanylin
`
`Zhiping Li*, Ashley G. Perkins, Matthew F. Peters, Michael J. Campa, Michael F. Goy
`
`Department of Physiology and Center for Gastrointestinal Biology and Disease at the University of North Carolina, Chapel Hill NC 27599, USA
`
`Received 29 June 1996; revised 17 October 1996; accepted 31 October 1996
`
`Abstract
`
`Guanylin, a peptide purified from rat jejunum,is thought to regulate water and electrolyte balance in the intestine. We show here, using
`a combination of Northern blots, Western blots, and functional assays, that guanylin and its receptor (GCC)are not distributed in parallel
`within the rat intestine. To investigate the possibility that there might be a secondintestinal peptide that serves as a ligand for GCC, we
`assayed tissue extracts for the ability to stimulate cyclic GMP synthesis in a GCC-expressingcell line. Duodenal extracts display a peak
`of biologicalactivity that is not present in colon and that does not comigrate with guanylin or proguanylin. The activity co-purifies with a
`novel peptide (TIATDECELCINVACTGC) that has high homology with uroguanylin, a peptide initially purified from human and
`opossum urine. A rat uroguanylin cDNA clone was found to encode a propeptide whose C-terminus corresponds to our purified peptide.
`Northern blots with probes generated from this clone reveal that prouroguanylin mRNAis strongly expressed in proximal small intestine,
`but virtually absent from colon, corroborating our biochemical measurements. Taken together, these studies demonstrate an intestinal
`origin for uroguanylin, and show that within the intestine its distribution is complementary to that of guanylin. © 1997 Elsevier Science
`BV.
`
`Keywords: GCC; STa receptor; CFTR; Guanylin; Uroguanylin
`
`1. Introduction
`
`A considerable body of evidence supports a role for the
`cyclic GMP pathway in the control of ion transport in the
`gastrointestinal tract. Elevation of intracellular cyclic GMP
`levels in intestinal epithelial cells enhances secretion of
`chloride into the intestinal
`lumen [1], and diminishes
`absorption of sodium and chloride [2]. The combination of
`increased secretion and decreased absorption elevates the
`osmolarity of the lumen, and drives the luminal accumula-
`tion of water. This mechanism was initially identified
`because it can be induced by heat-stable enterotoxin
`(STa)', an 18 amino acid peptide secreted by pathogenic
`
`*Corresponding author. Tel.: +1 919 9666993;
`9666927; e-mail: zpli@med.unc.edu
`enterotoxin; GCA,
`"The
`abbreviations used are: STa, heat-stable
`guanylate cyclase type A; GCB, guanylate cyclase type B; GCC,
`guanylate cyclase type C; HBSS, Hanks’ buffered salt solution; IBMX,
`3-isobutyl-1-methylxanthine; TFA,trifluoroacetic acid; TCA, trichloro-
`acetic acid; BSA, bovine serum albumin; RIA, radioimmunoassay.
`
`fax: +1 919
`
`0167-0115/97/$17.00 © 1997 Elsevier Science BY. All rights reserved
`PIT 80167-0115(96)02103-9
`
`strains of Escherichia coli [3-5]. Exposure to high levels
`of toxin, as occurs during acute bacterial
`infections,
`triggers non-physiological movementof electrolytes, and
`produces a watery diarrhea that can lead to dehydration
`and death.
`When the STa receptor was cloned from a small
`intestinal cDNA library [6], it was found to belong to a
`family of receptors that contain endogenous guanylate
`cyclase (GC) activity. Two other members of this family
`are the natriuretic peptide receptors, GCA and GCB [7].
`Because the STa receptor was the third such receptor
`cloned, it was named GCC. All members of this family
`contain:
`(a) an intracellular catalytic domain responsible
`for the conversion of GTP to cyclic GMP, (b) an intracellu-
`lar regulatory domain that controls the activity of the
`catalytic domain, (c) a single transmembrane domain, and
`(d) an extracellular receptor domain that provides an
`agonist binding site [7].
`These findings led to the hypothesis that GCC serves as
`a receptor for one or more endogenousligands in the GI
`
`MSN Exhibit 1006 - Page 1 of 12
`MSNv. Bausch - IPR2023-00016
`
`

`

`46
`
`Z. Li et al. / Regulatory Peptides 68 (1997) 45-56
`
`opossum uroguanylin) than to guanylin (50% amino acid
`identity with rat guanylin). (c) Cloning of a rat uroguanylin
`cDNA confirms that this second peptide represents the rat
`isoform of uroguanylin. (d) Northern blots show relatively
`selective expression of uroguanylin mRNA in proximal
`small intestine.
`
`tract, by analogy to the way that the other membrane-
`associated guanylate cyclases (GCA and GCB) serve as
`receptors for peptide ligands produced bytissues like the
`heart and brain [7]. Recently, a candidate intestinal pep-
`tide, called guanylin, was purified from jejunal extracts [8].
`Guanylin is believed to be a natural
`ligand for GCC
`because it (a) elevates intracellular cyclic GMP levels in
`GCC-expressing cells
`[8],
`(b) competes with STa for
`binding to GCC [8-10], and (c) stimulates the secretion of
`chloride by intestinal epithelial cells [9,11—15]. One key
`feature shared by guanylin and STa is a set of four
`conserved cysteines connected by specific disulfide bonds;
`Tissues were removed from Sprague—Dawleyrats (250—
`this provides the secondary structure required for bio-
`275 g) under urethane anesthesia (1.6 g urethane/kg
`logical activity [8,16].
`administered via i.p.
`injection). For RNA isolation,
`the
`The mRNAencoding GCCis strongly expressed in the
`whole intestinal tract was removed and put in ice cold
`intestine [6,17], exclusively in epithelial cells [18,19]. No
`Ringer’s-glucose (130 mM Na’, 120 mM Cl, 25 mM
`other tissue tested displays significant
`levels. However,
`HCO,, 1.2 mM Mg”", 1.2 mM Ca”, 2.4 mM K,HPO,,
`binding studies indicate that receptors for guanylin are
`present
`in the kidney [20,21]. Furthermore, STa and
`0.4 mM KH,PO,, 10 mM glucose). Tissues were then
`isolated as rapidly as possible, frozen on dry ice, and
`guanylin can stimulate sodium and potassium excretion by
`stored at — 80°C until used for RNA purification.
`the isolated, perfused kidney [22,23] and can elevate cyclic
`For the preparation of peptide-containing extracts for
`GMP in organ cultured kidney slices
`[21,24]. These
`Western blots,
`the mucosal
`layer of each tissue was
`observations suggest
`that guanylin, or a guanylin-like
`stripped free of the muscle layers, and homogenized in a
`peptide, mayplay a role in regulating kidney function. One
`buffer containing protease inhibitors (25 mM HEPES
`possibility is that guanylin is produced locally in the
`pH=7.4,
`1 mM phenylmethylsulfonylfluoride, 10 w~M
`kidney. Indeed, low levels of guanylin probe hybridization
`trans-epoxysuccinyl-L-leucylamido(4-guanidino)butane, 10
`have been reported on Northern blots of mRNA isolated
`
`
`from rat kidney [25]. However, this observation was not 1 mM_ben-pg/ml aprotinin, 10 pg/ml leupeptin,
`
`
`
`confirmed in comparable studies with mouse [26] or
`zamidine). After homogenization, the extracts were cen-
`human [10] kidney, and, furthermore, only small amounts
`trifuged at 10 000 X g for 20 min and insoluble material
`of guanylin-like bioactivity are present
`in rat kidney
`was discarded. The protein concentration was determined
`extracts [8,27].
`by the bicinchoninic acid method (BCA kit obtained from
`Pierce).
`A secondpossibility is that guanylin or a related peptide
`is delivered to the kidney from some other source. In an
`For the preparation of peptide-containing extracts for
`effort to identify such a peptide, Hamraet al. [28] and Kita
`HPLCfractionation, rat intestines were frozen on dry ice,
`and kept at — 80°C until used. After thawing, the intes-
`et al.
`[29] analyzed opossum and human urine for the
`presence of guanylin-like peptides. They found small
`tines were divided into regions corresponding to duodenum
`and colon. Duodenal tissue was taken as the 4 to 5 cm
`amounts of guanylin, and larger amounts of a second,
`structurally-similar
`peptide,
`which
`they
`named
`uroguanylin. Uroguanylin can bind to the STa/guanylin
`receptor:
`its EC, for activating cyclic GMP synthesis in
`GCC-expressing cells is
`intermediate between that of
`guanylin and STa [28,29], and it can competitively dis-
`place '**I-STa binding [28,29]. Uroguanylin is similar in
`amino acid sequence to guanylin and STa, and it retains
`their characteristic disulfide bond structure. Thus, guanylin
`and uroguanylin define a family of naturally occurring
`peptides that are structurally and functionally related.
`These studies leave open the question of what tissue
`serves as the biological source of uroguanylin found in the
`urine. In our current study, we show that small intestine is
`potentially one such source:
`(a) Rat duodenal extracts
`contain a peptide that displays guanylin-like bioactivity,
`but is chromatographically distinct from guanylin. (b) This
`second
`peptide
`is much more
`closely related
`to
`uroguanylin (80% amino acid identity with human and
`
`to the stomach.
`segment of intestine immediately distal
`Colon was taken to include both proximal and distal
`segments of the large bowel (posterior to the caecum and
`anterior to the sigmoid colon). Tissue pieces were split
`lengthwise and rinsed with normalsaline. The tissue was
`minced in 10 volumes | M acetic acid, placed in a boiling
`water bath for 5 min, and then homogenized. Boiled
`extracts were centrifuged at 4°C for 20 min at 230 000 X g.
`The resulting supernatant fractions were filtered through
`Whatman No.2 paper and applied to Waters C,, Sep-Pak
`cartridges. Unbound and weakly bound material was
`washed through with a solution of 10% acetonitrile—0.1%
`trifluoroacetic acid (TFA) in water. Tightly bound material
`was eluted with 60% acetonitrile—0.1% TFA. The eluted
`material was dried under vacuum and reconstituted either
`
`2. Experimental procedures
`
`2.1. Tissue and Extract Preparation
`
`in bioassay medium (Hanks’ buffered salt solution con-
`taining 1 mM IBMX)for bioassay on T84 cells, or in 10%
`acetonitrile/0.1% TFA for HPLC analysis (see below).
`MSNExhibit 1006 - Page 2 of 12
`MSNv. Bausch - IPR2023-00016
`
`

`

`Z. Li et al. / Regulatory Peptides 68 (1997) 45-56
`
`47
`
`For measuring region-specific responses to STa, seg-
`ments of tissue approximately 4 cm in length were placed
`in a dissecting pan filled with 37°C Ringer’s-glucose
`bubbled continuously with 95% O,-5% CO, Tubular
`sections were cut longitudinally, exposing a flat luminal
`surface, and luminal contents were discarded. Individual
`pieces of tissue (0.5 cm X 0.5 cm) were excised, placed in
`shell vials containing standard Ringer’s-glucose solution at
`37°C, and exposed to test solutions (see below).
`
`2.2. Northern blots
`
`Selected regions of the uroguanylin, guanylin, and GCC
`genes were amplified by PCR and subcloned into plasmid
`vectors (pBS, Stratagene), as described below and in a
`previous publication [18]. The cDNAinserts were isolated
`and used as templates for the synthesis of randomly-
`primed [°’P]-labeled cDNA probes (DECAprime II kit,
`Ambion).
`Total RNA was isolated by standard techniques [30],
`fractionated on a 1% agarose formaldehyde gel (2.2 M
`formaldehyde), and transferred to a nylon membrane
`(ICN). Membranes were treated with 10 ml prehybridiza-
`tion solution (50% formamide, 5X SSPE, 5X Denhardt’s,
`0.25mg/ml sperm DNA, 0.5% SDS) for 3 h and then
`hybridized in 10 ml Northern blot hybridization solution
`(50% formamide, 5X SSPE, 1X Denhardt’s, 0.1 mg/ml
`sperm DNA, 0.1% SDS, 10% dextran sulfate with 10°
`cpm/ml of each [*’P]-labeled probe) at 42°C for 24 h
`(SSPE = 150 mM NaCl, 11.5 mM NaH,PO,,
`1 mM
`EDTA, pH 7.4). Membranes were then washed twice (15
`min each) at room temperature in 2X SSC with 0.1% SDS
`(SSC = 150 mM NaCl, 15 mM Nacitrate, pH 7), followed
`by two 30 min washes at 55°C in 0.1X SSC with 0.1%
`SDS.
`
`2.3. Western blots
`
`Samples of tissue extracts or HPLC fractions were dried
`under vacuum,boiled in electrophoresis sample buffer, and
`fractionated on 15% SDS-—polyacrylamide gels made
`according to standard procedures
`[31] except
`for
`the
`composition of the electrode buffers (upper = 100 mM
`Tris-OH, 100mM tricine, 0.1% SDS, pH 8.3;
`lower =
`200mM Tris-Cl, pH 8.9). After electrophoresis, samples
`were transferred to 0.1 wm nitrocellulose membranes
`(Schleicher and Schuell) using a TE 22 transphor ap-
`paratus (Hoefer Scientific). The membranes were blocked
`with 3% BSA, washed, and incubated for 1 h at room
`temperature with a 1:500 dilution of antiserum 2538, a
`polyclonal antiserum that was raised against a 14 amino
`acid synthetic peptide whose sequence appears near the
`amino terminus of the rat guanylin prohormone [32]. The
`membranes were then washed and incubated with a
`
`secondary antibody (horseradish peroxidase-conjugated
`sheep-antirabbit IgG diluted 1:10 000, Boehringer-Mann-
`
`heim) for 60 min at room temperature. After an additional
`wash,
`the membranes were treated with chemilumines-
`cence reagent as specified by the manufacturer (Boeh-
`ringer-Mannheim), and exposed to Kodak XAR-5 film.
`
`2.4. Bioassay
`
`The studies described below have made use of two
`
`different bioassay procedures. Thefirst procedure was used
`to evaluate the relative ability of colon and duodenum to
`respond to STa (Fig. 1b). Shell vials containing 1 ml
`standard Ringer’s-glucose solution with 0.5 mM IBMX
`were placed in a 37°C water bath. Vials were continuously
`bubbled with 95% O,-5% CO,
`throughout the experi-
`ment. During the period of temperature and gas equilibra-
`tion, 100 units/ml STa (Sigma) was added to the appro-
`priate vials. Pieces of tissue from each region of the gut
`were then placed in the vials for 30 min. Reactions were
`stopped by removing the tissues from the vials and quick
`freezing them on a metalplate resting on a bed ofdry ice.
`The frozen tissue was then homogenized in 6% trichloro-
`acetic acid (TCA) and centrifuged to separate TCA-insolu-
`ble protein from TCA-soluble cyclic GMP. The protein
`pellet was dissolved by heating (37°C) in 1 M NaOH.
`Cyclic GMP was quantitated by RIA [33,34] and protein
`was determined using a dye-binding assay (Bio-Rad
`Bradford Assay) with BSA as
`standard. Results are
`reported as pmol cyclic GMP/mgprotein.
`The second procedure was used to evaluate HPLC
`fractions for guanylin-like activity (Figs. 2-5). This bioas-
`say, based on the method of Currie et al. [8], employs a
`reporter cell line (T84 cells) to detect the presence of GCC
`ligands, as manifested by an increase in intracellular cyclic
`GMP levels. T84 cells are derived from a human colon
`
`carcinoma, and retain many properties of crypt epithelium,
`including expression of high levels of GCC [35].
`In
`addition, because T84 cells express very low levels of the
`other known membrane cyclases [8,36], they provide a
`relatively specific bioassay system for ligands that target
`GCC. The cells were grown to confluency in 12- or
`24-well plastic trays. Cells were then washed two times
`with HBSS and incubated for 10 min at 37°C in HBSS
`
`containing 1 mM IBMX. HPLCfractions were reconsti-
`tuted in bioassay medium, then applied to the cells for an
`additional 30 min. The reaction was stopped by removing
`the incubation solution and replacing it with 0.5 ml ice
`cold 6% TCA.Cells were scraped andtransferred to plastic
`microcentrifuge tubes and the wells were rinsed with an
`additional 0.5 ml ice cold 6% TCA. Protein and cyclic
`GMP content were determined as above. Because the
`
`is quite uniform for T84 cell
`protein content per well
`cultures within a single plating, experimental results are
`reported as pmol cyclic GMP/well rather than pmol cyclic
`GMP/mgprotein.
`Activation of HPLC fractions: In our initial studies, the
`biological activity of our HPLC fractions was quite low.
`MSNExhibit 1006 - Page 3 of 12
`MSNv. Bausch - IPR2023-00016
`
`

`

`48
`
`Z. Li et al. / Regulatory Peptides 68 (1997) 45-56
`
`2.4
`
`1.35
`
`15
`
`—i °
`
`uo
`
`8
`Se
`SE
`gE
`@27
`°
`
`b
`
`B_
`as
`SE
`BE
`Qe
`ge22
`
` a
`blot
`
`1.0
`0.8
`
`SS
`0.6 g
`0.4
`02
`,
`
`Western
`oe 2 blot
`Sanaa
`5
`10
`15
`20
`25
`30
`0
`fraction number
`
`1.0
`0.8
`0.6 0
`g
`0.4 <
`0.2
`,
`
`Western
`
`¢
`30
`
`0
`
`5
`
`20
`15
`10
`fraction number
`
`25
`
`0.78
`
`
`response(pmolcGMP/mgprotein)
`bioassay
`
`0SSaeO)S)aaaII1IIInho==nN8onoF&© oon
`
`
`
`signaldensity(arbitraryunits)
`nmnNaOoua
`
`Cc
`
`kDa
`
`35
`30
`
`15
`
`10
`
`Cc
`D
`
`Fig. 1. Distribution of guanylin and GCC along the rostrocaudal axis of
`the GItract. (a) Northern blot of poly A” RNA (10 pg/lane) hybridized
`with radiolabeled guanylin and GCC probes. RNA wasisolated from rat
`stomach (S), duodenum (D), jejunum (J), ileum (1), or colon (C). All
`lanes are from a single membrane hybridized simultaneously with probes
`for GCC (upper), and guanylin (lower). Because the guanylin transcriptis
`much more abundant than the GCC transcript, the lowerhalf of the blot is
`shownafter a 14 h exposure, and the upper half after a 4 day exposure.
`Size standards, in kb, are indicated on the right. Tissue was pooled from
`three animals to obtain the RNA for this blot. Comparable results (not
`shown) have been obtained with RNA isolated from individual animals.
`(b) Cyclic GMP levels in duodenum (D) and colon (C) after incubation
`with (+) or without (—) STa at 100 units/ml. Each bar indicates the
`mean and standard error of nine determinations. (c) Western blot of tissue
`extracts performed with an antibody raised against an amino terminal
`domain of the guanylin prohormone. The autoradiograms (left) are of
`representative duodenal (D) and colonic (C) samples obtained from a
`single animal, with the positions of molecular weight standards as
`indicated. The bar graph (right) presents the densitometrically-determined
`mean (+SEM)ofdata obtained from four separate animals.
`
`(a) Two
`Fig. 2. HPLC comparison of colonic and duodenal extracts.
`extracts (generated from 15 and 30 colons, respectively) were indepen-
`dently analyzed by reverse phase HPLC,as described in Section 2. Upper
`trace: representative UV absorbance profile from one of the extracts.
`Middle trace (dashed line): the gradient of acetonitrile used to elute the
`column. Lower trace: bioassay responses to individual column fractions
`(mean+range from the two column runs, normalized as a percentage of
`the maximum response to correct for differences in the potencies of the
`extracts). Fractions were preincubated at 37°C before the bioassay was
`performed,
`in order to enhance their activity (see Section 2). At the
`bottom of the panel are Western blot data for HPLC fractions from one
`column run (pooled in pairs, except for fractions 19-26, which were
`analyzed individually) using an antibody that recognizes the N-terminus
`of the guanylin prohormone. An abbreviated region of the gel containing
`the immunoreactive proguanylin band is shown. The black and white
`arrows indicate the retention times of synthetic rat guanylin and opossum
`uroguanylin standards, respectively. (b) Three extracts (each generated
`from 15-30 duodena) were independently chromatographed under con-
`ditions identical to those employed for colonic extracts. UV absorbance,
`elution profile, and bioassay responses are plotted as in panel (a).
`
`Therefore, in order to obtain duplicate bioassay measure-
`ments, we dried each HPLCfraction, reconstituted it in
`bioassay buffer, applied the entire fraction to one well of
`T84 cells for 30 min, and then transferred it to a second
`well for 30 min. We noticed that the response of the cells
`in the second well was alwayssignificantly greater than the
`response of the cells in the first well, suggesting that
`something was happening during the first incubation to
`activate the sample. Further experiments showed that
`samples could be activated simply by incubating them at
`37°C for 30 min. As described in Section 3, we believe
`that this is due to proteolytic cleavage by a contaminating
`protease. We have used this activation procedure to evalu-
`ate samples in all of the HPLC analyses presented below;
`in some cases the activation was achieved by incubating
`the sample on a well of T84 cells at 37°C, in others by
`MSNExhibit 1006 - Page 4 of 12
`MSNv. Bausch - IPR2023-00016
`
`

`

`Z. Li et al. / Regulatory Peptides 68 (1997) 45-56
`
`49
`
`0.01 0.00
`
`0.04
`0.03
`
`0.02 A280
`
`oOo oa
`
`°°
`
`bioassayresponse(pmol
`
`cGMP/well)
`
`0
`
`5
`
`20
`15
`10
`fraction number
`
`25
`
`30
`
`Fig. 4. Preincubation at 37°C leadsto a shift in the retention time of the
`duodenal peptide. An extract of 30 duodena was fractionated as in Fig.
`2b. Fractions 23 and 24 were pooled, dried under vacuum (to remove
`HPLCsolvents), resuspended in bioassay medium, incubated for 60 min
`at 37°C, and rechromatographed on the same column using the same
`elution profile. The upper trace shows UV absorbance; the dashed line
`showsthe elution profile; the lower trace shows bioassay responses to
`individual fractions. Biological activity now elutes in fraction 16, while
`most of the UV absorbing material continues to elute in fractions 23 and
`24. The large peak marked with the asterisk is due to IBMX, which was
`present in the bioassay medium.
`
`=2
`
`control
`(n =3)
`
`figure 2a
`fraction 21
`(n =2)
`
`figure 2b
`fraction 23/24
`(n= 6)
`
`guanylin
`(n=3)
`
`uroguanylin
`(n=2)
`
`STa
`(n= 2)
`
`figure 4
`fraction 16
`(n=1)
`
`0
`
`0.5
`
`1
`
`20
`
`40
`
`60
`
`80
`
`cGMP pmol/well
`
`0.02 0.01
`
`$=
`Fig. 3. Preincubation at 37°C enhances the activity of HPLC column
`5g
`fractions, but not of synthetic guanylin or uroguanylin. The bars show the
`£2
`
`mean level of cyclic GMP (+SEM orrange) in T84 cells after exposure
`
`
`zoaS
`to the indicated stimuli. Each stimulus was either held at 4°C (C4) or
`incubated at 37°C (Ml)prior to applyingit to the cells. The numbers used
`gs
`ge o°
`to identify the HPLC fractions correspond to (i) the figure in which the
`0 TTTTTTT
`pertinent HPLC run is shown and(ii) the appropriate fraction number(s)
`0
`20
`40
`60
`80
`100
`120
`from that column run.
`fraction number
`
`0.00
`
`1.0
`
`incubating it without cells at 37°C. We have the impression
`that the activation process occurs more efficiently if T84
`cells are present, but we have not compared the two
`procedures in enough detail to be certain ofthis.
`
`2.5. HPLC purification
`
`2.5.1, Step 1
`Reconstituted Sep-Pak fractions were fractionated on a
`PepRPC HRS5S/5 (C,,) column (Pharmacia), pre-equili-
`brated with 10% buffer B (buffer A= 0.1% TFA; buffer
`B= 99.9% acetonitrile + 0.1% TFA). After sample appli-
`cation, the column waseluted isocratically for 5 min with
`10% buffer B, followed bya linear gradient to 50% buffer
`B over 43 min, and a final 12 min elution at 100% buffer
`B. One ml fractions were collected at a flow rate of 0.5
`ml/min, while absorbance was monitored at 214 and 280
`nm. After chromatography, a portion of each fraction was
`lyophilized (to eliminate the acetonitrile and TFA), re-
`suspended in bioassay medium, incubated at 37°C for 30
`
`Fig. 5. Final step in purification of the duodenal peptide. Fraction 16 from
`the column run illustrated in Figure 4 was dried under vacuum,
`resuspended in 10% acetonitrile + 0.1% TFA, and rechromatographedas
`described in Section 2. The upper trace shows UV absorbance; the dashed
`line showsthe elution profile; the lower trace shows bioassay responses to
`individual fractions.
`
`min to activate the peptide (see above), and tested for
`activity on T84 cells. Typical examples of such chromato-
`grams are shown in Fig. 2.
`
`2.5.2. Step 2
`fractions with activity were combined for
`Duodenal
`further purification. The pooled fractions were incubated at
`37°C for 60 min, then reapplied to the PepRPC HR5/5
`column and eluted without modification of the protocol
`described in step 1. The duodenal peptide now elutes
`earlier than in step 1, whereas most contaminants retain
`their original
`retention time. The fractions containing
`activity were identified by bioassay, as shownin Fig. 4.
`MSNExhibit 1006 - Page 5 of 12
`MSNv. Bausch - IPR2023-00016
`
`

`

`50
`
`Z. Li et al. / Regulatory Peptides 68 (1997) 45-56
`
`2.5.3. Step 3
`Active fractions from step 2 were pooled and applied to
`a Cy, analytical column (Vydac), pre-equilibrated with
`10% buffer B (buffer A, 10 mM ammonium acetate pH
`6.2; buffer B, 99% acetonitrile +10 mM ammonium
`acetate pH 6.2). The column waseluted isocratically for 10
`min with 10% buffer B, followed by a linear gradient to
`40% buffer B over 160 min, and a final 10 min elution at
`100% buffer B. Fractions (1.5 ml) were collected at a flow
`rate of 1 ml/min, and analyzed by bioassay, as shown in
`Fig. 5.
`Synthetic 15 amino acid rat guanylin and opossum
`uroguanylin (used as HPLC standards) were generously
`provided by Drs. Ding Chang (Peninsula Labs) and
`Leonard Forte (University of Missouri), respectively.
`
`2.6. Peptide characterization
`
`Sequencing was performed on a sample of peptide
`adsorbed to a 0.22 micron PVDF membrane, using a
`Perkin Elmer/ABI model 491 sequencer with on-line PTH
`amino acid detection.
`
`2.7. CDNA analysis
`
`Rat duodenal RNA wasprepared as described previous-
`ly [18], and reverse transcribed with SuperScript II using
`an oligo dT primer (Gibco BRL). The resulting cDNA was
`then subjected to PCR, using degenerate primers designed
`from areas of high homology in published sequences of
`human and opossum uroguanylin [37,38]. The sense
`primer, TACATCCAGTA(CT)(GC)A(AG)GCCTTCC,and
`antisense primer, GCAGCC(GT)GTACA(GC)GC(AC)-
`ACGTT,correspond, respectively, to base pairs 104—124
`and 334-354 of the human transcript
`[38]. PCR was
`performed for 40 cycles (denaturation for 1 min at 94°C,
`annealing for 1 min at 55°C, and extension at 72°C for 2
`min) followed by a final extension for 10 min at 72° C
`using Taq polymerase (Boehringer Mannheim). A 250 bp
`product was amplified, as expected. This PCR product was
`subcloned into the pBluescript II SK vector using the
`T/A cloning method [39]. DNA was sequenced at
`the
`UNC-CH Automated DNA Sequencing Facility on a
`Model 373A DNA Sequencer (Applied Biosystems) using
`the Taq DyeDeoxy™ Terminator Cycle Sequencing Kit
`(Applied Biosystems). Sequence analysis revealed a high
`degree of homology at the nucleotide level to the human
`and opossum forms of uroguanylin (72 and 60%, respec-
`tively), confirming that we had amplified the appropriate
`target sequence.
`To determine the full-length uroguanylin cDNA se-
`quence, we used the PCR product to produce a random-
`primed probe for screening a rat duodenal cDNAlibrary
`constructed in AGT-11 [40]
`(a gift
`from Dr. Andrew
`Leiter). Phage plaques were adsorbed onto nitrocellulose
`filters, hybridized with the probe in 50% formamide, 0.8 M
`
`NaCl, 20 mM PIPES pH 6.5, 0.5% SDS and 100 pg/ml
`salmon sperm DNA at 42°C, and washed in 0.1X SSC/
`0.1% SDS at 55°C. Positive plaques were purified by
`sequential low-density plating and bacteriophage DNA was
`isolated with a Qiagen Lambda kit (Qiagen). The cDNA
`insert was excised with EcoRI and subcloned into the
`
`pBluescript II SK vector for sequencing, as described
`above.
`
`3. Results
`
`3.1. Guanylin and its receptor are not distributed in
`parallel in the GI tract
`
`Previous studies have shown that guanylin mRNA is
`expressed in a rostrocaudal gradient, ranging from quite
`low in duodenum to quite high in colon [9,32,41,42]. In
`contrast, as shown in Fig.
`la, the mRNA encoding the
`guanylin receptor
`(GCC)
`is expressed at high levels
`throughout the GI tract. These data reveal a nonparallel
`distribution of guanylin and its receptor, with the mismatch
`particularly evident in duodenum.
`in
`To verify that
`the high level of GCC transcript
`duodenum corresponds to a high level of functional
`receptor, we applied a sub-saturating dose of a GCC-
`specific ligand (STa) to excised pieces of tissue in organ
`culture, and measured intracellular levels of cyclic GMP
`after 30 min of exposure (a time at which the responseis
`still proceeding linearly). This experiment was performed
`in the presence of a phosphodiesterase inhibitor, isobutyl
`methylxanthine (IBMX), to minimize the effects of phos-
`phodiesterase enzymes on cyclic GMP metabolism. The
`agonist-dependent
`increase in cyclic GMP levels (the
`amount of cyclic GMP in STa-stimulated tissue minus the
`amountin unstimulated control tissue) is slightly greater in
`duodenum than it
`is in colon (Fig. 1b),
`indicating that
`duodenum does express substantial
`levels of functional
`receptor.
`To verify that the low level of guanylin transcript in
`duodenum corresponds to a low level of guanylin prop-
`eptide, we performed Western blots on extracts of
`duodenum and colon, using an antibody that recognizes the
`N-terminus of the propeptide [32]. Extracts were prepared
`from four separate animals;
`in each case the level of
`proguanylin in duodenum, as measured densitometrically,
`was less than 11% of the level
`in colon (mean = 8.7%
`+£1.0% (SEM)) (Fig. Ic).
`
`3.2. HPLC analysis of guanylin-like peptides in
`intestinal extracts
`
`The ligand/receptor mismatch described above suggests
`the possibility that duodenum might produce a ligand that
`resembles guanylin in its
`receptor specificity, but
`is
`biochemically distinct. In order to compare guanylin-like
`MSNExhibit 1006 - Page 6 of 12
`MSNv. Bausch- IPR2023-00016
`
`

`

`Z. Li et al. / Regulatory Peptides 68 (1997) 45-56
`
`51
`
`peptides expressed in duodenum to those expressed in
`colon, we prepared aqueous extracts of each tissue, frac-
`tionated the extracts by HPLC, and assayed an aliquot of
`each fraction for
`its ability to stimulate cyclic GMP
`synthesis in cultured T84 cells, as described in the Section
`2. In parallel with these bioassays, we performed Western
`blots on a sample of each HPLC fraction, using the
`antibody described in Fig. 1c. For convenience, only the
`region of the blot containing immunoreactive proguanylin
`is shown in the figures below.
`Colonic extracts contain both guanylin and proguanylin.
`Fig. 2a illustrates the HPLC analysis of a colonic extract.
`The Western blot inset at the bottom of the figure shows
`that proguanylin elutes in fraction 21. When the fractions
`are bioassayed, a peak of cyclic GMP-promotingactivity is
`also seen in fraction 21. In the course of characterizing this
`peak, we noticed that its biological activity increases with
`timeif it is incubated at 37°C (Fig. 3). Such an increase is
`not
`observed when
`synthetic
`guanylin,
`synthetic
`uroguanylin, or commercially-purified STa are incubated
`under similar conditions (Fig. 3). We do not yet know the
`mechanism by which this time-dependent enhancement of
`activity occurs; however, as discussed below, we consider
`it likely that proguanylin, which is biologically relatively
`inactive
`[10,25,42,43], has co-eluted with a protease
`capable of converting it to a smaller, more active peptide.
`Hamra et al. have previously demonstrated that specific
`proteases enhancethe activity of proguanylin [44].
`Duodenal extracts contain biologically-active material
`that is distinct from both guanylin and proguanylin. When
`duodenal extracts are analyzed by HPLC, proguanylin-like
`immunoreactivity is again observed in fraction 21 (West-
`ern blot inset, Fig. 2b), though, as expected, the amountis
`muchless than can be seen in comparable colonic extracts
`(Western blot inset, Fig. 2a). The T84 cell assay confirms
`that this duodenal proguanylin is associated with a small
`peak of biological activity, which, as above, can be
`enhanced by preincubation at 37°C. However,
`the most
`conspicuousaspect of the chromatogram is the presence of
`a second, much larger peak of activity in fractions 23 and
`24. This material differs from proguanylin in two ways: it
`is retained more tightly by the column,andit fails to react
`with our proguanylin-specific antibody. In addition,
`the
`retention time of the material is significantly different from
`that of synthetic rat guanylin (fraction 13) or synthetic
`opossum uroguanylin (fraction 15). The novel duodenal
`peak does, however, share one property with proguanylin:
`its activity is also enhanced by preincubation at 37°C (Fig.
`3). If an intestinal protease is indeed responsible for this
`activation phenomenon,then the data suggest that such a
`protease is likely to come off the HPLC column in a broad
`peak that overlaps both proguanylin (fraction 21) and the
`material in fractions 23-24.
`In order to test whether the duodenal material
`
`is a
`
`peptidase, and S. aureus V8 protease). After a 2 h
`incubation,its biological activity was completely destroyed
`(data not shown). This, in conjunction with the purification
`and sequencing studies described below, confirms the
`peptide nature of the active material in fractions 23-24.
`
`3.3. Preincubation alters the chromatographic properties
`of the duodenal peptide
`
`If a proteolytic mechanism is responsible for the activa-
`tion of the duodenal peptide, then the size of the molecule
`should be altered once it has been activated, and this
`should be reflected by a change in its chromatographic
`properties. To test this prediction, we partially purified the
`duodenal peptide on our standard reverse phase HPLC
`column (as in Fig. 2b), allowed it to preincubate at 37°C,
`rechromatographed it on the same HPLC column, and
`bioassayed the resulting fractions