Immunochemical characterization of gastrinlike and cholecystokininlike peptides released in dogs in response to a peptone meal

GASTROENTEROLOGY

1984:87:323-34

Immunochemical Characterization of Gastrinlike and Cholecystokininlike Peptides Released in Dogs in Response to a Peptone Meal M.

MICHAEL

Department

WOLFE

of Medicine.

and JAMES

University

of Florida

This study, was designed to examine and compare the nature of gastrinlike and cholecystokininlike peptides released into the portal and peripheral venous circulation in response to a peptone meal. Six dogs were prepared with portal venous catheters and gastric fistulas. Portal and peripheral venous sera were obtained before and after gastric infusion of a 10% peptone meal. Serum levels of gastrin and cholecystokinin immunoreactive peptides were determined by radioimmunoassay using two distinct peptide region-specific antibody preparations. These separate antibody preparations demonstrated specificity for (a) C-terminal tetradecapeptide gastrin (417hG17], heptadecapeptide gastrin (Gl7), and big gastrin /G34) (gastrin antibody); and (b) a11 biologically active forms of gastrin and cholecystokinin (gastrin-cholecystokinin antibody). Using the antibody preparation with specificity for gastrin and not the mean basal immunoreactive cholecystokinin, gastrin from portal (8.33 ? 2.4 fmoliml, mean 2 SEM) and from peripheral (6.19 + 0.9 fmoJ/mJ) venous sera both increased after peptone infusion, with an early peak (2 min in portal and 4 min in peripheral serum) and a second peak at 30 min in both circulations. Measurements using antibodies with specificity for both cholecystokinin and gastrin yielded strikingly different results. The portal venous serum peptide concentration (49 2 10 fmoliml) Received September 17. 1981. Accepted February 13. 1984. Address requests for reprints to: lames E. McGuigan. M.D.. Department of Medicine. t!niversity of Florida. Gainesville. Flnrida 32610. This study was supported by Grants AM 07209 and AM 13711 from the National Institutes of Health. Dr. Wolfe is a recipient of Clinical Investigator Axvard 00147 from the National Institutes of Health. The authors thank Dr. M. Hocking and Dr. 1. Hawkins for technical assistance and J. Wilhite for assistance in the preparation of the manuscript. C 1984 by the American Gastroenterological 0016-5085/84/$3.00

Association

E. McGUIGAN College

of Medicine,

Gainesville.

Florida

increased sharply within 30 s after peptone infusion to a single peak at 2 min (139 2 37 fmoliml). The basal peripheral venous serum peptide concentration (43 ? 8.8 fmoJ/mJ) increased more gradually to Studies a single peak at 8 min (78 5 14 fmollml]. with Sephadex (G-50 superfine) gel chromatography indicated that gastrin released in response to the peptone meal was primarily Gl7. However, of the peptides released in response to the peptone meal that were recognized by the gastrin-cholecystokinin antibody, >80% were shown to be distinct from gastrin. Gel chromatographic studies demonstrated that peptone meal-stimulated immunoreactive choJecystokinin release consisted of two major peaks, eluting in positions identical to those of intact cholecystokinin (CCK33) and the octapeptide of choJecystokinin (CCK8). Gastrin and cholecystokinin (CCK) are structurally related in that both contain an identical carboxylterminal pentapeptide amide. There is evidence that both gastrin and CCK are present in tissues and in the circulation in multiple molecular forms of varying peptide length. Previous studies have revealed that the primary form of gastrin in antral mucosa is heptadecapeptide gastrin (G17), whereas the predominant circulating form is big gastrin (G34) (1). After its release into the splanchnic venous effluent, in response to various physiologic stimuli, gastrin traverses the liver, where smaller forms, i.e.. those less than nine amino acids in length, can be efficiently cleared (2). It is widely recognized that ingested protein and products of protein digestion are the most potent stimuli for gastrin release (3). In the present study we examined release of gastrinlike and CCK-like peptides in response to intragastric infusion of a peptone Abbreviations synthetic human

used in this paper: gastrin I.

CM.

c:holec:~~stokinin:

SHG.

324

WOLFE

AND

McGLJIGAN

GASTKOENTEROLOGY

Antibody

56-02

(Gostrin

Vol.

87. No

2

antibody)

A

Antibody Figure

1

06

( Gostrin/CCK

antibody)

Region-specific antibodies for gastrinlike and CCK-like peptides produced by immunization ot rabbits w,ith hllnlilll gastrin peptides covalently conjugated to bovine serum albumin. Peptide regions within brc~ckrts indicate peptide segments conjugated to carrier protein for use as iniInunogen in producing the trvo tlistinct entitlotl~ preparations.

meal by simultaneous sampling of portal and peripheral blood in alert, nonanesthetized dogs using peptide region-specific antibodies with differential recognition properties for gastrin and CCK peptides.

101, Bayer, Leverkusen, Federal Republic of Germany). samples were centrifuged at 1000 g at 4°C. and serum stored at -20°C until assayed.

Gastrointestinal

Materials and Methods Preparation

of Animals

Six healthy mongrel dogs (4 females and 2 males), weighing 15-20 kg, received midline laparotomg incisions under halothane-induced general anesthesia. A radiopaque Teflon catheter, 5 mm in diameter, -25 cm in length, with one end and three side portals, was introduced through a branch of the terminal ileal venous arcade. The catheter was then advanced through the superior mesenteric vein into the portal vein. A stainless steel stylus was inserted to occlude the catheter lumen, which ensured its patency and permitted a completely sealed system. The catheter and stylus were externalized through a stab incision in the right lower quadrant of the dog and secured on the skin surface, where a stainless steel removable cap was sutured over the catheter hub. This cap prevented premature removal of the catheter by the animal. After successful placement of the portal vein catheter. a gastrocutaneous fistula was prepared by insertion, in the most dependent portion of the stomach, of a stainless steel cannula (13 mm in diameter) which was brought out through the skin. Investigative

Procedure

After a minimum recovery period of 7 days following surgery, each dog was fasted overnight and placed in a harness. The portal vein catheter stylus was removed and a 0.9% NaCl solution was infused through the catheter to maintain patency and permit adequate portal venous blood sampling throughout the duration of the experiment. An intravenous catheter was placed in a cephalic vein: its patency was also maintained by infusion of O.gl$, NaCl. Simultaneous basal samples of portal and cephalic venous blood were obtained, immediately after which a 10% peptone meal (150 ml, pH 7) was infused over 30 s through the gastric fistula into the stomach. After infusion of the peptone meal, the fistula was capped. and portal and peripheral venous blood samples were obtained at 0.5, 2, 4, 8, 15, 30, 45, 60, and 90 min. Each sample of venous blood was collected in icechilled glass tubes containing 1000 KIU aprotinin (Trasy-

The was

Peptides

Synthetic human gastrin I (SHG. hGl7). used both for antibody production and assay standard. was obtained from Imperial Chemical Industries, Ltd., Cheshire, England. Human big gastrin I (hG34), the amino-terminal 0 hGl7lSHG) 0034

04-17017 V I-13017

4cG17

014~17017

. CCK 33 ??CCKG

I .oJ

QO 0.80.7. :: v) 068 -

0.5

mo ’ 04 4 g

03s 0.2Ol-

T

I

I

IO

1

I&

PEPTIDE

Figure

I

1

I03

’ lb4

’ lb5

lfmol)

2. Calibration curve examining immunologic. (.ross-ream.tivity of gastrin and CCK peptides rvith gastrin antiserurn (56-02). The antibodies are immunoreacti\,e with 1. 17hG17, hG17. and hG34 (human) and cCl7 (canine) on an equimolar basis. No immunologic. cross-reactivity with l-13hG17 and CCK33 is detected. There is minimal immunoreactivity with CCK8 and G-l The index of inhibition (the ratio of conr.entrations of standard and peptide required for 40% inhibition of antibody binding) of lZ’I-SHG (synthetic. human gastrin) binding by gastrin antibodies by CCK8 is 0.0418: the index of inhibition of IL’I-SHC binding b!, gastrin antibodies by G-4 is 0.0015. CCK = c.holrc.ystokinin.

GASTRIN-

August 1984

?? hGl7(SHGl

0 4-17617 V I-13617 0 14-17617

OG34 AcG17

I

.

o-

-43La

.

ACCK33

?? CCK6

.

.

-

09-

.

Gostrin/CCK

v

Antlbody

(06)

.

0s

0 0.75 fi’ 06-

. ??

g 0 5 a z 1 048

lJ ?! ?? 9

i

03-

A

i

0.2-

An

B

.‘ o A

Ol-

\ 1

1

I

IO

1

I

IO2 PEPTIDE

Figure

:j.

1

K

IO3

I

IO4

1

105

(fmol)

examining immunologic: cross-~a(.tivity of gastrin and CCK peptidrs 12rith gastrin-CCK antiserum (OG). The antibodies exhibit near trquimolar ilnmul7oreactivit~ with all known biologically active forms of gastrin and CCK (containing the carboxvlterminal tetrapeptide ~ornmon to gastrin ant1 CCK). No immunoreactivity \vith I-13hG17 is detectable. CCK = cholrc.\~stokinirl. hlibration

curve

tridecapeptide of human gastrin (I-13hG17). the carboxylterminal tetradecapeptide amide of human gastrin (4 17hG17). and canine gastrin I (cG17) were generous gifts from Professor R.A. Gregory, the Physiological Laboratory. University of Liverpool, Liverpool, England. The carboxylterminal tetrapeptide amide common to both gastrin and CCK (G4, CCK4) was a gift from Dr. John S. Morley, Imperial Chemical Industries, Ltd., Cheshire, England. Porcine cholecystokinin-pancreozymin (99”io pure) (CCK33) was obtained from Dr. Viktor Mutt, the Gastrointestinal Hormone (GIH) Research Laboratory, Karolinska Institute, Stockholm, Sweden. Synthetic cholecystokinin-octapeptide (sulfated and nonsulfated) (CCK8) was obtained from Squibb Chemical Company, New Brunswick, N.J.

Determination Cholecystokininlike

of Serum

Gastrinlike

Peptide

AND CCK-LIKE

PEPTIDES

IN DOGS

325

immunoreactive with 4-l 7hGl7, G17 [canine and human), and human G34 on an equimolar basis (Figure 2). Immunoreactivity with the carboxyl-terminal tetrapeptide amide (G4) of gastrin and cholecystokinin and with CCK is minimal, i.e., ~5%) (Figure 2) (4). The presence or absence of a sulfated tyrosyl residue, as determined by examination of sulfated and nonsulfated forms of hG17, does not alter the immunoreactivity of this antibody. Antibody 5602 is designated as gastrin antibody. Antiserum 06 was produced by immunization using the carboxyl-terminal tetrapeptide amide of gastrin (and CCK) as immunogen. This antibody exhibits near equimolar immunoreactivit) with all known biologically active forms of gastrin and CCK (Figure 3) (5). It is designated as gastrin-CCK antibody. The presence or absence of a sulfated tyrosyl residue, as determined by examination of sulfated and nonsulfated forms of CCK8, does not alter the immunoreactivity of this antibody. Both portal and peripheral venous serum samples from 5 dogs obtained at each time point were examined by radioimmunoassay using gastrin antibody. Peptide concentrations were also measured in portal and peripheral serum samples from 4 dogs at each time point by radioimmunoassay using gastrin-CCK antibod!,. All radioimmunoassay measurements were performed using modifications of previously described methods (4,5). Dilutions of assay reagents were prepared in 0.02 M sodium barbital buffer (pH 8.4) containing 2% normal human serum. All calibration curve standards and serum samples were included as loo-p1 aliquots in a total incubation volume of 2 ml in 10 x 75-mm glass tubes. To each gastrin standard and sample were added 100 ~1 of diluted antiserum (56-02 or 061, 100 ~1 of ‘2”I-SHG (-1500 counts per minute) prepared by a modification of the chloramine-T method of Hunter and Greenwood (6,7), and the previously mentioned 0.02 IM sodium barbital buffer, to a total volume of 2 ml. Antibody- 52-02 was used at a

I o-

and 0,

Concentrations

0

5

IO SHG

This laboratory has produced two different antisera with specificity for defined peptide regions of the gastrin molecule (Figure 1) by immunization with gastrin pep tides covalently conjugated to bovine serum albumin. Antiserum 56-02 contains antibodies produced by immunization of rabbits wifh SHG amino acids 2-17 and is

Figure

50

100

(fmoll

4. Calibration curves sho\ving mrnpetition of synthetic. human gastrin (SHG) and radiolabeled SHG for binding dncI gastrin-CCK antiby antibodies (56-02) (O--O) body 06 (0-O). The values depicted on the abscissa represent femtomoles of SHG per inc.uhation tube. CCK = cholecystokinin

326

GASTKOENTEKOLOGY Vol. 87. No. 2

WOLFE AND McGUIGAN

me

‘251Na

Dextran

1

50:

64

4

04 0

20

40

ELUTION

Figure

5

60 VOLUME

80

100

the assays using antiserum 06 was 0.38 fmol per incubation tube. All samples were measured in one assay (for each antibody] to eliminate interassay variation. Figure 4 depicts representative calibration curves using serial dilutions of G17 (SHG) and utilizing the two antisera used in the present study and illustrates virtually superimposable curves. Affinity chromatography was used to further characterize antiserum 06 (gastrin-CCK). The IgG immunoglobulin fraction of the antiserum was separated by application and elution (0.02 M potassium phosphate buffer, pH 7.45) of 1.0 ml of antiserum from a column (0.9 X 11 cm) containing diethylaminoethanol Affi-Gel Blue (Bio-Rad Laboratories, Richmond, Calif.). The antibody-containing immunoglobulin fraction was then conjugated to l,l’-carbonyldiimidazole-activated 6% agarose [Reactigel (6x), Pierce Chemical Co., Rockford, Ill.] in 0.1 M sodium borate buffer, pH 8.5, at 4°C for 48 h to yield an antibody-gel suspension. Sera from 2 dogs were obtained in the fasting state and 5 and 20 min after completion of a standard meal (454 g) of canned dog food. Serum samples (0.5 ml) at each time point from both dogs were incubated with 0.5 ml of a 40% antibody-gel suspension in 0.05 M potassium phosphate buffer, pH 7.45, on a rotary wheel at 4°C for 48 h. After centrifugation at 2000 g at 4°C for 30 min, supernates were separated and assayed for gastrinlike and CCK-like peptides with antibodies 56-02 and 06, as described previously. No immunoreactive gastrinlike or CCK-like peptides were detectable in any of the serum samples after

(%I

Elution profiles of gastrin and CCK peptides used in characterization of Sephadex (;-SO superfine column. Values depicted on the ordinate represent the COIICBIItrations in femtomoles per milliliter of peptides immunoreactive lvith gastrin-CCK antibody (06). Elution volume represents precentage of elution betlveen blue: dextran (V,,) and “‘INa. CCK = cholecystokinin.

final dilution of 1: 3,000,OOO and antibody 06 at a final dilution of 1:320,000. Assay tubes were vortexed and incubated at 4°C for 48 h. At that time, separation of antibody-bound from antibody-free ““I-SHG was performed using Amberlite resin CG-4B 200-400 mesh, lot #lo5286 (Chemical Dynamics Corporation, South Plainfield, N.J.). A 10% suspension of the resin in 0.02 M sodium barbital buffer (pH 8.4) was prepared and 200 ~1 was added to each incubation tube. To each standard tube 100 ~1 of normal human serum was added immediately before addition of the resin. After centrifugation at 2000 g for 10 min, the precipitates (containing free ‘““I-SHG) and supernates were counted wih a Tracer Analytic Model 1185 (Tracer Analytic, Inc., Elk Grove Village, Ill.] automatic y-spectrometer. Gastrinlike and CCK-like peptide concentrations were calculated from calibration curves using SHG standard concentrations of 238-1.9 fmoliml (11.9-0.095 fmol per incubation tube). Logit transformation with a Wang 700 advanced programming calculator (Wang Laboratory, Lowell, Mass.) was used in the determination of “‘I-SHG bound/free ratios and peptide concentrations. Sensitivity of assays using antibody 56-02 was 0.19 fmol of gastrin per incubation tube. Sensitivity of

20

4

1 CCK33

I5

+ 0

IO

20

30

ELUTION

40 VOLUME

50

60

70

80

90

IO0

I % i

Elution protiles of (XX33 ,rlltl (:(X8 used it] I.II~IT~I( II:I‘c.olurnn. Cholrc:vxization ot Sephadex G-50 superlinr tokinin 33 and CCK8 (250 fmol c~c.h) \\‘(:r(: added separately to canine portal serum and \vere applird II) and eluted from the c.olumn. Thta I rllu~s on tht, orclinate represent concentrations in tc:mtomolw per rnilliliter of CCK33 and (:CKX imlunnoreac.tivr ~vith gastrillCCK antibody (06]. Volumr! is c?sprrssed as perwntagv of elution betlveon bluc~ dwtran (\‘,,I itncl “‘liXcr. b:lution performed with 0.02 L1 sodium barbitcil llulfer (pH human sorum. (:(:K = 8.4) containing 2”,, normal cholecystokinin.

August

GASTRIN- AND CCK-LIKE PEPTIDES IN DOGS

1984

Gel Chromatography Gastrinlike Released

IO

20 TIME

xl

40

50

60

70

eo

90

(mtnutes)

Figure 7. Portal (0) and peripheral (0) venous serum gastrin concentrations after intragastric infusion of a 10% peptone meal, as measured by radioimmunoassay using gastrin-specific antibody (56-02). Data are expressed as femtomoles per milliliter of serum [mean t SEM).

incubation with the antibody-gel suspension, indicative of virtually complete binding of gastrin and CCK peptides by antibody 06, which, as indicated, is specific for the carboxyl-terminal tetrapeptide amide shared by gastrin and cholecystokinin.

200

Into

Separation

of

Cholecystokininlike the

Portal

Venous

Peptides Circulation

Further definition of released gastrinlike and CCKlike peptides was accomplished by means of gel chromatography, using a 0.9 x zoo-cm column containing Sephadex G-50 superfine (Pharmacia, Fine Chemicals AB, Uppsala, Sweden). Elution was performed at 4°C using 0.02 M sodium barbital buffer, pH 8.4, containing 2% normal human serum, at a flow rate of 4 ml/h. The column was calibrated by application and elution of blue dextran for definition of the void volume and with ‘““INa, CCK33. G34, hG17, cGli’, 4-17hG17, CCK8, and G4 to determine their positions of elution. Recovery of applied peptides was 81.4% * 6.1% as determined by measurement of known quantities of gastrinlike and CCK-like peptides applied to and eluted from the column. Figure 5 illustrates the elution profiles of the peptides used in the characterization of the Sephadex G-50 column. To further characterize the Sephadex G-50 superfine column, CCK33 and CCK8 (250 fmol each) were added to portal venous serum and were applied to and eluted from the column [Figure 6). One-milliliter samples of portal venous serum from each of 5 dogs obtained from 2 to 4 min after peptone meal infusion were applied separately to the column and z-ml fractions were collected. Radioimmunoassay was performed using O.l-ml aliquots of calibration curve stan-

5 0 1f/, 0

and

327

I GosI~IwCCK

175

Antibody

(06)

0 E

Figure 8. Portal (0) and peripheral (0) venous serum concentrations of all peptides containing the carboxyl-terminal tetrapeptide common to gastrin and CCK after intragastric infusion of a lo”/,, peptone meal. as measured by radioimmunoassay using gastrin-CCK antibody (06). Data are expressed as femtomoles per milliliter of serum (mean + SEM). CCK = cholecystokinin.

‘f 125

bO

10

20

30 TIME

40 ( minutes

50

1

60

70

80

90

328

GASTROENTEROLOGY

WOLFE AND McGUIGAN

Vol. 87, No. 2

which it was placed in a boiling waterbath for YImin in 13 x loo-mm glass culture tubes. The tissue \vas again homogenized for 1 min at 4°C. After centrifugation at 10,000 g at 4°C for 30 min, the supernates (containing gut extracts) were removed and frozen at -40°C until subsequent chromatographic separation. At that time l-ml samples from the midantrum and duodenum were applied to the Sephadex G-50 superfine column and eluted at 4°C with 0.02 M sodium barbital buffer, pH 8.4, containing 2% normal human serum. Two-milliliter fractions were collected and each eluted fraction was measured by both the gastrin and gastrin-CCK antibody preparation, as described previously.

200-

175-

Statistical

Analysis

The Student’s t-test for paired values was used to compare portal and peripheral venous serum peptide concentrations after test meal infusion with peptide levels obtained in the basal state. E 3 2

Results

25-

OLA

Serum Concentrations Cholecystokininlike I

0

I

IO 20

,

TIME

Figure

1

30 40

I

50 60

I

90

Wnutes)

dards and the gastrin-CCK antibody preparation, as described previously; fraction samples were examined as 0.3-ml aliquots in a total incubation volume of z ml. In order to measure gastrin with the region-specific gastrin antibody, two consecutive eluted fractions from each individual dog were pooled, and samples were measured by radioimmunoassay, using O.l-ml aliquots of calibration curve standards, as described previously: pooled fractions were examined as l.O-ml aliquots in a total incubation of 2 ml and were assayed in duplicate.

Gel Chromatography

and From

Separation Cholecystokininlike Canine

and

f

70 80

9. Portal (0) and peripheral (C) venous serum CLK cow centrations after intragastric infusion of a lWr peptone meal. Values were obtained by subtracting peptide concentrations obtained by measurement with gastrinspecific antibody (56-02) from those obtained by measurement with gastrin-CCK antibody (06). Data are expressed as femtomoles per milliliter of serum (mean 2 SEMI. CCK = cholecystokinin.

Gastrinlike Extracted

of Gastrinlike Peptides

of Peptides

Gut

The antrum and proximal 10 cm of duodenum were immediately removed from a dog killed by intravenous sodium pentobarbital injection and were placed on dry ice. Peptide extraction was accomplished by boiling tissue samples at neutral pH, as described by Ryder et al.

(8). While still frozen, tissue samples from the midantrum and duodenum (10 cm from the pylorus) were placed in distilled water at 4°C. The tissue was then homogenized in a tissue grinder (Inframo, Wayne, N.J.) for 30 s at 4”C, after

Mean basal portal and peripheral vein serum gastrin concentrations, as determined by radioimmunoassay with gastrin antibody (56-OZ), which measures Gl7, G34, and 4-17G17, but does not significantly detect CCK or CCK fragments, were 8.33 * 2.4 and 6.19 2 0.9 fmol/ml [mean 2 SEM), respectively. Gastric infusion of the peptone meal was followed by biphasic gastrin responses in both the portal and in the peripheral venous sera (Figure 7). The first portal venous serum gastrin peak, 23.33 t 9.5 fmoliml, was almost immediate (2 min); this represented a 188% ? 48% increase over the basal serum gastrin level (p < 0.05). The initial peripheral venous serum gastrin peak was also prompt (4 min) increasing to 15.71 t 1.9 fmol/ml, which was 154% -+ 27% greater than the basal serum gastrin concentration (p < 0.01). The second gastrin peak was noted at 30 min in both portal venous serum (16.67 ? 6.7 fmoliml, increase over basal of 106% 2 42%, p < 0.02) and in peripheral venous serum (13.81 I 3.8 fmoliml, increase over basal of 123% k 66%, p = 0.06). Radioimmunoassay using the gastrin-CCK antibody (06) indicated portal and peripheral serum peptide concentrations that were substantially greater than serum immunoreactive gastrin levels determined by radioimmunoassay using gastrin-specific antibody (56-02). Figure 8 illustrates the responses observed when portal and cephalic venous serum samples from each time point were examined bl radioimmunoassay with gastrin-CCK antibody (06). Mean basal concentrations of peptides containing the common carboxyl-terminal tetrapeptide amide

GASTRIN-

AND

CCK-LIKE

PEPTIDES

IN DOGS

329

this immunoreactive material in the portal venous serum increased virtually immediately, with a sharp peak at 2 min after the peptone meal to 139 ? 37 fmoliml (increase over basal 194% 2 59.1%, p < 0.05). In the peripheral venous serum, the peak was gradual, increasing to a concentration of 78 -C 14 fmoliml at 8 min (increase over basal of 87% 2 25.3%, p < 0.05). A gradual return to baseline by 60 min was observed in both portal and peripheral venous sera. Serum immunoreactive CCK concentrations were estimated by subtraction of serum gastrin from total immunoreactive gastrin-CCK peptides recognized by antibodies with specificity for the carboxyl-terminal tetrapeptide amide common to CCK and gastrin (gastrin-CCK antibody]. Figure 9 illustrates the CCK responses in portal and peripheral blood after infusion of the peptone meal. (Values were obtained by subtracting the peptide concentrations obtained by measurement with antibody 56-02 from those obtained by measurement with antibody 06.) Results depicted in Figure 9 demonstrate immediate monophasic CCK peptide release after intragastric peptone infusion. Elution Profiles of Gastrinlike and Cholecystokininlike Peptides Released the Portal Venous Circulation

20

30

40

ELUTION

Figure

io VOLUME

.30

9b

I60

(%)

10. Elution diagram of Sephadex G-50 superfine column obtained by application of a portal venous serum sample (containing 39.5 fmol of gastrin and 49 fmol of CCK) from one representative dog procured 2-4 min after the peptone meal. Values depicted on the ordinate are concentrations (femtomoles of peptide per milliliter of eluate) of gastrin using gastrin-specific antibody (56-02) (upper panel), gastrin-CCK peptides using gastrin-CCK antibody (06) (middle panel), and CCK, determined by subtracting immunoreactive gastrin from CCK-gastrin peptides (lower panel). Volume is expressed as percentage of elution between blue with dextran (V,) and ““INa. Elution was performed 0.02 M sodium barbital buffer (pH 8.4) containing 2% normal human serum. CCK = cholecystokinin.

were 49 2 10 and 43 + 8.8 fmoliml in portal and peripheral venous sera, respectively. After infusion of the peptone meal, only a single peak response was observed in both circulations. The concentration of

Into

Further definition of portal serum gastrinCCK peptides was accomplished by elution of portal venous serum samples obtained after peptone infusion using Sephadex G-50 superfine gel chromatogthe peptide elution raphy. Figure 10 demonstrates profile of a portal venous serum sample from one representative dog obtained 2-4 min after the peptone meal. Examination of the eluted column fractions with antibody 56-02 (specific for gastrins with minimal recognition for CCK peptides) revealed one major peak at the position corresponding to canine Gl7, with a smaller peak (
330

WOLFE

AND

GASTROENTEROLOGY

McGUIGAN

4.17617

CCK33 ““;I

hGrIcj17

Vol.

87.

No.

2

I-13Gl7 ,

CC/K”

G~l2+

600 GASTRINICCK

PEPTIDES

Figure

CL GASTRIN

L

500-

? t:

400-

<

PEPTIDES

II.

Elution diagram of Sephadex G-50 superfine obtained b!; applying canine antral mucosal extract. The values depicted on the ordinate represent concentrations in femtomoles per milliliter of eluate of peptides immunoreactive with gastrinCCK antibody (06) (upper panel) and gastrin-specific antibody (St?02) (lower panel]. \‘olume is expressed as percentage of elution between blue dextran (V,,) and ‘L’INa. Elution was performed with 0.02 M sodium barbital buffer (pH 8.4) con taining 2% normal human serum. CCK = cholecystokinin.

300

E 5 200

IO (i

0

10

20

30 ELUTION

40

50

VOLUME

70

60

80

90

100

W)

ed column fractions. The profile demonstrates two major peaks corresponding to positions identical for CCK33 and CCK8. (A third, and much smaller, intermediate peak was observed that may represent a CCK fragment somewhat larger in size than CCK8.) EJution Profiles of Gastrinlike and Cholecystokininlike Peptides Extracted Canine Gut

From

Figure 11 compares the elution profiles of peptides extracted from canine antrum using the gastrin antibody (56-02) and using the gastrin-CCK antibody (06). With each antibody the major immunoreactive peptide peak was identified in the position of elution of canine G17, with smaller peaks corresponding to the elution position of G34 and a smaller gastrin peptide in the elution position of CCK8-G4. Examination of eluted column fractions from canine duodenal mucosal extracts, again using

both antisera, revealed an elution profile (Figure 12) that differed from that obtained with examination of antral mucosal extracts. With the gastrin specific antibody (56-02) two peaks were identified: they were located at positions corresponding to G34 (17%) and Gl7 (83%). However, when the gastrinCCK antibody (06) was used, three major peaks of immunoreactivity were demonstrated. They corresponded with elution positions of G34iCCK33, G17, and CCK8.

Discussion In 1964 Gregory and Tracy (9) successfully isolated two gastrin peptides from hog antral mucosa and described their structures. Each of these gastrins was a heptadecapeptide (G17) and differed only in the absence or presence of a sulfated ester on the tyrosine residue at position 12. With the introduction of gastrin radioimmunoassay, an additional

August

GASTRIN-

1984

AND CCK-LIKE PEPTIDES IN DOGS

4.17617

CCK33 6311

vo

hGr

1

331

I-13Gl7 crl7

1

CCIfj

IT”‘“”

250GASTRINICCK

Figure

12. Elution diagram of Sephadex G-X superfine obtained by applying canine duodenal mucosal extract. ‘The values depicted on the ordinate represent concentrations in femtomoles per milliliter of eluate of peptides immunoreactive with gastrin-CCK antibody (06) (upper ponef) and gastrin-specific antibody [56-02) (lon,er panel). 1’01. ume is expressed as percentage of elution between blue dextran (V,,) and “‘INa. Elution was performed with 0.02 M sodium barbital buffer (pH 8.4) containing 2% normal human serum. CCK = cholecvstokinin.

2

PEPTIDES

IOO-

2

W is

::

50-

E G

04

i

c

250-

::

GASTRIN

PEPTIDES

c” El 200-

d 5

150-

0” 2

too-

50-

0

IO

20

30

ELUTION

larger and more basic form of gastrin was demonstrated. This was designated “big gastrin” by Yalow and Berson in 1970 (10). It was shown subsequently to contain 34 amino acids (G34) including the G17 heptadecapeptide covalently linked at its aminoterminus to a structurally distinct heptadecapeptide (11,lZ). As with G17, G34 also exists naturally in both sulfated and nonsulfated forms (2). Although previous studies suggested that the potency of circulating G17 in the stimulation of gastric acid secretion was six to eight times that of G34 (13). more recently the two have been found to be equipotent on a molar basis (14). In antral mucosa, G17 is the predominant form, with only a small proportion of G34 being present (12). With caudal progression down the gastrointestinal tract the ratio of G34 to Cl7 increases (2), although the total amount of immunoreactive gastrin decreases. The relationship between G34 and Gl7 remains speculative. Gl7 has been proposed as a potential storage form of gastrin,

40

50

VOLUME

60

70

60

90

100

t%)

with conversion to G34 required for secretion. Another theory suggests that both are derived by hydrolytic cleavage from a common and larger prohormone (15). Other circulating forms of gastrin have been described, which occur in low concentrations or do not increase postprandially (16-18). The carboxyl-terminal tetradecapeptide amide (Cl4 or 4-l7Gl7) and the amino-terminal tridecapeptide (l-13G17) of gastrin have been found in serum and tumor extracts from patients with gastrinoma (19-21). Small amounts of l-l3Gl7 have also been found in human serum after a mixed meal (22). Whereas, Gl4 is present in the sera of cats in significant quantities, its concentration is extremely low in the sera of other species. including humans (23,24). High resolution gel and ion-exchange chromatography have suggested the presence of at least 20 different peptide forms immunoreactive with antibodies to gastrin. most in very small quantities, in normal human serum (25).

332

WOLFE AND McGLJIGAN

Rehfeld (26) has reported that the carboxyl-terminal tetrapeptide that is common to both gastrin and CCK was present in large quantities as the dominant gastrin-CCK peptide in mucosal extracts from porcine antrum and small intestine. Ryder et al. (8), however, reported that the common carboxyl-terminal tetrapeptide amide (G4 and CCK4) was not found in significant quantities in extracts of rat small intestinal mucosa, which contained principally CCK33 and small CCKlike peptides (larger than CCK4). Recently Gregory et al. (27) examined porcine antral mucosa using ionexchange and high performance liquid chromatography. In their studies, no carboxyl-terminal tetrapeptide could be unequivocally identified; however, -2% of extracted gastrin eluted in a position identical to the carboxyl-terminal hexapeptide of gastrin (27). The proposed presence of small gastrin and CCK peptides in antral and small intestinal mucosa is of interest, as gastrin and CCK released into the splanchnic venous circulation in response to physiologic stimuli must first enter the portal blood and traverse the liver before gaining access to the systemic circulation. It has been shown in dogs that gastrin fragments with eight or fewer amino acid residues are >90% inactivated by hepatic transit (3). Thus, if small gastrin and CCK peptides were released from antral and duodenal mucosa and reached the portal venous circulation, efficient hepatic clearance would preclude their significant contribution to gastric acid secretion and other proposed physiologic functions, unless extremely large amounts of the peptides were released into the splanchnic venous effluent. Cholecystokinin and gastrin share the same carboxyl-terminal pentapeptide amide (28) and, analagous to the molecular heterogeneity of gastrin, CCK is also known to exist in tissues in at least three recognized forms, CCK39. CCK33, and CCK8 (26), and probably in additional forms, including a form larger than CCK39, recently described as CCK58 (29). and a form slightly larger than CCK8 (estimated to be CCK12) (26,28-30). Development of accurate, reliable, and direct radioimmunoassay techniques for CCK have proven difficult. Reasons for this difficulty include the fragility of CCK during the oxidative conditions of radioiodination, impedence of radioiodination by sulfation of the tyrosyl residue located at position seven (from the carboxyl-terminus), and the structure of CCK itself. Antibodies produced by immunization with intact porcine CCK, by convention CCK33, demonstrate immunologic specificity for either the carboxyl-terminal region or for various portions of the amino-terminal peptide region of the CCK molecule that differs in its amino acid sequence from gastrin. Antibodies with specificity for the amino-terminal peptide region of porcine CCK do not bind effectively to nonporcine CCK molecules. This observation strongly supports the probability

GASTROENTEROLOGY

Vol. 87, No. z

that the amino acid sequence of the amino-terminal region of porcine CCK differs substantially from that of various nonporcine species (including human, dog, and rat). Antibodies with specificity for the carboxyl-terminal portion of CCK bind and measure gastrin approximately as well as CCK. Several investigators have been successful in determining CCK concentrations by radioimmunoassay both directly by using antibodies with varying degrees of specificity for CCK (31~34), and indirectly by using antibodies with selective specificity for limited regions of gastrin/CCK peptides (8,35,36), as was done in the present study. The major physiologic stimulus to release of gastrin is feeding, particularly of meals containing protein and products of protein digestion. The results depicted in Figure 7 demonstrate clearly that after a peptone meal, release of gastrin (as measured by the gastrin antibody) was almost immediate and was biphasic. In contrast, radioimmunoassay measurement of all peptides containing the carboxyl-terminal tetrapeptide amide common to gastrin and CCK (Figure 8) yielded strikingly different responses from those obtained with measurement using the gastrinspecific antibody. After infusion of the peptone meal, when the gastrin-CCK antibody (06) was used only a single sharp immunofor radioimmunoassay, reactive peptide peak was noted at 2 nun in portal venous serum. (A more gradual peak occurred at 8 min in peripheral venous serum.) Sephadex gel chromatography was used to further characterize the peptides containing the carboxylterminal tetrapeptide amide common to gastrin and CCK that were released in response to the peptone meal; G17, G34, and Gl4 gastrins were excluded as principal components because peptides in this peak, immunoreactive with the gastrin-CCK antibody, failed to react with the gastrin-specific antibody (5602) that readily measures these peptides (on an equimolar basis). The amino-terminal tridecapeptide portion of gastrin (l-13Gl7) was excluded as a contributor to the large peptide peak detected by the gastrin-CCK antibody, as it does not contain the common carboxyl-terminal tetrapeptide amide and is not immunoreactive with either the gastrin-CCK antibody (OS) or the gastrin-specific antibody (5602). Thus, the principal component(s) of the peptide peak released into the portal circulation of the dog after the peptone meal that shares the common gastrin-CCK carboxyl-terminal tetrapeptide amide does not contain gastrin-specific antigcnic determinants. The most probable remaining known candidates were intact CCK (CCK33 or CCK39, or both), CCK8, and other small CCK and gastrin peptides. Cholectystokinin peptides in the portal venous serum peak (2-4 min after the peptone meal). when applied to the Sephadex G-50 superfine column,

August

1984

GASTRIN-

eluted with two major peaks: one was in the position of elution of CCK33 and the second was in the elution position of CCK8. A third, much smaller peak, representing a CCK fragment somewhat larger than CCK8 (possibly CCKl2) was also identified (Figure 10). This elution profile of portal venous CCK-like peptides is similar to the profile that was obtained with extracts of canine duodenal mucosa (Figure 12), in which the two major eluted peptide peaks correspond to CCK33 and CCK8. Examination of eluted Sephadex G-50 column fractions with the gastrin-specific antibody (56-02) indicated that Gl7 accounted for >90% of immunoreactive gastrin released initially after the peptone meal in the dog. This correlates with the elution profiles obtained with both canine antral and duodenal mucosal extracts, in which the majority (809& 90%) of gastrin peptide eluted in the region of canine G17, with smaller amounts eluting in the position of G34 (Figures 11 and 12). In addition, a small peak of immunoreactivity (<5%) representing a gastrin fragment somewhat larger than the carboxyl-terminal tetrapeptide (possibly G6) was identified in antral mucosa (Figure 11. lower panel). We found that canine Gl7 eluted from the Sephadex column somewhat later than human G17 (Figure 5), perhaps due, at least in part, to differences in net electrical charge resulting from alanine substitution in canine G17 for a glutamic acid residue in human G17 (37). These findings are consistent with those recently obtained by Dockray et al. (38), who fractionated canine portal venous plasma and canine antral mucosa extracts on a Sephadex G-50 superfine column and found that the majority of gastrin eluted slightly beyond the elution position of human G17. The predominance of G17 in portal venous serum immediately after the peptone meal (2-4 min) supports the hypothesis that antral mucosal G17 may constitute an “acute, releasable pool” of gastrin (39). In conclusion, this study in the alert, nonanesthetized dog demonstrates that after a peptone meal (a) gastrin release is almost immediate and is biphasic: (b) there is almost immediate monophasic release of larger quantities of peptides that contain the carboxvl-terminal tetrapeptide amide common to gastrin and CCK, but lack antigenic determinants specific for gastrin(s): and (c) Sephadex gel chromatography studies support the conclusion that these peptides are primarily intact CCK33 and CCK8.

References 1. McGuigan )I<. Gastrin. Vitam Harm 2. Strrmz UT. Thompson MR. Elashoff inactivation of gastrins of various Gastroenterc,lng!; 1978:74:550-3.

CCK-LIKE

PEPTIDES

IN DOGS

333

3. Korman MG, Soveny C. Hansky J. Effect of food on serum gastrin by radioimmunoassay. Gut t971:12:619-24. 4. McGuigan JE, Wolfe MM. Gastrin radioimmunoassav, Clin Chem 1982;28:368-73. JE. Antibodies to the C-terminal tetrapeptide am5. McGuigan ide of gastrin-assessment of antibody binding to cholecystokinin-pancreozymin. Gastroenterology 1968:54:1012-7. FC. Preparation of iodine-131 labelled 6. Hunter WM. Greenwood human growth hormone of high specific activity. Nature 1962; 194:495-6. 7. Rehfeld JF, Stadil F. Rubin B. Production and evaluation of antibodies for the radioimmnnoassa!, of gastrin Stand 1Clin Lab Invest 1972;30:221-32. extraction and 8. Ryder S. Eng J, Straus E, Yalow KS. Alkaline characterization of choler:~stokinin-immurloreactivit! from rat gut. Gastroenterology 1981;81:267-75. 9. Gregory RA, Tracy HJ. The constitution and properties of two gastrins extracted from hog antral mucosa. Gut 1964:5:10317. 10.

11.

12. 13

14.

15

16.

17 18

19.

20. 21.

Yalow RS. Berson SA. Size and c:harge distinctions between endogenous human plasma gastrin in peripheral blood and heptadecapeptide gastrins. Gastroenterology 1970;58:609~15. Yalow RS. Berson SA. Further studies on the nature of immunoreactive gastrin in human plasma. Gastroentrrolog> 1971:60:203-14. Gregory RA. Tracy HJ. Big gastrin. Mt Sinai ] Med 1973: 40:359-64. L’Valsh JH. Isenberg Jl, Ansfield J. CIearanc.e and acid stimulating action of human big and little gastrins in duodenal ulcer subjects. J Clin Invest 1976:57:1125-3 I. Eysslein VE, Maxwell V, Reddy T. Wuensch E. Walsh JH. Similar acid stimulatory potencies of synthetic human big and little gastrins in humans (abstr). Gastroenterology 1983; 84:1147. Agarwal KL, Noyes BE. Studies on gastrin messenger KNIZan illustrative approach. In: Bloom SK. Polak JM. eds. Gut hormones. Edinburgh: Churchill Livingstone. 1981:49-54. Stadil F. Rehfeld JF, Christiansen LA, MalstrGm J. Patterns of gastrin components in serum during feeding in normal subjects and duodenal ulcer patients. Stand J Gastroenterol 1975;10:862-8. Yalow RS. Berson SA. And IIO~V“big. big” gastrin. Biochem Biophys Res Commun 1971;48:391-5. Rshfeld JF. Schwartz TW. Stadil F. Irnmunoc:h~~mical studies on macromolecular gastrins. Gastroenterology 1977;73:46977. Rehfeld JF, Stadil F. Gel filtration studies on immunoreactive gastrin serum from Zollinger-Ellison patients. Gut 1973;14: 369-73. Gregory RA, Tracy HJ. Isolation of the two minigastrins from Zollinger-Ellison tumour tissue. Gut 1974;15:683-5. Dockray GH. Walsh JH. Identification of a N-terminal fragment of heptadecapeptide gastrin in the serum of patients with the Zollinger-Ellison syndrome (ZES) (abstr). Gastroenterology 1974;66:874.

22.

Dockray GJ. Taylor IL. Heptadecapeptitlc ment in blood by specific. radioimmunoassay. ogy 1976:71:971-7.

23.

Taylor IL. Dockray CJ. (:alarn J. Walker RJ. Big and gastrin responses to food in normal and ulcer subjects. 1979:20:957-62.

24.

Blair EL, Hamill A, Jackson B. Lund PK, Nicholson E. Sanders DJ. The effects of stimulation hy meat on gastrin in pyloric antral mucosa of anesthetized c.ats. J Physiol (Land) 1979: 295:201-15.

25.

Rehfeld IF. Heterogeneity of gastrointestinal hormones. In: Glass. GBJ ed. Gastrointestinal hormones. NRM. York: Raven. 1980:433-49.

1974:32:47?88.

J. Grossman MI. Hepatic. chain lengths in dogs.

AND

gastrin: measureGastroenterollittle (cut

334

WOLFE

AND

McCUIGAN

26. Rehfeld JF. The predominating antral gastrin and intestinal cholecystokinin is the common COOH-terminal tetrapeptide amide. In: Rehfeld JF. Amdrup E. eds. Gas&ins and the vagus. London: Academic, 1979:85-94. 27. Gregory RA. Dockray GJ. Reeve JR. Shively JE. Miller C. Isolation from porcine antral mucosa of a hexapeptide corresponding to the C-terminal sequence of gastrin. Peptides 1983;4:319-23. 28. Mutt V, Jorpes JE. Isolation of aspartyl-phenylalanine amide from cholecystokinin-pancreozymin. Biochem Biophys Res Commun 1967:26:392. 29. Epsslein VE, Reeve jR. Shivelg JE. Hawke D. \Valsh JH. Partial structure of a large canine cholecystokinin (CCK58): amino acid sequence. Peptides 1982;3:687-91. 30. McCuigan JE. Immunochemical studies with synthetic human gastrin. Gastroenterology 1968;54:1005-11. 31. Byrnes DJ, Henderson L, Borody T, Rehfeld JF. Radioimmunoassay of cholecystokinin in human plasma. Clin Chim Acta 1981;111:81-9. 32. Burhol PG, Jenssen TH. Lygren 1. Schulz TB. Jorde R, Waldum HL. Iodination with Iodo-gen and radioimmunoassay of cholecystokinin (CCK) in acidified plasma. CCK release. and

GASTROENTEROLOGY

33. 34. 35.

36.

37. 38.

39.

Vol.

87. No. 2

molecular CCK components in man. Digestion 1982:23:15668. Jansen JBMJ, Lamers CBHW. Bombesin releases cholecystokinin in man (abstr). Gastroenterology 1982:82:1093. Chang T-M. Chey WY. Radioimmunoassay of cholecystokinin. Dig Dis Sci 1983:28:456-68. Calam J, Ellis A. Dockrag GJ. Identification and measurement of molecular variants of cholecvstokinin in duodenal mucosa and plasma. J Clin Invest 1982;69:218-25. Walsh JH. Lamers CB. Valenzuela JE. Cholecystokinin-octapeptidelike immunoreactivity in human plasma. Gastroenterology 1982:82:438-44. Gregory RA. A review of some recent developments in the chemistry of the gastrins. Biol Chem 1979;8:497-511. Dockray GJ, Gregory RA, Tracy HJ, Zhu W-Y. Postsecretory processing of heptadecapeptide gastrin: conversion to C-terminal immunoreactive fragments in the circulation of the dog. Gastroenterology 1982;83:224-32. MalstrGm J. Stadil F. Gastrin content and gastrin releasestudies on the antral content of gastrin and its release to serum during stimulation by food. Stand J Gastroenterol 1976:37(Suppl):71-6.