Molecular and Cellular Biochemistry 112: 163-171, 1992. © 1992 Kluwer Academic Publishers. Printed in the Netherlands.

Characterization of gastrin binding to colonic mucosal membranes of guinea pigs

Satya Narayan, Louis Chicione and Pomila Singh Department of Surgery, The University of Texas Medical Branch, Galveston, Texas 77550, USA

Received 20 August 1991; accepted 15 January 1992

Abstract

Gastrin has significant growth and metabolic effects on colonic mucosal cells. It is, however, not known if gastrin receptors are present on colonic mucosal cells that may directly mediate the reported biological effects of gastrin. In the present studies, the presence of specific gastrin binding sites on colonic mucosal membranes was investigated and the binding sites were further characterized. Crude membranes from colonic mucosa of guinea pigs were analyzed for specific binding to gastrin by our published procedures. A significant number (14.7 _ 1.8 fmoles/mg protein) of high affinity gastrin binding sites (Kd = 0.49 +_ 0.05 mM) were measured, that were specific for binding gastrin/CCK related peptides and demonstrated negligible binding affinity for all other unrelated peptides examined. In addition a large number of low-affinity (Kd = -- 1/zM) binding sites were present. In order to further characterize the molecular size of gastrin binding proteins, we used the chemical cross-linking methods, and observed at least four bands of gastrin binding proteins (GBPs) ( - 3 3 , 45, 80 and 250 KDa), both under reducing and non-reducing conditions, indicating that these proteins were not sub-units of forms linked by disulfide bonds. Interestingly, majority of the specific gastrin binding sites ( - 70%) were present on the 45 KDa protein, unlike other target cells of gastrin. The presence of N- and O-linked glycosylated moieties were indicated on the 45 KDa protein, based on enzymatic de-glycosylation studies. The relative binding affinity (RBA) of gastrin and a closely related peptide, cholecystokinin octapeptide (CCK), for GBPs on colonic mucosal membranes was measured in order to determine if GBPs were similar to the CCK-A or CCK-B binding proteins reported in literature. The RBA of gastrin and CCK for displacing the binding of gastrin to the 33, 45, 80 and 250 KDa GBPs on colonic mucosal membranes were calculated to be 39, 100, 78 and 70% (gastrin), and 5.4, 2.9, 3.9 and 2.0% (CCK), respectively, wherein the binding affinity of gastrin for the 45 KDa protein was arbitrarily taken as 100%. Based on the RBA values, it appears more likely that the GBPs on colonic mucosal membranes are more akin to the unique GBPs described on colon cancer cells, and do not represent either the CCK-A or CCK-B binding sites. Based on the cross-linking studies we were not able to determine if the high- and low-affinity binding sites were differentially distributed on the different molecular forms of GBPs measured on the colonic mucosal membranes. The above studies thus indicate for the first time that specific gastrin binding proteins (receptors) are present on colonic mucosal membranes and that these receptor proteins may be directly mediating the observed effects of gastrin on colonic mucosal cells. (Mol Cell Biochem 112: 163-171, 1992)

Key words: gastrin, receptors, colon

Address for offprints: P. Singh, University of Texas Medical Branch, Department of Surgery, 6.202 Old John Sealy Rt. E32, Galveston, Texas 77550, USA

164

Introduction

The gastrointestinal peptide, gastrin, has significant se- cretory and growth stimulatory effects on the mucosal cells lining the gastrointestinal tract [1]. While specific, high affinity binding sites for gastrin have been demon- strated on the mucosal cells from the fundic stomach [2-5], no information is available on gastrin binding to colonic mucosal cells. The growth promoting and secre- tory effects of gastrin in the stomach are believed to be mediated via direct interaction of gastrin with the spe- cific high affinity binding sites present on the fundic mucosal cells [1, 2, 6]. Gastrin receptor proteins of at least two or more molecular sizes, with molecular mass ranging from 30-200 KDa, have been described on the fundic mucosal and pancreatic membranes of rodent and porcine species [5, 7, 8] with the major form being 70-80 KDa. Only 70-75 KDa gastrin binding proteins (GBP) were measured on parietal cells from canine stomach [9], indicating that different molecular forms of the receptor proteins may be present on different target cells in the gastrointestinal mucosa. This possibil- ity is highlighted by the fact that colon cancer cells (which perhaps represent a more homogeneous pop- ulation of growth responsive target cells), demonstrate the presence of only - 3 3 KDa GBPs [10], versus the presence of multiple molecular forms of GBPs on nor- mal target tissues. Freshly dispersed colonic mucosal cells were reported to respond to the growth effects of gastrin, in vitro [11]. In colonic mucosal explants, gas- trin stimulated the activity of tyrosine kinase and or- nithine decarboxylase [12]. Increase in protein synthesis by colonic mucosal cells, in response to gastrin, was also recently reported [13], indicating that gastrin receptors may be present on colonic mucosal cells that may be directly mediating the biological effects of gastrin. The purpose of the present study was therefore to demon- strate the presence of specific gastrin binding sites on colonic mucosal membranes and to further characterize the binding proteins.

While several molecular sizes of GBP on fundic mu- cosal cells have been defined, it is not known if binding affinity and binding specificity of these different forms of GBPs are similar or different, and if the proteins are glycosylated. It is of particular importance to determine if cholecystokinin (CCK), a closely related gastrointes- tinal peptide hormone, that shares the C-terminal tetra- peptide with gastrin, demonstrates a similar or a differ- ential binding affinity for the different molecular forms

of GBPs. In the present studies, the relative binding affinity of gastrin and CCK to the gastrin binding pro- teins on the colonic mucosal membranes was therefore further examined, and the possible presence of N-link- ed glycosylated moieties on the binding proteins deter- mined.

Materials and methods

Collection of tissue Male, young guinea pigs (150-200 g body weight), were obtained from Harlan Rich Glo, El Campo, TX, and fed purina chow and water ad libitum. The guinea pigs were maintained at 22°C for one week in dark/light cycles of 12 h each. They were killed by decapitation and the full length colon immediately excised, cut open along the length, cleaned and washed in ice-cold tissue collection buffer A (10 mM Tris, 137 mM NaCI, 5 mM KC1, 2 mM CaCI2, 2.5 mM MgC1,_ and 0.25 M sucrose, pH 7.4) containing 0.1% bovine serum albumin (BSA), 0.1% soybean trypsin inhibitor, 0.1% bacitracin and l mM phenylmethylsulfonyl fluoride (PMSF). After washing the mucosal side thoroughly, the mucosa was scraped with the help of ice-cold glass slides, and the colonic mucosa immediately frozen in liquid nitrogen. Crude membranes were prepared from colon mucosa as published previously [14].

Binding of gastrin to mucosal membranes Synthetic human gastrin and human gastrin-(2-17)-I (Research Plus, Bayone, N J) were iodinated with Iodo- Gen (Pierce Laboratory, Rockford, IL) as described previously [4], and specific activity calculated to be

1,000-2,000 cpm/fmole. For the purpose of measur- ing binding affinity and specificity of GBP, aliquots of membrane suspension were incubated with either in- creasing concentrations of [125I)-gastrin + 1000 fold ex- cess of unlabeled gastrin for 30 min at 30 ° C in binding buffer B (25mM HEPES, 2.5mM MgC12, 0.1% BSA, 0.1% soybean trypsin inhibitor, o.1% bacitracin and 100units/ml of Trasylol and lmM PMSF, pH7.4), or 0.5-1.0nM concentration of 125I-gastrin + increasing concentrations (1 nM to 10/xM) of unlabeled peptides (gastrin, CCK, bombesin, insulin and luteinizing hor- mone releasing hormone [LHRH]). For the purpose of cross-linking studies, aliquots of membrane suspension

in Buffer B, containing 150/xg protein, were incubated with 1-10pmoles t2SI-gastrin [2-17] + increasing con- centration of either gastrin or CCK for 30 min at 30 ° C. At the end of the reaction, the binding assay tubes were chilled on ice and excess ice cold Buffer B added. The peptide hormone bound to the membrane was quickly separated from excess unbound hormone by centrifu- gation in a microcentrifuge (Model 235B, Fisher Scien- tific, Houston, TX) at 4°C and the membrane pellet washed two times with the binding buffer in the pres- ence (affinity measurement studies) or absence (cross- linking studies) of BSA. The washed samples from the affinity measurement experiments were counted for [125I] in a gamma counter (Beckman Minaxq 5000 series, Houston, TX) with ~ 75% counting efficiency. Mem- branes labeled with p25I] gastrin [2-17] were subjected to cross-linking as published previously [10], and briefly described below.

Molecular weight determination of GBP by cross-linking studies Aliquots (1 ml) of membranes labeled with [125I]-gastrin [2-17], as given above, were incubated with 1 mM dis- uccinimidyl suberate (DSS, Sigma) for 15 min at 22 ° C. DSS was dissolved in dimethyl sulfoxide (DMSO), im- mediately prior to use and added to the substrate to give a final concentration of 1% DMSO in the binding buff- er. The reaction was terminated by rapid centrifuga- tion. The membranes were washed with excess binding buffer, devoid of BSA. The washed membrane pellets in the reaction tubes were solubilized by boiling at 100°C for 3-5min in 0.2M Tris HCI buffer, pH6.8, containing 6% sodium dodecyl sulfate (SDS, w/v), 2mM EDTA, 10% glycerol (v/v) in the presence or absence of 4% [3-mercaptoethanol (v/v). The super- natant was then subjected to SDS polyacrylamide gel electrophoresis according to the method of Laemmli [15] using 9.5% acrylamide in the separating gel and 4% acrylamide in the stacking gel. The gels were stained, destained, dried and exposed to Kodak XAR-5 film for 7-14 days at -70°C. The autoradiograms were scanned using a double-beam densitometer and the ar- ea under the peaks measured with a Hewlett-Packard digitizer.

Enzymatic de-glycosylation studies Solubilized membrane proteins, cross-linked to [125I]- gastrin [2-17] as given above were run on SDS-PAGE under reducing conditions and the 45 KDa and 33 KDa protein bands electroeluted separately using Centriul-

165

tot (Amicon Division, Danvers, MA). The protein con- centration of the electroelute was measured and the samples were treated with or without glycosidases as given below. In the present study, the three glycopepti- dases used were: 1) Endo-c~-N-acetylgalactosaminidase (Endo-N-Gal, Boehringer Mannheim Biochemicals, Indianapolis, IN), which liberates the disaccharide Gal ([31--+3) GalNac from O-glycans, which is bound to serine or threonine as a common core unit; 2) Endogly- cosidase H (Endo-H, Boehringer Mannheim Biochem- icals), which preferentially hydrolyzes N-glycans of the high mannose type; and 3) N-Glycanase (N-GIy, Gen- zyme Corporation, Boston, MA), which hydrolyzes as- paragine-linked oligosaccharides from glycoproteins and glycopeptides to give free oligosaccharide and a peptide/protein containing aspartic acid at the glycosy- lation site. The enzymatic de-glycosylation was carried as per the instructions of the manufacturers using opti- mal enzyme concentration and the suggested buffers for 18h at 37 °C. The reactions were stopped by adding 50/xl of SDS-sample buffer and boiling for 3 rain. The samples were re-analyzed on SDS-PAGE under reduc- ing conditions as described above and autoradiograph- ically examined.

Protein measurements Protein was measured by the method of Lowry et al. [16].

Results

Binding of gastrin to colonic mucosal membranes Specific binding of [12sI]-gastrin to crude membranes prepared from colonic mucosa was examined as a func- tion of dose (Fig. 1A). Specifically bound gastrin in- creased linearly with increasing doses of [~25I]-gastrin to

0.5-1.0nM concentration and appeared to plateau off after that. At doses higher than 2 nM, the concentra- tion of specifically bound gastrin once again increased steeply, in a dose-dependent manner, and the binding sites were apparently unsaturated even at 10 nM gastrin (Fig. 1A). Based on these results, it appears that there are two classes of gastrin binding sites on colonic muco- sal membranes, arbitrarily named Type I and Type II sites. Type I sites were apparently saturated at

1-2 nM gastrin, while Type II sites remained unsat- urated even at 10 nM gastrin (Fig. 1A). Specific gastrin binding data to Type I sites from Fig. 1A was re-plotted as a Scatchard plot (Fig. 1B), and equilibrium dissocia-

166

60-

50-

i --

":" 40-

. .L

.0"--

20- c}.. o9

10-

A

0 . 0 0 4

B / / /

0.003- ~ . ~ ttt " - o , . /

o oo2- \k, '

I~3 0.001-

o , , "~

\\\\ [ I I I ' I 1 \ ~ - 1 2 3 4 5 6 \ \

0 - - 1 0 10

'~Sl-gastrin (nM) Fig. 1. Specific binding of [l-~5I]-gastrin to guinea pig colonic mucosal membranes. Membrane aliquots (- 200 mg) were incubated with increasing concentrations of [~25I]-gastrin + 1000 fold excess of unlabeled gastrin at 30°C for 30 min. At the end of the incubation, fmoles of [)25I]-gastrin, specifically bound were determined and plotted against the M concentration of [~-5I]-gastrin used for incubation (A). Each point is mean of two observations from one experiment and is representative of three different experiments, Specific binding data from Fig. 1A, to high-affinity binding sites (apparently saturated by 1-2 nM gastrin), were re-plotted as a Scatchard plot (B), as described under the text.

tion constant (Kd) determined to be 0.49_+ 0.05nM, (n = 6), with a binding capacity of 14.7 + 1.8 fmoles/mg protein. In order to determine the concentration range of gastrin required to displace the specific binding of [L25I]-gastrin to Type I vs. Type II sites, colonic mucosal membranes were incubated with 0.5 nM [125I]-gastrin + increasing concentrations of gastrin (10-1°-10-4M), and the % inhibition of [125I]-gastrin binding plotted as a function of the log-dose of gastrin used for displacing maximum binding of [L2sl]-gastrin (Fig. 2). The binding of 0.5 nM [125I]-gastrin to Type I sites was apparently displaced in the dose-range of 10-J°-10-9M, while the binding of Type II sites was apparently displaced linea- rly in the dose-range of 10-7-10 -5 M. Thus in order to define the binding affinity of Type II sites, the above experiment was repeated by using a 10 point displace- ment curve in the dose-range of 10-s-10 -3 M, and the percent inhibition of [125I]-gastrin binding plotted as a function of the log-dose of gastrin used for displacing maximum binding of [125I]-gastrin (Fig. 3A). Since the

high-affinity (Type I) sites can be expected to be com- pletely displaced by doses < 10 nM gastrin, it is presum- ed that the displacement curve in Fig. 3A essentially represents displacement of Type II sites. As a corollary, specific binding presented in Fig. 1, at concentrations < 1 nM gastrin, can be expected to represent essentially Type I sites, since gastrin concentrations used were too low for occupying low affinity, Type II sites. The linear part of the displacement curve in Fig. 3A was re-plotted as a Scatchard plot (Fig. 3B) and the equilibrium dis- sociation constant of Type 1I binding sites determined to be 2,6 ± 0.3 ~M, with a binding capacity of 90__+ 17 pmoles/mg protein (n = 4).

Specificity of gastrin binding sites In order to determine the specificity of the gastrin bind- ing sites, the binding of [~25I]-gastrin to the colonic mu- cosal membranes, was measured in the presence or absence of increasing concentrations of various peptide hormones, as shown in Fig. 3A. Other than the closely

167

• .~20 4o

o=~ 60

• 80

100 lO

Gastrin (-Log)

Fig. 2. Dose-dependent inhibition of ['25I]-gastrin binding to colonic mucosal membranes. Membrane aliquots were incubated with l n M [~25I]-gastrin +_ increasing concentrations of gastrin, and the data plotted as percent inhibition of specific binding in relation to log M dose of gastrin used. The data represents mean of three observations from one experiment, wherein the variation between the data was < 5%.

related peptide, CCK, no other peptide tested (includ- ing bombesin, insulin, LHRH) demonstrated any sig- nificant binding affinity for the gastrin binding sites, indicating that the binding sites being measured by us were specific for the gastrin/CCK-like peptides.

Molecular weight of gastrin binding proteins The molecular weight of GBPs on colonic mucosal membranes was determined by cross-linking of [1251]- gastrin [2-17] to the colonic mucosal membranes under optimal conditions (as described under Materials and methods), in the presence or absence of increasing con- centrations of gastrin or CCK. The total specific bind- ing of - 1 pmole [125I]-gastrin was completely displaced in the presence of 2000 fold excess unlabeled gastrin, with EDs~ of 0.2nmoles. GBPs, cross-linked to [125I]- gastrin [2-17], separated out into four distinct bands on 9.5% polyacrylamide gels, as can be seen from the autoradiograms presented in Fig. 4A. The molecuar mass of the four bands was calculated to be ~ 33, - 45,

80 and - 2 5 0 KDa. The majority of gastrin binding sites (50 + 6%) were present on the 45KDa GBP, followed by 21 + 2, 19 _+ 1, and 9 _+ 1% on the 250, 33 and 80 KDa proteins, respectively (n = 3). The aut- oradiographic profile was similar under reducing and non-reducing conditions (Fig. 4A), indicating that the binding proteins did not represent sub-unit proteins linked by disulfide bonds. The relative binding affinity (RBA) of gastrin and CCK, for displacing binding of ['25I]-gastrin [2-17] to the four bands was determined by constituting displacement curves, based on densitomet- ric analysis of autoradiograms (presented as bar graphs in Fig. 4B), in the presence or absence of the various concentrations of gastrin and CCK. The RBA of gastrin

and CCK for displacing the binding of [12sI]-gastrin [2- 17] to the four bands of GBPs was calculated from the linear part of the displacement curves, computated from the data presented in Fig. 4B. The RBA of gastrin and CCK for displacing the binding of gastrin to the 33, 45, 80 and 250 KDa GBPs were calculated to be 39,100, 78 and 70% (gastrin), and 5.4, 2.9, 3.9 and 2.0% (CCK), respectively, wherein the binding affinity of gastrin for the 45 KDa protein was arbitrarily assigned 100%. Gastrin was therefore most potent towards dis- placing the binding of [125I]-gastrin to the 45 KDa pro- tein, while CCK was least potent towards displacing [125I]-gastrin binding to 45 KDa protein; CCK, on an average, was -15 -30 x less potent than gastrin to- wards displacing the binding of [125I]-gastrin [2-17] to the four protein bands, on a molar basis.

Glycosylation of gastrin binding proteins The two major gastrin binding proteins, 45 KDa and 33 KDa forms, were further analyzed for the possible presence of glycosylated moieties using three glycopep- tidases as described under Methods. None of the glyco- peptidases used in the present study resulted in any significant shift in the molecular weight of either the 33 or the 45KDa proteins (Figs 5A and B). However, there was a significant loss (> 50%) in the intensity of the 45 KDa protein on treatment with either of the three glycosidases (Fig. 5B), indicating that the 45 KDa pro- tein probably represented a glycoprotein that has both O- and N-linked carbohydrate moieties. The 33 KDa protein, on the other hand, did not appear to be a glycoprotein since there was no significant loss in the intensity of the 33 KDa band on treatment with either of the three glycosidases (Fig. 5A).

B

A 0 e..

e -

20 .., "~

i 40

~" E 60

= ,-~ 80 O

-~ 100 I I I

"~ 6 5 4

Peptide Concentration (-Log M)

168

0.005 -

0

0.004 -

0.003 -

0.002 -

0 . 0 0 1 -

0 0 2.5 5.0 7.5

pmoles specifically bound

Fig. 3. Dose-dependent inhibition of [J2~I]-gastrin binding to colonic mucosal membranes. Membrane aliquots were incubated with 1 nM [L~sI]-gastrin _+ increasing concentrations of gastrin (©), CCK (0), bombesin ([]), insulin (~) and LHRH (A) and the data plotted as percent inhibition of specific binding in relation to log M dose of peptides used (A). The data represents mean of two observations from one experiment and is representative of 2-6 separate experi- ments. The data for specific displacement of [t25I]-gastrin binding by unlabeled gastrin, shown in Fig. 3A, was re-plotted as a Scatchard plot (B), in order to determine the binding affinity of the low-affinity (Type II) gastrin binding sites as discussed in the text. Protein con- centration per assay point was 120tzg. The RBA of CCK for [~2~I]- gastrin binding was about 3% as compared to 100% of gastrin.

Discussion

The present studies demonstrate for the first time that a significant number of high affinity binding sites, that are specific for binding gastrin/CCK related peptides, are present on the colonic mucosal membranes of guinea pigs, and that the majority of the sites are present on

45 KDa proteins. Besides the 45 KDa gastrin binding

proteins (GBP) at least three other molecular forms of GBP were observed on the colonic mucosal mem- branes, ranging in mass from - 3 0 - 2 5 0 KDa. In other gastrin responsive target tissues/cells, the molecular mass of GBPs have also been reported to vary from 30-200 KDa [5-10], but the mass of the major band has been reported to be either - 7 0 KDa (fundic mucosal membranes, parietal cells [5, 7], or ~ 33 KDa (mouse and human colon cancer cells [10]). These studies thus indicate that there is a significant variation in the size of the GBPs on different target tissues/cells. The possibil- ity that different molecular forms are expressed by spe- cific target cell populations, is supported by the finding of only - 70 KDa GBP on canine parietal ceils [9], that are known to respond to the secretory effects of gastrin, and only - 3 3 KDa GBP on mouse and human colon cancer cells [10], that respond to growth effects of gas- trin [11, 17-19]. Colonic mucosal cells respond to growth effects of gastrin, in vitro [11-13]. Other possible biological effects of gastrin (including ion transport, absorptive and endocrine effects) on colonic mucosal cells have yet not been examined to our knowledge. It is possible that the 33 KDa binding protein observed on colonic mucosal membranes, in the present study, are present on growth responsive (stem?) cells, while other molecular forms of GBP are present on other cell pop- ulations in the colonic mucosa; this possibility needs to be further investigated by using pure cell populations from colonic mucosa, which to-date has been a difficult task to accomplish.

In the present studies, we observed the presence of two classes of binding sites for gastrin, one with a very high binding affinity (Kd = < 1 nM) and the other with a very low binding affinity (Kd = -- 1-3 tzM), which we arbitrarily described as Type I and Type II sites, respec- tively. Both the binding sites were apparently specific for binding gastrin/CCK-like peptides, and demonstrat- ed no significant binding affinity for all the unrelated peptides examined in the present study. On colon can- cer cell lines [20, 21] and human colon cancers [22], we and others have similarly described the presence of high (Type I) and low (Type lI) affinity gastrin binding sites. To-date only high-affinity gastrin binding sites have been described on normal fundic mucosal cells in the gastrointestinal tracts [2-5]; this is the first study that describes the additional presence of low-affinity sites on at least the normal colonic mucosal cells, similar to that present on colon cancer cells [10-21]. The biological significance of the low-affinity sites remains unknown.

A closely related peptide, cholecystokinin (CCK),

169

_Cs~trin ~ ~C~K-8

250 KDa 70

o,~ 0

°lL L ~: 0 T 1 5t050 I 5t050

CCK-8

T t 5t050 t 5 t050

Peptide Concentration (ninnies/aliquot)

Fig. 4. (A) Autoradiographic profile of [L~q]-gastrin [2-17] binding to colonic mucosal membranes. Membranes were incubated with p2sI]-gastrin [2-17] _+ increasing concentrations of unlabeled gastrin or CCK and cross-linked with l m M DSS as described in Materials and methods. Solubilized membrane proteins were subjected to SDS--PAGE and analyzed autoradiographically. The data is representative of three separate experiments. Lanes a-i = binding under reducing conditions (+ 1 mM DTT) and Lane j = total binding under non-reducing conditions. Lanes b-e = p25I]-gastrin [2-17] binding in the presence of 1,5, 10 and 50 nmoles of gastrin, Lanes f-i = p2sI]-gastrin [2-17] binding in the presence of 1,5, 10 and 50 nmoles of CCK. (B) Inhibition of [~25I]-gastrin binding to different forms of GBPs on colonic mucosal membranes by gastrin and CCK. The data is derived from the densitometric scanning of the autoradiographs shown in A. The percent inhibition of [~25I]-gastrin binding to each molecular form of GBP was determined by considering the densitometric reading of Lane 2 (Fig. 4A) as 100%, and the readings from other lanes were plotted as a percent of it. Data presented are mean _+ SEM from three separate experiments.

has been previously reported to demonstrate an equiv- alent binding affinity for GBPs on parietal cells [3], pancreatic acinar cells [23] and a pancreatic acinar can- cer cell line (AR42J) [24], and these sites were termed CCK-B/gastrin binding sites [25]. CCK-A binding sites reportedly bind CCK with many fold higher binding affinity compared to gastrin and are predominately pre- sent on pancreatic acinar cells [26]. The GBPs on colon cancer cells, on the other hand, demonstrate negligible binding affinity for CCK [10, 27] and several CCK receptor antagonists [27, 28], and therefore represent neither CCK-A nor CCK-B/gastrin binding sites. In the present studies, GBPs on colonic mucosa demonstrated significant binding affinity for CCK, but the relative binding affinity (RBA) of CCK was 15-30 fold lower than that of gastrin, indicating that majority of GBPs on

colonic mucosal cells are similar to that described on colon cancer cells [10].

It was interesting to note in the present studies that the different molecular forms of GBPs on colonic muco- sal cells demonstrated differential binding affinities for gastrin and CCK. The apparent RBA of gastrin for the 33, 45, 80, and 250 KDa binding proteins was in the order of 45 > 80 > 250 > 33, while the RBA of CCK for the four GBPs was in the order of 33 > 80 > 45 > 250, indicating that gastrin had the highest binding affinity for the 45 KDa GBP, while CCK had the high- est affinity for the 33KDa GBP. Based on the dis- placement of [~z5I]-gastrin binding to the cross-linked GBPs by increasing doses of gastrin in the dose-range used in the present studies, it was not possible to deter- mine if the high- and low-affinity sites were prefer- entially present on specific molecular forms of GBPs.

170

Fig. 5. Autoradiograpbic profile of [125I]-gastrin [2-17] binding to 33 KDa (A) and the 45 KDa (B) proteins from the colonic mucosal membranes. The 33 and 45 KDa protein bands, cross-linked to [~25I]-gastrin [2-17], as given above under Fig. 4 were electroeluted from the gels and subjected to enzymatic de-glycosylation with various glycopeptidases as described under Materials and methods. The enzymatically treated samples were re-run on SDS-PAGE under reducing conditions and autoradiographically analyzed. The data is representative of two separate experiments. Lanes 1-4 in Fig. 5A and 5B represent samples that were treated with the following glycopeptidases. Lane 1 = control samples incubated under similar conditions in the absence of glycopeptidases; Lane 2 = samples treated with Endo-aN-acetylgalactosaminidase; Lane 3 = samples treated with Endoglycosidase H; and Lane 4 = samples treated with N-Glycanase.

It, however , appea red likely that the high-affinity sites

were preferent ia l ly present on the 4 5 K D a proteins (based on the R B A measurements ) , while the low- affinity sites were ei ther equal ly distr ibuted amongst all the four forms or preferent ial ly present on the small G B P molecules . Based on our initial de-glycosylat ion

studies, it appears likely that the 45 K D a protein has N- and O- l inked glycosylated moieties, but the exact mole- cular cont r ibut ion of the glycosylated moiet ies to the 45 K D a protein was not clear f rom the present studies. The 33 K D a prote in on the o ther hand did not appear to

be significantly glycosylated, that could be detected in

the present studies. Since the 80 and the 250 K D a pro- teins were < 10-15% of the total GBPs present on the colonic mucosal membranes , it was not possible to con- duct meaningful de-glycosylat ion studies with these prote in bands, and hence the glycosylated nature o f the

80 and 250 K D a proteins remains to be investigated by using more sensitive methodologies in the future.

The present studies thus demons t ra te that a signif- icant n u m b e r of high affinity sites, that bind gastrin/ C C K peptides specifically, are present on the colonic

m u c o s a l cel ls and t h a t t h e m a j o r i t y o f t h e si tes a re

p r e s e n t on a - 45 K D a p r o t e i n . A l a rge n u m b e r o f low

af f in i ty gas t r in b i n d i n g s i tes a r e a lso p r e s e n t . W h i l e t he

p h y s i o l o g i c a l ro l e o f t he l ow-a f f in i t y b i n d i n g si tes, re-

m a i n s u n k n o w n , t h e h igh -a f f i n i t y b i n d i n g s i tes m a y be

i n v o l v e d in m e d i a t i n g t h e b i o l o g i c a l e f fec t s o f gas t r in on

c o l o n i c m u c o s a l cel ls , r e p o r t e d in l i t e r a t u r e [11-13].

Acknowledgements

S u p p o r t e d by a g r a n t f r o m t h e N a t i o n a l I n s t i t u t e s o f

H e a l t h ( C A 38651).

References

1. Johnson LR: Effect of exogenous gut hormones on gastrointesti- nal mucosal growth. Scand J Gastroent Supp117 (74): 8%92, 1982

2. Takeuchi K, Speir GR, Johnson LR: Mucosal gastrin receptor. I. Assay standardization and fulfillment of receptor criteria. Am J Physiol 237: E284-E294, 1979

3. Soil AH, Amirian DA, Thomas LP, Reedy TJ, Elashoff JD: Gastrin receptors on isolated canine parietal cells. J Clin Invest 73: 1434-1447, 1984

4. Singh P, Rae-Venter B, Townsend CM Jr, Khalil T, Thompson JC: Gastrin receptors in normal and malignant gastrointestinal mucosa. Age-associated changes in gastrin receptors. Am J Phy- siol 249: G761-G769, 1985

5. Ramani N, Praissman M: Molecular identification and character- ization of the gastrin receptor in guinea pig gastric glands. En- docrinology 124: 1881-1887, 1989

6. Singh P, Chicione L, Guo Y-S, Narayan S, Rajakumar G, Parekh D, Townsend CM Jr: Gastrointestinal hormone receptors and receptor-regulation. In: JC Thompson, C Cooper, GH Greeley Jr, PL Rayford, P Singh, CM Townsend Jr (eds) Gastrointestinal Endocrinology: Receptor and Post-Receptor Mechanisms. Aca- demic Press, New York, 1990, pp 13-33

7. Baldwin GS, Chandler R, Scanlon DB, Weinstock J: Identifica- tion of gastrin binding protein in porcine gastric mucosal mem- branes by covalent cross-linking with iodinated gastrin 2,17. J Biol Chem 261: 12252-12257, 1986

8. Fourmy D, Zahidi A, Fabre R, Guidet M, Pradayrol L, Ribet A: Receptor for cholecystokinin and gastrin peptides display specific binding properties and are structurally different in guinea-pig and dog pancreas. Eur J Biochem 252: G143-G147, 1987

9. Matsumoto M, Park J, Yamada T: Gastrin receptor character- ization: affinity cross-linking of the gastrin receptor on canine gastric parietal cells. Am J Physiol 252: G143-G147, 1987

10. Chicione L, Narayan S, Townsend CM Jr, Singh P: The presence of a 33-40 KDa gastrin binding protein on human and mouse colon cancer. Biochem Biophys Res Commun 164: 512-519, 1989

11. Sirinek KR, Levine BA, Moyer M: Pentagastrin stimulates in

vitro growth of normal and malignant human colon epithelial cells. Am J Surg 149: 35-39, 1985

171

12. Majumdar APN: Role of tyrosine kinases in gastrin induction of ornithine decarboxylase in colonic mucosa. Am J Physiol 259: G626-G630, 1990

13. Yassin RB, Clearfield HR, Katz SM, Murthy SNS: Gastrin in- duction of mRNA expression in rat colonic epithelium in vitro.

Peptides 12: 63-69, 1991 14. Narayan S, Draviam E, Rajaraman S, Singh P: High-affinity

binding sites for bombesin on mouse colonic mucosal mem- branes. Mol Cell Biochem 106: 31-39, 1991

15. Laemmli UK: Cleavage of structural proteins during the assemb- ly of the head of bacteriophage T4. Nature 227: 680-685, 1970

16. Lowry OH, Rosebrough N J, Farr AL, RandaLl R J: Protein mea- surement with the Folin phenol reagent. J Biol Chem 193: 265- 275, 1951

17. Singh P, Walker JP, Townsend CM Jr, Thompson JC: Role of gastrin and gastrin receptors on the growth of a transplantable mouse colon carcinoma (MC-26), in Balb/c mice. Cancer Res 46: 1612-1616, 1986

18. Watson SA, Durrant LG, Morris DL: Growth-promoting action of gastrin on human colonic and gastric tumor cells cultured in

vitro. Br J Surg 75: 342-345, 1988 19. Guo Y-S, Baijal M, Jin G-F, Townsend CM Jr, Singh P: Growth-

promoting effects of gastrin in vitro: Absence of autocrine ef- fects. In Vitro 26: 871-877, 1990

20. Singh P, Le S, Beauchamp RD, Townsend CM Jr, Thompson JC: Inhibition of pentagastrin-stimulated up-regulations of gastrin receptors and growth of mouse colon tumor (MC-26), in vivo by proglumide, a gastrin receptor antagonist. Cancer Res 47: 5000- 5004, 1987

21. Weinstock J, Baldwin CS: Binding of gastrin~7 to human gastric carcinoma cell lines. Cancer Res 48: 932-937, 1988

22. Upp JR Jr, Singh P, Townsend CM Jr, Thompson JC: The clinical significance of gastrin receptors in human colon cancers. Cancer Res 49: 488-492, 1989

23. Chevner JH, Sutliff VE, Grybowski DM, Jensen RT, Gardner JD: Functionally distinct receptors for cholecystokinin and gas- trin on dispersed chief cells from guinea pig stomach. Am J Physiol 254: G151-G155, 1988

24. Scemama JL, Fourmy D, Zahidi A, Pradyrol L, Susini C, Ribet A: Characterization of gastrin receptors on a rat pancreatic aci- nar cell line (AR42J). A possible model for studying gastrin mediated cell growth and proliferation. Gut 28: 233-236, 1987

25. Lin CW, Holladay MW, Barrett RW, Wolfram CA, Miller TR, Witte D, Kerwin JF Jr, Wagenaar F, Nadzan AM: Distinct requirements for activation of CCK-A and CCK-B/gastrin recep- tors: studies with a c-terminal hydrazide analogue of cholecysto- kinin tetrapeptide (30-33). Mol Pharmacol 36: 881-886, 1989

26. Yu DH, Huang SC, Wank SA, Mantey S, Gardner JC, Jensen RT: Pancreatic receptors for cholecystokinin: evidence for three receptor classes. Am J Physiol 258: G86-G95, 1990

27. Singh P, Kumar S, Townsend CM Jr, Thompson JC: Structural requirements for the binding of gastrin to mouse colon cancer cells. Surg Forum 38: 171-173, 1987

28. Baldwin GS, Casey A, Mantamadiotis T, McBride K, Sizeland AM, Thumwood CM: PCR cloning and sequence of gastrin mRNA from carcinoma cell lines. Biochem Biophys Res Com- mun 170: 691-697, 1990