WO1995021380A1

WO1995021380A1 - Assay for screening of receptor antagonists

Info

Publication number: WO1995021380A1
Application number: PCT/US1995/001375
Authority: WO
Inventors: Graham Sherard Baldwin
Original assignee: Ludwig Institute For Cancer Research
Priority date: 1994-02-02
Filing date: 1995-02-02
Publication date: 1995-08-10

Abstract

The invention provides a method of identifying a compound having the ability to block the low-affinity gastrin-cholecystokinin type C receptor, comprising the step of measuring the ability of said compound to block the binding to gastrin-binding protein of a compound selected from the group consisting of a gastrin-related peptide, a cholecystokinin-related peptide, an antagonist of gastrin or of cholecystokinin, an acyl CoA, an enoyl CoA, an antibody to gastrin, and an antibody to cholecystokinin. Several ways in which the method of the invention can be carried out are described. Thus, the invention provides a general screening method for identifying antagonists of the gastrin-binding protein interaction, which are useful for treatment of diseases involving rapidly proliferating cells, and for controlling gastric acid secretion. Compositions and methods of treatment are also claimed.

Description

ASSAY FOR SCREENING OF RECEPTOR ANTAGONISTS

This invention relates to a method for screening of putative antagonists of binding of gastrin to its receptor, in particular the low-affinity gastrin-cholecystokinin C receptor, more particularly the gastrin-binding protein (GBP), and also of inhibitors of the enzyme activities associated with the GBP.

Compounds having the ability to block binding of gastrin to GBP or to block the enzyme activities of the GBP are useful in control of cellular proliferation, especially in treatment of neoplastic disease, and are also useful in the control of acid production in the stomach.

BACKGROUND OF THE INVENTION

The hormone gastrin has been known for over 80 years to be a stimulant of acid secretion by the parietal cells of the stomach. Gastrin is a small polypeptide hormone which participates in a feedback system for regulating the environment in the lumen of the alimentary tract, and is thought to affect other target organs, including the brain and pancreas. More recently, it has been recognised that gastrin also acts as a growth factor for the gastric mucosa. Both of these functions are presumably mediated by the binding of gastrin to specific receptors on the target cells in the gastrointestinal mucosa. A variety of methods for detection of such

receptors in tissue, cell, and membrane preparations from various tissues has been demonstrated.

Gastrin/CCK-C receptors have been identified on cell lines of a variety of tumour cell lines of epithelial and non-epithelial origin, and on several cell lines of lymphocytic origin. In particular, significant levels of gastrin/CCK-C receptors have been observed with cell lines established from carcinomas of the stomach, colon, breast, and vulva, and from malignant melanoma: in addition, lower levels of binding were observed with T- and B-cell lymphoma cell lines, while still smaller levels of binding were observed with promyelocytiσ and myeloid leukaemia cell lines (3).

Three classes of receptors for gastrin and for the related hormone cholecystokinin (CCK) have been described. The CCK-A receptor on the pancreatic acinar cell, and the gastrin/CCK-B receptor in the gastric mucosa and brain, both belong to the family of receptors with 7 transmembrane segments. These two receptors can be readily distinguished with the CCK-A receptor-selective antagonist 1.364, 718, and the gastrin/CCK-B receptor-selective

antagonist 1.365,260. Characteristics of the three types of receptors are summarised in Table 1 below.

Evidence that a gastrin-like peptide acts as an autocrine growth factor in colorectal carcinoma has been accumulating steadily. Gastrin mRNA has been demonstrated in both normal and neoplastic colorectal mucosa by Northern blotting, and in colon carcinoma cell lines by the

polymerase chain reaction. An increase in progastrin production in colorectal tumour tissue compared with normal mucosa has been demonstrated in tumour sections by

immunohistochemistry, and in tissue extracts by

radioimmunoassay. Progastrin, but not mature amidated gastrin, was also produced by all (5/5) colonic carcinoma cell lines tested. Exogenous gastrin₁₇ or its derivatives enhanced the growth of xenografts of 60% of colon

carcinomas tested. However, many colon carcinoma cell lines do not increase either DNA or protein synthesis in vitro in response to exogenous gastrin, perhaps because they are already maximally stimulated by an autocrine gastrin-like peptide. In some cases previously

unresponsive cells have been rendered responsive by

synchronisation.

The most convincing evidence for an autocrine role for a gastrin-like peptide has been provided by studies with gastrin/cholecystokinin (CCK) receptor

agonists and antagonists. The antagonist proglumide, which does not discriminate clearly between the three known classes of gastrin/CCK receptors, increases the survival time of mice bearing tumours derived from the mouse colon carcinoma line MC26 (1). Proglumide also reverses the increase in tumour volume observed in tumour-bearing mice after treatment with pentagastrin. Proliferation of 6 colon carcinoma cell lines in vitro is inhibited by the non-selective antagonists proglumide and benzotript, with IC₅₀ values in the mM range (2). Furthermore, in the case of the HCT 116 cell line proliferation is also inhibited by an anti-gastrin antiserum (2). The high concentrations of gastrin required to reverse inhibition by both antagonists and antibodies suggest that neither pancreatic CCK-A nor gastric gastrin/CCK-B receptors are involved, but are consistent with the involvement of the low affinity

gastrin/CCK-C receptor which we have identified on the surface of several gastric and colonic carcinoma cell lines (2). The failure of the selective antagonists 1.364,718 and L365,260 to inhibit proliferation of HCT 116 cells at concentrations as high as 1 μM confirms that neither CCK-A nor gastrin/CCK-B receptors participate in the autocrine loop (4).

We have identified a gastrin-binding protein (GBP) in detergent extracts of porcine gastric mucosal membranes. The GBP has a molecular weight of 78kDa, and does not appear to have any other disulphide bonded subunits (5). Purification of the porcine GBP by a

combination of lectin and ion exchange chromatography followed by preparative gel electrophoresis has permitted determination of N-terminal and internal amino acid

sequences of the naturall.v occurring protein (Baldwin et al: Int. J. Biochem 26 529-538). These sequences have been used to generate probes which have enabled cloning of the cDNA encoding the porcine GBP from both c-DNA and genomic libraries.

The sequence of porcine GBP, deduced from the cDNA sequence, is related to those of enzymes having enoyl CoA hydratase activity and to enzymes having 3-hydroxy-acyl-CoA dehydrogenase activity. These enzymes are

involved in two consecutive steps in the synthesis and degradation of fatty acids during metabolism located in the α-subunit of a mitochondrial trifunctional protein (MTP). The N-terminal half of the GBP is related to the enoyl-CoA hydratase family, while the C-terminal half is related to the 3-hydroxy-acyl-CoA dehydrogenase family (6). In particular, the sequence of the porcine GBP is closely related to the sequence of the α-subunit of a rat

mitochondrial trifunctional protein (MTP), which catalyses both activities. We have found that the sequence of mature human GBP is 90% identical to that of the porcine GBP

(Zhang and Baldwin, submitted for publication).

To investigate the possibility that the GBP might be involved in the autocrine effects of gastrin-like peptides on colorectal carcinoma cells, we have now

examined the effect of agonists and antagonists on the interaction between the GBP and gastrin. Since colon carcinoma cell lines synthesise progastrin, but fail to process the prohormone to mature C-terminally amidated gastrin₁₇, a candidate receptor for autocrine gastrin should not require an amidated C-terminus for binding. We have therefore also examined the structural requirements for binding of C-terminally extended gastrins to the GBP.

We have now found that covalent cross-linking of ¹²⁵I-[Nle¹⁵]-gastrin_2,17 to the 78 kDa GBP is inhibited by crotonyl CoA and by acetoacetyl CoA. The finding that enoyl CoA and acyl CoA derivatives inhibit binding of gastrin to GBP is surprising. According to the autocrine hypothesis, the reverse effect would be expected. Gastrin, CCK, and their analogues also inhibit cross-linking, and the spectrum of analogue affinities correlates well with the values previously reported for binding to the

gastrin/CCK-C receptor, but poorly with the values reported for binding to either the CCK-A or gastrin/CCK-B receptors. The good correlation observed between the IC₅₀ values for antagonists in the GBP cross-linking assay and the IC₅₀ values for inhibition of either binding of gastrin to colon carcinoma cells or of cell proliferation suggests that the GBP is the target for the inhibitory effects of gastrin/CCK receptor antagonists on cell growth. Cross-linking is also inhibited by proglumide and benzotript, but no inhibition is seen with either the CCK-A receptor-selective antagonist L364,718 or the gastrin/CCK-B receptor-selective antagonist L365,260. The affinities of antagonists for the GBP correlate well with their affinities for the gastrin/CCK-C receptor, and with their potencies for inhibition of colon carcinoma cell growth. In addition to gastrin/CCK receptor antagonists, we have found that other classes of compounds inhibit the cross-linking. For example, non-steroidal anti-inflammatory drugs inhibit not only the cross-linking, but also proliferation of a human colon carcinoma cell line.

To investigate further the postulated autocrine loop for gastrin we have transfected two colon cell lines with a plasmid encoding a gastrin cDNA in the antisense orientation. We have observed inhibition of cell

proliferation, providing clear evidence for the presence of an autocrine loop. Moreover the failure to reverse

inhibition with exogenous gastrin₁₇ suggests that the loop in these cells is intracellular.

We conclude that the 78 kDa gastrin-binding protein is

(i) a member of the hydratase/dehydrogenase

family of fatty acid oxidation enzymes; (ii) the gastrin/CCK-C receptor;

(iii) the target for the anti-proliferative

action of gastrin/CCK receptor antagonists and compounds such as non-steroidal anti- inflammatory drugs;

(iv) a potential target for the autocrine

effects of progastrin; and (v) a target for anti-sense gastrin RNA regulation of proliferation of colon carcinoma cells.

SUMMARY OF THE INVENTION

In one aspect, the present invention provides a method of identifying a compound having the ability to block the low-affinity gastrin-cholecystokinin type C receptor, comprising the step of measuring the ability of said compound to block the binding to gastrin-binding protein of a compound selected from the group consisting of a gastrin-related peptide, a cholecystokinin-related peptide, an antagonist of gastrin or of cholecystokinin, an acyl CoA, an enoyl CoA, an antibody to gastrin, and an antibody to cholecystokinin.

The binding may be reversible or irreversible, but is preferably irreversible.

In a preferred embodiment, inhibition of irreversible binding is assessed by measuring the ability of the compound to block cross-linking of a gastrin

analogue to gastrin binding protein.

Cross-linking may be effected by an agent which is homobifunctional, heterobifunctional, or photoreactive. The person skilled in the art will be aware of a number of suitable reagents. Examples of commercially-available homobifunctional cross-linking reagents include

disuccinimidyl suberate, disulfosuccinimidyl suberate, and ethylene glycol bis-(succinimidyl)-succinate. Heterobi-functional cross-linking reagents include succinimidyl-4-(p-maleimidophenyl) butyrate, which has been found to be equally effective in cross-linking the gastrin peptide to the gastrin binding protein. Suitable photoreactive reagents include N-hydroxysuccinimidyl-4-azidobenzoate.

The gastrin peptide is labelled with a detectable marker, such as a radioactive label, a fluorescent label, or biotin. For example, the gastrin peptide may be

iodinated, or may be synthesised using one or more amino acids comprising a radioactive marker such as ¹⁴C or ³H

Native gastrin, gastrin₁₇, contains no free amino groups, and is therefore incapable of reacting with the cross-linking reagent. The minimum gastrin sequence required for binding to the gastrin/CCK-C receptor is known to be the C-terminal tetrapeptide, so N-truncated peptides are expected to be suitable. For example, [Leu¹⁵] gastrin₁₇ and [Nle¹⁵] gastrin_2,17 have equal ability to bind to isolated canine parietal cells or parietal cell plasma membranes: substitution of methionine by leucine isomers does not interfere with the binding or biological activity of gastrin. A particularly preferred gastrin peptide is [Nle¹⁵] gastrin_2,17, a derivative which lacks the amino-terminal pyroglutamate residue. Substitution of nor-leucine for the naturally occurring methionine at position 15 prevents formation of oxidised methionine derivatives during iodination. Although [Met¹⁵]-gastrin_2,17 could be used, there would be some loss of activity for this reason. It is contemplated that gastrin peptides extended at the C-terminus could also be used.

In another preferred embodiment, a displacement-type assay using reversible binding is employed. The ability of the putative antagonist to prevent binding of a known agonist or antagonist of very high affinity is measured, said known agonist or antagonist being labelled with a detectable marker. Preferably the known agonist or antagonist has at least the binding affinity of benzotript. The known antagonist is suitably labelled with ³H.

The method of the invention provides a general screening method for identifying antagonists of the

gastrin-gastrin binding protein interaction, which

compounds are useful for treatment of diseases, especially neoplastic diseases, involving rapidly proliferating cells. Such antagonists are also useful for controlling gastric acid secretion.

One preferred group of compounds useful for prevention and treatment of neoplastic diseases is non- steroidal anti-inflammatory drugs (NSAIDS), which are especially useful for treatment of neoplastic diseases of the gastrointestinal tract, in particular colorectal cancer. However, we have shown that GBP is widespread in the body, and consequently treatment of cancers of non-gastrointestinal origin is also within the scope of this invention. It is preferable, but not essential, that the NSAID is an inhibitor of prostaglandin activity.

In another preferred embodiment, inhibition of the enoyl CoA hydratase and/or 3-hydroxyacyl CoA

dehydrogenase activities intrinsic to the GBP is assayed. Either activity can be measured in either direction, but is conventionally assayed spectrophotometrically by the decrease in enoyl CoA absorption on hydration or by the decrease in NADH absorption on oxidation. Alternative assays could employ radiochemically labelled substrates.

Gastrin-binding protein suitable for use in the method of the invention may be partially purified or purified naturally occurring GBP, or may be recombinant GBP.

As an alternative to using a high-affinity labelled antagonist in a reversible binding assay, a high-affinity antibody, such as a monoclonal antibody, labelled with a detectable marker may be used. For example, it is contemplated that an antibody capable of blocking binding of an acyl CoA or of gastrin to GBP may be used.

The invention in another aspect therefore provides a method of treating neoplastic disease,

comprising the step of administering to a mammal in need of such treatment an effective amount of a compound having the ability to inhibit binding of a gastrin peptide to gastrin-binding protein.

Preferably the neoplastic disease is selected from the group consisting of gastrointestinal cancers such as colon carcinoma and gastric carcinoma, mammary

carcinoma, malignant melanoma, tumours of epithelial origin, and lymphomas. In yet another a act, the invention provides a pharmaceutical composition of a compound with the ability to inhibit the binding of a gastrin analogue to the low-affinity gastrin-cholecystokinin type C receptor,

comprising a pharmaσeutically-effective amount of said compound and a pharmaceutically-acceptable carrier.

In one preferred embodiment the compound is a non-selective antagonist of the gastrin-cholecystokinin receptor. Preferably the antagonist will have a IC₅₀ value" in the millimolar range or lower.

The compositions of the invention may comprise an antagonist either alone, or in combination with other pharmaceutically-active molecules such as antacid compounds or anti-cancer agents. Anti-cancer agents may include cytokines. Methods and pharmaceutical carriers for

preparation of pharmaceutical compositions are well known in the art, as set out in text books such as Remington's Pharmaceutical Sciences, 17th Edition, Mack Publishing Company, Easton, Pennsylvania, USA.

In a fourth aspect, the invention provides a method of treatment of a neoplastic disease, comprising the step of inducing production of antisense gastrin mRNA in neoplastic tissue of an animal in need of such treatment. Tissue-specific production of anti-sense mRNA can be achieved by targeting of corresponding DNA to the desired tissue, using tissue-specific promoters.

DETAILED DESCRIPTION OF THE INVENTION

The invention is illustrated by reference to the following non-limiting examples, and to the figures, in which:

Figure 1 shows gastrin analogues and antagonists inhibit covalent cross-linking of ¹²⁵ I-[Nle¹⁵]-gastrin_2,17 to the 78kDa GBP.

Cross-linking of ¹²⁵I- [Nle¹⁵]-gastrin_2,17 to the 78 kDa GBP in the presence of increasing concentrations of (A) gastrin analogues (closed symbols) and (B) gastrin/CCK receptor antagonists (open symbols) was measured by phosphorimager (insets), and expressed as a percentage of the value obtained in the absence of competitor. The following values for IC₅₀ and for the predicted ordinate intercept were obtained with the program LIGAND: gastrin₁₇ (♦), 0.11 μM, 102%; CCK₈ (■), 5.6 μM, 98%; CCK_eSO₄ (•), 3.55M, 125%; pentagastrin (▲), 200 μM, 96%; gastrin₄ (▼), 320 μM, 98%; proglumide (□), 4.1 mM, 94%; benzotript (◯), 250 μM, 114%. Lines of best fit are not shown for CCK₈ or gastrin₄ for clarity. No inhibition was observed with the CCK-A receptor-selective antagonist L364,718 (data not shown), or with the gastrin/CCK-B receptor-selective antagonist L365,260 (◊).

Figure 2 shows binding of fatty acyl CoA derivatives to the 78 kDa GBP.

Cross-linking of ¹²⁵I-[Nle¹⁵]-gastrin_2,17 to the 78 kDa GBP was measured by phosphorimager (insets) in the presence of increasing concentrations of crotonyl CoA, a substrate of enoyl CoA hydratase, or of acetoacetyl CoA, a product of 3-hydroxyacyl CoA dehydrogenase, and expressed as a percentage of the value obtained in the absence of competitor. Lines of best fit for inhibition by

acetoacetyl CoA and crotonyl CoA were obtained with the program LIGAND. Values for IC₅₀ and for the predicted ordinate intercept were as follows: Crotonyl CoA (•), 43 μM, 76%; acetoacetyl CoA (■), 74 μM, 70%. No inhibition was observed with NADH (▲), a cofactor for 3-hydroxyacyl CoA dehydrogenase.

Figure 3 shows that anti-proliferative gastrin/CCK receptor antagonists target the 78 kDa GBP. The mean IC₅₀ values (Table 1) for the inhibition of cross-linking of ¹²⁵I-[Nle¹⁵]-gastrin_2,17 to the 78 kDa GBP by gastrin analogues (closed symbols as in Figure 2) and antagonists (open symbols as in Figure 2) were determined as described in the legend to Figure 2, and compared with the IC₅₀ values for (A) inhibition of binding of ¹²⁵I-CCK₈ labelled using Bolton-Hunter reagent to the cloned human gastrin/CCK-B receptor expressed in COS cells (analogues) (8) and of ¹²⁵I-CCK₃₃ to the gastrin/CCK-B receptor on guinea pig gastric glands (antagonists) (9), or (B) inhibition of binding of ¹²⁵I-gastrin₁₇ to the gastrin/CCK-C receptor on HCT 116 colon carcinoma cells (2). Lines of best fit were obtained by linear regression.

Figure 4 illustrates processing and sequences of gastrin-related peptides.

The C-terminal sequence of human progastrin is shown using the one letter code. The corresponding region of porcine progastrin is only shown where it differs from human progastrin. The indicated peptides (numbered from the N-terminus of gastrin_1-17) were synthesised according to the human sequence, with the N-terminal glutamic acid of gastrin_1-4, gastrin_1-11, gastrin_1-17 and gastrin_1-17G cyclized to pyroglutamate. The C-termini of gastrin_1-17 and

gastrin_14-17 were amidated.

Figure 5 shows binding of gastrin-related peptides to the 78 kDa GBP.

Cross-linking of gastrin-related peptides was measured as described above, and the cross-linking

quantitated by phosphorimager, as shown in the inset panels, and lines of best fit calculated and plotted against peptide concentration.

Figure 6 illustrates how expression of antisense gastrin inhibits cell proliferation.

Proliferation of YAMC (A) or LIM 1215 (B) cells transfected with plasmid alone in the absence (black bars) or presence (light grey bars) of 1 μM gastrin₁₇, or with a plasmid expressing antisense gastrin mRNA in the absence

(dark grey bars) or presence (white bars) of 1 μM gastrin₁₇ was measured with the MTT assay at the indicated times after seeding in media containing 1% fetal calf serum. The means of triplicate absorbance readings were expressed as a percentage of the mean absorbance reading obtained on day 1 for the YAMC cells grown in the absence of gastrin₁₇; error bars represent the standard error of the mean. Statistical significance was assessed by Student's t-test: p < 0.05, *; p < 0.02, **.

Figure 7 shows the inhibition of YAMC cell proliferation by gastrin/CCK receptor antagonists.

Proliferation of YAMC cells in the presence of increasing concentrations of the non-selective gastrin/CCK receptor antagonists proglumide (●) or benzotript (■) was measured with the MTT assay at two days after seeding.

Antagonists were added one day after seeding. The means of triplicate absorbance readings are shown as a percentage of the mean absorbance reading obtained on day 1 for the cells grown in 1% fetal calf serum; error bars represent the standard error of the mean. Lines of best fit were

obtained with the programs EBDA and LIGAND as described previously (6).

Figure 8 shows the inhibition of growth of

LIM 1215 and LIM 2412 cells by proglumide.

Uptake of ³H-thymidine (○) by, and viability (□) of, cells was measured after 17 hours treatment with the concentration of proglumide indicated. LIM 1215 (A) and LIM 2412 (B) cells were grown in the absence of serum.

Points are the mean values of 5 samples, and lines of best fit were obtained with the program LIGAND. Bars represent 1 SD.

Figure 9 illustrates proliferation of LIM 1215 colon carcinoma cells in the presence of increasing

concentrations of NSAIDS, measured with the MTT assay as described for Figure 7. Cells were grown in a medium containing 10% foetal calf serum, and NSAIDS were added one day after seeding. The means of triplicate absorbance readings were expressed as a percentage of the mean

absorbance reading obtained for cells grown in the absence of NSAIDS; error bars represent the standard error of the mean. Lines of best fit were obtained with the programs EBDA and LIGAND as described previously (6). Values for IC₅₀ and for the predicted ordinate intercept were as follows: indomethacin, 91 μM, 137% (t); sulindac sulphoxide, 358 μM, 132% (■); aspirin, 3.22 mM, 134% (▲). Similar results were obtained in 3 separate experiments.

Figure 10 shows that the inhibition of proliferation of LIM 1215 cells by NSAIDS is not correlated with the ability of the NSAIDS to inhibit either

cyclooxygenase I (Panel A) or cyclooxygenase II (Panel B). The IC₅₀ values for inhibition of proliferation of LIM 1215 cells were measured as described for Figure 8, and are summarized in Table 4. IC₅₀ values for inhibition of cyclooxygenase I and cyclooxygenase II were taken from

Mitchell et al (Proc. Natl. Acad. Sci. USA., 1994 90. 11693-11697).

Figure 11 shows the ability of four different NSAIDS to block cross-linking of gastrin_2,17 to GBP in the assay described above.

Figure 12 shows the correlation between the ability of NSAIDS to inhibit proliferation of LIM 1215 cells and their ability to inhibit cross-linking of gastrin_2/17 to GBP.

Abbreviations used herein are as follows:

ASP - aspirin

CCK - cholecystokinin

CHO - Chinese hamster ovary

DIC - diclofenac

DIF - diflunisal

FCS - foetal calf serum

GBP - gastrin-binding protein

IBU - ibuprofen

IC₅₀ - concentration required for 50%

inhibition of cross-linking

IFN - niflumic acid

IND - indomethacin

MEC - meclofenamic acid

MEF - mefenamic acid

MTT - 3-(4,5-dimethylthiazol-2-yl)-2-,5- diphenyltetrazolium bromide

NAP - naproxen PHE - phenylbutazone

PIR - piroxicam

SAL - salicylic acid

SUL - sulindac sulphoxide

SUS - sulindac sulphide

TIA - tiaprofenic acid

TOL - tolmetin

YAMC - young adult mouse colon.

MATERIALS AND METHODS

Gastrin₁₇ , CCK₈, and CCK₈SO₄ were from Research

Plus, Bayonne, NJ. Pentagastrin, gastrin₄, proglumide (4-(benzoylamino)-5-(dipropylamino)-5-oxopentanoic acid, acetoacetyl CoA, crotonyl CoA, pepstatin, benzamidine, hexamethylphosphoramide, aprotinin and NSAIDS were from Sigma, St.Louis, MO. Benzotript (N-(p-chlorobenzoyl)-L-tryptophan) was from Calbiochem, San Diego, CA. The antagonists L364,718 and L365,260 were generous gifts from Dr. V.J.Lotti, Merck, Sharp and Dohme, West Point, PA.

Na¹²⁵I was from NEN, North Ryde, Australia.

1²⁵1- [Nle¹⁵]-gastrin_2,17 was prepared and iodinated as described in reference 5, the whole contents of which publication are herein incorporated by reference.

The 78 kDa GBP was partially purified from detergent extracts of porcine gastric mucosal membranes (5), and crosslinked to ¹²⁵I-[Nle]¹⁵-gastrin_2,17 with

disuccinimidylsuberate as previously described (5). The following protease inhibitors were included in all buffers to prevent proteolysis: pepstatin, 1 μM; benzamidine, 1 mM; hexamethylphosphoramide, 0.1% (w/v); aprotinin, 500 units/ml. The products of the cross-linking reaction were separated by polyacrylamide gel electrophoresis, and radioactivity associated with the 78 kDa GBP was detected and quantitated with a phosphorimager (Molecular Dynamics, Sunnyvale, CA). Initial estimates of IC₅₀ values, and of the levels of ¹²⁵I-gastrin_2,17 in the absence of competitor, were obtained by fitting the data with the program EBDA, and were refined with the program LIGAND (3).

The purity and composition of gastrin-related peptides was assessed by the manufacturer (Chiron

Mimotopes, Clayton, Australia) by reversed phase high pressure liquid chromatography and by ion spray mass spectrometry.

Human gastrin cDNA in the plasmid pBR322 (18) was kindly provided by Dr. E. Boel, University of Aarhus, Denmark. The conditionally immortalised young adult mouse colon cell line (YAMC), which was derived from a mouse transgenic for a temperature sensitive SV40 large T antigen (19), and the human colon carcinoma cell line LIM 1215, which was derived from a patient with inherited

nonpolyposis colorectal cancer (20), were obtained from Dr. R.H. Whitehead, Ludwig Institute for Cancer Research, Melbourne.

Cell Proliferation Assay.

A colorimetric assay (21) was used to measure cell proliferation. Briefly, 10⁴ cells were seeded in a 96 well plate in medium containing 1% fetal calf serum. The medium was replaced the next day with fresh medium

containing 1% fetal calf serum and the substance under investigation. For each time point 10 μl of 5 mg/ml MTT

(3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, Sigma, St.Louis, MO) was added to each well, and the plate was incubated for 4 hrs at 37°C before the medium was discarded. 200 μl 0.04M HCl in isopropanol was added to lyse the cells, and the absorbance at 560 nm was read on a Titertek Multiscan MCC 1340 (Labsystems, Helsinki,

Finland). Student's t-test was used for statistical analysis.

In one set of experiments with the human colon carcinoma cell lines LIM 1215 and LIM 2412, cell

proliferation was measured using a ³H-thymidine uptake assay, using a standard method. LIM 2412 cells were grown in RPM1 1640 medium, and LIM 1215 cells in the same medium containing 1 mM thioglycerol, 25 units/ml insulin, 1 mg/ml hydrocortisone, 5 mg/ml bovine serum albumin, 10 μg/ml iron-saturated transferrin and 1.5 mg/ml glutamine. Both cell lines were cultured in the absence of serum.

Proglumide (200 mM, Sigma, St. Louis, MO) was made up in 50 mM Na⁺ HEPES, pH 7.6, readjusted to pH 7.6 with 10 M NaOH, and added at concentrations ranging from 1 to 50 mM.

Control wells received the same volume of 50 mM Na⁺ HEPES, pH 7.6, as this buffer was sometimes found to inhibit [³H]-thymidine uptake. The cells were then incubated in a total volume of 100 μl at 37°C in a humidified atmosphere of 10% CO₂. After 17 hours the cells were pulsed with [³H]-thymidine (0.5 μCi per ml, NEN-DuPont, Boston, MA) for four hours. The medium was removed, and the cells were lysed with a mixture of 30 μl lysis buffer (30 mM Tris HCl, pH 7.6, 100 mM EDTA, 1% w/v Sarcosyl (Sigma, St. Louis, MO)) and 70 μl PBS, harvested and counted using a Beta-Plate harvester and liquid scintillation counter (LKB-Wallace, Turku, Finland). Cell viability in a duplicate plate was measured by Trypan Blue exclusion.

Radioimmunoassay

The gastrin concentrations in conditioned media and cell extracts were determined by radioimmunoassay against an amidated gastrin₁₇ standard curve with ¹²⁵I-amidated gastrin₁₇ as label (22). The calculated 50% inhibitory dose for antiserum 8017 was 1.6 ± 0.1 fmol/tube (n=13). The intraassay variation was <6% (n=6) and the interassay variation was less than 20% (n=13). Levels of amidated gastrin were assayed with antiserum 1296, which detects all amidated carboxyl terminal fragments greater than the pentapeptide, and does not cross-react with glycine-extended forms. Levels of unprocessed, partially processed and mature amidated forms of gastrin (i.e. all progastrin-derived peptides) were measured with antiserum 8017, which detects the amino terminal portion of

gastrin₁₇, after digestion of samples with trypsin (23). Samples were treated with an equal volume of 50 μg/ml trypsin in 20 mM veronal, pH 8.2, containing 0.005% sodium azide and 2% charcoal-stripped plasma. After incubation at 37°C for 2 hr, the samples were boiled for 2 min, cooled on ice for 2 min, and frozen until assay.

Gastrin Binding Assays

Gastrin/CCK-B receptors were measured by the method of Kopin et al (24). 10⁴ cells/well were seeded in a 24 well plate and grown for 2 days at 33°C in RPMI medium containing 10 mM thioglycerol, 100 units/ml insulin, 50 mg/ml hydrocortisone and 10% fetal calf serum. Cells were washed in phosphate-buffered saline, and incubated for 80 min at 37°C in Hank's balanced salt solution containing ¹²⁵I-CCK₈ (10,000 cpm, 2.9 fmol, Amersham, Bucks., UK), 16 μM PMSF and 0.1% BSA. Cells were then washed twice with PBS and lysed with NaOH. Lysates were counted in a γ-counter (Packard, Downer's Grove, IL) at 77% efficiency.

Gastrin/CCK-C receptors were measured by the phthalate oil centrifugation method (3) with ¹²⁵l-gastrin₁₇ (1.5 × 10⁵ cpm, 44 fmol (Amersham, Bucks., UK)) as ligand. Example 1 Definition of the Gastrin Binding Site

The ability of gastrin peptides and of non-specific and specific CCK receptor antagonists to inhibit cross-linking of GBP to a labelled gastrin peptide was tested as described above. The results, expressed as IC₅₀ values, are set out in the far right hand column of Table 1, and are compared with IC₅₀ values previously reported in the literature for binding of gastrin agonists or

antagonists to tissues or tumour cells, or for inhibition of proliferation of tumour cell lines. Values are for the human receptors unless indicated as follows: ^a guinea pig; ^b rat. ND, not determined. Reference numbers are given in brackets.

IC₅₀ values (mean ± SD, calculated from at least 3 sets of data) for inhibition of cross-linking of

¹²⁵I-[Nle¹⁵]-gastrin_2,17 to the 78 kDa GBP in detergent extracts of porcine gastric mucosal membranes were

determined from the data presented in Figure 1 and from similar experiments by least squares fitting with the programs EBDA and LIGAND.

Table 1

Comparison of the 78 kDa GBP with Human Gastrin and Cholecystokinin Receptors.

IC₅₀ for Binding IC₅₀ for Cell Proliferation IC₅₀ for Cross

(μM) (μM) linking (μM)

Receptor CCK-A Gastrin/CCK-B Gastrin/CCK- 78 kDa GBP

Tissue Pancreas, Stomach, C HCT 116 Cells HT 29 Cells Stomach

Gall Bladder Brain HCT 116

Cells

Agonists

Gastrin₁₇ 0.08 6.4 × 10^-3 1.6 - - 0.23 ± 0.15

1.8 9.4 × 10^-4

2.6 × 10^-3

CCk₈ 0.14^a 1.9 × 10^-3 13.2 - - 4.5 ± 1.0

4.7 × 10^-3

I

CCK₈SO₄ 2 × 10^-5 3 × 10^-3 ND - - 5.8 ± 3.2

3 × 10^-3 1.4 × 10^-4

1.1 × 10^-4

Pentagastrin ND ND 100 - - 210 ± 90

Gastrin₄ ND 3.2 × 10^-2 ND - - 370 ± 280

3.0 × 10^-2

Antagonists

Proglumide 600^b 900^a 8600 5000 14000 5100 ± 3600

Benzotript 102^b 59^a 400 500 3400 200 ± 120

L364,718 1 × 10^-4 0.5 ND 4 >100 >200

8 × 10^-4 0.15

L365,260 9 1.0 × 10^-2 ND 15 22 >200

0.02 3.8 × 10^-3

Data in first 5 columns summarised from reports in the literature.

Gastrin₁₇ and its derivatives pentagastrin and gastrin₄ all inhibited cross-linking of ¹²⁵I-[Nle¹⁵]-gastrin_2,17 to the 78 kDa GBP, as illustrated in Figure IA. The observed IC₅₀ values, which are dependent on chain length, are in good agreement with the IC₅₀ values reported previously for inhibition of binding of ¹²⁵I-gastrin₁₇ to the gastrin/CCK-C receptor on the HCT 116 colon carcinoma cell line (see Table 1). In contrast, the IC₅₀ values for the 78 kDa GBP differ markedly from the IC₅₀ values for the CCK-A and gastrin/CCK-B receptors shown in Table 1. The correlation between the IC₅₀ values for the 78 kDa GBP and the IC₅₀ values for the gastrin/CCK-C receptor, and the lack of correlation between the IC₅₀ values for the 78 kDa GBP and the IC₅₀ values for the gastrin/CCK-B receptor, is particularly apparent when the data are presented

graphically (Figures 3A and 3B).

Example 2 Inhibition by Gastrin/CCK Receptor

Antagonists .

The non-selective antagonists proglumide and benzotript both inhibited cross-linking of ¹²⁵I-gastrin_2,17 to the GBP, as shown in Figure IB, with IC₅₀ values of 5.1 mM and 195 μM respectively (Table 1). As with gastrin and its derivatives, a good correlation is observed between the IC₅₀ values for the 78 kDa GBP and the IC₅₀ values for inhibition of ¹²⁵I-gastrin₁₇ binding to the gastrin/CCK-C receptor, while the correlation between the IC₅₀ values for the 78 kDa GBP and the IC₅₀ values for the gastrin/CCK-B receptor is poor. These results are summarised in Figure 3. Furthermore, neither the CCK-A receptor antagonist L364,718 nor the gastrin/CCK-B receptor antagonist L365,260 had any effect on cross-linking at concentrations as high as 200 μM (Figure 1B). A good correlation was also observed between the IC₅₀ values for the 78 kDa GBP and the IC₅₀ values for inhibition of HCT 116 and HT 29 colon carcinoma cell proliferation by proglumide and benzotript (Table 1). Example 3 Inhibition by Fatty Acid Derivatives.

A substrate for enoyl CoA hydratase, crotonyl CoA, and a product of 3-hydroxyacyl CoA dehydrogenase, acetoacetyl CoA, both inhibited cross-linking of

1²⁵I-gastrin_2,17 to the GBP, as summarised in Figure 2. The mean IC₅₀ values from 3 separate experiments were 45 ± 26 μM for crotonyl CoA and 180 ± 150 μM for acetoacetyl CoA. No inhibition was observed with NADH, the cofactor of the dehydrogenase reaction, at concentrations as high as

1.5 mM.

Example 4 Amidation of Gastrin₁₇ is not Required for

GBP Binding

The sequences of gastrin related peptide and glycine-extended peptides are shown in Figure 4.

Processing of progastrin to form gastrin begins by cleavage by a dibasic-specific endopeptidase at the sites marked by arrowheads. Removal of the basic amino acids by carboxypeptidase E yields glycine-extended

gastrin₁₇ (gastrin_1-17G), which is transamidated by

peptidyl-glycine α-amidating monooxygenase to form mature amidated gastrin_1-17.

The effect of extending the C-terminus of

gastrin₁₇ on binding to the GBP was investigated by

comparing the potencies of glycine-extended gastrin₁₇ and gastrin₁₇ as inhibitors of cross-linking of [¹²⁵I]-Nle¹⁵-gastrin_2-17 to the GBP.

IC₅₀ values (mean ± SEM of at least 3 separate determinations) for the inhibition of cross-linking of

[¹²⁵I] -Nle¹⁵-gastrin_2-17 to the 78 kDa gastrin-binding protein by the indicated peptides were determined as described above, and are summarised in Table 2. IC₅₀ values for gastrin_12-17GRRSAE, gastrin_1-4 and YFGRRSAEEGD represent estimates only, because the limiting amounts of peptide available precluded assay at concentrations greater than 100, 400 and 100 μM respectively. IC₅₀ values for the gastrin/CCK-B receptor on canine parietal cells, and for the binding of gastrin_1-17 and gastrin_14-17 to the GBP, are taken from reference 24 and Example 1 respectively. The amino acid sequences of C-terminal extensions are shown in the one letter code; the peptide sequences are shown in full in Figure 4.

Table 2

Affinities of Gastrin-Related Peptides

for Gastrin Receptors

IC₅₀ (μM)

Peptide Purity 78 kDa Gastrin- Gastrin/CCK-B *

(%) Binding Protein Receptor

Gastrin_1-17 82 0.23 ± 0.15 5 × 10^-4

Gastrin_1-17G 78 0.19 ± 0.10 ND

Gastrin_5-17G 93 26 ± 15 ND

Gastrin_5-17GRRSAE 87 20 ± 11 ND

Gastrin_12-17G 81 31 ± 17 14 ± 71

Gastrin_12-17GRR ND ND 44 ± 21

Gastrin_12-17GRRSAE 91 760 ± 550 ND

Gastrln_14-17 95 370 ± 280 0.1

Gastrin_1-11 97 21 + 9 ND

Gastrin_1-4 98 >1000 ND

YFGRRSAEEGD 62 105 ± 21 ND

ND : not determined * Values taken from Reference 11

The IC₅₀ value for glycine-extended gastrin₁₇ was slightly lower than the IC₅₀ value for gastrin₁₇, as shown in Figure 5 and Table 2. Addition of a further 5 residues to the C-terminus of glycine-extended gastrin_5-17 did not alter the affinity of the resultant peptide for the GBP, while addition of 5 residues to the C-terminus of glycine-extended gastrin_12-17 actually resulted in a 25-fold

reduction in affinity. Example 5 Gastrin₁₇ Binding Determinants

The regions of gastrin₁₇ contributing to GBP binding were investigated by testing the effect of N- and C-terminal deletion on the potency of glycine-extended gastrin₁₇ as an inhibitor of cross-linking of [¹²⁵I]-Nle¹⁵-gastrin_2-17 to the GBP. Results are shown in Figure 5.

Cross-linking of ¹²⁵I-[Nle¹⁵]-gastrin_2-17 to the 78 kDa GBP was measured as described above in the presence of increasing concentrations of gastrin-related peptides .

Duplicate samples were subjected to electrophoresis on NaDodS0₄-polyacrylamide gels, and the radioactivity

associated with the 78 kDa GBP was quantitated by

phosphorimager scanning and expressed as a percentage of the value obtained in the absence of competitor. The following values for IC₅₀ and for the predicted ordinate intercept were obtained by computer fitting as described herein, and used to construct the indicated lines of best fit: (A) gastrin_1-17 (138 nM, 115.5%) or glycine-extended gastrin_1-17 (98 nM, 90.0%), (B) glycine-extended gastrin_5-17 (11 μM, 93.8%) or glycine-extended gastrin_12-17 (40 μM,

99.6%), or (C) gastrin_1-4 (1.96 mM, 101.7%) or gastrin_1-11 (31 μM, 101.3%).

Comparison of the affinities of gastrin analogues for the 78 kDa GBP revealed that both ends of gastrin₁₇ contribute to binding. Thus removal of 4 residues

(pyroglutamylGPW) from the N-terminus of glycine-extended gastrin₁₇ resulted in a 140-fold decrease in affinity, while removal of 7 residues (YGWMDFG) from the C-terminus of glycine-extended gastrin₁₇ resulted in a 110-fold decrease in affinity (Table 2). Removal of a further 7 residues from the C-terminus of glycine-extended gastrin₁₇ resulted in almost complete loss of binding affinity. In reciprocal experiments the binding of the deleted peptides to the GBP was also measured. As expected, gastrin_14-17 bound weakly to the GBP, but binding of the N-terminal tetrapeptide pyroglutamylGPW was barely detectable. Since benzotript, which we have shown in Example 1 to be the best available antagonist for the GBP, is an acylated tryptophan derivative (N-4-chlorobenzoyl-L-tryptophan), it seems likely that the tryptophan residue of either the N-terminal or C-terminal tetrapeptide, or both, makes a significant contribution to binding.

The role of the sequence of five glutamic acid residues of gastrin₁₇ in GBP binding is not clear at present. The similar affinities of glycine-extended gastrin_5-17 and glycine-extended gastrin_12-17 for the GBP (Table 2) suggest that this polyglutamate sequence makes no contribution. However, deletion of the five glutamate residues from gastrin_1-11 to yield gastrin_1-4 resulted in a 60-fold decrease in affinity.

In contrast to the present results with the GBP, two lines of evidence demonstrate that the gastrin/CCK-B receptor does not recognise the N-terminus of gastrin₁₇. Firstly, no binding of the N-terminal tridecapeptide to the gastrin/CCK-B receptor on canine parietal cells was

observed even at concentrations as high as 100 μM (11). Secondly, removal of 4 residues from the N-terminus of gastrin₁₇ had no effect on binding to rabbit parietal cells (12). The observation that removal of a further 5

residues, including 4 of the 5 glutamic acid residues, reduced binding to rabbit parietal cells by 100-fold additionally implied that the pentaglutamic acid sequence contributed to gastrin/CCK-B receptor binding (16). The gastrin binding sites of the GBP and the gastrin/CCK-B receptor are thus clearly distinct.

Example 6 Anti-Sense Progastrin RNA Inhibits

Expression of Progastrin and Gastrin

A cDNA fragment encoding human progastrin (18) was cut from pBR322 with Pstl and cloned into pGEM-3Z

(Promega, Madison, WI). An expression plasmid encoding gastrin mRNA in the sense orientation was then constructed by cloning a Bam HI/SphI fragment from pGEM-3Z into

pCDNA1amp (Invitrogen, San Diego, CA). The corresponding antisense plasmid was constructed by cloning an XbaI/SphI fragment from pGEM-3Z into pCDNAlamp, and a XbaI/BstXI fragment thence into pCDNA1neo (Invitrogen).

For transient transfections 10⁷ COS cells were incubated with 2.5 μg/ml plasmid DNA, 100 μM chloroquine and 200 μg/ml DEAE-dextran in RPMI medium containing 2 mM glutamine and 10% heat-inactivated NuSerum (Gibco/BRL, Gaithersburg, MD) for 4 hr. at 37°C, grown for 3 days at 37°C in RPMI medium containing 2 mM glutamine and 10% fetal calf serum, and harvested. The lipofection method (25) was used to generate stable cell lines. Half-confluent YAMC cells were incubated with 20 μg/ml plasmid DNA for 4-6 hrs at 33°C in serum-free RPMI medium containing 20 μg/ml lipofectin (Gibco/BRL), 10 mM thioglycerol, 100 units/ml insulin and 50 mg/ml hydrocortisone. Transfected cells were grown at 33°C in RPMI medium containing 10 mM

thioglycerol, 100 units/ml insulin, 50 mg/ml hydrocortisone and 10% fetal calf serum for 2 days, and selected for 14 days in the same medium containing 100 μg/ml neomycin.

No progastrin or amidated gastrin was detected in

YAMC or LIM 1215 cell extracts and conditioned media. In contrast, progastrin-derived peptides were readily detected in YAMC cell extract (20 fmol/10⁷ cells) and conditioned medium (66 fmol/10⁶ cells) following transfection with a sense gastrin construct; however, amidated gastrin was not detected in these cells. No progastrin-derived peptides were detected in YAMC or LIM 1215 cells transfected with an antisense gastrin construct.

To confirm that expression of antisense gastrin mRNA was able to reduce levels of progastrin-derived peptides, COS cells were transiently transfected with a plasmid encoding sense gastrin mRNA or with plasmids encoding both sense and antisense gaε crin mRNA. The levels of progastrin-derived peptides in cell extracts (760 pmol/10⁷ cells) and conditioned media (1540 pmol/10⁷ cells) were reduced by 37 and 48% respectively in the doubly transfected cells. Amidated gastrin was not detected in either case. Similar results were obtained in 3 separate experiments with YAMC, and two experiments with LIM 1215.

We have previously demonstrated the presence of very low levels (2.1 molecules/1000 cells) of gastrin mRNA in LIM 1215 cells by quantitative PCR (15). Unfortunately we have been prevented from quantitating the amount of gastrin mRNA in YAMC cells by PCR by absence of a published mouse gastrin cDNA sequence. However, we have been unable to detect any progastrin-derived peptides or mature amidated gastrin by radioimmunoassay in either cell extracts or media conditioned by YAMC or LIM 1215 cells. Based on the sensitivity of the assay (0.4 fmol), we should have detected levels of progastrin-derived peptides as low as 1 fmol/10⁶ cells. Presumably the discrepancy between our results and the observation that progastrin-derived peptides were present in all 5 colon carcinoma cell lines tested at levels between 17 and 54 fmol/10⁶ cells (13) is a reflection of the fact that LIM 1215 cells had the lowest levels of gastrin mRNA of 7 colon carcinoma cell lines tested (15). Because of the absence of detectable levels of progastrin-derived peptides in YAMC and LIM 1215 cells, we were unable to demonstrate any reduction in levels following expression of antisense gastrin mRNA. However, we were able to demonstrate such a reduction in COS cells transiently transfected with plasmids expressing both sense and antisense gastrin mRNA, compared with COS cells transiently transfected with plasmids expressing sense gastrin mRNA only. We therefore postulate that the observed inhibition of proliferation of YAMC and LIM 1215 cells is a consequence of a reduction in the level of progastrin-derived peptides.

Example 7 Expression of Antisense Gastrin Reduces

Proliferation

Expression of antisense gastrin mRNA

significantly reduced proliferation of both YAMC cells and the colon carcinoma cell line LIM 1215 (Figure 6). In neither case was the inhibition reversed by the addition of gastrin₁₇ to the medium.

Example 8 Gastrin Receptors in YAMC and LIM 1215

cells

The total amount of ¹²⁵I-CCK₈ (10,000 cpm, 2.9 fmol) bound to gastrin/CCK-A + -B receptors on YAMC, LIM 1215 and AR4-2J cells was measured by the method of Kopin et al (24). The AR4-2J cell line was derived from a rat pancreatic carcinoma, and expresses both gastrin/CCK-A + -B receptors. The amount of ¹²⁵I-gastrin₁₇ (150,000 cpm, 44 fmol) bound to gastrin/CCK-C receptors on YAMC and LIM 1215 cells was measured by the phthalate oil centrifugation assay described by Weinstock and Baldwin (3). Background binding was measured in the presence of 1 μM (A + B) or 10 μM (C) gastrin₁₇. The results are summarised in Table 3.

Table 3

Presence of Gastrin/CCK-C, but not Gastrin/CCK-A or -B

Receptors on YAMC and LIM 1215 Cells

Cell line Cell No. ¹²⁵I-CCK₈ Bound Background

(cpm) (cpm)

Gastrin/CCK-A + -B Receptors

YAMC 10⁴ 201 ± 22 171 ± 12

LIM 1215 10⁴ 1663 ± 89 1604 ± 106

AR4-2J 10⁴ 10567 ± 1341 1866 ± 252

Gastrin/CCK-C Receptor;

YAMC 4.2 × 10⁵ 1002 ± 72 420 ± 88

LIM 1215 10⁶ 5407 ± 800 700 ± 89 No CCK-A or gastrin/CCK-B receptors were detected on YAMC or LIM 1215 cells by binding of ¹²⁵I-CCK₈, even though gastrin/CCK-A and -B receptors were readily detected on AR4-2J cells. The lack of involvement of either CCK-A or gastrin/CCK-B receptors in the autocrine growth of YAMC cells was confirmed by demonstrating that neither the CCK- A-receptor selective antagonist L364,718, nor the

gastrin/CCK-B receptor-selective antagonist L365,260, inhibited YAMC cell proliferation at concentrations

(100 nM) sufficient to saturate their respective receptors (data not shown).

Gastrin/CCK-C receptors were readily detected on both YAMC and LIM 1215 cells using the phthalate oil centrifugation assay and ¹²⁵I-gastrin₁₇ as ligand.

Example 9 Inhibition of Cell Proliferation by

Receptor Antagonists

The effect of non-selective gastrin/CCK receptor antagonists on the proliferation of YAMC cells was

measured. The non-selective gastrin/CCK receptor

antagonists proglumide and benzotript were both able to inhibit proliferation of YAMC cells, as assessed using the MTT assay at two days after seeding of the cells. The results of one experiment are shown in Figure 7.

Values for IC₅₀ and for the predicted ordinate intercept were as follows: proglumide, 7.4 mM, 119% ;

benzotript, 13.7 mM, 113%. Concentrations of benzotript higher than 3 mM were toxic to YAMC cells. No inhibition of proliferation was observed with the CCK-A receptor-selective antagonist L364,718 or the gastrin/CCK-B

receptor-selective antagonist L365,260 at 100 nM. Similar results were obtained in 2 further experiments.

Our overall results of inhibition of

proliferation of YAMC by proglumide (IC₅₀ value of 7.2 ± 0.4 mM) and by benzotript (IC₅₀ value of 8.0 ± 8.1 mM) were consistent with the involvement of the gastrin/CCK-C receptor in the autocrine growth of YAMC. Similarly, we have found that proliferation of LIM 1215 cells and

LIM 2412 cells is inhibited by proglumide, with IC₅₀ values of 2.2 ± 1.0 mM, and 2.9 ± 1.3 mM respectively. No correlation was observed between the degree of inhibition by proglumide and the presence or absence of binding sites for ¹²⁵I-[Nle¹⁵]-gastrin₁₇ determined previously. Even though gastrin₁₇ breakdown did not exceed 30% during a 24 hour incubation in the absence of serum and a non-saturating concentration of inhibitor was chosen to facilitate ready reversal of inhibition, no reversal of the inhibition of LIM 1215 or LIM 2412 cells by proglumide was detected with gastrin₁₇ concentrations as high as 50 μM. These results are shown in Figure 8.

Example 10 Inhibition of Proliferation of LIM 1215

Cells by Non-Steroidal Anti-Inflammatory Drugs

NSAIDS have been shown to inhibit colorectal tumour growth in several experimental models, and have been shown to reduce the size and number of colorectal polyps in patients with familial adenomatous polyposis (26), and to reduce the death rate from cancer of the eosophagus, stomach, colon, and rectum (27). It has been widely accepted that the effect of NSAIDS on colon cancers has resulted from the ability of these agents to inhibit cyclooxygenases, which are key enzymes in the synthesis of prostaglandins, and in particular it has been proposed that the anti-tumour activity is correlated with inhibition of cyclooxygenase-2 (28).

We have compared the ability of a variety of NSAIDS to inhibit the proliferation of the colon carcinoma cell line LIM 1215 with their ability to inhibit

cyclooxygenase I and cyclooxygenase II. Cell proliferation was measured using the colorimetric assay described above. For these experiments, the medium was RPMI 1640 containing 10 μM thioglycerol, 25 units/ml insulin, 1 mg/ml

hydrocortisone and 10% foetal calf serum. After 24-hours, fresh medium containing the test drug was added. Otherwise the assay was as described above. Figure 9 shows results for indomethacin, sulindac sulphoxide and aspirin. All of these agents inhibited proliferation of LIM 1215 cells. In fact, all of the NSAIDS tested inhibited proliferation of LIM 1215 cells, with IC₅₀ values obtained by computer curve fitting with the programmes EBDA and LIGAND ranging from 120 μM for meclofenamic acid and sulindac sulphide to 21 mM for acetaminophen. These results are summarized in Table 4, and are also compared for some agents with IC₅₀ values for cyclooxygenase I and cyclooxygenase II, which have been reported in the literature.

Table 4

IC₅₀ Values for Inhibition of Colon Carcinoma Cell Proliferation by NSAIDS

IC₅₀ (μM)

NSAIDS

LIM 1215 GBP Cross-linking Cyclo-oxygenase I Cyclo-oxygenase II Cell Proliferation

Acetaminophen 21000 ± 14000 10000 17 .9* 132*

Aspirin 1700 ± 640 9300 ± 4700 1 .7 278

Diclofenac 120 ± 110 160 ± 90 1 .6 1.1

Ibuprofen 290 ± 160 2300 ± 600 4 .8 73

Indomethacin 200 ± 130 150 ± 40 0 .03 1.7

Naproxen 660 ± 460 1200 ± 400 9 .6 5.7

Salicylate 1750 ± 400 3300 ± 1400 219 625

Diflunisal 160 ± 80 80 ± 30

Meclofenamate 120 ± 70 90 ± 50

Mefenamate 200 ± 60 220 ± 60

Niflumate 130 ± 100 290 ± 180

Phenylbutazone 600 ± 180 2200 ± 400

Piroxicam 690 ± 390 2600 ± 500

Sulindac S 120 ± 120 40 ± 10

Sulindac 330 ± 170 580 ± 260

Tiaprofenate 440 ± 230 1400 ± 400

Tolmetin 1900 ± 170 1400 ± 100

* IC₃₀ values are shown because 50% inhibition of cyclooxygenase II was not achieved by concentrations as high as 1 mg/ml (28) .

Figure 10 compares IC₅₀ values for LIM 1215 cell proliferation and for inhibition of cyclooxygenases I and II in intact bovine aortic endothelial cells and murine macrophages respectively, and shows that there is no correlation between inhibition of cell proliferation and enzyme inhibition for either enzyme. In particular, the potent cyclooxygenase inhibitor indomethacin, which has an IC₅₀ value of 30 nM for cyclooxygenase I and 1.7 μM for cyclooxygenase II, was only able to inhibit LIM 1215 cell proliferation to a moderate extent, having an IC₅₀ value of 200 μM. Moreover, acetaminophen , which does not decrease the risk of gastrointestinal tract cancer (27), was even less effective in inhibiting LIM 1215 cell proliferation, having an IC₅₀ of 21 mM. This agent is more potent than aspirin as an inhibitor of cyclooxygenase II, with an IC₅₀ value of 130 μM (28). Our results provide further evidence that the anti-proliferative affects of NSAIDS on colon cancers are not primarily mediated by cyclooxygenase II.

Example 11 Non-Steroidal Anti-Inflammatory Drugs

Inhibit Cross-Linking of Gastrin Analogue to Gastrin-Binding Protein

Using the cross-linking assay described above, the ability of several NSAIDS to inhibit the cross-linking of ¹²⁵I-gastrin_2,17 to gastrin-binding protein was examined. The results for sulindac sulphide, indomethacin, ibuprofen and aspirin are presented in Figure 11. All four agents inhibited the cross-linking in a concentration-dependent manner, with the order of activity being sulindac

sulphide > indomethacin > ibuprofen > aspirin.

The correlation between ability to inhibit in the gastrin-gastrin binding protein cross-linking assay and the ability to inhibit proliferation of LIM 1215 cells was therefore investigated, using sixteen different NSAIDS. The results are shown in Figure 12. Anti-proliferative activity was strongly correlated with the IC₅₀ value in the gastrin binding protein cross-linking assay, with a correlation coefficient of 0.87.

Again the results suggested that the strongest activity was shown by sulindac sulphide, and the weakest by aspirin. No inhibition was detected with acetaminophen at concentrations as high as 10 mM.

We conclude that the gastrin-binding protein cross-linking assay is useful, either alone or in

conjunction with the cell proliferation assay, to identify NSAIDS useful for treatment of diseases involving rapidly proliferating cells, especially neoplastic diseases. It is considered that use of both assays for a given candidate comopund gives a more accurate prediction of activity than the use of either assay alone.

CONCLUSIONS

Identity of the 78 kDa GBP and the Gastrin/CCK-C Receptor

The IC₅₀ values determined for inhibition of cross-linking of ¹²⁵I-[Nle¹⁵]-gastrin_2,17 to the 78 kDa GBP by gastrin₁₇, CCK₈ and their analogues correlate well with the values previously reported for the gastrin/CCK-C receptor on gastrointestinal cell lines (2,3), as shown in Figure 3B. In contrast, the correlation with the values reported for binding to the CCK-A and gastrin/CCK-B

receptors (Figure 3A) is poor. Similarly, for the

antagonists proglumide and benzotript a good correlation is observed in the case of the gastrin/CCK-C receptor only (Figure 3B).

Anti-proliferative Gastrin/CCK Receptor Antagonists Target the Gastrin/CCK-C Receptor

A good correlation is also observed between the IC₅₀ values for inhibition of cross-linking of

1²⁵I-[Nle¹⁵]-gastrin_2,17 to the 78 kDa GBP by the

non-selective gastrin/CCK receptor antagonists proglumide and benzotript, and the IC₅₀ values for inhibition of HCT 116 colon carcinoma cell proliferation by the same

antagonists (2; Table 1). We have previously reported that the CCK-A receptor-selective antagonist L364,718 and the gastrin/CCK-B receptor-selective antagonist L365,260 had no effect on colon carcinoma cell proliferation at

concentrations as high as 1 μM. Furthermore, the

inhibition observed at higher concentrations of L364,718 was not reversed by addition of gastrin₁₇ or CCK₈SO₄, and hence was not caused by binding to cell surface gastrin or CCK receptors (4).

We have now found that neither L364,718 nor

L365,260 has any effect on cross-linking of

¹²⁵I-[Nle¹⁵]-gastrin_2,17 to the 78 kDa GBP at concentrations up to 200 μM. The effects of antagonists on cross-linking are consistent with the conclusion that binding of

antagonists to the gastrin/CCK-C receptor inhibits cell growth.

NSAIDS inhibit cell proliferation via the GBP

CyclooxygenaseS are involved in prostaglandin synthesis, and utilize fatty acid derivatives as

substrates; they are widely thought to be the targets for NSAIDS. The rat mitochondrial trifunctional protein (MTP), the α-subunit of which has strong amino acid sequence homology to the porcine GBP, also utilizes fatty acid derivatives as substrates. We have shown that NSAIDS, with the exception of acetaminophen, inhibit cross-linking of iodinated gastrin to the porcine GBP, and conclude that GBP may be the target of the anti-proliferative effect of these agents on colon cancer.

Hereditary defects in the MTP can cause death in early childhood (29). A wide range of diagnostic symptoms has been reported, including diarrhoea, vomiting,

hypoketotic hypoglycaemic coma, myopathy and

cardiomyopathy. Pathological examination reveals

widespread fat deposition in liver, heart and kidneys.

Similar symptoms and pathology have been described in

Reye's syndrome, which affects children treated with aspirin during viral infection (30). We suggest that the previously unsuspected link between these syndromes is the MTP. Limitation of caloric intake during illness would result in an increased reliance as a metabolic fuel on acetyl CoA, which is one product of the reaction catalysed by the β-subunit of the MTP. Blockade of fatty acid oxidation by the MTP, either by mutation or by inhibition with aspirin, would then result in hypoketotic

hypoglycemia. Hence the observation of a negative

correlation between acetaminophen usage and Reye's syndrome is consistent with the failure of acetaminophen to inhibit cross-linking of gastrin to the GBP at concentrations as high as 10 mM. This hypothesis predicts that NSAIDS will not inhibit the proliferation of cell lines derived from patients with MTP mutations, and enables rational design of more potent and selective MTP/GBP antagonists for use in novel therapies for colon cancer.

Expression of gastrin CCK-C receptors by YAMC and LIM 1215 cells

We have previously reported that LIM 1215 colon carcinoma cells express cell-surface gastrin/CCK-C

receptors (3). This observation was consistent with our observation that proliferation of LIM 1215 cells was inhibited by the non-selective gastrin/CCK receptor

antagonist proglumide (Example 9, and our unpublished results 10), and with the lack of inhibition by the CCK-A receptor-selective antagonist L364,718 or the gastrin/CCK-C receptor-selective antagonist L365,260 at concentrations as high as 1 μM (4).

We have now found that the conditionally immortalised mouse colon cell line YAMC also expresses cell surface gastrin/CCK-C receptors, but not CCK-A or

gastrin/CCK-B receptors. YAMC cell proliferation is unaffected by the selective antagonists L364,718 and

L365,260 at concentrations as high as 0.1 μM, but is inhibited by the non-selective gastrin/CCK receptor antagonists proglumide and benzotript. Moreover the IC₅₀ values which we have observed for these antagonists are consistent with the hypothesis that the gastrin/CCK-C receptor is the target for the inhibitory effects (6).

Role of the Gastrin/CCK-C Receptor in Autocrine Growth

Without wishing to be bound by any proposed mechanism for the observed effect, our observations provide a possible explanation for the inhibitory effects of gastrin/CCK receptor antagonists on cell proliferation. Competition between antagonists and gastrin, and between acyl CoA derivatives and gastrin, in the GBP cross-linking assay implies that antagonists will compete for substrate binding to the 78 kDa GBP. If the GBP possesses enoyl CoA hydratase and 3-hydroxyacyl CoA dehydrogenase activities with longer chain acyl CoA derivatives as substrates, as suggested by the binding of short chain acyl CoA

derivatives (Figure 1) and by sequence comparisons with other members of the hydratase/dehydrogenase family (7), then presumably inhibition of fatty acid metabolism by antagonist will block cell growth, because of a reduction either in energy supply or in the availability of essential lipids. Although the competition observed between gastrin and acyl CoA derivatives suggests that gastrin might also inhibit one or both activities, gastrin is able to reverse the inhibition of HCT 116 cell proliferation by proglumide (2), implying that gastrin itself does not block any essential function of the 78 kDa GBP. Direct measurements of the effect of gastrin on the enzyme activities of the purified 78 kDa GBP are required to resolve this

inconsistency.

The inhibition of HCT 116 cell proliferation by proglumide and benzotript, and the reversal of inhibition by concentrations of gastrin in the mM range, is consistent with the 78 kDa GBP being a component of an extracellular autocrine loop involving gastrin. However, no such

reversal of proglumide inhibition by gastrin is observed with LIM 1215 or LIM 2412 cells (10), although the observed competition between antagonists and gastrin for the 78 kDa GBP implies that reversal should be observed if the GBP is accessible to exogenous gastrin. One possible explanation for this discrepancy is that the 78 kDa GBP, the mRNA for which is expressed in all colon carcinoma cell lines tested so far, may appear on the cell surface in only a subset of colon carcinoma cell lines. The observation that of 5 colon carcinoma cell lines tested only HCT 116 and LoVo cells secreted progastrin (1) is also consistent with the hypothesis that only a subset of cell lines utilise gastrin in an extracellular autocrine loop.

Inhibition of proliferation of both YAMC and LIM 1215 cells by expression of antisense gastrin mRNA provides clear evidence for the presence in both cell lines of an autocrine loop involving progastrin-derived peptides. Our evidence implicates the gastrin/CCK-C receptor in the autocrine loop in both YAMC and LIM 1215 cells. While binding studies clearly localise a population of the gastrin/CCK-C receptor to the cell surface in both YAMC and LIM 1215 cells (Table 3), subcellular fractionation has localised a second population of the gastrin/CCK-C receptor to mitochondrial membranes in rat liver (31). Furthermore the failure to reverse antisense inhibition by exogenous gastrin₁₇ indicates that the autocrine loop is exclusively intracellular. We therefore postulate that a novel intracellular autocrine (i.e. intracrine) loop involving amidated gastrin and the gastrin/CCK-C receptor

contributes to proliferation of both YAMC and LIM 1215 cells.

It may therefore be necessary to screen biopsy samples for the presence of cell-surface GBP in order to determine the suitability of a cancer for treatment with an antagonist according to the method of the present

invention. Such screening may be carried out using antibody to GBP, labelled with a detectable marker, and measuring the binding of the labelled antibody to dispersed tumour cells. Alternatively screening may be carried out as described by Weinstock and Baldwin (3).

The use of immunochemical assays of inhibition of binding of labelled antibody, or assays of inhibition of binding of labelled high-affinity antagonists, provide the opportunity for automated assays which can be used for large scale screening of putative antagonists. Assays of the enzyme activities intrinsic to the GBP may also be used for large-scale screening of GBP antagonists. Covalent cross-linking of ¹²⁵I-gastrin_2,17 to the purified 78 kDa GBP provides a convenient assay with which to isolate more potent and selective GBP antagonists capable of blocking cell growth.

References cited herein are listed on the following pages. it will be clearly understood that the invention in its general aspects is not limited to the specific details referred to herein.

REFERENCES

1. Baldwin, G.S. and Whitehead, R.H.

Bailliere's Clin. Endocrinol. and Metab., 1994 8

185-214.

2. Hoosein, N.M., Kiener, P.A., Curry, R.C.,

Rovati, L.C., McGilbra, D.K. and Brattain, M.G.

Cancer Res., 1988 48 7179-7183.

3. Weinstock, J. and Baldwin, G.S.

Cancer Res., 1988 48 932-937.

4. Thumwood, C.M., Ji, H. and Baldwin, G.S.

Exp. Cell Res., 1991 192 189-192.

5. Baldwin, G.S., Chandler, R., Scanlon, D.B. and

Weinstock, J.

J. Biol. Chem., 1986 261 12252-12257.

6. Mantamadiotis, T., Sobieszczuk, P., Weinstock, J.

and Baldwin, G.S.

Biochim. Biophys. Acta, 1993 1170 211-215.

7. Baldwin, G.S.

Comp. Biochem. Physiol., 1993 104B 55-61

8. Lee, Y.M., Beinborn, M., McBride, E.W., Lu, M.,

Kolakowski, L.F. and Kopin, A.S.

J. Biol. Chem., 1993 268 8164-8169.

9 . Chang, R.S.L., Lotti, V.J., Monaghan, R.L.,

Birnbaum, J., Stapley, E.O., Goetz, M.A.,

Albers-Schonberg, G., Patchett, A.A., Liesch, J.M.,

Hensens, O.D. and Springer, J.P.

Science, 1985 230 177-179.

10. Murphy, V.J., McBride, K., Ji Hong and Baldwin, G.S.

Gastrointestinal Cancer, 1994 In press.

11. Matsumoto, M., Park, J., Sugano, K. and Yamada, T.

"Biological activity of progastrin posttranslational processing intermediates"

Am. J. Physiol., 1987 252 G315-G319.

12. Magous, R. and Bali, J.-P.

"High-affinity binding sites for gastrin on isolated rabbit gastric mucosal cells"

Eur. J. Pharmacol., 1982 82 47-54. 13. Van Solinge, W.W., Nielsen, F.C., Friis-Hansen, L., Falkmer, U.G. and Rehfeld, J.F.

"Expression but incomplete maturation of progastrin in colorectal carcinomas"

Gastroenterology, 1993 104 1099-1107.

14. Baldwin, G.S., Casey, A., Mantamadiotis, T.,

McBride, K., Sizeland, A. and Thumwood, C.M. "PCR cloning and sequence of gastrin mRNA from carcinoma cell lines"

Biochem. Biophys. Res. Commun., 1990 170 691-697.

15. Baldwin, G.S. and Zhang, Q.-X.

"Measurement of gastrin and transforming growth factor α messenger RNA levels in colonic carcinoma cell lines by quantitative polymerase chain

reaction"

Cancer Res., 1992 52 2261-2267.

16. Finley, G.G., Koski, R.A., Melhem, M.F., Pipas, J.M.

and Meisler, A.I.

"Expression of the gastrin gene in the normal human colon and colorectal adenocarcinoma"

Cancer Res., 1993 53 2919-2926.

17. Guo, Y.-S., Baijal, M., Jin, G.-F., Thompson, J.C,

Townsend, CM. and Singh, P.

"Growth-promoting effects of gastrin on mouse colon cancer cells in vitro: Absence of autocrine effects"

In Vitro Cell. Dev. Biol., 1990 26 871-877.

18. Boel, E., Vuust, J., Norris, F., Norris, K.,

Wind, A., Rehfeld, J.F. and Marcker, K.A.

"Molecular cloning of human gastrin cDNA: Evidence for evolution of gastrin by gene duplication"

Proc. Natl. Acad. Sci. USA., 1983 80 2866-2869.

19. Whitehead, R.H., VanEeden, P.E., Noble, M.D.,

Ataliotis, P. and Jat, P.S.

"Establishment of conditionally immortalized epithelial cell lines from both colon and small intestine of adult H-2K^b-tsA58 transgenic mice" Proc. Natl. Acad. Sci. USA, 1993 90 587-591. 20. Whitehead, R.H., Macrae, F.A., St. John, D.J.B. and Ma, J.

"A colon cancer cell line (LIM 1215) derived from a patient with inherited nonpolyposis colorectal cancer"

J. Natl. Cancer Inst., 1985 74 759-765.

21. Mossman, T.

"Rapid colorimetric assay for cellular growth and survival: Application to proliferation and

cytotoxicity assays"

J. Immunol. Methods, 1983 65 55-63.

22. Ciccotosto, G. and Shulkes, A.

"Pharmacokinetics and organ specific metabolism of glycine-extended and amidated gastrin in sheep" Am. J. Physiol., 1992 263 G802-G809.

23. Rehfeld, J.F. and Bardram, L.

"Processing-independent analysis (PIA) - A new diagnostic tool"

Scan J. Clin. Lab. Invest., 1991 204 9-16.

24. Kopin, A.S., Lee, Y.-M., McBride, E.W.,

Miller, L.J., Lu, M., Lin, H.Y., Kolakowski, L.F. and Beinborn, M.

"Expression cloning and characterization of the canine parietal cell gastrin receptor"

Proc. Natl. Acad. Sci. USA., 1992 89 3605-3609.

25. Feigner, P.L., Gadek, T.R., Holm, M., Roman, R.,

Chan, H.W., Wenz, M., Northrop, J.P., Ringold, G.M. and Danielsen, M.

"Lipofection: A highly efficient, lipid-mediated DNA-transfection procedure"

Proc. Natl. Acad. Sci. USA, 1987 84 7413-7417.

26. Giardiello, F.M., Hamilton, S.R., Krush, A.J.,

Piantadosi, S., Hylind, L.M., Celano,

P., Booker, S.V., Robinson, CR. and

Offerhaus, G.J.A.

New Eng. J. Med., 1993 328 1313-1316 27. Thun, M.J., Namboodiri, M.M., Calle, E.E. ,

Flanders, W.D. and Heath, C.W.

"Aspirin use and risk of fatal cancer"

Cancer Res., 1993 53 1322-1327

28. Mitchell, J.A., Akarasereenont, P., Thiemermann, C. ,

Flower, R.J. and Vane, J.R.

"Selectivity of nonsteroidal antiinflammatory drugs as inhibitors of constitutive and inducible

cyclooxygenase"

Proc. Natl. Acad. Sci. USA., 1994 90 11693-11697

29. R.J.A. Wanders et al

Lancet, 1989 ii 52

30. K.M. Starko, C.G. Ray, L.B. Dominguez,

W.L. Stromberg, D.F. Woodall

Pediatrics, 1980 66 859

31. Uchida, Y., Izai, K., Orii, T. and Hashimoto, T.

"Novel fatty acid β-oxidation enzymes in rat liver mitochondria. II. Purification and properties of enoyl-coenzyme A (CoA) hydratase/3-hydroxyacyl-CoA dehydrogenase/3-ketoacyl-CoA thiolase trifunctional enzyme"

J. Biol. Chem., 1992 267 1034-1041.

Claims

CLAIMS :

1. A method of identifying a compound having the ability to block the low-affinity gastrin-cholecystokinin type C receptor, comprising the step of measuring the ability of said compound to block the binding to gastrin-binding protein of a compound selected from the group consisting of a gastrin-related peptide, a cholecystokinin-related peptide, an antagonist of gastrin or of

cholecystokinin, an acyl CoA, an enoyl CoA, an antibody to gastrin, and an antibody to cholecystokinin.

2. A method according to Claim 1 in which the binding is irreversible.

3. A method according to Claim 2 in which inhibition of irreversible binding is assessed by measuring the ability of the compound to block cross-linking of a gastrin

analogue to gastrin binding protein.

4. A method according to Claim 3 in which the cross-linking is effected by a homobifunctional cross-linking agent.

5. A method according to Claim 4 in which the

homobifunctional cross-linking agents is selected from the group consisting of disuccinimidyl suberate,

disulfosuccinimidyl suberate, and ethylene glycol bis-(succinimidyl)-succinate.

6. A method according to Claim 3 in which the cross-linking is effected by a heterobifunctional cross-linking agent.

7. A method according to Claim 6 in which the

heterobifunctional cross-linking agent is succinimidyl-4-(p-maleimidophenyl) butyrate.

8. A method according to Claim 3 in which the cross-linking is effected by a photoreactive cross-linking agent.

9. A method according to Claim 8 in which the

photoreactive agent is N-hydroxysuccinimidyl-4-azidobenzoate.

10. A method according to any one of Claims 3 to 9 in which the detectable marker is biotin or is a radioactive or fluorescent label.

11. A method according to Claim 1 in which the binding is reversible.

12. A method according to Claim 11 in which the ability of the putative antagonist to prevent binding of a known agonist or antagonist of very high affinity is measured, said known agonist or antagonist being labelled with a detectable marker.

13. A method according to Claim 12 in which the known agonist or antagonist has at least the binding affinity of benzotript.

14. A method according to Claim 12 or Claim 13 in which the detectable marker is ³H.

15. A method according to Claim 11 in which the ability of the putative antagonist to prevent binding of a high affinity antibody is measured, said antibody being labelled with a detectable marker.

16. A method according to Claim 15, in which the

antibody is capable of blocking binding of an acyl CoA or of gastrin to gastrin-binding protein.

17. A method according to Claim 16 in which the

antibody is a monoclonal antibody.

18. A method according to Claim 1 in which the

inhibition of the enoyl CoA hydratase and/or 3-hydroxyacyl CoA dehydrogenase activity intrinsic to the GBP is assayed.

19. A method according to Claim 18 in which the

hydratase or dehydrogenase activity is measured

spectrophotometrically or using a substrate labelled with a radioactive marker.

20. A method according to Claim 19 in which the activity is measured spectrophotometrically by the decrease in enoyl CoA absorption on hydration or by the decrease in NADH absorption on oxidation.

21. A method according to any one of Claims 1 to 20 in which the gastrin-related peptide is [Leu¹⁵]gastrin₁₇ or

[Nle¹⁵]gastrin_2,17.

22. A method according to Claim 21 in which the gastrin-related peptide is [Nle¹⁵]gastrin_2,17.

23. A pharmaceutical composition of a compound with the ability to inhibit the binding of a gastrin analogue to the low-affinity gastrin-cholecystokinin type C receptor, comprising a pharmaceutically-effective amount of said compound and a pharmaceutically-acceptable carrier, with the proviso that the compound is not idomethacin, sulindac, piroxicam, aspirin, proglumide or benzotript.

24. A composition according to Claim 23, in which the compound is a non-selective antagonist of the gastrin-cholecystokinin receptor.

25. A composition according to Claim 24 in which the antagonist has an 1C₅₀ value in the millimolar range or lower.

26. A composition according to any one of Claims 23 to 25 which additionally comprises another pharmaceutically-active molecule.

27. A composition according to Claim 26 in which the other pharmaceutically-active molecule is an antacid.

28. A composition according to Claim 26 in which the other pharmaceutically-active molecule is an anti-cancer agent.

29. A composition according to Claim 28 in which the anti-cancer agent is a cytokine.

30. A method of controlling gastric acid secretion, comprising the step of administering to a mammal in need of such treatment an effective amount of a compound having the ability to inhibit binding of a gastrin peptide to gastrin-binding protein.

31. A method according to Claim 30, comprising the step of administering a composition according to Claim 27.

32. A method of treating neoplastic disease comprising the step of administering to a mammal in need of such treatment an effective amount of a compound having the ability to inhibit binding of a gastrin peptide to gastrin-binding protein, with the proviso that the compound is not indomethacin, sulindac, piroxicam, aspirin, proglumide or benzotript.

33. A method according to Claim 32, in which the neoplastic disease is a cancer of the gastrointestinal tract.

34. A method according to Claim 32, in which the neoplastic disease is selected from the group consisting of colon carcinoma, gastric carcinoma, mammary carcinoma, malignant melanoma, tumours of epithelial origin, and lymphomas .

35. A method according to any one of Claims 32 to 34, comprising the step of administering a composition

according to Claim 28 or Claim 29.

36. A method according to any one of Claims 32 to 35, in which the compound having the ability to inhibit binding of the gastrin peptide to gastrin-binding protein is a non-steroidal anti-inflammatory compound.

37. A method according to Claim 36 in which the non-steroidal anti-inflammatory compound is an inhibitor of prostaglandin activity.

38. A method of treatment of a neoplastic disease, comprising the step of inducing production of antisense gastrin mRNA in neoplastic tissue of an animal in need of such treatment.

39. A method according to Claim 38, in which tissue-specific production of anti-sense mRNA is achieved by targeting of corresponding DNA to the desired tissue, using tissue-specific promoters.

40. A method according to Claim 38 or Claim 39, in which the neoplastic disease is selected from the group

consisting of colon carcinoma, gastric carcinoma, other gastrointestinal cancers, mammary carcinoma, malignant melanoma, tumours of epithelial origin, and lymphomas.