Graham S. Baldwin

The role of gastrin and cholecystokinin in normal and neoplastic gastrointestinal growth

Abstract Gastrin and cholecystokinin (CCK) act as growth factors for the gastric mucosa and the pancreas, respectively. CCK is also responsible, via the CCK‐A receptor, for the pancreatic hyperplasia observed following the feeding of protease inhibitors or pancreaticobiliary diversion. Hypergastrinaemia does not increase the incidence of spontaneous gastrointestinal carcinoma, but does stimulate the proliferation of gastric enterochromaffin‐like cells via the gastrin/CCK‐B receptor, with a consequent increase in the incidence of gastric carcinoids. Whether gastrin influences mutagen‐induced gastrointestinal carcinogenesis is still controversial, but CCK clearly enhances the induction by carcinogens of acinar tumours in the pancreas. While gastrin increases xenograft growth of 50% of gastrointestinal tumours tested, effects on the proliferation of gastrointestinal tumour cell lines in vitro have been more difficult to demonstrate, perhaps because many cell lines are already maximally stimulated by autocrine gastrin. Gastrin mRNA and progastrin, but not mature amidated gastrin, have been detected in all gastrointestinal cell lines tested. Although cell proliferation is inhibited by gastrin/CCK receptor antagonists, the spectrum of antagonist affinities is not consistent with binding to either CCK‐A or gastrin/CCK‐B receptors. Definition of the molecular structure of the receptor involved in the autocrine loop may lead to novel therapies for gastrointestinal cancer.

Comparison of transferrin sequences from different species

1. Amino acid sequences of transferrins from eight species, from human to tobacco hornworm, have been compared. Eighty-four amino acids (12%) are invariant, including three of the four ligands for the N-terminal Fe3+ ion. 2. The most highly conserved regions of both lobes of transferrin are the internal beta-sheets of domains 1 and 2, and helices 5 and 7 which abut the Fe3+ binding site. Two small patches of conserved surface residues, which may be involved in receptor binding, have also been identified. 3. Phylogenies have been deduced from pairwise alignment of the sequences of the N- and C-terminal lobes independently. The phylogenies are consistent with the evolutionary tree derived from the fossil record, and with the observation that the gene duplication which created the N- and C-terminal lobes of transferrin occurred before the divergence of the mammalian and insect lines. 4. The phylogenies predict that the lactotransferrin family diverged some 200 Myr ago, after the separation of the lines leading to mammals and birds. In contrast, the phylogenies predict that melanotransferrin diverged before the separation of the mammalian and avian lines. 5. Sequence comparisons also suggest that the stoichiometry of the transferrin receptor:transferrin complex is 2:1.

Expression of gastrin, gastrin/CCK-B and gastrin/CCK-C receptors in human colorectal carcinomas.

To investigate further the presence of an autocrine proliferative loop involving gastrin in colorectal carcinomas and to clarify the receptor responsible, 102 human colorectal carcinomas and 10 hepatic metastases were investigated for the expression of the genes encoding gastrin, the gastrin/CCK-B receptor and the gastrin/CCK-C receptor. Levels of RNA expression were assayed by RNase protection assay. In addition, gastrin/CCK receptors on crude membranes of tumour tissue were assayed by radioligand binding. High-affinity gastrin/CCK-B receptors were not detected in any of the carcinomas investigated, whereas in 36% low-affinity binding was observed, consistent with the expression of the gastrin/CCK-C receptor. RNase protection assay detected the RNA for the gastrin/CCK-B receptor in 11% of the carcinomas investigated, whereas the RNA for the gastrin/CCK-C receptor was demonstrated in 75% and the RNA for gastrin in 86% of the carcinomas investigated. These results confirm the recent demonstration of progastrin fragments in colorectal carcinomas. One possible explanation for progastrin expression is that such progastrin fragments may participate in an autocrine proliferative loop. The receptor involved in this loop is more likely to be the low-affinity gastrin/CCK-C receptor rather than the gastrin/CCK-B receptor, which is rarely expressed in colorectal carcinomas.

BINDING OF PROGASTRIN FRAGMENTS TO THE 78 KDA GASTRIN-BINDING PROTEIN

The non‐selective gastrin/cholecystokinin receptor antagonists proglumide and benzotript inhibit colon carcinoma cell proliferation by binding to the 78 kDa gastrin‐binding protein (GBP) (Baldwin, Proc. Natl. Acad. Sci. USA, 91 (1994) 7593–7597). However, although most colon carcinoma cell lines synthesize progastrin, production of mature amidated gastrin17 has not been observed. In order to define the structural requirements for the binding of gastrin to the GBP the affinities of various fragments of amidated and C‐terminally extended gastrin17 for the GBP have been measured. The results indicate that the GBP recognizes both N‐ and C‐termini of gastrin17. Moreover since C‐terminal amidation is not a prerequisite for binding of gastrin to the GBP, the GBP is a potential target for the autocrine effects of progastrin.

PCR cloning and sequence of gastrin mRNA from carcinoma cell lines

The hypothesis that a gastrin-like peptide is acting as an autocrine growth factor in gastric and colonic carcinoma cell lines requires that the cells should synthesize a gastrin-like mRNA. Although no gastrin mRNA was observed in the gastric line Okajima or the colonic lines HCT 116 or LIM 1215 by Northern blotting, gastrin mRNA was detected by application of the polymerase chain reaction. Two products were observed corresponding to mRNA with and without a 130 bp intron. The sequences of both products were identical to the sequences predicted from the normal human gastrin gene.

Inhibition of cell proliferation by the cholecystokinin antagonist L-364,718

Specific cholecystokinin (CCK) receptor and gastrin receptor antagonists were used to assess what role, if any, these receptors have in autocrine cell growth. Although the cholecystokinin receptor antagonist, L-364,718, inhibited cell proliferation in a broad spectrum of cell lines, the gastrin antagonist, L-365,260, had no effect on cell proliferation. In addition neither added gastrin17, nor sulfated cholecystokinin8, could reverse the inhibitory action of L-364,718. It is proposed that L-364,718 inhibits cell proliferation independently of classical gastrin/CCK receptors.

Structures of the human cDNA and gene encoding the 78 kDa gastrin-binding protein and of a related pseudogene

The nucleotide sequence encoding the human 78 kDa gastrin binding protein (GBP) has been deduced from overlapping fragments generated by the polymerase chain reaction with oligonucleotides based on the sequence of the porcine GBP (Mantamadiotis, T. et al. (1993) Biochim. Biophys. Acta 1170, 211-215) and cDNA from the colonic carcinoma cell line LIM 1215 as template. The mature human GBP is 90% identical to the porcine GBP. Clones encoding the human GBP gene, which contains 19 exons, have been isolated from human genomic libraries. The positions of the exon/intron junctions are completely different from the junctions in the gene encoding the related peroxisomal trifunctional enzyme. Clones encoding a related pseudogene have also been isolated and sequenced.

Comparison of sequence of the 78 kDa gastrin-binding protein and some enzymes involved in fatty acid oxidation

1. 1. The amono acid sequences of enzymes possessing enoyl-CoA hydratase activity have been aligned with each other and with the sequences of related proteins. 2. 2. The amino acid sequences possessing 3-hydoxyacyl-CoA dehydrogenase activity have been similarly aligned. 3. 3. The N-terminal half of the 78 kDa gastrin-binding protein (Baldwin et al., J biol. Chem.261, 12,252–12,257, 1986) is related to the enoyl-CoA hydratase family, while the C-terminal half is related to the 3-hydroxyacyl-CoA dehydrogenase family. 4. 4. Evolutionary trees for the two families are presented.

Phosphorylation of a synthetic gastrin peptide by the tyrosine kinase of A431 cell membranes.

Abstract The peptide Arg.Arg.Leu.Glu.Glu.Glu.Glu.Glu.Ala.Tyr.Gly was synthesized as an analogue of residues 20–30 of human gastrin 34. The epidermal growth factor-stimulated tyrosine kinase of A431 cell membranes phosphorylated the peptides single tyrosine residue. K m values of 0.11 and 0.61mM and V max values of 1.71 and 0.68nmol/min/mg were obtained in the presence and absence of epidermal growth factor respectively. This is the first report of phosphorylation of tyrosine in a sequence related to a human hormone.

Cyclooxygenase-2-selective antagonists do not inhibit growth of colorectal carcinoma cell lines

Non-steroidal anti-inflammatory drugs (NSAIDs) reduce the incidence of colorectal carcinoma. We now report that the potent cyclooxygenase-1 inhibitor indomethacin had no effect on the growth of human colorectal carcinoma cell lines in vitro at concentrations up to 30 microM. The selective cyclooxygenase-2 inhibitors L-745337 and NS-398 reduced cyclooxygenase activity, but had no effect on cell growth at concentrations as high as 100 microM. Our results provide direct evidence that inhibition of cyclooxygenase activity does not necessarily inhibit the growth of colorectal carcinoma cell lines and imply that the growth-inhibitory effects of NSAIDs in vitro are not mediated by inhibition of cyclooxygenases.