Abstract
Structural requirements for binding of gastrin to the human CCK2 receptor and the existence of receptors for gastrin peptides on human colon cancer cells were investigated. CCK2 receptor studies focused on the importance of the C-terminal Phe and putative P- tum region of gastrin 17 (G17). Gastrin analogs were synthesized in which either Phe at position 17 was removed or replaced with various amino acids or Gly at position 13 was replaced with constrained amino acids. Binding of the peptides to CHO cell membranes expressing the human CCK2 receptor was assessed in competition with [3H]CCK8. The biological activity of peptides that competed with ['HJCCK8 was determined by their ability to stimulate guanine nucleotide exchange on CHO cell membranes. The secondary structure of peptides modified at position 13 was investigated using MD, ECD Spectroscopy, and VCD Spectroscopy.|The role of gastrin peptides in colon cancer was investigated using the human colon cancer cell lines DLD-1 and HT-29. Binding of CCK8 and gastrin peptides to membrane preparations and whole cells, in competition with [3H][Leul5]G17-Gly and [l0I][Leu‘5]G17-Gly, was assessed. Binding of ['HJCCK8 to DLD-1 membranes was also examined. The ability of peptides to stimulate the growth of the cancer cells was determined by cell counting.|Removal of Phe at position 17 of G17 resulted in complete loss of binding to the human CCK.2 receptor. A side chain larger than Leu is required for binding but aromaticity is not important. Additionally, neither the size nor the electron withdrawing/donating properties of the para-substituent of Phe affected binding. Altering side chain or backbone orientation either by side chain constraint or D-amino acid substitution results in a marked decrease in binding. The binding affinity is strongly correlated with the biological activity of each analog, demonstrating little role for the residue at position 17 in receptor activation. Based on these results, the use of des-Phe analogs of gastrin is not an appropriate strategy for the development of human CCK2 receptor antagonists.|Replacement of Gly13 with Aib, Acc, y-Lac, and Pro was well tolerated with little detriment to binding and biological activity in the case of the first three replacements and an order of magnitude increase in binding and biological activity with incorporation of Pro. MD simulations of [Leu'3]G(l 1-17), [Aib13, Leul5]G(l 1-17), and [Pro13, Leu1 ]G( 11-17) demonstrated more turn or bend character for the former peptide than for the other two. The analogs produced class C CD spectra in TFE and in the presence of SDS micelles suggesting the presence of either a type I or type III P-tum. The CD spectra of [Pro13, Leul:!]G(l 1-17) in the above-mentioned solvents shared many characteristics to the spectrum under purely aqueous conditions suggesting that a similar conformation is present in all three solvents. The spectra of the other peptides in various solvents shared fewer similarities. In addition, the near UV spectrum of [Pro13, Leul5]G(l 1-17) in TFE is nearly identical to that of [Leu^]G(6-l7) suggesting that the C- terminal region of these peptides is folded in a similar manner. The amide I' VCD spectra of [Pro13, Leul5]G(l 1-17) in SDS/D20, TFE-D,, and DMSO-D6 shared similar characteristics suggesting that a common secondary structure is present in all three solvent systems. The amide I’ VCD spectra of the other analogs varied considerably depending on the solvent. Comparison of the amide I' VCD spectra of [Pro13, Leu15]G(l 1-17) with the spectra of peptides known to form either type I and type III P- tums suggests that [Pro13, Leu1:,]G(l 1-17) forms a stable type I P-tum. The conformational stability of the [Pro13, Leul5]G(l 1-17) might account for its higher binding affinity and biological activity. These results suggest that gastrin’s biologically active conformation involves a type I P-tum that is important for the full biological activity of this peptide at the human CCK2 receptor.|[Leul3]G17 and [Leul;i]G17-Gly competed with [3H][Leul5]G17-Gly at both high and low affinity sites on DLD-1 membranes but only at a low affinity site on HT-29 membranes. [3H]-CCK-8 did not bind to either site on DLD-1 membranes and CCK-8 was unable to displace [H][Leu' ]G17-Gly from DLD-1 membranes and HT-29 membranes. G17 and G17-Gly competed with [125I][Leul5]G17-Gly at both a high and low affinity site on DLD-1 and HT-29 whole cells. Both peptides significantly stimulated the growth of HT-29 and DLD-1 cells in a dose dependent and biphasic manner with maximal stimulation observed at nanomolar concentrations of peptide. This investigation has helped to clarify the controversy surrounding the affinity of the G17- Gly receptor by demonstrating the existence of both high and low affinity receptors on human colon cancer cells. These results emphasize the importance of using appropriate radioligands and experimental conditions for detection of receptors at both ends of the affinity spectrum.