Specific binding of cholecystokinin, estradiol and somatostatin to human pancreatic cancer xenografts.

We recently reported that human pancreatic cancers differentially respond to the growth inhibitory effects of an estradiol (E2) receptor antagonist, tamoxifen, and a long-acting analogue of somatostatin, Sandostatin. In the present study two human pancreatic cancers, established as xenografts in nude mice, were examined as representative of cancers that respond to either tamoxifen (PGER) or Sandostatin (SKI), for the presence of binding sites for various hormones. Male nude mice were inoculated with either SKI or PGER, by passage of tumor chunks (3 mm2) to the interscapular region. Tumors, obtained from mice after approximately 30 days of in vivo growth, were analyzed for binding to cholecystokinin-octapeptide (CCK), somatostatin and E2, by published procedures, using either crude tumor membranes (CCK), cytosol and nuclear fractions (E2), or cryostat sections of whole tumors (somatostatin). SKI was highly positive for high-affinity (Kd = approximately 1 nM) CCK binding sites at the time of resection with a binding capacity of approximately 1000 fmol/mg protein. With increasing passages, the total number of high-affinity binding sites for CCK, were reduced to non-detectable levels in SKI tumors, while non-saturable binding (Kd = greater than 10 nM) became increasingly evident. Early passages of PGER tumors were similarly positive for high-affinity binding sites for CCK, that steeply declined with increasing passages. Specific binding sites for E2, were observed only in the cytosolic fractions of PGER, with a high binding affinity (Kd = approximately 0.05 nM) and a low binding capacity (15 +/- 3 fmol/mg cytosolic proteins), at all passages examined; E2 binding sites were not detected in cytosolic and nuclear fractions of SKI and in the nuclei of PGER, at all passages. SKI and PGER at different passages were examined for somatostatin binding, and both the early and late passages of PGER were devoid of somatostatin binding sites, while SKI tumors were positive for them. Based on the above results, it appears likely that Sandostatin directly inhibited the growth of SKI tumors, since SKI was positive for somatostatin binding sites; it appears less likely that Sandostatin indirectly mediated its inhibition by attenuating possible stimulatory effects of CCK. Growth inhibitory effects of tamoxifen on PGER were apparently via E2 binding sites, since only the tumors positive for E2 binding sites (PGER) responded to tamoxifen; it remains to be determined if tamoxifen can exert additional effects independent of E2 binding sites on pancreatic cancers.(ABSTRACT TRUNCATED AT 400 WORDS)