The biological activity of 16-epoxy side-chain analogs of 1 alpha,25-d
ihydroxyvitamin D-3 (1 alpha,25(OH)(2)D-3) was evaluated in vitro and
in vivo. Compared to 1 alpha,25(OH)(2)D-3, all analogs had lower affin
ities for the pig duodenal vitamin D receptor and also for the human s
erum vitamin D binding protein. The in vitro effects on cell prolifera
tion or differentiation of human promyeloid leukemia (induction of sup
eroxide production in HL-60 cells), human osteosarcoma MG-63 cells (os
teocalcin secretion), or human breast cancer cells (incorporation of t
hymidine in MCF-7 cells), was markedly inhibited by several epoxy anal
ogs, compared to 1 alpha,25(OH)(2)D-3, but the rank order of their act
ivity widely varied among different cancer cells. The most potent anal
ogs (24S,25S-24-hydroxy-25,26-epoxy-22-ene-1 alpha-OHD3, 25,26-epoxyl-
23-yne-1 alpha-OHD3 and 25,26-epoxy-23-yne-20-epi-1 alpha-OHD3 or comp
ounds 16, 5, and 7, respectively) were equipotent (16 and 5) or 30-fol
d (compound 7 on MG-63 cells) to 40-fold (compound 7 on MCF-7 cells) m
ore active than 1 alpha,25-(OH)(2)D-3. These analogs were nevertheless
poorly antirachitic (<3%) when tested in vitamin D-deficient chicks (
using serum and bone calcium, serum osteocalcin and duodenal calbindin
D-28K, as end points). Compound 7 was also 100-fold more active than
1 alpha,25-(OH)(2)D-3 in inhibition of proliferation of human foreskin
keratinocytes. Some epoxy analogs of 1 alpha,25-(OH)(2)D-3 thus displ
ay interesting dissociations between their receptor affinity and their
potency to induce cell differentiation, whereas their effect on cell
proliferation/differentiation exceed their calcemic effects more than
100- to 1000-fold.