Inhibition of prostate cancer metastasis in vivo: a comparison of 1,23-dihydroxyvitamin D (calcitriol) and EB1089.

The steroid hormone 1,25-dihydroxyvitamin D [1,25(OH)2D, also known as calcitriol] is known to inhibit the proliferation and to promote the differentiation of human prostate cancer cells. Additionally, we showed that 1,25(OH)2D markedly inhibits the invasiveness of human prostate cancer cells in vitro (G. G. Schwartz et al., Cancer Epidemiol. Biomark. Prev., 6: 727-732, 1997). These properties support the use of 1,25(OH)2D as differentiation therapy in prostate cancer. However, the use of 1,25(OH)2D in vivo is limited by the risk of hypercalcemia. We therefore compared the effects of 1,25(OH)2D and of EB1089, an analogue of 1,25(OH)2D with reduced calcemic effects, in an in vivo model of androgen-insensitive metastatic prostate cancer, the rat Dunning MAT LyLu prostate cancer model. Tumor growth and metastasis were studied using Copenhagen rats given s.c. injections of MAT LyLu cells. Fifty male rats were divided into five groups of 10 rats each. Four experimental groups received i.p. injections of low and high doses of 1,25(OH)2D and EB1089 (0.5 and 1.0 microg/kg, low and high, respectively). A control group received injections of vehicle only. Tumor volumes were measured three times per week. Rats were weighed weekly. The number of metastases to the lungs and the extent of hypercalcemia were evaluated. Compared with controls, tumor volumes were significantly smaller in all experimental groups. Similarly, the number of lung metastases (number of foci/lung) was reduced markedly by both 1,25(OH)2D and EB1089. Control rats developed 22.7 (+/- 1.98 SE) tumor foci per lung. Rats treated with 1,25(OH)2D and with EB1089 (1.0 microg/kg) developed 10.4 (+/- 2.81) and 7.70 (+/- 1.29) tumor foci, respectively (P < 0.001 and P < 0.0001, respectively; drug versus control). Compared with controls (10.79 +/- 0.1 mg/dl), serum calcium levels were significantly elevated in both 1,25(OH)2D and EB1089-treated rats (P < 0.01). However, EB1089 was significantly less calcemic than 1,25(OH)2D (12.59 +/- 0.21 mg/dl versus 14.47 +/- 0.46 mg/dl; 1.0 microg/kg; P < 0.001). Rats treated with 1,25(OH)2D showed marked weight loss: 20.0 +/- 1.9% and 26.3 +/- 1.7% of their initial weight (low and high doses, respectively, P < 0.001). Weight loss was significantly lower in rats treated with EB1089 at the high dose 8.4 (+/- 2.9) %. Moreover, rats treated with low-dose EB1089 gained 5.2 (+/- 3.7) % of their initial weight. In conclusion, 1,25(OH)2D and EB1089 showed marked and equivalent inhibition of prostate cancer metastasis in vivo. EB1089 was significantly less calcemic than 1,25(OH)2D and did not induce severe weight loss. This is the first report of a vitamin D analogue that significantly inhibits prostate cancer metastasis in vivo and that does so without producing cachexia or unacceptable hypercalcemia.

Antiproliferative effect of 1alpha,25-dihydroxyvitamin D3 in human prostate cancer cell line LNCaP involves reduction of cyclin-dependent kinase 2 activity and persistent G1 accumulation.

1Alpha,25-dihydroxyvitamin D3 (1,25 D), the most active metabolite of vitamin D3, exerts antiproliferative and prodifferentiating effects on some human prostate cancer cell lines. We previously reported an inverse relationship between functional vitamin D receptor (VDR) levels and antiproliferative response to 1,25 D in two human prostate cancer cell lines, LNCaP and ALVA 31. Although LNCaP cells are far more sensitive to growth inhibition by 1,25 D than ALVA 31 cells, LNCaP express approximately half the number of VDR as ALVA 31. Two other human prostate cancer cell lines studied, PC3 and DU145, express lower levels of functional VDR and are relatively insensitive to growth inhibition by 1,25 D. In this report, we investigated potential mechanisms of the variable antiproliferative activity of 1,25 D. In PC3 cells stably expressing VDR [PC3(VDR)] at levels comparable to LNCaP, 1,25 D treatment resulted in only moderate growth inhibition. These results further support the contention that VDR expression, although required, is not sufficient for maximal growth suppression by 1,25 D, as is exhibited by LNCaP cells. We did not detect 1,25 D-mediated DNA fragmentation after 4 days of 1,25 D treatment in either LNCaP or ALVA 31 cells. This result suggests that variability in 1,25 D sensitivity does not derive from differences in the capacity of these cells to undergo apoptosis in response to 1,25 D. Flow cytometry of propidium iodine-stained cells revealed that 48 h 1,25 D treatment of LNCaP cells resulted in a 2-fold decrease of cells in G2/M plus S phases and accumulation of LNCaP cells in the G1/G0 phase. This effect persisted for 72 h after 1,25 D removal. In contrast, 1,25 D did not significantly alter the cell cycle distribution of ALVA 31 or PC3(VDR) cells. Consistent with accumulation of cells in G1/G0, 1,25 D treatment of LNCaP cells resulted in decreased retinoblastoma protein phosphorylation, repressed E2F transcriptional activity, increased levels of the cyclin-dependent kinase (CDK) inhibitor p21(WAF1, CIP1), and decreased CDK2 activity. However, p21 messenger RNA levels were not altered, suggesting translational or posttranslational regulation of p21 by 1,25 D. In contrast, p21 was not detected in ALVA 31 or PC3(VDR) and was not induced by 1,25 D, consistent with the failure of 1,25 D to influence cell cycle distribution in these cells. These results suggest that variability in sensitivity to the antiproliferative effects of 1,25 D among prostate cancer cells is dependent, at least in part, on the integrity of the retinoblastoma pathway and in particular on p21 expression and 1,25 D regulation of CDK2 activity.

Vitamin D receptor content and transcriptional activity do not fully predict antiproliferative effects of vitamin D in human prostate cancer cell lines.

Prostate cancer cell lines exhibit variable growth suppression by the hormonal form of vitamin D3, 1,25-Dihydroxyvitamin D3 [1,25 (OH)2D] (1,25 D3). To understand the molecular basis for this differential sensitivity to 1,25 D3, we compared growth response to 1,25 d3, vitamin D receptor (VDR) content and VDR transcriptional activity in four well-characterized human prostate cancer cell lines: LNCaP, DU145, PC-3 and ALVA-31. In PC-3 and DU145 cells, relative lack of growth inhibition by 1,25 D3 (< 10% inhibition) correlates with very low levels of VDR (9-15 fmol/mg protein) compared to classical vitamin D3 target tissues (approximately 75-200 fmol/mg protein). Transfection of DU145 and PC-3 cells with a VDR cDNA expression vector is sufficient to establish growth sensitivity to 1,25 D3, suggesting that low VDR levels are responsible for the failure of these cell lines to respond to 1,24 D3. LNCaP cells are highly sensitive to growth inhibition by 1.25 D3 (approximately 55% inhibition) and contain approximately 2-3-fold more VDR (25 fmol/mg) than the relatively 1,25 D3-insensitive PC-3 and DU145 cell lines. However, ALVA-31 cells display less than 20% growth inhibition to 1.25 D3 although they contain the highest levels of VDR (45 fmol/mg) of the four cell lines. Thus, sensitivity to growth inhibition by 1,25 D3 does not correlate with VDR content in ALVA-31 and LNCaP cells. This lack of correlation between VDR density and growth responses to 1,25 D3 led us to investigate VDR-mediated gene transcription in these cell lines. We employed two different naturally occurring vitamin D response elements (VDREs) linked to a reporter gene. Reporter gene activation by 1,25 D3 correlated well with VDR content in all four cell lines. Therefore, compared to LNCaP cells, decreased sensitivity of ALVA-31 to growth inhibition by 1,25 D3 is not due to a decrease in the general transcriptional activity of VDR. We conclude that growth inhibition by 1,25 D3 in prostate cancer cells requires VDR but that this response is modulated by non-receptor factors that are cell line-specific.