Antitumour activity of melatonin in a mouse model of human prostate cancer : relationship with hypoxia signalling

Melatonin is known to exert antitumour activity in several types of human cancers, but the underlying mechanisms as well as the efficacy of different doses of melatonin are not well defined. Here, we test the hypothesis whether melatonin in the nanomolar range is effective in exerting antitumour activity in vivo and examine the correlation with the hypoxia signalling mechanism, which may be a major molecular mechanism by which melatonin antagonizes cancer. To test this hypothesis, LNCaP human prostate cancer cells were xenografted into seven-wk-old Foxn1nu/nu male mice that were treated with melatonin (18 i.p. injections of 1 mg/kg in 41 days). Saline-treated mice served as control. We found that the melatonin levels in plasma and xenografted tissue were 4× and 60× higher, respectively, than in control samples. Melatonin tended to restore the redox imbalance by increasing expression of Nrf2. As part of the phenotypic response to these perturbations, xenograft microvessel density was less in melatonin-treated animals, indicative of lower angiogenesis, and the xenograft growth rate was slower (P < 0.0001). These changes were accompanied by a reduced expression of Ki67, elevated expression of HIF-1α and increased phosphorylation of Akt in melatonin than saline-treated mice. We conclude that the beneficial effect of melatonin in reducing cancer growth in vivo was evident at melatonin plasma levels as low as 4 nm and was associated with decreased angiogenesis. Higher HIF-1α expression in xenograft tissue indicates that the antitumour effect cannot be due to a postulated antihypoxic effect, but may stem from lower angiogenesis potential.