Accelerated tumor growth under intermittent hypoxia is associated with hypoxia-inducible factor-1-dependent adaptive responses to hypoxia

Mounting evidence has revealed a causative role of intermittent hypoxia (IH) in cancer progression in mouse models of obstructive sleep apnea (OSA), but most studies have focused on the effects of IH following tumor implantation using an exposure to single IH frequency. Thus, we aimed to investigate 1) the potential effect of IH on the initial tumor growth in patients with OSA without an interaction with other mechanisms induced by IH in mice and 2) the influence of the IH frequency on tumor growth, which were tested using pre-conditioning with IH (Pre-IH) and 2 different IH frequencies, respectively. Pre-IH was achieved by alternatively maintaining melanoma cells between normoxia (10 min, 21% O(2)) and hypoxia (50 min, 1% O(2)) for 7 days (12 cycles per day) before administering them to mice. The conditions for IH-1 and IH-2 were 90 s of 12% FiO(2) followed by 270s of 21% FiO(2) (10 cycles/h), and 90 s of 12% FiO(2) and 90 s of 21% FiO(2) (20 cycles/h), respectively, for 8 h per day. Tumor growth was significantly higher in the Pre-IH group than in the normoxia group. In addition, the IH-2 group showed more accelerated tumor growth compared to the normoxia and IH-1 groups. Immunohistochemistry and gene-expression results consistently showed the up-regulation of molecules associated with HIF-1α-dependent hypoxic adaptation in tumors of the Pre-IH and IH-2 groups. Our findings reveal that IH increased tumor progression in a frequency-dependent manner, regardless of whether it was introduced before or after in vivo tumor cell implantation.