Characterizing OXPHOS inhibitor-mediated alleviation of hypoxia using high-throughput live cell-imaging.

Background: Hypoxia is a common feature of many solid tumors and causes radiotherapy and immunotherapy resistance. Pharmacological inhibition of oxidative phosphorylation (OXPHOS) has emerged as a therapeutic strategy to reduce hypoxia. However, the OXPHOS inhibitors tested in clinical trials caused only moderate responses in hypoxia alleviation or trials were terminated due to dose-limiting toxicities. To improve the therapeutic benefit, FDA approved OXPHOS inhibitors (e.g. atovaquone) were conjugated to triphenylphosphonium (TPP) to preferentially target cancer cell's mitochondria. In this study, we evaluated the hypoxia reducing effects of several mitochondria-targeted OXPHOS inhibitors and compared them to non-mitochondria-targeted OXPHOS inhibitors using newly developed spheroid models for diffusion-limited hypoxia.

Methods: B16OVA murine melanoma cells and MC38 murine colon cancer cells expressing a HIF-Responsive Element (HRE)-induced Green Fluorescent Protein (GFP) with an oxygen-dependent degradation domain (HRE-eGFP-ODD) were generated to assess diffusion-limited hypoxia dynamics in spheroids. Spheroids were treated with IACS-010759, atovaquone, metformin, tamoxifen or with mitochondria-targeted atovaquone (Mito-ATO), PEGylated mitochondria-targeted atovaquone (Mito-PEG-ATO) or mitochondria-targeted tamoxifen (MitoTam). Hypoxia dynamics were followed and quantified over time using the IncuCyte Zoom Live Cell-Imaging system.

Results: Hypoxic cores developed in B16OVA.HRE and MC38.HRE spheroids within 24 h hours after seeding. Treatment with IACS-010759, metformin, atovaquone, Mito-PEG-ATO and MitoTam showed a dose-dependent reduction of hypoxia in both B16OVA.HRE and MC38.HRE spheroids. Mito-ATO only alleviated hypoxia in MC38.HRE spheroids while tamoxifen was not able to reduce hypoxia in any of the spheroid models. The mitochondria-targeted OXPHOS inhibitors demonstrated stronger anti-hypoxic effects compared to the non-mito-targeted OXPHOS inhibitors.

Conclusions: We successfully developed a high-throughput spheroid model in which hypoxia dynamics can be quantified over time. Using this model, we showed that the mitochondria-targeted OXPHOS inhibitors Mito-ATO, Mito-PEG-ATO and MitoTam reduce hypoxia in tumor cells in a dose-dependent manner, potentially sensitizing hypoxic tumor cells for radiotherapy.