Atovaquone-induced oxidative stress activates pentose phosphate pathway and immunogenic cell death in ovarian cancer

Atovaquone, an FDA-approved oxidative phosphorylation (OXPHOS) inhibitor, has shown promise in treating epithelial ovarian cancer (EOC), the deadliest gynecologic malignancy. However, the precise mechanisms underlying its antitumorigenic effects remain unclear. We employed a longitudinal transcriptomic approach to characterize the molecular effects of atovaquone on EOC cells. Our findings demonstrate that atovaquone disrupts cellular homeostasis and metabolism, activates stress responses, and primes immune recognition. We observed temporal downregulation of genes and pathways involved in key cellular processes, such as cell cycle and DNA replication, correlating with reduced proliferative capacity. Atovaquone also downregulated both OXPHOS and glycolysis, while upregulating the pentose phosphate pathway, suggesting a metabolic shift toward redox homeostasis restoration in response to severe oxidative stress. Consistent with oxidative stress, we found that atovaquone activated endoplasmic reticulum (ER) stress, which is linked to immunogenic cell death. During ER stress, calreticulin, a Damage-Associated Molecular Pattern (DAMP), translocates to the plasma membrane, where it promotes immune recognition. We observed that calreticulin was upregulated on the plasma membrane of atovaquone-treated EOC cells. Additionally, we detected increased levels of other DAMPs, such as High Mobility Group Box 1 (HMGB1) and Mitochondrial Transcription Factor A (TFAM), in the supernatant of atovaquone-treated cells, indicating the release of immunogenic molecules. Moreover, increased expression of ligands for activating receptors of NK cells was observed and co-culture experiments showed enhanced NK cell activity toward atovaquone-treated cells. These results highlight atovaquone’s potential as an adjuvant therapy that activates immune responses, offering a new avenue for combination therapies in EOC treatment. Bulk RNA-seq profiling of OVCAR3 cancer cells in the presence of DMSO, 30μM Atovaquone, and 45μM Atovaquone at 1, 8, 24, and 48 hours, as well as untreated OVCAR3 cells at 24 hours. Samples were sent in three batches at different times. All observations at 48 hours were in the third batch, batch C, and all samples in batch C were taken at 48 hours. This batch had 3 replicates each of DMSO-, 30μM-, and 45μM-treated samples. The first and second batch, A and B respectively, contained each of 1, 8, and 24 hours combined with each of 30μM, 45μM, and DMSO, as well as untreated cells observed at only 24 hours. For each such experimental condition, batch A had one biological replicate, and batch B had two.