Abstract
We determined whether the approved myelofibrosis drug ruxolitinib (Jakafi(®)), an inhibitor of Janus kinases 1/2 (JAK1 and JAK2), could be repurposed as an anti-cancer agent for solid tumors. Ruxolitinib synergistically interacted with dual ERBB1/2/4 inhibitors to kill breast as well as lung, ovarian and brain cancer cells. Knock down of JAK1/2 or of ERBB1/2/3/4 recapitulated on-target drug effects. The combination of (ruxolitinib + ERBB1/2/4 inhibitor) rapidly inactivated AKT, mTORC1, mTORC2, STAT3, and STAT5, and activated eIF2α. In parallel, the drug combination reduced expression of MCL-1, BCL-XL, HSP90, HSP70, and GRP78, and increased expression of Beclin1. Activated forms of STAT3, AKT, or mTOR prevented the drug-induced decline in BCL-XL, MCL-1, HSP90, and HSP70 levels. Over-expression of chaperones maintained AKT/mTOR activity in the presence of drugs and protected tumor cells from the drug combination. Expression of dominant negative eIF2α S51A prevented the increase in Beclin1 expression and protected tumor cells from the drug combination. Loss of mTOR activity was associated with increased ATG13 S318 phosphorylation and with autophagosome formation. Autophagosomes initially co-localized with mitochondria and subsequently with lysosomes. Knock down of Beclin1 suppressed: drug-induced mitophagy; the activation of the toxic BH3 domain proteins BAX and BAK; and tumor cell killing. Knock down of apoptosis-inducing factor (AIF) protected tumor cells from the drug combination, whereas blockade of caspase 9 signaling did not. The drug combination released AIF into the cytosol and increased nuclear AIF: eIF3A co-localization. A 4-day transient exposure of orthotopic tumors to (ruxolitinib + afatinib) profoundly reduced mammary tumor growth over the following 35 days. Re-grown tumors exhibited high levels of BAD S112 phosphorylation and activation of ERK1/2 and NFκB. Our data demonstrate that mitophagy is an essential component of (ruxolitinib + ERBB inhibitor) lethality and that this drug combination should be explored in a phase I trial in solid tumor patients.