ARF6 IS A CORE DRIVER OF MESENCHYMAL METASTASIS AND DRUG RESISTANCE OF CANCERS UNDER RTK AND GPCR SIGNALIN

The acquisition of malignant phenotypes by cancer cells often involves their transition into mesenchymal-type properties. Such mesenchymal-type malignancy involves resistance against anoikis to maintain cell viability during the distant metastasis, whereas mesenchymal phenotypes on its own includes highly invasive and metastatic potentials. Moreover, acquisition of mesenchymal properties appears to be closely related to therapeutic resistance. However, proteins that execute the mesenchymal malignancy, including therapeutic resistance, still remain largely elusive.

A series of our studies established that a small GTPase Arf6 and its signaling pathway employing AMAP1 drive invasion and metastasis under receptor tyrosine kinases in breast cancers, to be statistically correlated with poor overall survival of patients.

Here, we first examined whether the Arf6-AMAP1 pathway is linked to a mesenchymal property in order to investigate whether this pathway contributes to therapeutic resistance. We also investigated events that may initially trigger the generation of mesenchymal properties. Furthermore examined was whether clear cell renal cell carcinomas (ccRCCs), which like breast cancer are mostly originated from tubular epithelia, also utilize the Arf6-based pathway for malignancy.

We found that a mesenchymal-specific protein EPB41L5 is an integral binding partner of AMAP1 for invasion and metastasis, and also for therapeutic resistance of breast cancer cells. ccRCCs also utilized the same Arf6-based mesenchymal pathway, while in ccRCCs this pathway was activated by G-protein coupled receptors for lysophosphatidylic acid (LPA) via GTP-Ga12 and its binding to EFA6. The robust presence of this pathway in primary ccRCCs statistically correlated with poor overall survival of the patients. Moreover, the mevalonate pathway (MVP), which is activated such as by estrogen and by oncogenic p53, was essential for Arf6 activation by receptor tyrosine kinases. Consistently, the Arf6-based malignancy could be effectively blocked by MVP inhibitors, such as statins, in breast cancers, while it was inhibited by Ki16425, an inhibitor of LPAR1-3, in ccRCCs. Furthermore, not EPB41L5 on its own, but the whole Arf6 pathway containing EPB41L5 was required for therapeutic resistance. We identified a precise mechanism therein involved.

Our results establish essential roles of the Arf6-based mesenchymal pathway in the overall mesenchymal malignancy of breast cancers and ccRCCs; and demonstrate molecular bases regarding the usefulness of inhibitors of MVP and LPAR in their therapeutics. Furthermore, LPA can be produced extracellularly by autotaxin from lysophosphatidylcholine, which is abundant in body fluids. LPA is well known to activate Rho and Rac GTPases, and has long been believed to thereby promote cancer malignancy. We identified that LPA is a major factor promoting mesenchymal malignancy of large populations of primary ccRCCs, most of which face the body fluids; and that LPA does not require Rho, but activates another small-GTPase Arf6 in the malignancy development.