Abstract P2-05-04: Differential potency of TORC1/C2 (INK/MLN-128) and pan-PI3K (GDC0941) inhibitors on breast cancer polysome composition and phosphoprotein response biomarkers

The PI3K (phosphoinositide-3 kinase)/AKT (protein kinase B) pathway is dysregulated in over 70% of breast cancers, including more than 40% of non-basal breast cancers that possess activating PIK3CA mutations and 15% of basal breast cancers with functional loss of PTEN (n = 357 TCGA samples), suggesting that inhibitors of this signaling pathway may be clinically beneficial. Moreover, the recently reported BOLERO-2 trial (N Engl J Med, 2012) demonstrated the potential clinical benefit of using a TORC1 (target of rapamycin complex I) inhibitor, Everolimus, to treat patients with advanced non-basal, ER-positive breast cancers presumeably enriched with PIK3CA mutations. Among the new PI3K/AKT/TOR pathway-targeted therapeutics now under clinical development are the dual TORC1/C2 inhibitor, INK/MLN-128, and the pan-PI3K inhibitor, GDC0941. Despite having comparable nanomolar target affinities, it is unclear whether similar subsets of breast cancers will respond similarly to these agents, or if targeting this PI3K/AKT/TOR pathway either upstream (e.g. by GDC0941) or downstream (e.g. by INK/MLN-128) will yield comparable antitumor effects. Many of the downstream cellular responses regulated by this pathway, including tumor cell survival, growth, metabolism, invasiveness and angiogenic potential, are mediated by phosphorylation of polysome effector proteins, including the known TORC1 regulated eukaryotic translation initiation factor 4 binding protein (4eBP1) and 40S ribosomal protein S6. Therefore, we compared the effects of INK/MLN-128 and GDC0941 on cell growth and polysome protein phosphorylation profiles using a model breast cancer cell line (SKBr3) with amplified upstream receptor (HER2) activation of the PI3K/AKT/TOR pathway. Continuous dosing of these two agents against SKBr3 cultures (0–625 nM x5 days) revealed a 50-fold greater anti-proliferative efficacy for INK/MLN-128, with IC 50 = 9nM compared to the GDC0941 IC 50 = 479nM. This difference in antiproliferative activity correlated with their 50-fold different doses required to inhibit polysome formation (relative to free 40S, 60S, and 80S ribosomes) and polysomal 40S Rack1 and 60S RPL24 content. Likewise, INK/MLN-128 more potently inhibited the phosphorylation of two polysome-associated proteins of 32 KD and 60 KD size, recognized by an antibody toward a motif phosphorylated by AKT and related kinases (RXXS*/T*); the 32 KD substrate has been identified as S6. With equipotent dosing (0.1 mM INK/MLN-128 and 5 mM GDC0941), these two agents showed comparable inhibitory effects on polysome formation and protein phosphorylation. In summary, in a breast cancer model bearing upstream receptor activation of the PI3K/AKT/TOR pathway, distal pathway inhibition of TORC1/C2 by INK/MLN-128 appeared more potent and effective than more proximal inhibition of pan-PI3K by GDC0941. Furthermore, polysome inhibitory effects following INK/MLN-128 exposure were apparent within 8 h of treatment, and the most sensitive polysome response biomarker for this agent appeared to be phospho(T389)-S6 kinase, providing rationale for the use of this response biomarker in future clinical trials. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P2-05-04.