Genomic and Pathway Connectivity Analyses Identify Novel Strategies to Overcome mTOR Inhibitor Resistance In DLBCL

Abstract 436

mTOR inhibitors have been used with clinical success in solid tumors and non-Hodgkin lymphoma (NHL), and are attractive therapeutic options for DLBCL (diffuse large B-cell lymphoma, which has been shown to have constitutively active mTOR signaling). However, resistance to this class of agents remains problematic, and mechanisms of resistance are poorly understood. We performed candidate drug discovery using connectivity mapping and global gene expression profiling (GEP) to understand the pathways and genes responsible for resistance to the mTOR inhibitor Rapamycin (Sirolimus), which is the active metabolite of several clinically available mTOR inhibitors (eg, Temsirolimus, Everolimus). Treatment of DLBCL cell lines by Rapamycin at varying doses permitted stratification of cell lines into 2 groups of 3 cell lines each: sensitive (SU-DHL6, WSU-NHL, and Karpas-422) and resistant (SU-DHL4, OCI-Ly19, and Farage). Using the Affymetrix Human Gene 1.0 ST Array, we generated a profile of 1164 differentially-expressed genes (P<0.01) in the resistant cell lines. Pathway analysis of this particular gene expression signature enriched most strongly for the networks “EIF2 signaling” and “Regulation of eIF4 and p70S6K,” both of which are known to be involved in the PI3K and mTOR/AKT pathway. The genes thus identified present novel opportunities to understand and overcome resistance to mTOR inhibitors in DLBCL and other cancers. The Connectivity Map (CMAP) database contains a reference collection of more than 7,000 expression profiles from cultured human cells treated with bioactive small molecules, together with pattern-matching software to mine these data. We next analyzed the differentially-expressed genes associated with mTOR inhibitor resistance with the CMAP database in order to identify compounds likely to reverse the profile associated with resistance. From over 6,000 agents, the top 2% (by connectivity score) contained two PI3K inhibitors (Wortmannin and LY-294002), the protease inhibitor Saquinavir, and multiple HDAC inhibitors (including both Vorinostat and Trichostatin-A in the top 40 drugs). Among protease inhibitors, Nelfinavir (and to a lesser extent Saquinavir), has been shown to have potent cytotoxicity in a variety of solid tumors, by inhibition of the AKT signaling pathway. To validate the hypothesis that modulation of AKT might help overcome mTOR inhibitor resistance, we targeted AKT with two agents: Nelfinavir and MK-2206. We found that Nelfinavir demonstrated significant cytotoxicity at clinically achievable levels in all DLBCL cell lines tested (including those resistant to Rapamycin), and inhibited phosphorylation of AKT and downstream proteins (including p70S6 kinase; S6 ribosomal protein; 4-EBP-1) in a dose-dependent fashion. Baseline total AKT and phosphorylated AKT levels correlated with degree of sensitivity to Nelfinavir. Inhibition of downstream mTOR signaling by Rapamycin synergized with Nelfinavir in cell kill and inhibition of cell cycle progression. MK-2206, an AKT inhibitor which has shown success in early-phase clinical trials, was evaluated in the same panel of cell lines and likewise demonstrated synergism with Rapamycin in cytotoxicity and cell cycle inhibition. The degree of synergism between Rapamycin and either Nelfinavir and MK-2206, as calculated using the Chou-Talalay equation, was comparable. We have also demonstrated synergy between Nelfinavir and doxorubicin, a key component in commonly utilized regimens for AIDS lymphoma patients such as CHOP (Cytoxan, Adriamycin, Oncovin and Prednisone). We are now validating these in vitro results in a mouse xenograft model of DLBCL. In conclusion, our study demonstrates that AKT inhibition by Nelfinavir results in potent cytotoxicity in DLBCL cell lines at clinically relevant doses. Our results may have implications for combination therapy beyond NHL in non-hematologic malignancies where mTOR inhibitors and MK-2206 are being used independently with clinical success. Furthermore, the synergistic combination of either Nelfinavir or MK-2206, along with Rapamycin, may permit use of lower doses of each drug to therapeutically inhibit mTOR/AKT signaling while potentially reducing toxicity from off-target effects from the individual drugs. Finally, the use of Nelfinavir has particular relevance in AIDS patients with DLBCL, where the drug has both anti-viral and anti-lymphoma potential.