SF1126, a Novel PI3K Inhibitor Results in Downstream Inhibition of the PI3K Axis and Displays Sequence Specific Synergy When Combined with Bortezomib in Multiple Myeloma Cells

Aim and Background: Two PI3K inhibitors frequently used in laboratory evaluations are LY294002 and Wortmannin. Both inhibit all class I PI3K enzymes but are poorly soluble and have off-target activity. This led to the development of SF1126, a soluble conjugate of LY294002 that localizes to the tumor vasculature in vivo, inhibits dex resistant xenografts (Garlich, Cancer Res. 2008) and SF-1126 is currently in Phase I trials. Combination therapy with PI3K inhibitors and conventional anti-myeloma therapies are intriguing since preclinical studies suggest that PI3K inhibitors exhibit synergistic effect with low concentrations of vincristine, a microtubule destabilizing agent by dephosphorylating GSK-3β in tumor cells (Fujiwara, Mol. Cancer Ther. 2007) in the induction of apoptosis. Thus, it is imperative to perform pre-clinical studies in order to optimize the clinical use of agents with broad effects such as SF1126.

Methods: Human myeloma cell lines and primary patient tumor cells were used in evaluating downstream effects of SF1126 as well as to evaluate the effect of combination therapy. There cell lines were used to evaluate the effect of SF1126 on global gene expression profiles. We have validated a set of RNA microarray analysis data obtained by SF1126 treatment alone or with bortezomib in myeloma cells and compared the inhibition of the upregulated or downregulated targets by timeline measurements of additions of SF1126 with bortezomib. RNA from MM cells was run on Illumina Human WG-6 v3 BeadChips at Emory microarray facility. MTT assays were used for the cell viability. AnnexinV staining was used to assess the level of apoptosis. Western blotting was performed using antibodies on various targets.

Results: Treatment of MM.1R, OPM1, MM.1S myeloma cells with clinically achievable concentrations of SF1126 (2–14 uM) showed a response rate of cell killing in the range of 20% by 72hrs. Bortezomib alone at 4nM showed a response rate of cell killing in less than 20%. When they were combined the cell killing increased to 50% or more demonstrating additive benefit. To maximize the additive benefit, this led us to combine them in a sequential manner. In sequence assays, drug adminstrations was staggered by SF1126 and bortezomib. Both concurrent and addition of bortezomib prior to SF1126 were superior to SF1126 given to bortezomib in the MM.1R cell line. Western blotting data suggests that adminstration of SF1126 first increased the expression of p21Waf1/Cip1 protein with a modest reduction of phos-Akt in all the MM cells tested. Sequence data were consistent among several cell lines and consistent with apoptosis assays and also in primary CD38⁺ myeloma cells. Interestingly, there was little overlap in the number of genes upregulated or downregulated by SF1126 and bortezomib suggesting complementary molecular profiles. Microarray data based on MM.1R, OPM1, H929 cell lines revealed that p21Waf/Cip1 was among the upregulated genes upon the SF1126 treatment. Additionally, other off-target genes such as BCl2, genes in Wnt-pathway as well as in Akt pathway responsible for the antiapoptotic response during sequence administration are under investigation. Combination of SF1126 with conventional agents such as melphalan. dexamethasone, and the Imid lenalidoamide also demonstrate encouraging combination effects in myeloma cells.

Conclusions: Sequence and combination administration could shed light in the preclinical setting to help guide the clinical evaluation of SF1126 as a single agent and in combination with bortezomib, dexamethasone, melphalan and lenalidoamide in myeloma.