Inhibition Of PI3K-δ and -γ Isoforms By IPI-145 In Chronic Lymphocytic Leukemia Overcomes Signals From PI3K/AKT/S6 Pathway and Promotes Apoptosis

The functional relevance of the B cell Receptor (BCR) pathway and identification of protein kinases as therapeutic targets have recently shifted the paradigm for treatment of B cell malignancies. Inhibition of protein and lipid kinases (Bruton Tyrosine Kinase [BTK] and phosphoinositide 3-kinase [PI3K]) with ibrutinib and GS-1101 has been shown to be active in treatment of chronic lymphocytic leukemia (CLL). Importantly, differential expression and function of PI3K isoforms support isoform-selective inhibition of this kinase in CLL. Whilst PI3K-α and PI3K-β are ubiquitously expressed, PI3K-δ and PI3K-γ are primarily restricted to leukocytes. Since CLL cells generally express high levels of active PI3K-δ, great interest has been focused on inhibition of PI3K-δ. However, given the distinct and non-overlapping roles of PI3K-δ and PI3K-γ in immune cells, exploration of the therapeutic potential of combined inhibition of both PI3K-δ and PI3K-γ in CLL patients is warranted.

IPI-145 is a potent, orally bioavailable, inhibitor of PI3K-δ and PI3K-γ isoforms with K_D values of 0.023 nM and 0.24 nM, respectively. Treatment of primary CLL cells (n=51) with IPI-145 (1 µM) resulted in significant apoptosis (median 33%; range 12 – 40%). Patients with mutated (n=13) or unmutated IGHV gene status (n=13), previously untreated (n=21) or treated (n=8), displayed no significant difference in apoptosis from IPI-145. Samples with different prognostic markers such as 13q (del) or FISH negative samples were equally sensitive to IPI-145. Side by side studies of IPI-145 with ibrutinib and GS-1101, revealed that IPI-145 is comparatively potent (IC₅₀ 7.6 µM, compared to >10 µM) in promoting apoptosis. Crosslinking with anti-IgM enhanced the survival of primary CLL cells in association with activation of PI3K-δ,γ/AKT^Ser473/pBad^Ser136/S6^Ser235/236 pathway, which was in turn mitigated upon treatment with IPI-145 (n=9). Consistent with cell death, cleavage of PARP and decrease in anti-apoptotic protein Mcl-1 (but not Bcl-2 or Bcl-xL) was observed. Measurement of the C-C chemokine, CCL3, a biomarker for BCR signaling inhibition in CLL, demonstrated 15 to 48 – fold increase upon anti-IgM stimulation, which was reversed when cells were treated with 1 µM IPI-145 (10 to 80-fold decrease; n=6). Alternatively, co-culturing CLL primary cells with bone marrow stromal cells to mimic the leukemic microenvironment induced the protein levels of all four Class I PI3K isoforms and downstream PI3K/AKT/S6 signaling axis, which was significantly attenuated by IPI-145.

To mimic the proliferative state in lymph node pseudofollicles, CLL cells were stimulated to proliferate with CD40L/IL-2/IL-10 and the effect of IPI-145 was measured. Both pAKT and Ki-67 expression were markedly inhibited in primary CLL cells at concentrations of IPI-145 in the low nanomolar range (EC50<10nM; n=2), suggesting a potent anti-proliferative effect of IPI-145 on CLL cells in the nodal environment. Given the significant role of the chemo-attractant, SDF-1, in the directed migration of B-cells, chemotaxis assay demonstrated reduction in migration of CLL cells towards SDF-1 in presence of IPI-145 (% control reduction - median 23%; range 2-42%; n=8). Furthermore, IPI-145 treatment enhanced production of reactive oxygen species (n=6). Taken together, these results demonstrate the potential of combined inhibition of the PI3K-δ and -γ isoforms in CLL, and support clinical investigation of IPI-145 in B-cell malignancies, including CLL.