Dysregulation of the PI3K/AKT/mTOR pathway and histone deacetylases (HDACs) has been described in diffuse large B-cell lymphoma (DLBCL). Previous studies demonstrated that combined treatment with PI3K/AKT pathway signaling inhibitors (e.g., LY294002, or perifosine) and histone deacetylase inhibitors (HDACis, e.g., sodium butyrate, vorinostat) resulted in a dramatic induction of apoptosis in human myeloid leukemia cells (Rahmani et al., Oncogene 22:6231-42, 2003; Rahmani et al., Cancer Res 65:2422-32, 2005). Here we examined interactions between the clinically relevant dual PI3K/mTOR inhibitor BEZ235 and the pan-HDACi panobinostat (both Novartis) in DLBCL cells. Notably, combined treatment with BEZ235 (25-200 nM) and panobinostat (5-15 nM) resulted in sharp decreases in cell growth and viability and profound induction of mitochondrial dysfunction and apoptosis. These events were observed in various DLBCL cell lines including both GC-DLBCL (SUDHL4, SUDHL16, OCI-LY7) and ABC-DLBCL (HBL-1, TMD8) as well as in mantle cell lymphoma cells (JeKo-1). Enhanced lethality of this regimen was accompanied by a marked increase in cytochrome c and AIF release into the cytosol, caspase-3 activation, and PARP cleavage. It was also associated with down-regulation of Mcl-1, a pronounced increase in H3 and H4 acetylation, and up-regulation of phospho- gH2AX, an indicator of DNA damage (e.g., DNA double-strand breaks). In addition, panobinostat robustly induced p21CIP1 accumulation in DLBCL cells, an event that was largely abrogated by BEZ235. Of note, treatment with BEZ235 alone or in combination with panobinostat triggered a decrease in GSK3 phosphorylation and levels of its downstream target, b-catenin consistent with GSK3 activation. Interestingly, inhibition of GSK3 by CHIR-98014 or the GSK3 inhibitor IX (BIO), but not its inactive analogue MeBIO, significantly diminished BEZ235/panobinostat lethality, arguing that GSK3 activation plays a significant functional role in lethality. Immunoprecipitation studies revealed that down-regulation of Mcl-1 was associated with enhanced binding of Bim to Bcl-xL and Bcl-2 and a marked decrease in Bak binding to Bcl-xL. In addition, knockdown of Bak also markedly diminished BEZ235/panobinostat-mediated lethality, as assessed by Annexin V/PI positivity. Together, these findings suggest that Bak plays a key functional role in the pronounced activity of BEZ235/panobinostat toward DLBCL cells. They also raise the possibility that BEZ235 may enhance panobinostat lethality by increasing Bim binding to Bcl-xL/Bcl-2, leading to the release of Bak/Bax from Bcl-2/Bcl-xL, culminating in apoptosis. Significantly, HS-5 stromal cell-conditioned media failed to protect DLBCL cells from combined panobinostat/BEZ235 treatment, suggesting that this strategy may be effective in circumventing microenvironmental forms of resistance. Combined treatment also exhibited robust activity against primary DLBCL cells, whereas exposure to the same regimens did not significantly reduce the viability of normal CD34+ progenitor cells nor did it reduce their clonogenic potential. Finally, in vivo studies utilizing a murine xenograft model bearing SUDHL4 cells revealed that co-treatment with BEZ235 and panobinostat markedly reduced in vivo tumor growth, whereas agents administered individually exhibited only modest effects. In addition, Kaplan-Meier analysis revealed that combined treatment significantly prolonged the survival of mice; in contrast, single agents were ineffective in increasing survival. Together, these findings suggest that the anti-DLBCL activities of combined BEZ235 and panobinostat treatment may involve multiple mechanisms, including Mcl-1 down-regulation, increased Bim binding to Bcl-2/Bcl-xL, release of Bak from Bcl-xL, abrogation of p21CIP1 accumulation, induction of DNA damage, and GSK3 activation, culminating in Bax/Bak activation and apoptosis. These findings raise the possibility that combining PI3K/AKT/mTOR inhibitors (e.g., BEZ235) and HDACis (e.g., panobinostat), may represent a novel and effective strategy against various DLBCL subtypes and possibly other hematologic malignancies.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution