Inhibition of PIM Kinases in Diffuse Large B-Cell Lymphoma Cells Targets MYC-Dependent Transcriptional Program, Increases CD20 Expression and Augments the Efficacy of Anti-CD20 Antibodies

R-CHOP immunochemotherapy remains standard frontline therapy for newly diagnosed diffuse large B-cell lymphoma (DLBCL) patients. However, this therapy is ineffective in approximately 1/3 of patients, underscoring the need for better treatment modalities. Targeting DLBCL oncogenic drivers is a promising strategy to improve the treatment efficacy and outcome. Although MYC transcription factor is one of the key oncogenes in DLBCL development, direct MYC targeting strategies have been largely ineffective, highlighting the need for other, indirect approaches. For example, MYC expression is stabilized by PIM serine-threonine kinases, indicating that PIM inhibition might be a rational approach to indirectly target MYC. In this study, we assessed the PIM-MYC relationship and the consequences of PIM inhibition in DLBCL. We first evaluated the expression of PIM1-3 and MYC proteins in 57 DLBCL diagnostic sections by immunohistochemistry. In this series, 70.17% of specimens were positive for at least one PIM isoform and 84.22% cases were MYC-positive. 100% of cases with high MYC expression (MYC present in ≥30% of the cells, n=35) were PIM-positive, whereas 86,36% of cases with undetectable or low MYC expression (MYC detected in ≤20% of cells, n= 22) were PIM-negative (Fisher's exact test, p<0.0001). Since the coexpression of MYC and PIMs highlights the functional link between these proteins in DLBCLs, we evaluated the expression of PIM kinases in cell lines following siRNA-mediated MYC knockdown or treatment with MYC-MAX dimerization inhibitor, 10058F4. The genetic or chemical MYC inhibition markedly decreased PIM1-3 expression in six GCB and ABC cell lines. Likewise, knockdown of all three PIM isoforms decreased MYC levels, attenuated proliferation and induced apoptosis. Similarly, PIM blockade with SEL24/MEN1703, a novel pan-PIM/FLT3 inhibitor tested currently in clinical trial in AML patients and exhibiting favorable safety profile, decreased the expression of multiple MYC-dependent genes. To assess the MYC role in PIM inhibitor-mediated toxicity, we generated DHL4 cells expressing degradation-resistant MYC_T58A mutant. MYC_T58A expression partially protected cells from PIM inhibitor-induced proliferation arrest and apoptosis, indicating that the inhibitor's toxicity is at least partially mediated by MYC depletion.

The MS4A1 gene, encoding CD20 surface antigen and rituximab target, is regulated by an upstream promoter containing potential MYC-binding sites (E-boxes). MYC association to these regions was confirmed in chromatin immunoprecipitation assays. As expected, in SEL24/MEN1703-treated cells, MYC occupancy at the MS4A1 promoter markedly decreased. To determine the consequences of MYC binding to the MS4A1 promoter, we assessed CD20 levels in a lymphoblastoid cell line carrying tetracycline-regulated (tet-off) MYC. MYC repression markedly elevated transcript and surface CD20 levels in a time-dependent manner, reaching 17.3-fold (transcript) and 3.82-fold (surface) inductions at 96 h. Consistently, the pan-PIM inhibitor decreased MYC expression in DHL4 and RAJI cells and resulted in increased surface CD20 levels up to 3.72-fold of baseline. In cells expressing the MYC_T58A mutant, PIM inhibition did not increase CD20 level, indicating that PIM kinases modulate CD20 surface expression via MYC. Importantly, PIM inhibitors increased CD20 levels also in primary, patient-derived DLBCL cells. These data suggest that indirect MYC targeting via PIM inhibition would lead to increased rituximab activity. Indeed, in PIM inhibitor-treated DHL4 and RAJI cells, rituximab triggered higher complement-dependent toxicity. Likewise, PIM inhibitor potentiated rituximab-dependent uptake of DHL4 and DHL6 cells by human monocyte-derived macrophages in antibody-dependent cellular phagocytosis assay.

Taken together, we characterize a PIM-MYC regulatory circuit promoting DLBCL growth and resistance to anti-CD20 antibody. We also demonstrate that PIM inhibition exhibits pleiotropic effects that combine direct cytotoxicity with increased surface CD20 levels and increased susceptibility to anti-CD20 antibody-based therapies.

Study supported by Foundation for Polish Science (POIR.04.04.00-00-5C84/17-00), Polish National Science Centre (2016/22/M/NZ5/00668 and 2017/26/D/NZ5/00561) and Ministry of Science and Higher Education in Poland (iONCO) grants.