BET Protein Antagonist-Based Therapy of Novel Models of Richter Transformation-Diffuse Large B-Cell Lymphoma (RT-DLBCL)

Abstract

Richter Transformation (RT) is the development of aggressive DLBCL (mostly ABC sub-type) in up to ~15% of patients with antecedent CLL. Approximately 80% of RT-DLBCLs are clonally-related (CLR) to the underlying CLL, with a median survival (MS) of one year. Alternatively, ~20% of RT-DLBCLs are clonally-unrelated (CLUR) to the underlying CLL, exhibiting a better MS of 5 years. Chemo-immunotherapy, or monotherapy with the Bruton's tyrosine kinase (BTK) inhibitor ibrutinib, anti-apoptotic BCL2 inhibitor venetoclax or with anti-PD1 therapy fails to achieve prolonged disease-free survival, with a majority relapsing with therapy-refractory disease. To develop and test novel targeted therapies for RT-DLBCL, we successfully established three patient-derived xenograft (PDX) models of luciferized RT-DLBCLs, including the CLR HPRT3 and CLUR HPRT2. HPRT3 and HPRT2 cells were of ABC sub-type of DLBCL, based on positive MUM1/IRF4 and negative CD10 and BCL6 expressions. The third RT-DLBCL (HPRT1) was a rarer GCB variety of DLBCL, displaying high CD10, BCL6, Ki-67 and c-Myc expressions. FISH analysis confirmed 5' MYC amplification in HPRT1 cells. Cytogenetic and array-CGH analyses showed large numbers of karyotypic and genetic alterations in HPRT3 and HPRT2 RT-DLBCLs. Low-pass whole genome sequencing showed that HPRT3 and HPRT2, but not HPRT1, exhibit large areas of amplification of DNA on chromosome 18, with copy gains at 18q21.1 involving the TCF4 gene. NextGen sequencing (NGS) of a panel of 300 genes (L-300 panel) showed more genetic mutations (at high % VAF) in HPRT3, as compared to HPRT2 or HPRT1 cells. These genetic alterations targeted transcription factors, epigenetic regulators, DNA damage/repair enzymes, signaling enzymes and their regulators. Utilizing ATAC-Seq, ChIP-Seq with H3K27Ac and BRD4 antibodies, and single cell (sc) RNA-Seq analyses, we evaluated active chromatin and gene-expressions in the three RT-DLBCLs. ChIP-Seq-determined signal-density of H3K27Ac and BRD4, as well as ATAC-Seq-determined peak-density were increased at active SE/E of BCL2, TCF4, PAX5 and IRF4 in HPRT3 and HPRT2 cells. Notably MYC, BCL6 and CDK6 SEs were active in HPRT1 cells. scRNA-Seq generated t-SNE plots showed that HPRT3 cells exhibited highest number of transcriptionally active cell-clusters, with high mRNA expressions of TCF4, PAX5 and IRF4 in HPRT3 and HPRT2 cells. HPRT1 cells overexpressed MYC mRNA. HPRT3 and HPRT2 were relatively more sensitive to venetoclax-induced lethality, which correlated with higher BCL2, BAX, BAK and BIM protein levels. Higher Bcl-xL levels correlated with increased sensitivity of HPRT3 and HPRT2 versus HPRT1 cells to a Bcl-xL-specific inhibitor A-1155463. Conversely, higher MCL1 levels correlated with greater sensitivity of HPRT1 cells to an MCL1-specific inhibitor AZD-5991. Notably, HPRT3 and HPRT2 versus HPRT1 cells were relatively resistant to ibrutinib treatment. This was due to activation of the alternative MAP3K14 (NIK kinase)-NFkB pathway with increased processing of p100 to p52. Treatment with CDK9 inhibitor NVP2 dose-dependently induced apoptosis of the 3 RT-DLBCL cells, associated with depletion of c-Myc, Bcl-xL, and MCL1 protein levels. BET inhibitor OTX015 and BET protein degrader ARV-771 induced more lethality in HPRT1, compared to HPRT3 and HPRT2 cells, which correlated with higher levels of BRD4, c-Myc and TCF4, but markedly lower levels of IRF4 in HPRT1 cells. CRISPR-Cas9-mediated knockout (KO) of IRF4 markedly depleted nuclear levels of IRF4 and c-Myc (a target of IRF4), which significantly increased sensitivity to OTX015 in IRF4 KO RT-DLBCL cells. Co-treatment with OTX015 and ibrutinib or venetoclax was synergistically lethal in all three RT-DLBCL cell-types (CI < 1.0). Co-treatment with BET protein degrader ARV-771 and ibrutinib or venetoclax exerted synergistic lethality in all three RT-DLBCL cell-types (CI < 1.0). Finally, following tail vein infusion and engraftment of HPRT3 cells, daily treatment for three weeks with ARV-771 and venetoclax compared to each drug or vehicle control, significantly reduced the RT-DLBCL burden, as well as improved survival without inducing toxicity in the NSG mice. These findings strongly support further testing of BET protein antagonist-based combinations with BH3 mimetics, BTK inhibitors, as well as with other novel agents, utilizing pre-clinical models of RT-DLBCL. Disclosures Green: KDAc Therapeutics: Current equity holder in private company.