PARP1 facilitates transcriptional activity and stability of the BCL6 protein and is a therapeutic target in BCL6 expressing lymphomas Free

BCL6, the “master regulator” of the Germinal Center (GC) reaction, is a transcriptional repressor implicated in lymphomagenesis. Determining BCL6 mechanisms of action is imperative for understanding GC biology, molecular mechanisms of lymphomagenesis and identifying novel therapeutic targets. We previously reported that symmetric arginine methyltransferase PRMT5 directly interacts with and methylates the BCL6 at arginine 305 (R305) in the RD2 domain, a modification necessary for BCL6 transcriptional repression (TR), GC formation and affinity maturation. We hypothesized that R305 may mediate interactions with additional proteins regulating BCL6 functions. Mass spectrometry analysis of BCL6-GFP wild type and R305K-GFP proteins immunoprecipitated from Raji whole cell lysates identified 325 BCL6 interacting proteins. In 19 proteins, the intensity of peptides derived from the wild type BCL6 was significantly enhanced compared to BCL6 R305K immunoprecipitants, and Poly (ADP-ribose) polymerase 1 (PARP1) exhibited the largest fold change and a statistically significant difference. The BCL6-PARP1 interaction was confirmed by reciprocal co-immunoprecipitation (Co-IP) in multiple lymphoma cell lines and isolated GC cells. This interaction was: a) preserved after benzonase pre-treatment, indicating that it is specific and not mediated by DNA and/or RNA; b) not dependent on PRMT5 or its catalytic activity; and c) localized in insoluble chromatin. Recombinant c-Myc-BCL6 and FLAG-PARP1 pull-down assays confirmed direct interaction.

PARP1 can function as a repressor or coactivator of transcriptional factors, and these effects may require or function independently of its PARylation activity. Using two transcriptional reporter systems in multiple cell lines we showed that BCL6-mediated TR was attenuated upon PARP1 knockdown with two different siRNAs, knockout via Crispr/Cas9 editing, or inhibition by Olaparib in a dose dependent manner, suggesting that PARP1 facilitates BCL6-mediated TR.

To examine whether this effect is PARylation-dependent, we analyzed BCL6 PARylation in IP experiments and in two vitro PARylation assays using recombinant BCL6, PARP1 and the PARylation inactive E988K PARP1 mutant. While we showed that BCL6 is PARylated, BCL6 PARylation was not necessary for BCL6 TR, since reconstitution of PARP1 Crispr/Cas9-edited cells with wild type PARP1 or E988K PARP1 mutant suppressed BCL6 reporter activity to levels observed in the non-manipulated 293T cells expressing BCL6. Since PARylation can regulate protein stability we performed cycloheximide (CHX) pulse-chase experiments showing progressive decrease in BCL6 protein levels over time, that was enhanced in the presence of Olaparib and rescued by proteasome inhibitor MG132. In vivo, Olaparib-treated mice immunized with sheep RBCs showed statistically significantly reduced BCL6 expression in GCs and a decreased number of GC cells compared to DMSO-treated controls, suggesting that PARP1 promotes BCL6 stability and GC maintenance.

ChIP-seq for BCL6 in the presence or absence of Olaparib performed in two independent biological experiments using OCI-LY1 cells demonstrated a global reduction of BCL6 binding to its transcriptional targets and a concordant decrease in the expression of BCL6 targets assessed by RNA sequencing after treatment with Olaparib for 24 hours. These findings were validated by quantitative real time PCR analysis of several known BCL6 targets in three BCL6 expressing DLBCL cell lines. We observed decreased interaction between BCL6 and PARP1 in Co-IP and decreased localization and binding of BCL6 to insoluble chromatin following treatment with Olaparib.

We examined whether PARP1 inhibition could suppress the growth of BCL6 expressing DLBCL cells and potentiate effects of BCL6 inhibitor. Olaparib and Veliparib induced a pronounced decrease in colony formation and cell viability as assessed by the MTS assay and potentiated similar effects of BCL6 inhibitor 79-6. Both PARP1 inhibitors and BCL6 inhibitor 79-6 induced apoptosis and cell death in a cell line specific manner. These combinations may be considered for clinical evaluation.

Overall, for the first time our studies show that PARP1 plays an essential and direct role in regulating transcription of BCL6 target genes and may be used for indirect therapeutic targeting of BCL6 in lymphomas.