PRMT5 interacts with the BCL6 oncoprotein and is required for germinal center formation and lymphoma cell survival

Abstract

The germinal center (GC) reaction plays an important role in generating humoral immunity and is believed to give rise to most B-cell lymphomas. GC entry and exit are tightly regulated processes, controlled by the actions of transcription factors such as BCL6. Herein, we demonstrate that protein arginine methyltransferase 5 (PRMT5), a symmetric dimethyl arginine methyltransferase, is also necessary for GC formation and affinity maturation. PRMT5 contributes to GC formation and affinity maturation at least in part through its direct interaction with and methylation of BCL6 at arginine 305 (R305), a modification necessary for the full transcriptional repressive effects of BCL6. Inhibition of PRMT5 in B-cell lymphoma lines led to significant upregulation of BCL6 target genes, and the concomitant inhibition of both BCL6 and PRMT5 exhibited synergistic killing of BCL6-expressing lymphoma cells. Our studies identify PRMT5 as a novel regulator of the GC reaction and highlight the mechanistic rationale of cotargeting PRMT5 and BCL6 in lymphoma.

Graphical abstract

Figure 1.

PRMT5 is required for GC formation and affinity maturation. (A) Immunohistochemistry of paraffin-embedded splenic tissue from Cγ1-Cre^−/− Prmt5^fl/fl and Cγ1-Cre^+/− Prmt5^fl/fl mice (n = 7 per group) immunized with SRBCs for 10 days. (B) Quantification of PNA⁺ clusters from panel A. (C) Representative flow cytometry plot showing percentage of GC B cells (GL7⁺ FAS⁺ or CD38^dim FAS⁺) gated on live B220⁺ splenocytes in Cγ1-Cre^−/− Prmt5^fl/fl and Cγ1-Cre^+/− Prmt5^fl/fl mice (n = 7 per group) immunized with SRBC as described in panel A. (D) Titers of low-affinity NP-specific immunoglobulin were measured using NP26-BSA in the serum of Cγ1-Cre^−/− Prmt5^fl/fl and Cγ1-Cre^+/− Prmt5^fl/fl mice (n = 6 per group) immunized with NP-CGG for 8 days. (E) Ratio of high- to low-affinity NP-specific immunoglobulin detected with NP4-BSA and NP26-BSA, respectively, in NP-CGG immunized Cγ1-Cre^−/− Prmt5^fl/fl and Cγ1-Cre^+/− Prmt5^fl/fl mice described in panel D.

Figure 2.

The arginine methyltransferase PRMT5 directly interacts with BCL6. (A) Co-IP experiments for PRMT5 and BCL6 in 293T cells transfected with BCL6. (B) Co-IP experiments for PRMT5 and BCL6 in SUDHL8 and U2932 DLBCL cell lines. (C) PRMT5 and BCL6 interact in normal CD77⁺ GC B cells, enriched as described in supplemental Materials and methods. (D) GST pull-down assay of recombinant GST-PRMT5 and BCL6-MYC/DDK proteins. Purified GST-PRMT5 or GST proteins were incubated with BCL6-MYC/DDK protein for 12 hours. The coprecipitated BCL6 and PRMT5 proteins were detected by western blot with anti-GST and anti-BCL6 antibodies.

Figure 3.

PRMT5 interacts with the RD2 and ZNF domains of BCL6 via its N terminus. (A) Schematic of plasmids encoding either the V5-tagged full-length BCL6 or domains of BCL6 used in panel B. (B) Co-IP experiments of PRMT5 with V5-tagged full-length or BTB/POZ, RD2 and ZNF domains of BCL6. (C) GST pull-down assay of recombinant GST-PRMT5 with BCL6 RD2-V5/His proteins. (D) V5 pull-down assay of recombinant GST-PRMT5 with BCL6 ZNF-V5 proteins. (E) RD2 domain of BCL6 interacts with the catalytically active N terminus of PRMT5 (residues 1-320).

Figure 4.

PRMT5 and MEP50 mediate the repressive activity of the BCL6 RD2 domain. (A) Schematic of plasmids encoding the GAL4 DNA-binding domain (GAL4^DBD) fused to either full-length BCL6 or domains of BCL6 used in panels B and D. (B) GAL4 luciferase reporter assays in 293T cells transfected with plasmids encoding GAL4^DBD-BCL6 and PRMT5/MEP50. Western blots demonstrate representative expression of GAL4^DBD-BCL6, PRMT5, and MEP50 following expression of corresponding expression vectors. All the experiments were repeated 3 times in triplicate. ***P < .001. (C) 4xBCL6 binding site luciferase reporter assays in 293T cells transfected with plasmids encoding BCL6 and PRMT5/MEP50. Western blots demonstrate representative expression of BCL6, PRMT5, and MEP50 following expression of corresponding expression vectors. All the experiments were repeated 3 times in triplicate. ***P < .001. (D) GAL4 luciferase reporter assays in 293T cells transfected with full-length GAL4^DBD-BCL6, GAL4^DBD-BTB/POZ, or GAL4^DBD-RD2 domains of BCL6 in the presence or absence of cotransfected PRMT5/MEP50 siRNAs from GE Dharmacon (Lafayette, CO). Western blots demonstrate representative expression of PRMT5 and BCL6 or MEP50 and BCL6 following transfection of increasing concentrations of PRMT5 siRNA-1 and MEP50 siRNA-1, respectively. All the experiments were repeated 3 times in triplicate. ***P < .001. Independent siRNA to PRMT5 and MEP50 shown in supplemental Figure 3. (E) 4xBCL6 binding site luciferase reporter assays in 293T cells transfected with plasmids encoding BCL6 alone or in the presence of GSK591 (*P < .05; **P < .01; ***P < .001). (F) 4xBCL6 binding site luciferase reporter assays in 293T cells transfected with plasmids encoding BCL6, PRMT5, or catalytically inactive PRMT5 mutant (G367A/R368A, muPRMT5). Western blots demonstrate representative expression of BCL6, PRMT5, and muPRMT5 (G367A/R368A) following expression of corresponding expression vectors. All the experiments were repeated 3 times in triplicate. *P < .05; **P < .01.

Figure 5.

PRMT5 directly dimethylates arginines of BCL6 in lymphoma. (A) IP for BCL6 and symmetric arginine dimethylation (SYM10) in lymphoma cell lines OCI-LY1 and Raji reveals symmetric arginine dimethylation of BCL6. (B) Knockdown of PRMT5 with specific siRNA decreases BCL6 symmetric arginine methylation. Raji cells were transfected with PRMT5 or control siRNAs followed by IP with symmetric arginine dimethylation (SYM10) antibody and immunoblotting with BCL6 antibody. Also shown are western blots with indicated antibodies from the same cells. (C) In vitro methyltransferase assay with recombinant PRMT5, MEP50, and BCL6 or H2A proteins. The proteins were blotted with the indicated antibodies. In addition, the reaction mixture was immunoprecipitated with BCL6 antibody and blotted with antibodies for symmetric arginine dimethylation (SYM10) and BCL6. (D) In vitro thymidine incorporation methyltransferase assay with recombinant PRMT5 and BCL6 or H2A proteins.

Figure 6.

PRMT5 dimethylates BCL6 at R305 to mediate the repressive activity of the BCL6 RD2 domain. (A) Mass spectroscopy of BCL6 protein following in vitro methyltransferase assay by PRMT5 identifies BCL6 methylation at R305. Representative tandem mass spectrometry spectrum of methylated R305 of BCL6 (P11482). Band corresponding to BCL6 was excised from the gel and subjected to trypsin digestion. Tryptic peptides were resolved on a reverse phase column, and high-energy collision dissociation spectra were obtained using Orbitrap Fusion Tribrid mass spectrometer. Data were searched against human protein database using Proteome Discoverer (v 1.4, ThermoScientific) by considering methylation on arginine as a potential modification. Results were filtered at 1% false discovery rate. A tandem mass spectrometry spectrum corresponding to ³⁰²EEErPSSEDEIALHFEPPNAPLNR³²⁵ of BCL6 (precursor [M+H]⁺⁴ = 698.3398 m/z. DPPM = 1.6, inset) is shown. Observed b- and y-ions are indicated. The lowercase “r” is the methylated arginine. (B) 4xBCL6 binding site luciferase reporter assays in 293T cells transfected with wild-type and R305K mutant BCL6 or control pcDNA3.1 vector in the presence or absence of cotransfected siRNAs to PRMT5. ***P < .001. Western blots in each experimental condition with the indicated antibodies. (C) Arginine dimethylation of wild-type or R305K BCL6 transfected into Raji cells. (D) Co-IP experiments of endogenous PRMT5 with wild-type or R305K BCL6 transfected into Raji and 293T cells. WT, wild-type.

Figure 7.

PRMT5 recruits BCL6 to its target genes to induce gene repression. (A) Enrichment of BCL6, SMRT, HA-tagged PRMT5, and IgG at BCL6 targets in OCI-LY1 cells treated with vehicle or 200 nM GSK591 for 72 hours. *P < .05; **P < .01. (B) Messenger RNA abundance of BCL6 target genes in OCI-LY1, OCI-LY7, SUDHL4, and SUDHL6 cells treated with vehicle or 200 nM GSK591 for 96 hours. *P < .05; **P < .01. (C) Immunoblot demonstrating inhibition of SDMA using the SYM10 antibody at 72 hours after treatment with 200 nM GSK591. IP demonstrating decrease in symmetric dimethylation of BCL6 shown in supplemental Figure 7A. (D) Pathway analysis of gene expression changes in SUDHL6 cells treated with vehicle or 200 nM GSK591 for 24 or 96 hours. The Fisher exact test was used to calculate enrichment P values for each gene set, and the Benjamini-Hochberg method was used for false discovery rate control. (E) GSK591 concentration that results in 50% growth inhibition (GI₅₀) of BCL6-dependent and BCL6-independent DLBCL cell lines treated with vehicle or increasing concentrations of GSK591 for 6 days. Raw growth inhibition curves of GSK591 alone are shown in supplemental Figure 8A. (F) Messenger RNA abundance of BCL6 targets in OCI-LY1 cells with combined treatment of 25 μM FX1 and 200 nM GSK591 for 48 hours. *P < .05; **P < .01 relative to vehicle. †P < 0.05; ††P < .01 relative to each drug alone. (G) Combination indexes of the BCL6 inhibitor FX1 with GSK591 after treating cells with increasing concentrations of GSK591 for 6 days and FX1 for 2 days. Data are representative of 3 triplicates ± standard error of the mean (SEM). Raw growth inhibition curves of each drug alone and in combination are shown in supplemental Figure 8C. (H) Mean fluorescence intensity of carboxyfluorescein diacetate succinimidyl ester of live CD19⁺ (GhostDye⁻ or DAPI⁻) human DLBCL samples on day 6 after cells were exposed to GSK591 on day 0 then treated with FX1 3 days later. Data representative of 3 triplicates ± SEM. *P < .05; **P < .01 relative to VEH. †P < 0.05; ††P < .01 relative to each drug alone. (I) Cell viability (GhostDye⁻ or DAPI⁻) of CD19+ human DLBCL from panel H. Data are representative of 3 triplicates ± SEM. *P < .05; **P < .01 relative to vehicle. †P < .05; ††P < .01 relative to each drug alone. FACS, fluorescence-activated cell sorting; FL, follicular lymphoma; MFI, mean fluorescence intensity; MHC, major histocompatibility complex; mRNA, messenger RNA; VEH, vehicle.