Abstract
The genetic hallmarks of follicular lymphomas include BCL2 translocations and somatic mutations of epigenetic modifier genes such as EZH2. Histologically, FLs typically feature a rich microenvironment, most notably featuring extensive follicular dendritic cell (FDC) networks with dendrites making extensive contact with lymphoma cells. In recent work we showed that the main effect of EZH2 gain-of-function mutations in GC B-cells is to enable them to become less dependent of T-cell help and strengthen their immune synapse formation with FDCs, which induces aberrant proliferation and survival of GC centrocytes and hence formation of a unique lymphoma-permissive immune niche. However, it is still unknown how interactions between EZH2 mutant GC B-cells and other immune cells change throughout the progression of the disease.
To evaluate the evolution of the tumor microenvironment (TME) throughout EZH2 mutant lymphoma progression, we developed a genetically engineered mouse model designed for conditional expression of gain-of-function Ezh2 mutant and BCL2 in GC B-cells, Rosa26 LSL.BCL2.IRES.GFP;Ezh2 Y641F;Cγ1Cre (hereafter BCL2/Ezh2 Y641F), in which Cγ1-driven Cre excision of a STOP cassette in the Rosa26 locus leads to overexpression of BCL2 and GFP, and expression of the endogenous Ezh2 mutant Y641F. This mouse model develops low-grade follicular-like lymphoma, characterized by expanded follicles composed largely of centrocyte neoplastic GC B-cells and extensive FDC meshwork, along with presence of CD4 +, TFH and regulatory T cells (FOXP3 +), as depicted by multiparametric in situ imaging and multiparametric flow cytometry. At cytological level, cells are predominantly small with condensed chromatin, irregular nuclei, scant cytoplasm, and inconspicuous nucleoli, without sheets of large cells. Over time, these low grade FLs progress to advanced grade, characterized by disruption of follicle structures, expansion of centroblast-like large tumor cells and reduced CD4 + T cell infiltration and FDC meshwork.
Furthermore, we have developed a murine transformed FL cell line by sequential passages of an original BCL2/Ezh2 Y641F low-grade FL into immunodeficient Rag1KO mice. We have transduced this BCL2/Ezh2 Y641F cell line with luciferase, and it successfully engrafted and homed to lymphoid organs when injected i.v. into immunocompetent C57BL6 mice. The cell line consists of high proliferative GC B-cells with multiple and irregular nuclei, open chromatin and prominent nucleoli that resemble DLBCL. The microenvironment of tumors developed in C57BL6 mice is characterized by disruption of follicle structures, severe reduction or absence of FDC meshwork, decreased CD4 +, CD8 + and Tregs, downregulation of MHC-I and MHC-II. Therefore, our mouse model also recapitulates progression and transformation stages.
Since MYC translocations are frequent events that occur during histological transformation of FL, we modeled this more aggressive EZB MYC subtype of DLBCL. For that, we further crossed our BCL2/Ezh2 Y641F mice to Rosa26 LSL.MYC.IRES.hCD2. We generated cohorts of transplanted mice by injecting bone marrow cells from BCL2, BCL2/Ezh2 Y641F, BCL2/MYC, and BCL2/Ezh2 Y641F/MYC mice into lethally irradiated C57BL6 recipients. Recipient mice (n=5 per group) were immunized with SRBC to induce formation of GCs and sacrificed 5.5 months post bone marrow transplant. All genotypes showed expansion of FAS +CD38 -BCL2 GFP+ GC B-cells in spleen and lymph nodes; however, the proportion of GC B-cells in Ezh2 Y641F was significantly increased compared with Ezh2 WT, in both presence and absence of MYC. In contrast, FAS -CD38 +BCL2 GFP+ and CD138 +BCL2 GFP+ cells were significantly increased in MYC + mice and decreased in Ezh2 Y641F, suggesting that Ezh2 mutation is required to maintain the GC phenotype. The acquisition of Ezh2 Y641F in BCL2/MYC mice induced a reduction of CD4 + and CD8 + in the TME, and decreased TFH without changes in TFR proportions in GCs. Strikingly, FDC meshwork was significantly shrank in MYC + cases, and partially restored by acquisition of Ezh2 Y641F, indicating that MYC overexpression may contribute to acquire FDC independency for the survival of lymphoma cells.
Our results indicate that progression of FL critically affects the TME and acquisition of additional mutations such as MYC alters the interactions between lymphoma cells and tumor supportive immune cells in TME.
Melnick: Epizyme: Consultancy; Janssen: Research Funding; KDAC Therapeutics: Current holder of individual stocks in a privately-held company; Celgene Corporation: Consultancy; Constellation Pharmaceuticals: Consultancy; Astra Zeneca: Consultancy; Daiichi Sankyo: Consultancy; Exo Therapeutics: Membership on an entity's Board of Directors or advisory committees; Sanofi US Services: Research Funding.