The Irf4 Gene, a Susceptibility Locus for Chronic Lymphocytic Leukemia (CLL), Controls Establishment of Follicular and Marginal Zone B Cell Compartments in Mice

Abstract 285

Chronic lymphocytic leukemia (CLL) originates from the malignant transformation of mature B cells. Recently, single nucleotide polymorphisms (SNPs) in the 3'UTR region of the Irf4 gene have been associated with an increased risk of developing CLL in independent patient cohorts. IRF4 is a member of the interferon regulatory factor (IRF) family of transcription factors, and in the B lineage is essential for plasma cell differentiation in T-dependent immune responses. Irf4 has been demonstrated to act as an oncogene in multiple myeloma. Conversely, evidence suggests that the corresponding SNPs in the Irf4 gene lead to a reduction of IRF4 mRNA expression, and it was reported that tumor cells in half of CLL cases show reduced IRF4 protein expression compared to normal B cells. Together, these observations suggest that aberrant downregulation of IRF4 expression in mature B cells may contribute to CLL development. However, the normal function of IRF4 in mature B cells is incompletely understood. In order to investigate how IRF4 deficiency affects the biology of mature B cells, we investigated the consequences of deleting Irf4 specifically in B cells in vivo using a conditional Irf4 knockout mouse line.

We and others had previously observed that Irf4^−/− mice develop an expansion of B cells with a marginal zone (MZ) phenotype (IgM^hiIgD^loCD23^–CD21⁺). By demarcating the MZ with the MOMA-1 marker, we here show that B cells in Irf4^−/− mice localized preferentially in the MZ area, causing MZ hyperplasia. In contrast, the area where follicular B cells normally home contained few B cells. B cell autonomy of the observed phenotype was ascertained by crossing a “floxed” Irf4 allele with CD19-Cre mice to achieve B cell conditional deletion. We then crossed the floxed Irf4 allele with a transgenic mouse that allows inducible deletion of Irf4 specifically in B cells. Whereas flow-cytometric analysis revealed an unchanged ratio between cells with a follicular (IgM^loIgD^hiCD23⁺CD21^int) vs. a MZ B cell surface phenotype upon Irf4 deletion, immunohistochemical (IHC) stainings of spleen sections for a marker protein (eGFP) that signals gene deletion demonstrated that the Irf4-deleted cells localized preferentially in the MZ, leading to MZ hyperplasia. Together, these results suggest that deletion of Irf4 in B cells alters chemokine responsiveness and migratory capacity. In agreement, global gene expression profile analysis of B cells purified from Irf4^−/− and Irf4^+/+ mice identified a set of differentially expressed genes with known functions in cell migration and homing. Notably, PLXND1 and the chemokine receptor CXCR7 showed 4.5 and 6.5-fold upregulation, respectively, while the G protein coupled receptor RGS13 and the adhesion molecule ALCAM (CD166) showed 3.5 and 5-fold downregulation in Irf4^−/− vs. Irf4^+/+ B cells.

Unexpectedly, we observed in the profiling analysis that expression of the known NOTCH target genes Deltex1 and Hes5 was significantly upregulated (3 and 5-fold) in Irf4^−/− vs. Irf4^+/+ B cells. Despite unchanged NOTCH1 and NOTCH2 mRNA levels, Western blot and immunofluorescence analysis showed that Notch2, a gene known to be indispensible for MZ B cell development, was strongly upregulated in Irf4^−/− B cells, suggesting that IRF4 is indirectly involved in NOTCH2 repression at a post-transcriptional level. Together with the altered migratory properties of Irf4^−/− vs. Irf4^+/+ B cells, these findings indicate that a balanced expression of IRF4 and NOTCH2 in B cells is required for establishing the follicular and MZ B cell compartments in mice, and suggest that IRF4 maintains the cellular identity of follicular B cells.

The results may have implications for understanding CLL pathogenesis, as both NOTCH1 and NOTCH2 transmembrane receptors were reported to be expressed and activated in CLL B cells, and since NOTCH1 was recently found to be aberrantly activated in a fraction of CLL cases due to genetic mutations. Alterations in the balance of the transcriptional network established by NOTCH and IRF4 either through mutations, polymorphisms, or microenvironmental factors may disrupt normal B cell physiology and thereby contribute to tumorigenesis by an as yet unknown mechanism. Recently, a small fraction of CLL patients were identified that have a recurrent heterozygous somatic mutation in exon 2 of Irf4, providing additional rationale for determining how alterations in IRF4 function may promote CLL pathogenesis.