Over-Expression of IRF4 (MUM1) Protein Predicts Poorer Outcome in Myeloma Patients

Abstract

IRF4 (also known as MUM1) is a member of the interferon regulatory factor family of transcription factors whose expression is critical for the transition from pre B-cell to immature B-cell and for differentiation of mature B-cell to plasma cell. It has previously been reported by Heintel el al (Leukemia (2008) 22, 441–445) that over-expression of IRF4 at the mRNA level is associated with a poor outcome in myeloma. Measurement of expression at the protein level using flow cytometry would remove the need for plasma cell purification and also allow evaluation of IRF4 in other cell populations. We have developed a multi-colour flow cytometry assay incorporating an indirect intracellular staining using the goat polyclonal antibody IRF4 (Santa Cruz). This permits simultaneous analysis of IRF4 in combination with five cell surface markers, allowing quantitation of the IRF4 levels present in specific B-cell and plasma cell populations. The aim of this study was to assess the expression profile of IRF4 in neoplastic and normal B-cells and plasma cells and evaluate the relationship between IRF4 protein expression and outcome in multiple myeloma patients. Leucocytes prepared from whole bone marrow using ammonium chloride lysis were incubated with surface markers CD52 PE, CD38 PerCPCy5.5, CD19 Pe-Cy7, CD138 APC and CD20 APC-Cy7. Cells were then fixed and permeabilised before incubation with unconjugated IRF4, washing and incubation with an Alexa488 secondary antibody. IRF4 expression showed the expected profile in samples with normal B-lineage cells. Low levels of expression were found in normal mature B-cells (mean MFI 1033, range 395–2265), higher levels in progenitor B-cells (mean 4.1-fold higher than mature B-cells, mean MFI 4252, range 612–22958) and strong uniform expression in normal BM plasma cells (mean 13.8-fold higher than mature B-cells, mean MFI 14250, range 7391–21914). Plasma cells from untreated myeloma patients showed a mean 1.9-fold higher IRF4 expression level than normal plasma cells (P=<0.001, normal n=16, neoplastic n = 19, mean MFI 27510, range 6807–53662). The myeloma patients were then separated into two groups; those with IRF4 expression in the normal range (n=6, mean MFI 12858, range 4182–21028) and those with high IRF4 expression (n=13, mean MFI 34501, range 24892–53662). After a median follow-up of 19 months, 6 patients had died in the high expression group (46.2%) compared with no deaths in the low expression group; this difference was statistically significant (Cox regression analysis, p=0.036). The assay developed can detect quantitative differences in IRF4 expression between B-progenitors, mature B-cells, normal plasma cells and neoplastic plasma cells. The ability to simultaneously analyse IRF4 expression with other cell surface markers allows rapid enumeration of IRF4 protein levels without plasma cell purification and greatly enhances the potential for IRF4 to be used as a prognostic marker. Neoplastic plasma cells can be categorised as having low or high levels of IRF4 expression with over-expression predicting a significantly shorter overall survival.