Epstein-Barr virus-induced gene 3 (EBI3): a novel diagnosis marker in Burkitt lymphoma and diffuse large B-cell lymphoma

Abstract

The distinction between Burkitt lymphoma (BL) and diffuse large B-cell lymphoma (DLBCL), two types of mature aggressive B-cell lymphomas that require distinct treatments, can be difficult because of forms showing features intermediate between DLBCL and BL (here called BL/DLBCL). They can be discriminated by the presence of c-myc translocations characteristic of BL. However, these are not exclusive of BL and when present in DLBCL are associated with lower survival. In this study, we show that Epstein-Barr virus-induced gene 3 (EBI3) is differentially expressed among BL and DLBCL. Analysis of gene expression data from 502 cases of aggressive mature B-cell lymphomas available on Gene Expression Omnibus and immunohistochemical analysis of 184 cases of BL, BL/DLBCL or DLBCL, showed that EBI3 was not expressed in EBV-positive or -negative BL cases, whereas it was expressed by over 30% of tumoral cells in nearly 80% of DLBCL cases, independently of their subtypes. In addition, we show that c-myc overexpression represses EBI3 expression, and that DLBCL or BL/DLBCL cases with c-myc translocations have lower expression of EBI3. Thus, EBI3 immunohistochemistry could be useful to discriminate BL from DLBCL, and to identify cases of BL/DLBCL or DLBCL with potential c-myc translocations.

Figure 1. Differential expression profile of EBI3 gene in BL and DLBCL defined by molecular profiling.

Individual data for EBI3 gene were retrieved from series GSE4732 (A, B), and pooled from series GSE4475 and GSE10172 , (C, D). Relative expression of EBI3 is shown using the authors' scale. On (A) and (C) EBI3 expression among the different molecular types defined by gene profiling is shown. On (B) and (D), the expression of EBI3 in mBL is shown according to their original pathological diagnosis. p values are indicated. The horizontal bar indicates the mean. NS: not significant; unclass: unclassifiable.

Figure 2. Immunohistochemical analysis of EBI3 expression in classic BL and DLBCL.

(A) Percentage of EBI3-positive tumoral cells in BL and DLBCL cases analyzed by EBI3 immunohistochemistry. (B) Frequency of EBI3-positive or -negative cases among BL and DLBCL, when positivity is defined by ≥30% stained tumoral cells. (C) Distribution of EBI3-positive and -negative cases among DLBCL of the GCB or non-GCB subtypes.

Figure 3. Analysis of EBI3 gene expression according to the presence or absence of c-myc translocation.

Individual data for EBI3 and c-myc were retrieved from GSE4475 and GSE10172 series , . On (A), the expression of EBI3 and of c-myc was analyzed according to the presence or not of c-myc translocations involving an Ig locus (Ig-myc) or not (non-Ig-myc). « others » designated lymphomas other than mBL, ie lymphomas classifed as non-mBL and lymphomas that could not be classified as mBL or non-mBL. On (B), EBI3 expression among all lymphomas other than mBL is represented according to that of c-myc. On (C), EBI3 expression among lymphomas that had initially received a pathological diagnosis of BL/DLBCL is shown according to that of c-myc. p values are indicated. NS: not significant.

Figure 4. In vitro effect of c-myc overexpression on EBI3 expression level.

Karpas 1106 cell line was transfected with a GFP reporter plasmid together with c-myc-ER expression vector or a control vector and cultured for various times in the absence or presence of 4-OHT. EBI3 expression in GFP-positive cells isolated from vector control transfected cells (Col) or c-myc-ER transfected cells (c-myc-ER) was analyzed by RTqPCR. In (A), transfected cells were cultured for 20 hours in the presence of 4-OHT (200 nM). Data are expressed as mean ± SEM from 3 independent transfections. p value is indicated. In (B), cells were cultured for 20 hours in the absence or presence of 4-OHT (200 nM). In (C), cells were cultured with increasing concentrations of 4-OHT. The expression of EBI3 in c-myc-ER transfected cells relative to that measured in control tranfected cells is represented. The percentage of decrease is indicated. A higher concentration of 4-OHT (1 µM) did not result in higher repression of EBI3 expression (not shown). In (D), 4-OHT was added to the transfected cells for the last 16, 24 or 40 h of the culture. In (B)–(D), a representative experiment is shown.

Figure 5. Effect of c-myc translocation on EBI3 expression in lymphomas.

(A) The distribution of cases with or without c-myc translocations among all BL/DLBCL and DLBCL cases, or among EBI3-negative versus EBI3-positive cases, is shown. EBI3 positivity (≥30% stained tumoral cells) was assessed by immunohistochemistry. (B) The relative expression of EBI3 and c-myc (normalized to ß2m expression) was determined by RTqPCR in frozen tissues from 10 cases of B-cell lymphomas without c-myc translocation that were positive for EBI3 by immunohistochemistry and 6 cases of B-cell lymphomas without c-myc translocation that were negative for EBI3. (C) The ratio of EBI3/c-myc expression in the cases analyzed is (B) is represented. p values are indicated.

Figure 6. Immunohistochemical analysis of EBI3 expression in BL, BL/DLBCL and DLBCL.

Sections from a case of BL (a), 2 cases of BL/DLBCL (b, c), and 2 cases of DLBCL (d, e) with or without c-myc translocation as indicated on each figure, were analyzed for EBI3 expression by immunohistochemistry. An inverse correlation between the presence of c-myc translocation and the positivity for EBI3 in tumoral cells was observed. Each picture is shown at the same magnification. The bar shown on (c) represents 50 µm.

Figure 7. Immunohistochemical analysis in a case of follicular lymphoma and its transformed counterpart.

Sections from a follicular lymphoma (a) that subsequently acquired a c-myc translocation and transformed into a DLBCL (b) were analyzed for EBI3 expression by immunohistochemistry. The bar represents 50 µm.