The B-cell lymphoma/leukemia 11A (BCL11A) gene was first identified in the rare t(2;14)(p16;q32.3) translocation in aggressive B-cell chronic lymphocytic leukemia and, together with REL, is a candidate for the 2p16 chromosomal amplifications in B-cell malignancies.1 BCL11A encodes a Krüppel zinc-finger transcription factor, and Evi9/Bcl11a/Ctip1 (mouse homologue of human BCL11A) has been shown to be essential for pre-B-cell development and thymocyte maturation.2 There is evidence that Bcl11a, and hence human BCL11A, may not only encode a dominant proto-oncogene but also function as a non-autonomous T-cell tumor suppressor gene.2 Alternative RNA splicing generates at least three BCL11A transcripts with restricted expression in bone marrow, lymphoid tissue and brain. BCL11AXL mRNA is the largest and the most abundant transcript.1 We produced a novel monoclonal antibody, BCL11A/123, raised against BCL11AXL recombinant GST-BCL11A fusion protein (aa 637–835), which was used in a variety of immunolabeling and Western blotting techniques3 to study BCL11AXL expression in a wide variety of normal and malignant human cells and tissues. The specificity of the antibody was confirmed using transfected COS cells expressing BCL11AS, BCL11AL, BCL11AXL or the highly homologous BCL11B protein (Figure 1a and b and Supplementary Information). BCL11AXL protein was detected in the nuclei of cells in the mantle zone, the majority of nuclei in the germinal centers and a subpopulation of cells in the interfollicular areas of tonsil and lymph node (Figure 1c). Nuclear staining of the splenic marginal zone was also observed. Double immunolabeling studies demonstrated most of the BCL11AXL-positive cells in the tonsil to be of B-cell origin (Figure 1d), whereas T cells (Figure 1e) and plasma cells (Figure 1f) were BCL11AXL-negative. Only rare BCL11AXL-positive cells were observed in the bone marrow and fetal thymus and these coexpressed either CD68 or CD79a antigen. Although high levels of BCL11A transcripts have been described in peripheral blood B cells (http://symatlas.gnf.org/), BCL11AXL protein expression was not detected in peripheral blood smears or purified CD20-positive B cells from normal subjects. Activation of purified CD20-positive B cells with Staphyloccus aureus and rIL-2 did, however, result in the induction of nuclear expression of BCL11AXL. Of note was the appearance of scattered cells with very strongly labeled nuclei in the interfollicular areas of the tonsil and lymph node (Figure 1c, inset). These cells were also labeled by all three commercially available anti-Ctip1/Bcl11a antibodies (Abcam, Cambridge, UK) but not with the anti-Bcl11b-specific polyclonal antibody. These distinctive cells were negative for CD20 (Figure 1g, left inset) and CD3 (Figure 1g, right inset), CD11c, CD31, CD32, CD34 and CD79a but expressed CD68 (Figure 1h) and CD4. Eighty-five to 95% of these BCL11AXL-positive cells also coexpressed the plasmacytoid dendritic cell (pDC) marker CD123 (Figure 1i). Although high levels of BCL11AXL transcripts have been reported in pDCs (http://symatlas.gnf.org/), this is the first report of BCL11AXL protein expression in human pDCs. Additional support for BCL11A expression in this subset of dendritic cells is provided by the recent report of Bcl11a transcripts in murine pDCs.4 Nuclear labeling for BCL11AXL was observed in a subpopulation of cells in thymus. Although the majority of the medullary BCL11AXL-positive cells were CD20-positive B cells, scattered CD20-negative cells with strongly stained nuclei, reminiscent of the pDCs described above, were also observed (Figure 1j). BCL11AXL protein was detected in a subpopulation of CD3-positive thymic T cells (Figure 1k). With the exception of weak nuclear labeling of neurons in the adult cerebral cortex, strong labeling of tonsillar epithelium and a subpopulation of epithelial cells in the gut and thymic cortex (Figure 1k, inset), BCL11AXL was not detected in non-hematopoietic cells present in a wide range of normal tissues. Immunolabeling and Western blotting studies on a range of tumor-derived cell lines demonstrated BCL11AXL expression at most stages of B-cell differentiation, ranging from pre-B to cells of germinal center or activated B-cell origin, but was absent in 5/6 plasma cell-derived lines studied (Supplementary Table and Supplementary Figure of cell line results). The results obtained from a study of a wide range of hematological tumors present on tumor microarrays (TMA) are summarized in Table 1. Nuclear BCL11AXL expression was observed in a wide variety of B-cell tumors but was undetectable in the myeloma cases studied and in Reed-Sternberg cells in the majority of cases of Hodgkin's lymphoma. With the exception of only 2/5 cases of T-cell acute lymphoblastic lymphoma/leukemia (T-ALL) and 1/3 cases of natural killer cell lymphoma, BCL11AXL protein was not found in the T-cell tumors studied. An immunolabeling study on a TMA of 107 cases of previously untreated diffuse large B-cell lymphoma (DLBCL) obtained from the British Columbia Cancer Agency5 demonstrated BCL11AXL expression in 82% (45/55) and 70% (35/50) of cases immunophenotyped previously as being of either germinal center (GC) or non-germinal center (ABC) origin, respectively. This lack of differential expression of BCL11AXL between GC- and ABC-derived tumors is in keeping with a recent report in which BCL11A transcripts were not influenced by 2p14–p16 gains.6 There was, however, correlation between BCL6 and BCL11AXL protein expression, with 79% of BCL6-positive cases also expressing BCL11AXL (P>0.010), but no correlation was observed with BCL2 (P>0.86). Although no relationship between BCL11AXL expression and failure-free survival was found, there was a tendency towards decreased survival in patients whose tumors lacked BCL11AXL (P>0.1) (Figure 1l). Further analyses of protein expression data may reveal additional correlations between BCL11AXL and...