Overexpression of BCL9, a Wnt/β-Catenin Transcriptional Co-Activator, in Transgenic Mice Promotes Spontaneous Development of B-Cell Malignancies

Dysregulation of the Wnt signaling pathway underlies the pathogenesis of a wide range of human cancers, including hematological malignancies such as multiple myeloma (MM). The terminal effector of this signaling pathway is a transcriptional complex formed by β-catenin and BCL9. This complex is of particular interest because the BCL9 locus resides on the frequently recurring 1q21 chromosomal amplification in MM, which has been linked to poor clinical prognosis and outcome. Our previous studies indicate that BCL9-mediated enhancement of β-catenin activity increases cells proliferation, migration, invasion, and the metastatic potential of MM cells. Therefore, in order to: (I) unequivocally determine the oncogenic role of BCL9, (II) better understand its mechanism of action, and (III) develop mouse preclinical model of cancer with dysregulated Wnt/β-catenin/BCL9 activity, we generated transgenic mouse models.

To overcome problems inherently related to embryonic lethality, we generated BCL9^fl/- conditional transgenic mice using site-specific transgene integration into the mouse ColA1 gene in embryonic stem cells. To remove the stop cassette and activate BCL9 expression in vivo, we generated AID-Cre^+/-; BCL9^fl/- and ERT2-Cre^+/-; BCL9^fl/- compound mice. Recombinase activity driven by AID (activation-induced cytidine deaminase) gene promoter or ER receptor after tamoxifen administration, caused removal of the stop cassette and expression of BCL9 in germinal center (GC) B cells or several tissues, respectively, as confirmed by immunoblot, immunohistochemical (IHC) and PCR analysis. Since BCL9 is a β-catenin co-activator, next we generated cohorts of AID-Cre^+/-; BCL9^fl/-; TCF/Lef1-lacZ^+/- and ERT2-Cre^+/-; BCL9^fl/-; TCF/Lef1-lacZ^+/- triple compound transgenic mice carrying the Wnt reporter system that expresses β-galactosidase (β-gal), to determine whether Wnt/β-catenin transcriptional activity is increased as a consequence of BCL9 overexpression in vivo. β-gal stain was increased in frequency and intensity in cells within GCs but not outside them in AID-Cre^+/-; BCL9^fl/-; TCF/Lef1-lacZ^+/- compared to control mice. In ERT2-Cre^+/-; BCL9^fl/-; TCF/Lef1-lacZ^+/- mice β-gal staining was primarily detected in cells outside the GCs, not within them. Overall, these results indicate that Wnt transcriptional activity is increased in B-cells as a consequence of Cre-induced expression of BCL9 and that AID-Cre^+/- and ERT2-Cre^+/- target expression of BCL9 to GC and non-GC B cells, respectively. Because BCL9 is involved in the pathogenesis of human cancers, we evaluated whether our transgenic mice develop hematological malignancies. Except for mild splenic enlargement, BCL9-transgenic mice were indistinguishable from control mice between 8 and 30 weeks of age as assessed by weight and posture. However, after 40 weeks of age and at variable times thereafter, 80% (32/40) of AID-Cre^+/-; BCL9^fl/- and 70% (28/40) of ERT2-Cre^+/-; BCL9^fl/- mice but none from control cohorts showed signs of disease. Gross pathologic examination of euthanized animals with BCL9 overexpression revealed enlargement of the spleen and LNs. Two distinct patterns of clonal hematological malignancies were identified after detailed histological, IHC and molecular examination. In AID-Cre^+/-; BCL9^fl/- mice tumors resembled human plasmacytomas (PCs), whereas in ERT2-Cre^+/-; BCL9^fl/- mice B-cell acute lymphoblastic leukemia (B-ALL). This later result is of particular interest, since BCL9 was first identified by cloning the t(1;14)(q21;q32) translocation from a patient with B-ALL. These findings indicate that BCL9 overexpression at different stages of B-cell development leads to distinct subtypes of B-cell malignancies. Finally, we investigated the BCL9 expression in human extramedullary plasmocytomas (EMP) and B-ALL. 32% of EMP cases analyzed by IHC expressed BCL9 at significant levels. Utilizing gene expression data available in the public domain we also showed that BCL9 is significantly overexpressed in ETV6-RUNX1 and TCF3-PBX1 subtypes of human B-ALL when compared to normal bone marrow counterparts, suggesting that BCL9 may play important roles in the pathogenesis of EMP as well as B-ALL in humans. Since BCL9 is highly expressed in tumors but not in the cells of origin and its interaction with β-catenin is specific, these results imply BCL9 as a promising candidate for targeted therapy.