Identification of BCL6 As a Therapeutic Target in RAS-Driven Acute Lymphoblastic Leukemia

Background & Hypothesis: In ~50% of cases of acute lymphoblastic leukemia, activating lesions in RAS pathway are found. These lesions are particularly frequent in relapse ALL and specifically acquired in the relapse clone (Irving et al., 2013). A previous study of senescence rescue screens identified the transcriptional repressor BCL6 as a key factor to overcome p53 dependent senescence and enable RAS-mediated transformation of mouse embryonic fibroblasts (Shvarts et al. 2002). Using peptide inhibitors and small molecules that interfere with the ability of BCL6 to recruit essential corepressors, we and others recently demonstrated that therapeutic targeting of BCL6 in BCL6-dependent malignancies is feasible. Here we tested the hypothesis that BCL6 represents a therapeutic target in multiple subsets of ALL, including ALL with RAS pathway lesions.

Results: In support of this hypothesis, we found that inducible expression of oncogenic NRAS^G12D increased BCL6 mRNA levels by ~350-fold (qRT-PCR) and protein levels by ~50-fold (Western blot). Upregulation of BCL6 in response to NRAS^G12D activation was sensitive to treatment with the MEK kinase inhibitor PD325901 and correlated with levels of phospho-ERK downstream of MEK. These findings suggest that BCL6 expression is a result of ERK activation downstream of oncogenic NRAS^G12D and MEK in ALL cells. To verify this observation in patient-derived cells, we compared ALL cells that were isolated at the time of initial diagnosis (D), and at the time of relapse (R) from the same patient. Interestingly, the patient had acquired a KRAS^G12V mutation at the time of relapse. As a likely consequence, we found both hyper-phosphorylation of ERK and overexpression of BCL6 in the relapse cells (KRAS^G12V), but not in diagnosis sample (KRAS wild-type). R-ALL cells harboring the KRAS^G12V mutation were more sensitive to the treatment with the MEK inhibitor PD325901 and the BCL6 peptide inhibitor RI-BPI than D-ALL cells. BCL6 inhibition also markedly increased survival rate of NOD/SCID mice xenografted with R-ALL cells. These results suggested BCL6 served as an important contributor to RAS -mediated transformation.

To further study the mechanistic role of BCL6 in RAS-mediated pre-B cell transformation, we tested its function in a mouse ALL model. Pre-B cells from both Bcl6+/+ and Bcl6-/- mice could be transduced by NRAS^G12D and achieved growth-factor independence under cell culture conditions. However, Bcl6-/- NRAS^G12D ALL cells failed to initiate fatal leukemia in NOD/SCID transplant recipient mice, whereas Bcl6+/+ NRAS^G12D ALL cells gave rise to lethal leukemia in all transplant recipients. Studying Cre-mediated deletion of Bcl6-fl/fl alleles in a complementary mouse model revealed that continuous presence of Bcl6 function is required for normal proliferation of ALL cells. Cre-mediated ablation of BCL6 in NRAS^G12D driven ALL induced rapid cell death and completely abrogated the ability of NRAS^G12D ALL cells to form colonies.

Conclusion: These results support that BCL6 is not only required for the initiation of RAS-transformed ALL in vivo but also for the maintenance of fully established RAS-driven leukemia. The findings provide genetic evidence for BCL6 function as a critical cofactor of RAS-mediated transformation in human ALL. Inhibition of BCL6 in RAS-driven ALL may be useful to prevent leukemia relapse after initial remission (Bcl6-dependent leukemia-initiation) and also to achieve remission by combining conventional cytotoxic therapies with currently available BCL6 inhibitors (e.g. RI-BPI peptide inhibitor or FX-1085 small molecule inhibitor).