BCL6 Is Required for Leukemia-Initiation and Self-Renewal Signaling in Chronic Myeloid Leukemia.

Abstract 2167

Poster Board II-144

The transcriptional repressor BCL6 was discovered as a potent protooncogene in diffuse large B-cell lymphoma (DLBCL) and is also expressed in normal germinal center (GC) B cells. BCL6 protects DLBCL and GC B cells against DNA damage-induced apoptosis by suppressing target genes including p53, p21, ATR and p27. In preliminary experiments for this study, we found that treatment of BCR-ABL1-driven Ph⁺ ALL cells with Imatinib results in a 60- to 90-fold up-regulation of BCL6, reaching protein levels as high as in DLBCL and GC B cells. Since the oncogenic BCR-ABL1 kinase also drives leukemic transformation in CML cells, we studied regulation of BCL6 in human CML cells upon TKI-treatment. Like in Ph+ ALL, TKI-treatment resulted in rapid BCL6 mRNA and protein upregulation in CML cells, albeit to lesser degree than in Ph⁺ ALL and DLBCL. We hypothesized that upregulation of BCL6 in CML cells in response to TKI-treatment serves the same function as in DLBCL and protects from p53-mediated apoptosis. Through transcriptional repression of p53, p21, ATR and p27 as in DLBCL, upregulation of BCL6 in response to TKI-treatment would represent a defense mechanism of CML cells to evade TKI-induced cell death.

We tested the role of BCL6 in CML in a genetic loss-of-function experiment: To this end, myeloid progenitor cells from bone marrow of BCL6^+/+ and BCL6^-/- mice were transformed with BCR-ABL1. Compared to their BCL6^+/+ counterparts, BCL6^-/- mouse CML cells were highly sensitive to Imatinib-treatment. In a complementary approach, we treated human CML cells with either Imatinib alone or with a novel BCL6 peptide inhibitor (Retro-inverso BCL6-peptide inhibitor, RI-BPI). Consistent with findings in BCL6^-/- mouse CML cells, also human CML cells were significantly sensitized to Imatinib by BCL6 inhibition (using 5 mmol/l RI-BPI). Of note, the BCL6 inhibitory peptide RI-BPI alone induced no significant toxicity in human CML cells but synergized with Imatinib-treatment.

Since RI-BPI alone had no cytotoxic effect on CML cells, we tested whether loss of BCL6 function affects self-renewal capacity of CML cells. This possibility was explored in two complementary colony formation assays: In one set of experiments, we tested the ability of BCL6^+/+ and BCL6^-/- mouse CML cells to form colonies in semisolid methylcellulose agar. The comparison between BCL6^+/+ (94 ± 11 colonies) and BCL6^-/- (<1 ± 1 colonies; p<0.004) mouse CML cells revealed that BCL6 represents an absolute requirement for self-renewal signaling in CML cells. In line with these results, we observed that BCL6^-/- mouse CML cells undergo replicative senescence and cell cycle arrest under cell culture conditions after ∼2 months, whereas leukemic growth proceeds undiminished in BCL6^+/+ CML cells. These results were confirmed studying BCL6 function in human CML cells using RI-BPI. Despite lack of toxicity, 5 mmol/l RI-BPI drastically reduced the number of colonies that were formed by human CML cells (980 ± 79 vs 262 ± 174; p= 0.001). Based on these findings we conclude that inhibition of BCL6 using a novel peptide inhibitor effectively disrupts self-renewal signaling and may be developed as a novel therapeutic approach for leukemia stem cell eradication in CML.