The BCL6 Oncogene Drives an Epigenomic Program Linking Stem Cells to the Pathogenesis of Human Diffuse Large B Cell Lymphoma in Mice.

Diffuse large-B-cell lymphoma (DLBCL) is the most common lymphoid malignancy in adults. The BCL6 oncogene, a transcriptional repressor, is frequently involved in DLBCL pathogenesis. Constitutive expression of BCL6 in GC cells leads to larger GCs and development of DLBCL in mice, although constitutive expression of BCL6 earlier in the B-cell lineage has the opposite effect and abrogates B-cell development. Limitations in our knowledge of how BCL6 mediates transcriptional programming to impose the malignant phenotype presents a barrier towards improved understanding and therapeutic targeting of B-cell lymphomas. Recent advances in stem cell biology suggest that many tumors are sustained by a pool of biological distinct cancer stem cells (CSCs), which may be uniquely susceptible to epigenomic programming by oncogenes and require specific therapeutic targeting. Although CSCs have not yet been identified in DLBCL, we hypothesized that such cells may exist in these tumors and their characterization could be critical for studies of lymphomagenesis and therapeutics. In order to test this hypothesis our group developed a mouse model where BCL6 is expressed transiently in the stem cell compartment at levels comparable to normal GC cells. These animals undergo completely normal hematopoiesis, except that 100% develop lymphomas within 20 to 28 weeks. Extensive histopathologic, molecular and expression array phenotyping demonstrate that these tumors are most similar to human DLBCLs. However, these tumor cells do not express BCL6, and so must be transcriptionally programmed at the stem cell level to behave in a malignant fashion. Moreover, transplantation of DLBCL cells does not propagate the tumor whereas transplantation of the CSCs rapidly re-establishes the DLBCL in recipient mice. In-depth microarray and QPCR transcriptional profiling demonstrate that BCL6 imposes a B cell lymphoma program in the stem cell compartment. This transcriptional programming subsequently becomes latent during hematopoiesis and reappears only in the mature B-cell fraction. We reasoned that this effect must be dependent on epigenetic programming imposed by BCL6 in CSCs. To determine whether this is the case we compared and contrasted the epigenetic signature of CSCs, and DLCBL tumors in the transgenic animals vs. those of normal stem cells and mature B-cells in control animals. This was performed using the HELP (HpaII tiny fragment enrichment by ligation mediated PCR) method with a custom designed high-density oligonucleotide microarray that allows quantitative detection of DNA methylation at 24,000 murine gene promoters. The data show that BCL6 imposes a unique DNA hypermethylation signature in CSCs that persists in the DLBCL cells. Taken together, our results suggest that BCL6 mediates transcriptional and epigenomic programming in stem cells that can latently persist during differentiation (thus allowing early B-cell development to proceed) and impose the DLBCL phenotype in their mature progeny. Since human lymphomas exhibit similar transcriptional programming, therapeutic targeting of the BCL6 epigenomic program might enhance the power of anti-lymphoma regimens to eradicate the disease.