Abstract A16: Master regulators of transformation of follicular lymphoma to diffused large B-cell lymphoma in patients defined by expression and activity of MYC

Abstract

Abstract Follicular lymphoma (FL) is the most common indolent subtype of non-Hodgkin's lymphoma with relatively long overall survival rate where median survival ranges from 6 to 10 years from the time of diagnosis. FL is a consequence of malignant transformation of germinal center (GC) derived B-cells and is associated with the translocation t(14;18)(q32;q21) resulting in overexpression of BCL2. FL transforms to more aggressive form of lymphoma, Diffused Large B-Cell lymphoma (DLBCL), with very high frequency (20-60% cases) causing the dramatic decrease in median survival rate to about 1.2 year from transformation. Several abnormalities were associated with transformation, such as alterations in TP53, MYC and BCL6, inactivation of p15 and p16, or amplification of REL, however they were observed only in a fraction of transformed lymphomas and the precise mechanism of transformation is still unknown. In the past, we have used a computational algorithm, called Master Regulator Inference algorithm (MARINa) which infers genes causally responsible for rendering a specific phenotype by interrogating cell context specific regulatory networks using gene signatures of disease progression. MARINa has been extensively validated at the experimental level leading to discovery of novel regulatory genes and mechanisms of disease progression in different cellular contexts. In this study, we applied MARINa algorithm to infer master regulators (MRs) of FL transformation to DLBCL using expression profiles from patient's biopsies at the time of FL diagnosis and subsequent biopsies at the time of transformation (1). Gene expression profile (GEP) analysis revealed two subtypes among the patients diagnosed with transformed FL, one with “high activity MYC” and another with “low activity MYC”. We performed MR analysis for both groups separately and found that MRs with high activity in one group were associated with low activity in the other group. In “high activity MYC” group we identified FOXM1, TFDP1, ATF5 and HMGA1 as best candidate MRs for therapeutic targeting of FL transformation. We show that targeting these MRs by small interfering RNA can reverse transformed signature. Furthermore, to infer novel small molecules to target transformed DLBCL cells study we used gene expression profiles from DLBCL cell lines treated with a library of FDA-approved and late-stage experimental 92 compounds and computed the similarity of these profiles with the profile obtained by silencing MR genes in DLBCL cell lines. Based on the gene signature similarity with gene profiles from DLBCL cell lines after MR silencing we predicted several compounds which in experimental settings recapitulated synergistic mechanism of MRs and induced cell death in transformed DLBCL cell lines but not in FL or control normal B-cells. In conclusion, our studies demonstrate that integration of computational predictions and experimental validations provide novel Master Regulators genes of FL transformation to DLBCL and small molecule inhibitors with anti-tumorigenic potential for DLBCL therapy.