Elucidation and Pharmacological Targeting of Novel Molecular Drivers of Follicular Lymphoma Progression

Abstract

Follicular lymphoma (FL), the most common indolent subtype of non-Hodgkin’s lymphoma, is associated with a relatively long overall survival rate ranging from 6 to 10 years from time of diagnosis. However, in 20–60% of FL patients, transformation to aggressive diffuse large B-cell lymphoma (DLBCL) reduces median survival to only 1.2 years. The specific functional and genetic determinants of FL transformation remain elusive, and genomic alterations underlying disease advancement have only been identified for a subset of cases. Therefore, to identify candidate drivers of FL transformation, we performed systematic analysis of a B-cell-specific regulatory model exhibiting FL transformation signatures using the Master Regulator Inference algorithm (MARINa). This analysis revealed FOXM1, TFDP1, ATF5, HMGA1, and NFYB to be candidate master regulators (MR) contributing to disease progression. Accordingly, validation was achieved through synthetic lethality assays in which RNAi-mediated silencing of MRs individually or in combination reduced the viability of (14;18)-positive DLBCL (t-DLBCL) cells. Furthermore, specific combinations of small molecule compounds targeting synergistic MR pairs induced loss of viability in t-DLBCL cells. Collectively, our findings indicate that MR analysis is a valuable method for identifying bona fide contributors to FL transformation and may therefore guide the selection of compounds to be used in combinatorial treatment strategies.

Cross-Cohort Analysis Identifies a TEAD4–MYCN Positive Feedback Loop as the Core Regulatory Element of High-Risk Neuroblastoma

“…The results of this study are also highly consistent with our proposed model of a recurrent tumor architecture responsible for canalizing the effect of multiple genomic alterations to implement critical programs necessary to tumor initiation and stability (15). While MR proteins were discovered in multiple prior study (10, 13, 87, 88) for instance, their modular regulatory architecture and the existence of core autoregulatory loops had not been previously reported. These findings, combined with previous MR protein studies, suggest that similar regulatory architectures may be responsible for the implementation and stability of other cancers.…”

“…Critically, aberrant activity of tumor checkpoint MRs arises from one or more genetic alterations in their upstream pathways (14, 17), thus providing a rationale for the heterogeneous mutational landscape of tumors with highly similar transcriptional states (15). MR proteins have been shown to represent critical molecular tumor vulnerabilities that can be effectively targeted pharmacologically in several human malignancies (10, 11, 16, 18, 19), thus providing novel therapeutic opportunities. Here, we sought to interrogate regulatory networks of high-risk neuroblastomas to identify and validate MR proteins representing highly conserved, tumor-specific vulnerabilities for specific subtypes of this malignancy.…”

The recurrent architecture of tumour initiation, progression and drug sensitivity

“…Recent results suggest that dystatic cell states may be implemented by a handful of tumour-specific MR proteins — working cooperatively within tight modular structures (tumour checkpoints) — the aberrant activity of which is necessary and sufficient for tumour phenotype presentation and maintenance 7,9–11,29,30 . If this hypothesis were further confirmed, these tumour checkpoints and associated MRS would constitute a novel class of mechanistic biomarkers and therapeutic targets 11,29–31 and would further link cancer and development.…”

“…MR analyses have helped elucidate proteins that regulate tumour-associated phenotypes as diverse as predisposition 40,44 , subtype-specific tumorigenesis 10,43,50,51 , progression to aggressive or metastatic disease 9,11,12,45,47,52 , stroma-specific regulation of tumour outcome 41 and drug resistance 29,30 , most of which have been experimentally validated. In addition, their use in non-cancer phenotypes has helped to elucidate an equivalent disease checkpoint architecture in neurological phenotypes — including amyotrophic lateral sclerosis (ALS) 53 , Alzheimer disease 7,54 , Parkinson disease 55 and alcohol addiction 56 — and in developmental phenotypes, from regulation of germinal centre formation 39 to stem cell pluripotency 57 .…”

“…1b). Thus, although individual MR proteins are highly tumour specific 9–11 , the resulting tumour checkpoint oncotecture is virtually tumour independent 7 .…”