Modeling The Role Of Progression In Follicular Lymphoma

Ronald Levy, MD, PhD (Lead)

Abstract

Follicular lymphoma (FL) is the most common indolent non-Hodgkin's lymphoma subtype in adults and remains incurable. Though initially having an indolent course, FL frequently transforms into a more aggressive disease by clonal evolution. Within this transformation process there is evidence for a tumor cell-intrinsic clock and for an interaction with the host immune system, and to deviations from the ordered development of maturing B lymphocytes as they gain antigen experience. We have recently discovered that each FL tumor contains a subpopulation of tumor cells that has defective signaling through their B cell antigen receptor. The size of this BCR insensitive tumor subpopulation expands over time, and this increase predicts poor response to chemotherapy and risk of death from disease. Separately, we have discovered a relationship between the expression of gene expression signatures reflecting self-renewal pathways of embryonic stem cells and the risk for histological transformation and death. The main question addressed by this proposal is "what is the relationship between the two subpopulations of tumor cells- the BCR sensitive and the BCR insensitive, and to developmental hierarchies including self-renewing populations?" Several models can be envisioned to explain the observations: 1. A rare self-renewing tumor initiating cell that gives rise to successive subpopulations of end stage tumor cells, 2. Parallel evolution of tumor subpopulations from a rare tumor initiating cell, and 3. Successive evolution of tumor subpopulations to acquire more and more self-renewal properties. Using approaches from systems biology, we will generate new data and apply integrative computational and experimental methods to predict and then to validate one or more of these models. The experimental component will focus on the characterization of the signaling responsiveness of subsets within FL tumors to BCR ligation, the prospective purification of these subsets for the assessment of somatic mutations in BCR signaling pathway, as well as the profiling of their genomes and transcriptomes. The computational component will focus on the characterization of determinants of BCR responsiveness in these subsets, including definition of its genetic and expression signatures, derivation of empirically refined signaling networks, construction of clonal hierachicies relating them, and relationships to programs of self-renewing cell populations. Validation will entail experimentally testing hypotheses generated from the computational models, and assessing robustness of models on publicly available data from patients with FL.