Oncogenic perturbations of the non-coding regulatory genome in classic Hodgkin lymphoma

BACKGROUND. Whereas treatment of classic Hodgkin lymphoma (cHL) has a major success story, and although cHL was the first lymphoid malignancy to be recognized, its biology has remained enigmatic for a long time because of two main reasons. First, the rareness of the tumor cells [i.e., Hodgkin and Reed-Sternberg (HRS) cells] in the affected lymph nodes, which has much hindered their molecular analysis. Second, the very unusual phenotype of HRS cells. Despite they originate from germinal center (GC) B-cells, HRS cells nearly completely lack the B-cell receptor and a range of classic B lineage markers that are essential for the survival and function of normal GC B-cells. Such gross reshapes of the GC B-cell program that characterize HRS cells is still enigmatic in its cause.

RATIONALE. Sequencing of the coding genome of cHL did not reveal somatic lesions in the B-cell transcription program or in the B-cell receptor pathway, but the non-coding genome of cHL is limitedly explored. Non-coding mutations can modify gene expression by affecting its proximal or distal cis-regulatory regions, or by introducing premature termination codons in abnormally spliced transcripts. Our preliminary data strongly support the hypothesis that non-coding mutations might be the cause of the gross reshape of the GC cellular program that characterizes HRS cells.

OBJECTIVES. We aim at identifying regulatory units of the HRS cell genome, mapping somatic mutation hotspots on them, and functionally testing whether non-coding regulatory mutations have direct cis-consequences on the expression of genes implicated in GC B-cell program, and if they are essential for HRS cell viability.

METHODS. We will identify non-coding mutations of cHL (130 cases in total) at the genome level by whole genome sequencing (60x) of genomic DNA of 10 cHL cell lines and of flow sorted HRS cells from 20 cHL cases, and by deep whole genome sequencing (200x) of high-tumor-fraction cell free DNA from 100 patients with cHL. Then we will further validate hot spots of non-coding mutations by CAPP-seq of low-tumor-fraction cell free DNA from 300 patients. To map the cis-regulatory network of cHL (and of normal GC B-cells for comparison) we will decorate chromatin regulatory domains, acting either in a proximal or a distal fashion, by ChIP-seq and Chromatin Interaction Analysis (ChIA-PET-seq) of histone marks in 10 cHL cell lines. ATAC-seq of flow sorted HRS cells from 20 patients will be used to validate the cis-regulatory network returned by cell lines in primary tumors. RNA-seq will be used to gain a glimpse on cognate gene transcription and splicing. By using biochemical and genome editing assays, we will determine whether somatic mutations of regulatory regions have cis-functional consequences on genes of clear predicted significance for the gross reshape of the GC cellular program that characterize HRS cells.

EXPECTED RESULTS. AID-mediated mutational misfiring affects non-coding regions of the HRS genome that are linked to important regulators involved in GC B-cell specification program in several ways: by creating de novo transcription factor-binding site, by altering a topologically associating domain, thus disrupting enhancer-promoter interaction, by disrupting transcription factor-binding or by creating new splice sites, which in turns result in improperly transcribed genes. Ultimately, non-coding mutations orchestrate a constitutive, dysregulated expression of a physiologic gene-expression program or promote an ectopic program that cause the gross reshape of HRS cell identity.

IMPACT. The project may have implication on the knowledge on the pathophysiology of cHL. It can also nominate biomarkers for the subclassification of cHL, which is currently morphological and has limited clinical impact, and nominate vulnerabilities that can be exploited to develop new therapeutic concepts.