8 A Precision Medicine Approach Using Whole Transcriptome Profiling by RNA-seq for B Cell Cancers. (11th January 2018)

• Record Type: Journal Article
• Authors: Andrews, Jared; Zaydman, Mark; Pyfrom, Sarah; Schmidt, Jennifer; Luo, Hong; Koues, Olivia; Oltz, Eugene; Payton, Jacqueline
• Is Part Of: American journal of clinical pathology. Volume 149(2018)Supplement 1
• Page Start: S167
• Page End: S167
• Publication Date: 2018-01-11
• Subjects: Diagnosis, Laboratory -- Periodicals; Pathology -- Periodicals; 616.07
• DOI: 10.1093/ajcp/aqx149.377
• Languages: English
• ISSNs: 0002-9173

Abstract

Abstract: B-cell lymphomas and leukemias have an annual incidence of ~95, 000 cases. They exhibit alterations in genes in "targetable" signaling pathways, enabling a precision medicine approach for therapeutic decisions. However, molecular genetic testing is not yet standard of care for non-Hodgkin lymphoma (NHL). Therefore, we have used whole transcriptome profiling by RNA-seq to simultaneously identify mutations and expression changes in B-cell cancers. We have profiled ~70 primary B-cell cancers and ~50 primary normal B-lymphocyte subsets (naïve, germinal center, memory, and in vitro activated). We first compared gene expression in tumor types to normal B-cell subsets. Our approach segregated major tumor subtypes (follicular [FL] and germinal center B [GCB] or activated [ABC] diffuse large B-cell lymphomas [DLBCLs], chronic lymphocytic leukemia [CLL]) and revealed two novel subtypes of FL. We also identified tumors with overexpression of genes that are associated with poor prognosis (BCL2, MYC). Variants were identified by at least two variant callers and met stringent quality filters; common SNPs (>1% frequency) were excluded. We identified mutations in chromatin modifier, transcription factor, and B-cell receptor signaling proteins. To evaluate the functional impact of these mutations, we designed a method to calculate the expressed mutant allele ratio (EMAR). In general, these ratios corresponded to the proportion of tumor cells. However, the EMAR of some genes was substantially higher than the proportion of tumor cells, which we termed skewed EMAR . This could be caused by a high mutational burden in non-tumor cells. Indeed, the genomes of normal and malignant B cells exhibit evidence of activation-induced cytidine deaminase (AID) activity, which mediates somatic hypermutation. However, AID activity is not limited to Ig genes; mutations in several lymphoma proto-oncogenes are induced by AID. In fact, some of the genes with skewed EMAR are known AID off-targets (BCL2, BCL6, PAX5). Notably, several other genes with a skewed EMAR have not previously been associated with AID activity. Finally, we have previously shown that sequence variants in non-coding gene regulatory regions promote tumorigenesis by deregulating expression. Therefore, we calculated the burden of sequence variation to one megabase resolution across the transcriptome. Strikingly, we identified several novel regions with significantly enriched mutation burden, the majority of which overlapped NHL enhancers. One of these is flanked by two immunoglobulin receptor genes, FCMR and PIGR, which are associated with CLL (FCMR) or autoimmune disease (PIGR). The mechanisms by which these genes are deregulated and promote disease have not been defined, but the non-coding regulatory mutations we identified may play a role. In summary, this study demonstrates the power of RNA-seq to discover novel mechanisms of lymphomagenesis and to simultaneously identify mutations and expression changes that stratify risk and suggest effective therapies.