Abstract
Mantle cell lymphoma (MCL) is an aggressive mature B-cell neoplasm molecularly characterized by the presence of the t(11;14) and expression of the transcription factor SOX11 in most of the patients. A subset of tumors do not express SOX11 and are frequently associated with hypermutated IGHV, a predominantly non-nodal leukemic disease and an indolent clinical course. To gain insight into the potential role of epigenetic alterations in the pathogenesis of MCL, we have extensively characterized the DNA methylomes of 86 tumors by high density DNA methylation arrays and whole-genome bisulfite sequencing (WGBS) of selected samples, using naïve and memory B cells (NBC and MBC, respectively) as controls.
An unsupervised principal component analysis showed the presence of two major MCL subgroups associated with different molecular features such as SOX11 expression levels and IGHV mutational status. The major subgroup (C1) comprised mostly SOX11 positive MCLs with low levels of IGHV somatic hypermutation (SHM), although some SOX11 positive cases with high levels of SHM also belonged to C1. Interestingly, this group also contained few SOX11 negative cases, all presenting low levels of SHM (IGHV germline identity ≥98%). The second, smaller subgroup (C2), was comprised almost entirely by SOX11 negative cases, all having high levels of SHM (IGHV germline identity <98%). These data suggest that at the DNA methylation level, a threshold of 98% identity seems to be more accurate to separate MCL cases according to IGHV mutational status.
We next compared the DNA methylation profiles of C1 and C2 and we observed a high level of de novo hypomethylation specific for C1, which was neither observed in C2 nor in the normal controls. This specific hypomethylation mainly targeted enhancers and transcriptionally active regions. Interestingly, the hypermethylation observed in C1 as compared to C2 seemed to be associated with DNA methylation changes occurring during normal B-cell differentiation. Considering the methylation levels of these CpGs, C1 cases have similar methylation patterns as pre-germinal center B cells (NBC), while C2 MCLs seem to be similar to post-germinal center B cells (MBC). These data indicate that MCL cases from C1 and C2 display epigenetic imprints of different B cells and suggest a different cell of origin (COO) for these two subsets of MCL.
Afterwards, we compared the two epigenetic MCL subgroups with its respective putative normal counterpart (i.e. C1 vs. NBC and C2 vs. MBC). C1 MCL showed massive changes as compared to NBCs (array: n=43,246 CpGs, 35,087 hypo- and 8,070 hypermethylated; WGBS: n=2,329,576 DMRs, 2,153,100 hypo-, 176,746 hypermethylated) 78% of which were shared with differences observed between NBC and MBC, indicating that many of the changes occurring in C1 cases where shared with the normal B-cell differentiation process. Hypomethylated CpGs occurred mostly in intragenic enhancers whereas hypermethylation targeted mainly polycomb repressed regions. Comparison of C2 MCL and MBC showed only about 2,000 differentially methylated CpGs (array: n=2,196 CpGs, 1,544 hypo- and 331 hypermethylated; WGBS: n=1,566,450 DMRs, 1,082,672 hypo-, 483,778 hypermethylated). Although changes occurred at lower scale, these CpG changes target similar chromatin states as those in C1 MCL cases.
In conclusion, our study shows the existence of two major MCL subgroups that, based on DNA methylation pattern, seem to have originated from B cells at different differentiation stages. In addition to this biological insight, we also identified epigenetic signatures in C1 and C2 that were associated with clinical outcome.
Klapper:Roche: Research Funding.
Author notes
Asterisk with author names denotes non-ASH members.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal