Frequent mutation of histone modifying genes in non-Hodgkin lymphoma

PMID: 21796119 · DOI: 10.1038/nature10351 · Journal: Nature (2011)

TL;DR

This study sequenced genomes, exomes and transcriptomes of 127 non-Hodgkin lymphoma (NHL) cases (DLBCL and FL) and identified 109 recurrently somatically mutated genes. The central finding was that histone-modifying genes are frequent mutation targets in NHL: KMT2D (MLL2) was mutated in 89% of FL and 32% of DLBCL cases, while MEF2B harboured recurrent missense mutations in 13-15% of cases. The work established disruption of chromatin biology as a core driver event in lymphomagenesis.

Cohort & data

• 127 B-cell NHL tumour samples plus 10 DLBCL cell lines (DLBCLNOS, FL).
• 14 cases with matched tumour/normal whole-genome or whole-exome sequencing; 113 additional cases with RNA-seq.
• KMT2D validation cohort: 89 cases (35 FL, 37 DLBCL, 17 DLBCL cell lines) via Illumina amplicon resequencing.
• MEF2B validation cohort: 261 FL, 259 DLBCL, 17 cell lines, 35 other NHL, 8 centroblasts via Sanger sequencing.
• Dataset: nhl_bcgsc_2011. SRA accession SRP001599 / dbGaP phs000235.v2.p1.
• Methods: whole-genome-seq, whole-exome-seq, rna-seq, sanger-sequencing.

Key findings

• 109 genes with confirmed somatic mutations in 2 or more NHL cases were identified from 651 candidate genes.
• 26 genes showed significant evidence of positive selection (FDR 0.03).
• KMT2D (MLL2): mutated in 89% (31/35) of FL, 32% (12/37) of DLBCL, and 59% (10/17) of DLBCL cell lines. 91% of mutations were inactivating (frameshift indels, nonsense, splice-site). Bi-allelic inactivation confirmed in 8 FL cases.
• MEF2B: recurrent non-synonymous mutations in 12.7% of DLBCL and 15.3% of FL. 59.4% of variants clustered at four residues (K4, Y69, N81, D83). Restricted to GCB subtype.
• GNA13: mutated in 18/106 DLBCL cases (17%), restricted to GCB subtype (P = 2.28e-4, Fisher exact test). Multiple nonsense mutations indicating tumour suppressor role.
• SGK1: mutated in 18/106 DLBCL cases, restricted to GCB subtype (P = 1.93e-3).
• EZH2: Y641 hotspot mutations with allelic imbalance favouring the mutant; additional novel sites A682G and A692V identified.
• Known oncogenes BCL2, CARD11, MYD88, CD79B confirmed with selection signatures.
• Novel hotspot mutations in FOXO1 (start codon, T24), CCND3, B2M (start codon), and BTG1.
• Mutual exclusion pattern between ABC-enriched mutations (MYD88, CD79B) and GCB-enriched mutations (EZH2, GNA13).

Genes & alterations

Gene	Alteration	Finding
KMT2D	Nonsense, frameshift indels, splice-site	Tumour suppressor; 89% FL, 32% DLBCL; bi-allelic loss
MEF2B	Missense (K4, Y69, N81, D83)	GCB-restricted; cooperates with CREBBP/EP300 in histone acetylation
EZH2	Y641 gain-of-function, A682G, A692V	Enhanced H3K27 trimethylation; GCB-enriched
CREBBP	Nonsense, missense	Inactivating; HAT activity disruption
EP300	Missense	HAT activity disruption
GNA13	Nonsense, missense (T203A)	Tumour suppressor; GCB-restricted; RhoA signalling
SGK1	Nonsense, missense	Tumour suppressor; GCB-restricted; PI3K-regulated kinase
TP53	Nonsense, missense	Tumour suppressor; GCB-enriched
TNFRSF14	Nonsense	Tumour suppressor; GCB-restricted
BTG1	Nonsense, missense	Apoptosis regulation
BCL2	Missense (somatic hypermutation)	Allelic imbalance favouring translocated allele
CARD11	Missense (F123I)	Oncogene; NF-kB signalling
MYD88	Missense	ABC-enriched; 39% of ABC DLBCL
CD79B	Hotspot	ABC-enriched; BCR signalling
TNFAIP3	Inactivating	ABC-enriched; 24% of ABC DLBCL
FOXO1	Start codon (M1), T24	N-terminal truncation; GCB-associated
CCND3	Hotspot	Novel mutation target
B2M	Start codon	Immune evasion
BCL10	Nonsense (truncating)	NF-kB positive regulator; oncogenic truncated products

Clinical implications

• KMT2D inactivation is as prevalent as t(14;18) in FL (89%), suggesting it is a foundational genetic event in FL pathogenesis and a potential diagnostic/prognostic marker.
• The mutual exclusion of ABC- versus GCB-associated mutations supports molecular subtyping as a framework for precision medicine in DLBCL.
• Disruption of histone acetylation (CREBBP, EP300, MEF2B) and methylation (KMT2D, EZH2) pathways identifies epigenetic therapy (e.g., HDAC inhibitors, EZH2 inhibitors) as a rational therapeutic strategy.
• MEF2B mutations restricted to GCB-subtype tumours may serve as a GCB-specific biomarker.

Limitations & open questions

• Functional consequences of MEF2B missense mutations (K4, Y69, N81, D83) were not experimentally resolved; mechanism of oncogenic contribution remains speculative.
• The cohort lacked ABC cases with MEF2B mutations, limiting power to definitively confirm GCB restriction.
• KMT2D target genes whose dysregulation drives lymphomagenesis were not identified.
• The study did not assess whether KMT2D or MEF2B mutations carry prognostic significance independently of COO subtype.
• Cooperative effects between KMT2D loss and EZH2 gain-of-function were hypothesized but not experimentally tested.

Citations from this paper used in the wiki

• “32% of DLBCL and 89% of FL cases had somatic mutations in MLL2, which encodes a histone methyltransferase.” (Abstract)
• “11.4% of DLBCL and 13.4% of FL cases had somatic mutations in MEF2B, a calcium-regulated gene that cooperates with CREBBP and EP300 in acetylating histones.” (Abstract)
• “GNA13 was affected by mutations in 22 cases including multiple nonsense mutations.” (Results, p.4)
• “both SGK1 and GNA13 mutations were found only in GCB cases (P = 1.93x10-3 and 2.28x10-4, Fisher exact test)” (Results, p.5)
• “We confirmed MLL2 mutations in 31 of 35 FL patients (89%), in 12 of 37 DLBCL patients (32%), in 10 of 17 DLBCL cell lines (59%)” (Results, p.6)
• “59.4% of detected variants affecting K4, Y69, N81 or D83” (Results, p.6)

This page was processed by paper-compiler on 2026-05-06.