Abstract
Abstract 675
The multiple myeloma SET domain (MMSET) gene is fused to the immunoglobulin locus in t(4;14)-associated multiple myeloma, and MMSET levels are elevated in these patients relative to other myeloma cases and normal cells. MMSET contains several domains commonly found in chromatin regulators including the PHD domain, PWWP domain and SET domain; responsible for histone methyl transferase (HMT) activity. What histone residues are methylated by MMSET in vivo has been uncertain. A well-folded, highly active form of the MMSET SET domain made in bacteria was promiscuous, methylating the H3K36, H3K27 and H4K20 residues of native histone as well as itself. To determine how MMSET affects chromatin in vivo and to identify genes regulated by MMSET, we engineered t(4;14)+ KMS11 cells with a tetracycline-inducible shRNA, leading to a >90% decrease in MMSET expression. Upon loss of MMSET expression, there was a striking decrease of trimethylated histone 3, lysine 36 (H3K36me3), a mark of transcriptional elongation and repression of intragenic transcription. At the same time, loss of MMSET expression was associated with a strong increase in H3K27me3, a chromatin mark associated with gene repression. For gain-of-function studies, the overexpressed MMSET allele in KMS11 cells was disrupted by homologous recombination (KMS11-TKO). KMS11-TKO cells, stably infected with a retrovirus carrying MMSET, displayed high levels of H3K36me3 and loss of H3K27me3. A specific mutation (Y1118A) in the SET domain of MMSET, predicted from crystal structure models to be required for histone binding, abrogated HMT activity of MMSET in vitro. Accordingly expression of MMSET Y1118A in KMS11-TKO cells failed to increase H3K36me3 levels. Another mutant, F1177A actually increased H3K36me3 levels above those stimulated by wild-type MMSET itself, most likely due to an expanded “pocket” within the SET domain that removed steric obstacles to the conversion of H3K36 from the mono-methyl to tri-methyl state. Collectively, these data indicate that H3K36 is a major methylation target of MMSET in vivo. To determine the genes regulated by MMSET and the importance of histone methylation in MMSET action, we profiled gene expression in both gain and loss-of-function systems using Illumina HumanWG-6 v3.0 expression arrays. We compared these gene lists with the top 2000 genes bound by MMSET as determined in a ChIP-on-chip assay using NimbleGen 2.7kb promoter arrays. MMSET knockdown affected expression of 1845 genes (FC>1.5, p<0.05); 931 were upregulated and 914 had reduced expression levels. Among these, 192 genes were also bound by MMSET. Re-expression of MMSET in KMS11-TKO cells led to increased expression of 749 genes while 788 genes were downregulated; 176 of these genes were also bound by MMSET. There were 38 genes that were bound by MMSET and regulated in both systems, including BMF, BTG2 and TP53INP1. These genes implicated in apoptosis represent potential direct transcriptional targets of MMSET. Furthermore, functional annotation of genes bound and regulated by MMSET in either the knockdown or repletion system, using Ingenuity Pathway Analysis, showed enrichment of genes implicated in the regulation of cell death and the p53 pathway (e.g. BAX, BCL2, CASP6), the cell cycle (CCNE2, E2F2, TP53INP1, CDC25A) and integrin-mediated signaling (ACTB, CDC42, ITGAL). The effect of MMSET on integrin signaling is of interest given that loss of MMSET expression or repletion of KMS11-TKO cells with MMSET altered the adhesive and growth properties of KMS11 cells. Finally, gene expression changes were contrasted between re-expression of wild-type MMSET and catalytically inactive MMSET Y1118A. Strikingly, the Y1118A mutant, which was deficient in altering cell adhesion and which did not change bulk histone methylation, altered expression of 1209 genes, 50% overlapping with those regulated by wild-type MMSET. Genes regulated by MMSET and the SET domain mutant were enriched mostly in cellular metabolism pathways (FDPS, IDI1, MVK) suggesting that effects on the cell cycle, adhesion and p53 pathways required the HMT activity of MMSET. These data indicate that MMSET can regulate genes in a HMT dependent and independent manner. Furthermore, MMSET target genes may be both activated and repressed upon changes in MMSET levels, indicating a complex interplay with the transcriptional machinery, likely through interactions with other transcriptional co-factors.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.