LAUSR

Background and Objective: Evidence suggests that alterations in histone modifications are crucial in cancer development and progression. Among these, histone 3 lysine 4 trimethylation (H3K4me3) which is one of the most common histone modifications, is found in promoters of active genes and is proposed to promote gene expression. Aberrant H3K4 methylation is association with cancer. In our previous study, we found a germline MLL3 stop codon mutation in a Chinese pedigree. MLL3 proteins act as methyltransferase and regulate H3K4 methylation. Our study aims to evaluate the H3K4me3 genome binding pattern in this cancer pedigree with MLL3 mutation and have further understanding about pathogenesis in this pedigree. Also we hope to find valuable epigenetic marker of cancer. Materials and Methods: To evaluate 3 cancer patients in a Chinese pedigree with a heterozygous stop codon mutation in MLL3 (coding a histone H3K4me3 methyltransferase) segregating with colorectal carcinoma (CRC) and acute myeloid leukemia (AML) and a mutation-free normal control, ChIP-chip analyses were performed to determine which promoters coprecipitated with the H3K4me3 antibody in both patients and control. The pulled-down genes were mapped to Kyoto Encyclopedia Of Genes And Genomes (KEGG) pathways. Results:We selected 1 AML patient, 2 CRC patients, and 1 normal control from a Chinese pedigree. A previous study showed that the 3 patients all harbored an MLL3 stop codon mutation. MLL3 belongs to the human TRX/MLL family and is an important mammalian H3K4 methyltransferase. In this study, using ChIP-chip, we detected the global promoter occupancy profile of H3K4me3 in the two different tumor types compared with a normal control. Compared with the normal control, H3K4me3 differentially bound to the promoter regions of 2425 (CRC), 22319 (CRC), and 2385 (AML) genes in the three patients, with an overlap of 877 genes between AML and CRC tumors. To our surprise, the total numbers of promoters pulled down by the H3K4me3 antibody were not significantly different between the 3 cancer patients and the mutation-free normal control. However, the H3K4me3 binding patterns were unique in the cancer patients compared with the normal control. We also noticed a subtle difference in H3K4me3 binding patterns between the AML patient and the 2 CRC patients. The mother-son pair (CRC-AML) and the sister pair (CRCs) each should have 50% genetic similarities (sex chromosome genes were excluded); however, given the same MLL3 nonsense mutation, the CRC mother-AML son pair showed greater differences in H3K4me3 pulled-down genes than did the CRC sister pair. To understand the possible biological effects of H3K4me3 in AML and CRC, we annotated the functions of the identified genes by mapping them to KEGG pathways. Although the number of identified genes was not significantly differ between patients and control, we found an unique profile of H3K4me3 in AML and CRC patients compared with control that the identified genes were concentrated in the carcinogenesis-associated pathways, including the TGF-beta, Wnt, and MAPK signaling pathways. Conclusions:The results indicate that the H3K4me3 patterns are differentially altered in different cancers, which may play a role in carcinogenesis, and the MLL3 mutation may influence the alteration of H3K4 methylation. Our studies make it is possible to link genetic changes and epigenetic modifications. Clearly, a large sample cohort study is needed to test H3K4me3 binding patterns in MLL3 mutation patients, as well as in all cancer patients and even normal individuals. No relevant conflicts of interest to declare.