Byungho Lim

Accurate quantification of transcriptome from RNA-Seq data by effective length normalization

We propose a novel, efficient and intuitive approach of estimating mRNA abundances from the whole transcriptome shotgun sequencing (RNA-Seq) data. Our method, NEUMA (Normalization by Expected Uniquely Mappable Area), is based on effective length normalization using uniquely mappable areas of gene and mRNA isoform models. Using the known transcriptome sequence model such as RefSeq, NEUMA pre-computes the numbers of all possible gene-wise and isoform-wise informative reads: the former being sequences mapped to all mRNA isoforms of a single gene exclusively and the latter uniquely mapped to a single mRNA isoform. The results are used to estimate the effective length of genes and transcripts, taking experimental distributions of fragment size into consideration. Quantitative RT–PCR based on 27 randomly selected genes in two human cell lines and computer simulation experiments demonstrated superior accuracy of NEUMA over other recently developed methods. NEUMA covers a large proportion of genes and mRNA isoforms and offers a measure of consistency (‘consistency coefficient’) for each gene between an independently measured gene-wise level and the sum of the isoform levels. NEUMA is applicable to both paired-end and single-end RNA-Seq data. We propose that NEUMA could make a standard method in quantifying gene transcript levels from RNA-Seq data.

Integrative genomics analysis reveals the multilevel dysregulation and oncogenic characteristics of TEAD4 in gastric cancer

Tumorigenesis is a consequence of failures of multistep defense mechanisms against deleterious perturbations that occur at the genomic, epigenomic, transcriptomic and proteomic levels. To uncover previously unrecognized genes that undergo multilevel perturbations in gastric cancer (GC), we integrated epigenomic and transcriptomic approaches using two recently developed tools: MENT and GENT. This integrative analysis revealed that nine Hippo pathway-related genes, including components [FAT, JUB, LATS2, TEA domain family member 4 (TEAD4) and Yes-associated protein 1 (YAP1)] and targets (CRIM1, CYR61, CTGF and ITGB2), are concurrently hypomethylated at promoter CpG sites and overexpressed in GC tissues. In particular, TEAD4, a link between Hippo pathway components and targets, was significantly hypomethylated at CpG site cg21637033 (P = 3.8 × 10(-) (20)) and overexpressed (P = 5.2 × 10(-) (10)) in 108 Korean GC tissues compared with the normal counterparts. A reduced level of methylation at the TEAD4 promoter was significantly associated with poor outcomes, including large tumor size, high-grade tumors and low survival rates. Compared with normal tissues, the TEAD4 protein was more frequently found in the nuclei of tumor cells along with YAP1 in 53 GC patients, demonstrating the posttranslational activation of this protein. Moreover, the knockdown of TEAD4 resulted in the reduced growth of GC cells both in vitro and in vivo. Finally, chromatin immunoprecipitation-sequencing and microarray analysis revealed the oncogenic properties of TEAD4 and its novel targets (ADM, ANG, ARID5B, CALD1, EDN2, FSCN1 and OSR2), which are involved in cell proliferation and migration. In conclusion, the multilevel perturbations of TEAD4 at epigenetic, transcriptional and posttranslational levels may contribute to GC development.

SERPINE1 Intron Polymorphisms Affecting Gene Expression Are Associated With Diffuse-Type Gastric Cancer Susceptibility

A primary inhibitor of plasminogen activators, SERPINE1 (serpin peptidase inhibitor 1, clade E, member 1, also known as plasminogen activator inhibitor type 1), is an important regulator in tumorigenesis and is highly expressed in many cancers.

Increased genetic susceptibility to intestinal‐type gastric cancer is associated with increased activity of the RUNX3 distal promoter

The runt‐related transcription factor RUNX3 plays essential roles in various types of tumors, including gastric cancer. Epigenetic changes in the methylation of the RUNX3 proximal promoter, but not common genetic changes in RUNX3, have been associated with both changes in the gene expression and development of the cancer.

Decrease of 5hmC in gastric cancers is associated with TET1 silencing due to with DNA methylation and bivalent histone marks at TET1 CpG island 3'-shore.

Recent evidence has shown that the level of 5-hydroxymethylcytosine (5hmC) in chromosomal DNA is aberrantly decreased in a variety of cancers, but whether this decrease is a cause or a consequence of tumorigenesis is unclear. Here we show that, in gastric cancers, the 5hmC decrease correlates with a decrease in ten-eleven translocation 1 (TET1) expression, which is strongly associated with metastasis and poor survival in patients with gastric cancer. In gastric cancer cells, TET1-targeted siRNA induced a decrease in 5hmC, whereas TET1 overexpression induced an increase in 5hmC and reduced cell proliferation, thus correlating decreased 5hmC with gastric carcinogenesis. We also report the epigenetic signatures responsible for regulating TET1 transcription. Methyl-CpG Binding Domain Sequencing and Reduced Representation Bisulfite Sequencing identified unique CpG methylation signatures at the CpG island 3′-shore region located 1.3 kb from the transcription start site of TET1 in gastric tumor cells but not in normal mucosa. The luciferase activity of constructs with a methylated 3′-shore sequence was greatly decreased compared with that of an unmethylated sequence in transformed gastric cancer cells. In gastric cancer cells, dense CpG methylation in the 3′-shore was strongly associated with TET1 silencing and bivalent histone marks. Thus, a decrease in 5hmC may be a cause of gastric tumorigenesis owing to a decrease in TET1 expression through DNA methylation coupled with bivalent marks in the 3′-shore of TET1.

Genetic alterations and their clinical implications in gastric cancer peritoneal carcinomatosis revealed by whole-exome sequencing of malignant ascites

Peritoneal carcinomatosis accompanied by malignant ascites is a major cause of death of advanced gastric cancer (GC). To comprehensively characterize the underlying genomic events involved in GC peritoneal carcinomatosis, we analyzed whole-exome sequences of normal gastric tissues, primary tumors, and malignant ascites from eight GC patients. We identified a unique mutational signature biased toward C-to-A substitutions in malignant ascites. In contrast, the patients who received treatment of adjuvant chemotherapy showed a high rate of C-to-T substitutions along with hypermutation in malignant ascites. Comparative analysis revealed several candidate mutations for GC peritoneal carcinomatosis: recurrent mutations in COL4A6, INTS2, and PTPN13; mutations in druggable genes including TEP1, PRKCD, BRAF, ERBB4, PIK3CA, HDAC9, FYN, FASN, BIRC2, FLT3, ROCK1, CD22, and PIK3C2B; and mutations in metastasis-associated genes including TNFSF12, L1CAM, DIAPH3, ROCK1, TGFBR1, MYO9B, NR4A1, and RHOA. Notably, gene ontology analysis revealed the significant enrichment of mutations in the Rho-ROCK signaling pathway-associated biological processes in malignant ascites. At least four of the eight patients acquired somatic mutations in the Rho-ROCK pathway components, suggesting the possible relevance of this pathway to GC peritoneal carcinomatosis. These results provide a genome-wide molecular understanding of GC peritoneal carcinomatosis and its clinical implications, thereby facilitating the development of effective therapeutics.

Genomic and epigenomic heterogeneity in molecular subtypes of gastric cancer

Gastric cancer is a complex disease that is affected by multiple genetic and environmental factors. For the precise diagnosis and effective treatment of gastric cancer, the heterogeneity of the disease must be simplified; one way to achieve this is by dividing the disease into subgroups. Toward this effort, recent advances in high-throughput sequencing technology have revealed four molecular subtypes of gastric cancer, which are classified as Epstein-Barr virus-positive, microsatellite instability, genomically stable, and chromosomal instability subtypes. We anticipate that this molecular subtyping will help to extend our knowledge for basic research purposes and will be valuable for clinical use. Here, we review the genomic and epigenomic heterogeneity of the four molecular subtypes of gastric cancer. We also describe a mutational meta-analysis and a reanalysis of DNA methylation that were performed using previously reported gastric cancer datasets.

Maspin genetically and functionally associates with gastric cancer by regulating cell cycle progression

Human SERPINB5, commonly known as maspin, has diverse functions as a tumor suppressor. In this study, we discovered that maspin has a novel role in cell cycle control, and common variants were discovered to be associated with gastric cancer. The genotypes of 836 unrelated Korean participants (including 430 with gastric cancer) were examined for 12 tag single-nucleotide polymorphisms (SNPs) and imputed for 178 SNPs in the maspin gene. Susceptibility to diffuse-type gastric cancer was strongly and significantly associated with several SNPs including rs3744941 (C>T) in the promoter (TT versus CC+CT, odds ratio = 0.56 [0.37-0.83], P = 0.0038) and rs8089104 (C>T) in intron 1 (TT+CT versus CC, odds ratio = 1.7 [1.2-2.5], P = 0.0021). No SNPs were associated with susceptibility to intestinal-type gastric cancer. A haplotype of three highly correlated promoter SNPs associated with higher cancer risk showed 40% of the activity of a non-risk-associated haplotype promoter in the diffuse-type gastric cancer cell line MKN45. Maspin downregulation achieved either by a short hairpin RNA targeting maspin or overexpression of the E2F1-DP1 complex in MKN45 cells dramatically accelerated cell cycle progression and caused an increase of active CDC25C levels and a decrease of inactive CDK1 levels. In contrast, maspin upregulation had the opposite effect, substantially retarding cell proliferation. Therefore, our results suggest that a maspin promoter haplotype that reduces maspin gene expression accelerates cell cycle progression and, consequently, is associated with increased susceptibility to diffuse-type gastric cancer. Furthermore, a novel maspin-related pathway is demonstrated to underlie gastric carcinogenesis.

Intrinsic Molecular Processes: Impact on Mutagenesis

Mutations provide resources for genome evolution by generating genetic variability. In addition, mutations act as a driving force leading to disease pathogenesis, and thus have important implications for disease diagnosis, prognosis, and treatment. Understanding the mechanisms underlying how mutations occur is therefore of prime importance for elucidating evolutionary and pathogenic processes. Recent genomics studies have revealed that mutations occur non-randomly across the human genome. In particular, the distribution of mutations is highly associated with intrinsic molecular processes including transcription, chromatin organization, DNA replication timing, and DNA repair. Interplay between intrinsic processes and extrinsic mutagenic exposure may thus imprint a characteristic mutational landscape on tumors. We discuss the impact of intrinsic molecular processes on mutation acquisition in cancer.

Variability in chromatin architecture and associated DNA repair at genomic positions containing somatic mutations

Dynamic chromatin structures result in differential chemical reactivity to mutational processes throughout the genome. To identify chromatin features responsible for mutagenesis, we compared chromatin architecture around single-nucleotide variants (SNV), insertion/deletions (indels), and their context-matched, nonmutated positions. We found epigenetic differences between genomic regions containing missense SNVs and those containing frameshift indels across multiple cancer types. Levels of active histone marks were higher around frameshift indels than around missense SNV, whereas repressive histone marks exhibited the reverse trend. Accumulation of repressive histone marks and nucleosomes distinguished mutated positions (both SNV and indels) from the context-matched, nonmutated positions, whereas active marks were associated with substitution- and cancer type-specific mutagenesis. We also explained mutagenesis based on genome maintenance mechanisms, including nucleotide excision repair (NER), mismatch repair (MMR), and DNA polymerase epsilon (POLE). Regional NER variation correlated strongly with chromatin features; NER machineries exhibited shifted or depleted binding around SNV, resulting in decreased NER at mutation positions, especially at sites of recurrent mutations. MMR-deficient tumors selectively acquired SNV in regions with high active histone marks, especially H3K36me3, whereas POLE-deficient tumors selectively acquired indels and SNV in regions with low active histone marks. These findings demonstrate the importance of fine-scaled chromatin structures and associated DNA repair mechanisms in mutagenesis. Cancer Res; 77(11); 2822-33. ©2017 AACR.