Wangxiong Hu

Dynamics of chloroplast genomes in green plants

Chloroplasts are essential organelles, in which genes have widely been used in the phylogenetic analysis of green plants. Here, we took advantage of the breadth of plastid genomes (cpDNAs) sequenced species to investigate their dynamic changes. Our study showed that gene rearrangements occurred more frequently in the cpDNAs of green algae than in land plants. Phylogenetic trees were generated using 55 conserved protein-coding genes including 33 genes for photosynthesis, 16 ribosomal protein genes and 6 other genes, which supported the monophyletic evolution of vascular plants, land plants, seed plants, and angiosperms. Moreover, we could show that seed plants were more closely related to bryophytes rather than pteridophytes. Furthermore, the substitution rate for cpDNA genes was calculated to be 3.3×10(-10), which was almost 10 times lower than genes of nuclear genomes, probably because of the plastid homologous recombination machinery.

MicroRNA mediates DNA methylation of target genes

Small RNAs represented by microRNA (miRNA) plays important roles in plant development and responds to biotic and abiotic stresses. Previous studies have placed special emphasis on gene-repression mediated by miRNA. In this work, the DNA methylation pattern of microRNA genes (MIRs) was interrogated. Full-length cDNA and EST were used to confirm the entity of pri-miRNA. In parallel, miRNA in 24 nucleotides (nt) was pooled to detect chromatin modification effect by using bisulfite sequencing data. 97 MIRs were supported by full-length cDNA and 30 more were hit by EST. Notably, methylation levels of conserved MIRs were significantly lower than the non-conserved at all contexts (CG, CHG, and CHH). Additionally, a substantial part of 24-nt miRNA was able to induce target site methylation, providing a broader perspective for researchers.

Flexible microRNA arm selection in rice.

MicroRNAs act at the post-transcriptional level and guide Argonaute proteins to cleave their corresponding target transcripts. However, little attention has been paid to arm selection in miRNA precursors. In this study, small RNA high-throughput sequencing data from 29 different rice libraries were pooled to investigate tissue- and abiotic stress-specific dynamic expression of miRNAs. We found that more than half of pre-miRNAs showed changes in arm selection in different tissues and/or under different abiotic stresses. Our findings suggest that miRNA selection is remarkably prevalent in plants, providing new insights into the role of miRNAs in plant growth and development.

Tumor heterogeneity uncovered by dynamic expression of long noncoding RNA at single-cell resolution

The expression of long noncoding RNA (lncRNA) is thought to be more cell-type specific than the expression of protein-coding genes. However, the expression profile of individual cells regarding lncRNA remains to be elucidated. Here, we comprehensively investigated the pattern of lncRNA expression across five glioblastoma tumors (414 cells) and two cell lines (GBM6 and GBM8, 127 cells). We found that there were more than 1,000 lncRNAs that varied between any two cells and that there was frequent gain and loss of lncRNA expression during tumor cell proliferation, suggesting a great heterogeneity in lncRNA expression across different single cells in glioblastoma.

Association mining of mutated cancer genes in different clinical stages across 11 cancer types

Many studies have demonstrated that some genes (e.g. APC, BRAF, KRAS, PTEN, TP53) are frequently mutated in cancer, however, underlying mechanism that contributes to their high mutation frequency remains unclear. Here we used Apriori algorithm to find the frequent mutational gene sets (FMGSs) from 4,904 tumors across 11 cancer types as part of the TCGA Pan-Cancer effort and then mined the hidden association rules (ARs) within these FMGSs. Intriguingly, we found that well-known cancer driver genes such as BRAF, KRAS, PTEN, and TP53 were often co-occurred with other driver genes and FMGSs size peaked at an itemset size of 3∼4 genes. Besides, the number and constitution of FMGS and ARs differed greatly among different cancers and stages. In addition, FMGS and ARs were rare in endocrine-related cancers such as breast carcinoma, ovarian cystadenocarcinoma, and thyroid carcinoma, but abundant in cancers contact directly with external environments such as skin melanoma and stomach adenocarcinoma. Furthermore, we observed more rules in stage IV than in other stages, indicating that distant metastasis needed more sophisticated gene regulatory network.

Pan-organ transcriptome variation across 21 cancer types

It is widely accepted that some messenger RNAs are evolutionarily conserved across species, both in sequence and tissue-expression specificity. To date, however, little effort has been made to exploit the transcriptome divergence between cancer and adjacent normal tissue at the pan-organ level. In this work, a transcriptome sequencing dataset from 675 normal-tumor pairs, representing 21 solid organs in The Cancer Genome Atlas, is used to evaluate expression evolution. The results show that in most cancer types, gene expression divergence and organ-specificity are reduced in cancer tissue compared to adjacent normal tissue. Furthermore, we observe that all cancers share cell cycle dysregulation through interrogating differentially expressed protein coding genes. Meanwhile, weighted correlation network analysis is used to detect of the gene module structure variation between cancer and adjacent normal tissue. And modules consisting of tightly co-regulated genes in cancer change substantially compared with those in adjacent normal tissue. We thus assume that the destruction of a coordinated regulatory network might result in tumorigenesis and tumor progression. Our results provide new insights into the complex cancer biology and shed light on the mysterious regulation mode for cancer.

Risk of eighteen genome-wide association study-identified genetic variants for colorectal cancer and colorectal adenoma in Han Chinese

Background Recent genome-wide association studies (GWAS) identified eighteen single-nucleotide polymorphisms (SNPs) to be significantly associated with the risk of colorectal cancer (CRC). However, overall results of the following replications are inconsistent and little is known about whether these associations also exit in colorectal adenomas (CRA). Methods The SNP genotyping was performed using a Sequenom MassARRAY to investigate the association of these eighteen SNPs with colorectal neoplasm in a case-control study consisted of 1049 colorectal cancers, 283 adenomas, and 1030 controls. Results Two of these SNPs, rs10505477 and rs719725, showed evidence of an association in both CRC and CRA in our study population. Besides, seven SNPs (rs10808555, rs7014346, rs7837328, rs704017, rs11196172, rs4779584, and rs7229639) were significantly associated with CRC, and another one SNP rs11903757 was over-represented in CRA compared with controls. The strongest association was provided by rs11196172 (OR = 2.02, 95% CI = 1.66 - 2.46, P < 0.0001) and rs11903757 (OR = 1.96, 95% CI = 1.28 - 3.00, P = 0.0026). Conclusion These results suggest that some previously reported SNP associations also have impact on CRC and CRA predispositions in the Han Chinese population. A part of genetic risk to CRC is possibly mediated by susceptibility to adenomas.

Multi-omics Approach Reveals Distinct Differences in Left- and Right-Sided Colon Cancer

Increasing evidence suggests that left-sided colon cancer (LCC) and right-sided colon cancer (RCC) are emerging as two different colorectal cancer types with distinct clinical characteristics. However, the discrepancy in the underlying molecular event between these types of cancer has not been thoroughly elucidated to date and warrants comprehensive investigation. To this end, an integrated dataset from The Cancer Genome Atlas was used to compare and contrast LCC and RCC, covering mutation, DNA methylation, gene expression, and miRNA. Briefly, the signaling pathway cross-talk is more prevalent in RCC than LCC, such as RCC-specific PI3K pathway, which often exhibits cross-talk with the RAS and P53 pathways. Meanwhile, methylation signatures revealed that RCC was hypermethylated relative to LCC. In addition, differentially expressed genes (n = 253) and differentially expressed miRNAs (n = 16) were determined between LCC and RCC. Especially for Prostate Cancer Susceptibility Candidate 1 (PRAC1), a gene that was closely associated with hypermethylation, was the top significantly downregulated gene in RCC. Multi-omics comparison of LCC and RCC suggests that there are more aggressive markers in RCC and that tumor heterogeneity occurs within the location-based subtypes of colon cancer. These results clarify the debate regarding the conflicting prognosis between LCC and RCC, as proposed by different studies. Implications: The underlying molecular features present in LCC and RCC identified in this study are beneficial for adopting reasonable therapeutic approaches to prolong overall survival and progression-free survival in colorectal cancer patients. Mol Cancer Res; 16(3); 476–85. ©2017 AACR.