Cuiyan Wu

Tissue-specific pathway association analysis using genome-wide association study summaries

Motivation: Pathway association analysis has made great achievements in elucidating the genetic basis of human complex diseases. However, current pathway association analysis approaches fail to consider tissue-specificity. Results: We developed a tissue-specific pathway interaction enrichment analysis algorithm (TPIEA). TPIEA was applied to two large Caucasian and Chinese genome-wide association study summary datasets of bone mineral density (BMD). TPIEA identified several significant pathways for BMD [false discovery rate (FDR) < 0.05], such as KEGG FOCAL ADHESION and KEGG AXON GUIDANCE, which had been demonstrated to be involved in the development of osteoporosis. We also compared the performance of TPIEA and classical pathway enrichment analysis, and TPIEA presented improved performance in recognizing disease relevant pathways. TPIEA may help to fill the gap of classic pathway association analysis approaches by considering tissue specificity. Availability and Implementation: The online web tool of TPIEA is available at https://sourceforge.net/projects/tpieav1/files. Contact: fzhxjtu@mail.xjtu.edu.cn Supplementary information: Supplementary data are available at Bioinformatics online.

Genome-wide DNA methylation profiling of articular cartilage reveals significant epigenetic alterations in Kashin-Beck disease and osteoarthritis

OBJECTIVE To determine genome-wide DNA methylation profiles of knee cartilage from patients with Kashin-Beck disease (KBD) and osteoarthritis (OA). METHOD Knee cartilage was collected from 14 grade III KBD patients, 5 primary OA patients and 13 healthy subjects. The genome-wide methylation profiles of 5 KBD cartilage, 5 OA cartilage and 5 normal cartilage were determined by Illumina HumanMethylation450 array. Illumina Methylation Analyzer package was employed for identifying differentially methylated CpG sites. Functional annotation and enrichment analysis of differentially methylated genes (DMG) were conducted using GeneRIF database, Ingenuity Pathway Analysis (IPA) and The Database for Annotation, Visualization and Integrated Discovery (DAVID). Mass spectrometry (MS) and immunohistochemistry (IHC) were conducted to validate the functional relevance of identified KBD associated gene. RESULTS We identified a total of 1212 differentially methylated CpG sites in KBD vs Normal, annotated to 264 hypermethylated and 368 hypomethylated genes. Comparing the DNA methylation profiles of KBD vs Normal and OA vs Normal detected overlap of 367 differentially methylated CpG sites (annotated to 182 genes) as well as 845 KBD-specific differentially methylated CpG sites (annotated to 471 unique genes). MS and IHC confirmed the hypermethylation status and decreased protein expression of HAPLN1 gene in KBD cartilage. CONCLUSION Our data implicate epigenetic dysregulation of a host of genes in KBD and OA. Furthermore, we observed common causal epigenetic changes shared by KBD and OA.

Gene expression profiles and molecular mechanism of cultured human chondrocytes' exposure to T-2 toxin and deoxynivalenol

ABSTRACT T‐2 toxin and deoxynivalenol (DON) are secondary metabolites produced by Fusarium fungi and are commonly found on food and feed. Although T‐2 toxin and DON have been suggested as the etiology of Kashin‐Beck disease (KBD), an endemic osteochondropathy, little is known about the mechanism when human chondrocytes are exposed to T‐2 toxin and DON. The purpose of this study is to identify the gene expression differences and underlying molecular changes modulated by T‐2 toxin and DON in vitro in human chondrocytes. After the experiments of cell viability, the gene expression profiles were analyzed in cells that were treated with 0.01 &mgr;g/ml T‐2 toxin and 1.0 &mgr;g/ml DON for 72 h by Affymetrix Human Gene Chip. The array results showed that 882 and 2118 genes were differentially expressed for T‐2 toxin and DON exposure, respectively. Enrichment analysis revealed that diverse cellular processes including DNA damage, cell cycle regulation and metabolism of extracellular matrix were affected when human chondrocytes were exposed to T‐2 toxin and DON. These results demonstrate the gene expression differences and molecular mechanism of cultured human chondrocytes exposure to T‐2 toxin and DON, and provide a new insight into future research in the etiology of KBD. HighlightsMicroarray analysis of gene expression profiles of human chondrocytes exposed to T‐2 toxin and DON was performed.Bioinformatic analysis showed the potential molecular mechanism.DNA damage, cell cycle regulation and metabolism of extracellular matrix were mainly affected.

The potential of induced pluripotent stem cells as a tool to study skeletal dysplasias and cartilage-related pathologic conditions

The development of induced pluripotent stem cells (iPSCs) technology has opened up new horizons for development of new research tools especially for skeletal dysplasias, which often lack human disease models. Regenerative medicine and tissue engineering could be the next areas to benefit from refinement of iPSC methods to repair focal cartilage defects, while applications for osteoarthritis (OA) and drug screening have evolved rather slowly. Although the advances in iPSC research of skeletal dysplasias and repair of focal cartilage lesions are not directly relevant to OA, they can be considered to pave the way to future prospects and solutions to OA research, too. The same problems which face the present cell-based treatments of cartilage injuries concern also the iPSC-based ones. However, established iPSC lines, which have no genomic aberrations and which efficiently differentiate into extracellular matrix secreting chondrocytes, could be an invaluable cell source for cell transplantations in the future. The safety issues concerning the recipient risks of teratoma formation and immune response still have to be solved before the potential use of iPSCs in cartilage repair of focal cartilage defects and OA.

The roles of selenium, insulin-like growth factor binding protein 2 and suppressor of cytokine signaling 3 in the pathogenesis of Kashin–Beck disease

Abstract We aimed to verify the levels of IGFBP2 and SOCS3 in cartilage and chondrocytes of Kashin–Beck disease (KBD) patients and the effects of different selenium concentrations on the protein expression levels. Chondrocytes were cultured with sodium selenite in vitro. Immunohistochemistry and western blotting were used to verify the protein expressions. IGFBP2 and SOCS3 were up-regulated in KBD chondrocytes and decreased with increasing selenium concentrations. IGFBP2 expressed highest in the middle zone of KBD cartilage, SOCS3 expressed higher in the middle and deep zone. IGFBP2 and SOCS3 may be the biomarkers for KBD diagnosis and evaluating the effect of selenium supplement.

Integrating genome-wide association study and expression quantitative trait loci data identifies multiple genes and gene set associated with neuroticism

ABSTRACT Neuroticism is a fundamental personality trait with significant genetic determinant. To identify novel susceptibility genes for neuroticism, we conducted an integrative analysis of genomic and transcriptomic data of genome wide association study (GWAS) and expression quantitative trait locus (eQTL) study. GWAS summary data was driven from published studies of neuroticism, totally involving 170,906 subjects. eQTL dataset containing 927,753 eQTLs were obtained from an eQTL meta‐analysis of 5311 samples. Integrative analysis of GWAS and eQTL data was conducted by summary data–based Mendelian randomization (SMR) analysis software. To identify neuroticism associated gene sets, the SMR analysis results were further subjected to gene set enrichment analysis (GSEA). The gene set annotation dataset (containing 13,311 annotated gene sets) of GSEA Molecular Signatures Database was used. SMR single gene analysis identified 6 significant genes for neuroticism, including MSRA (p value = 2.27 × 10− 10), MGC57346 (p value = 6.92 × 10− 7), BLK (p value = 1.01 × 10− 6), XKR6 (p value = 1.11 × 10− 6), C17ORF69 (p value = 1.12 × 10− 6) and KIAA1267 (p value = 4.00 × 10− 6). Gene set enrichment analysis observed significant association for Chr8p23 gene set (false discovery rate = 0.033). Our results provide novel clues for the genetic mechanism studies of neuroticism. HIGHLIGHTSWe conducted SMR software which is an integrative analysis of GWAS and eQTL data.GSEA Molecular Signatures Database was used.SMR single gene analysis identified 6 significant genes for neuroticism.Gene set enrichment analysis observed significant association for Chr8p23 gene set.The findings provide novel clues for the genetic mechanism studies of neuroticism.

Gene Expression Signature in Endemic Osteoarthritis by Microarray Analysis

Kashin-Beck Disease (KBD) is an endemic osteochondropathy with an unknown pathogenesis. Diagnosis of KBD is effective only in advanced cases, which eliminates the possibility of early treatment and leads to an inevitable exacerbation of symptoms. Therefore, we aim to identify an accurate blood-based gene signature for the detection of KBD. Previously published gene expression profile data on cartilage and peripheral blood mononuclear cells (PBMCs) from adults with KBD were compared to select potential target genes. Microarray analysis was conducted to evaluate the expression of the target genes in a cohort of 100 KBD patients and 100 healthy controls. A gene expression signature was identified using a training set, which was subsequently validated using an independent test set with a minimum redundancy maximum relevance (mRMR) algorithm and support vector machine (SVM) algorithm. Fifty unique genes were differentially expressed between KBD patients and healthy controls. A 20-gene signature was identified that distinguished between KBD patients and controls with 90% accuracy, 85% sensitivity, and 95% specificity. This study identified a 20-gene signature that accurately distinguishes between patients with KBD and controls using peripheral blood samples. These results promote the further development of blood-based genetic biomarkers for detection of KBD.

eQTLs Weighted Genetic Correlation Analysis Detected Brain Region Differences in Genetic Correlations for Complex Psychiatric Disorders

Background Psychiatric disorders are usually caused by the dysfunction of various brain regions. Incorporating the genetic information of brain regions into correlation analysis can provide novel clues for pathogenetic and therapeutic studies of psychiatric disorders. Methods The latest genome-wide association study (GWAS) summary data of schizophrenia (SCZ), bipolar disorder (BIP), autism spectrum disorder (AUT), major depression disorder (MDD), and attention-deficit/hyperactivity disorder (ADHD) were obtained from the Psychiatric GWAS Consortium (PGC). The expression quantitative trait loci (eQTLs) datasets of 10 brain regions were driven from the genotype-tissue expression (GTEx) database. The PGC GWAS summaries were first weighted by the GTEx eQTLs summaries for each brain region. Linkage disequilibrium score regression was applied to the weighted GWAS summary data to detect genetic correlation for each pair of 5 disorders. Results Without considering brain region difference, significant genetic correlations were observed for BIP vs SCZ (P = 1.68 × 10-63), MDD vs SCZ (P = 5.08 × 10-45), ADHD vs MDD (P = 1.93 × 10-44), BIP vs MDD (P = 6.39 × 10-9), AUT vs SCZ (P = .0002), and ADHD vs SCZ (P = .0002). Utilizing brain region related eQTLs weighted LD score regression, different strengths of genetic correlations were observed within various brain regions for BIP vs SCZ, MDD vs SCZ, ADHD vs MDD, and SCZ vs ADHD. For example, the most significant genetic correlations were observed at anterior cingulate cortex (P = 1.13 × 10-34) for BIP vs SCZ. Conclusions This study provides new clues for elucidating the mechanism of genetic correlations among various psychiatric disorders.

A large-scale integrative analysis of GWAS and common meQTLs across whole life course identifies genes, pathways and tissue/cell types for three major psychiatric disorders

&NA; Attention deficit hyperactivity disorder (ADHD), bipolar disorder (BP) and schizophrenia (SCZ) are complex psychiatric disorders. We conducted a large‐scale integrative analysis of genome‐wide association studies (GWAS) and life course consistent methylation quantitative trait loci (meQTLs) datasets. The GWAS data of ADHD (including 20,183 cases and 35,191 controls), BP (including 7481 cases and 9250 controls) and SCZ (including 36,989 cases and 113,075 controls) were derived from published GWAS. Life course consistent meQTLs dataset was obtained from a longitudinal meQTLs analysis of 1018 mother–child pairs. Gene prioritization, pathway and tissue/cell type enrichment analysis were conducted by DEPICT. We identified multiple genes and pathways with common or disease specific effects, such as NISCH (P = 9.87 × 10−3 for BP and 2.49 × 10−6 for SCZ), ST3GAL3 (P = 1.19 × 10−2 for ADHD), and KEGG_MAPK_SIGNALING_PATHWAY (P = 1.56 × 10−3 for ADHD, P = 4.71 × 10−2 for BP, P = 4.60 × 10−4 for SCZ). Our study provides novel clues for understanding the genetic mechanism of ADHD, BP and SCZ.

GWAS summary-based pathway analysis correcting for the genetic confounding impact of environmental exposures

Genome-wide association study (GWAS)-based pathway association analysis is a powerful approach for the genetic studies of human complex diseases. However, the genetic confounding effects of environment exposure-related genes can decrease the accuracy of GWAS-based pathway association analysis of target diseases. In this study, we developed a pathway association analysis approach, named Mendelian randomization-based pathway enrichment analysis (MRPEA), which was capable of correcting the genetic confounding effects of environmental exposures, using the GWAS summary data of environmental exposures. After analyzing the real GWAS summary data of cardiovascular disease and cigarette smoking, we observed significantly improved performance of MRPEA compared with traditional pathway association analysis (TPAA) without adjusting for environmental exposures. Further, simulation studies found that MRPEA generally outperformed TPAA under various scenarios. We hope that MRPEA could help to fill the gap of TPAA and identify novel causal pathways for complex diseases.