Yuanhao Zhang

Parallel comparison of Illumina RNA-Seq and Affymetrix microarray platforms on transcriptomic profiles generated from 5-aza-deoxy-cytidine treated HT-29 colon cancer cells and simulated datasets

BackgroundHigh throughput parallel sequencing, RNA-Seq, has recently emerged as an appealing alternative to microarray in identifying differentially expressed genes (DEG) between biological groups. However, there still exists considerable discrepancy on gene expression measurements and DEG results between the two platforms. The objective of this study was to compare parallel paired-end RNA-Seq and microarray data generated on 5-azadeoxy-cytidine (5-Aza) treated HT-29 colon cancer cells with an additional simulation study.MethodsWe first performed general correlation analysis comparing gene expression profiles on both platforms. An Errors-In-Variables (EIV) regression model was subsequently applied to assess proportional and fixed biases between the two technologies. Then several existing algorithms, designed for DEG identification in RNA-Seq and microarray data, were applied to compare the cross-platform overlaps with respect to DEG lists, which were further validated using qRT-PCR assays on selected genes. Functional analyses were subsequently conducted using Ingenuity Pathway Analysis (IPA).ResultsPearson and Spearman correlation coefficients between the RNA-Seq and microarray data each exceeded 0.80, with 66%~68% overlap of genes on both platforms. The EIV regression model indicated the existence of both fixed and proportional biases between the two platforms. The DESeq and baySeq algorithms (RNA-Seq) and the SAM and eBayes algorithms (microarray) achieved the highest cross-platform overlap rate in DEG results from both experimental and simulated datasets. DESeq method exhibited a better control on the false discovery rate than baySeq on the simulated dataset although it performed slightly inferior to baySeq in the sensitivity test. RNA-Seq and qRT-PCR, but not microarray data, confirmed the expected reversal of SPARC gene suppression after treating HT-29 cells with 5-Aza. Thirty-three IPA canonical pathways were identified by both microarray and RNA-Seq data, 152 pathways by RNA-Seq data only, and none by microarray data only.ConclusionsThese results suggest that RNA-Seq has advantages over microarray in identification of DEGs with the most consistent results generated from DESeq and SAM methods. The EIV regression model reveals both fixed and proportional biases between RNA-Seq and microarray. This may explain in part the lower cross-platform overlap in DEG lists compared to those in detectable genes.

Comparison of Fecal Microbiota in Children with Autism Spectrum Disorders and Neurotypical Siblings in the Simons Simplex Collection

In order to assess potential associations between autism spectrum disorder (ASD) phenotype, functional GI disorders and fecal microbiota, we recruited simplex families, which had only a single ASD proband and neurotypical (NT) siblings, through the Simons Simplex Community at the Interactive Autism Network (SSC@IAN). Fecal samples and metadata related to functional GI disorders and diet were collected from ASD probands and NT siblings of ASD probands (age 7–14). Functional gastrointestinal disorders (FGID) were assessed using the parent-completed ROME III questionnaire for pediatric FGIDs, and problem behaviors were assessed using the Child Behavior Check List (CBCL). Targeted quantitative polymerase chain reaction (qPCR) assays were conducted on selected taxa implicated in ASD, including Sutterella spp., Bacteroidetes spp. and Prevotella spp. Illumina sequencing of the V1V2 and the V1V3 regions of the bacterial 16S rRNA genes from fecal DNA was performed to an average depth of 208,000 and 107,000 high-quality reads respectively. Twenty-five of 59 ASD children and 13 of 44 NT siblings met ROME III criteria for at least one FGID. Functional constipation was more prevalent in ASD (17 of 59) compared to NT siblings (6 of 44, P = 0.035). The mean CBCL scores in NT siblings with FGID, ASD children with FGID and ASD without FGID were comparably higher (58–62 vs. 44, P < 0.0001) when compared to NT children without FGID. There was no significant difference in macronutrient intake between ASD and NT siblings. There was no significant difference in ASD severity scores between ASD children with and without FGID. No significant difference in diversity or overall microbial composition was detected between ASD children with NT siblings. Exploratory analysis of the 16S rRNA sequencing data, however, identified several low abundance taxa binned at the genus level that were associated with ASD and/or first order ASD*FGID interactions (FDR <0.1).

Differential expression of miRNAs in colon cancer between African and Caucasian Americans: Implications for cancer racial health disparities

Colorectal cancer (CRC) incidence and mortality are higher in African Americans (AAs) than in Caucasian Americans (CAs) and microRNAs (miRNAs) have been found to be dysregulated in colonic and other neoplasias. The aim of this exploratory study was to identify candidate miRNAs that could contribute to potential biological differences between AA and CA colon cancers. Total RNA was isolated from tumor and paired adjacent normal colon tissue from 30 AA and 31 CA colon cancer patients archived at Stony Brook University (SBU) and Washington University (WU)-St. Louis Medical Center. miRNA profiles were determined by probing human genome-wide miRNA arrays with RNA isolated from each sample. Using repeated measures analysis of variance (RANOVA), miRNAs were selected that exhibited significant (p<0.05) interactions between race and tumor or significant (fold change >1.5, p<0.05) main effects of race and/or tumor. Quantitative polymerase chain reaction (q-PCR) was used to confirm miRNAs identified by microarray analysis. Candidate miRNA targets were analyzed using immunohistochemistry. RANOVA results indicated that miR-182, miR152, miR-204, miR-222 and miR-202 exhibited significant race and tumor main effects. Of these miRNAs, q-PCR analysis confirmed that miR-182 was upregulated in AA vs. CA tumors and exhibited significant race:tumor interaction. Immunohistochemical analysis revealed that the levels of FOXO1 and FOXO3A, two potential miR-182 targets, are reduced in AA tumors. miRNAs may play a role in the differences between AA and CA colon cancer. Specifically, differences in miRNA expression levels of miR-182 may contribute to decreased survival in AA colon cancer patients.

Altered Interactions between the Gut Microbiome and Colonic Mucosa Precede Polyposis in APCMin/+ Mice

Mutation of the adenomatous polyposis coli (APC gene), an early event in the adenoma-carcinoma sequence, is present in 70-80% of sporadic human colorectal adenomas and carcinomas. To test the hypothesis that mutation of the APC gene alters microbial interactions with host intestinal mucosa prior to the development of polyposis, culture-independent methods (targeted qPCR assays and Illumina sequencing of the 16S rRNA gene V1V2 hypervariable region) were used to compare the intestinal microbial composition of 30 six-week old C57BL/6 APCMin/+ and 30 congenic wild type (WT) mice. The results demonstrate that similar to 12-14 week old APCMin/+ mice with intestinal neoplasia, 6 week old APCMin/+ mice with no detectable neoplasia, exhibit an increased relative abundance of Bacteroidetes spp in the colon. Parallel mouse RNA sequence analysis, conducted on a subset of proximal colonic RNA samples (6 APCMin/+, 6 WT) revealed 130 differentially expressed genes (DEGs, fold change ≥ 2, FDR <0.05). Hierarchical clustering of the DEGs was carried out by using 1-r dissimilarity measurement, where r stands for the Pearson correlation, and Ward minimum variance linkage, in order to reduce the number of input variables. When the cluster centroids (medians) were included along with APC genotype as input variables in a negative binomial (NB) regression model, four of seven mouse gene clusters, in addition to APC genotype, were significantly associated with the increased relative abundance of Bacteroidetes spp. Three of the four clusters include several downregulated genes encoding immunoglobulin variable regions and non-protein coding RNAs. These results support the concept that mutation of the APC gene alters colonic-microbial interactions prior to polyposis. It remains to be determined whether interventions directed at ameliorating dysbiosis in APCMin/+mice, such as through probiotics, prebiotics or antibiotics, could reduce tumor formation.

Establishment of Highly Tumorigenic Human Colorectal Cancer Cell Line (CR4) with Properties of Putative Cancer Stem Cells

Background Colorectal cancer (CRC) has the third highest mortality rates among the US population. According to the most recent concept of carcinogenesis, human tumors are organized hierarchically, and the top of it is occupied by malignant stem cells (cancer stem cells, CSCs, or cancer-initiating cells, CICs), which possess unlimited self-renewal and tumor-initiating capacities and high resistance to conventional therapies. To reflect the complexity and diversity of human tumors and to provide clinically and physiologically relevant cancer models, large banks of characterized patient-derived low-passage cell lines, and especially CIC-enriched cell lines, are urgently needed. Principal Findings Here we report the establishment of a novel CIC-enriched, highly tumorigenic and clonogenic colon cancer cell line, CR4, derived from liver metastasis. This stable cell line was established by combining 3D culturing and 2D culturing in stem cell media, subcloning of cells with particular morphology, co-culture with carcinoma associated fibroblasts (CAFs) and serial transplantation to NOD/SCID mice. Using RNA-Seq complete transcriptome profiling of the tumorigenic fraction of the CR4 cells in comparison to the bulk tumor cells, we have identified about 360 differentially expressed transcripts, many of which represent stemness, pluripotency and resistance to treatment. Majority of the established CR4 cells express common markers of stemness, including CD133, CD44, CD166, EpCAM, CD24 and Lgr5. Using immunocytochemical, FACS and western blot analyses, we have shown that a significant ratio of the CR4 cells express key markers of pluripotency markers, including Sox-2, Oct3/4 and c-Myc. Constitutive overactivation of ABC transporters and NF-kB and absence of tumor suppressors p53 and p21 may partially explain exceptional drug resistance of the CR4 cells. Conclusions The highly tumorigenic and clonogenic CIC-enriched CR4 cell line may provide an important new tool to support the discovery of novel diagnostic and/or prognostic biomarkers as well as the development of more effective therapeutic strategies.

Aberrant DNA Methylation: Implications in Racial Health Disparity

Background Incidence and mortality rates of colorectal carcinoma (CRC) are higher in African Americans (AAs) than in Caucasian Americans (CAs). Deficient micronutrient intake due to dietary restrictions in racial/ethnic populations can alter genetic and molecular profiles leading to dysregulated methylation patterns and the inheritance of somatic to germline mutations. Materials and Methods Total DNA and RNA samples of paired tumor and adjacent normal colon tissues were prepared from AA and CA CRC specimens. Reduced Representation Bisulfite Sequencing (RRBS) and RNA sequencing were employed to evaluate total genome methylation of 5’-regulatory regions and dysregulation of gene expression, respectively. Robust analysis was conducted using a trimming-and-retrieving scheme for RRBS library mapping in conjunction with the BStool toolkit. Results DNA from the tumor of AA CRC patients, compared to adjacent normal tissues, contained 1,588 hypermethylated and 100 hypomethylated differentially methylated regions (DMRs). Whereas, 109 hypermethylated and 4 hypomethylated DMRs were observed in DNA from the tumor of CA CRC patients; representing a 14.6-fold and 25-fold change, respectively. Specifically; CHL1, 4 anti-inflammatory genes (i.e., NELL1, GDF1, ARHGEF4, and ITGA4), and 7 miRNAs (of which miR-9-3p and miR-124-3p have been implicated in CRC) were hypermethylated in DNA samples from AA patients with CRC. From the same sample set, RNAseq analysis revealed 108 downregulated genes (including 14 ribosomal proteins) and 34 upregulated genes (including POLR2B and CYP1B1 [targets of miR-124-3p]) in AA patients with CRC versus CA patients. Conclusion DNA methylation profile and/or products of its downstream targets could serve as biomarker(s) addressing racial health disparity.

Identification of Candidate Adherent-Invasive E. coli Signature Transcripts by Genomic/Transcriptomic Analysis

Adherent-invasive Escherichia coli (AIEC) strains are detected more frequently within mucosal lesions of patients with Crohn’s disease (CD). The AIEC phenotype consists of adherence and invasion of intestinal epithelial cells and survival within macrophages of these bacteria in vitro. Our aim was to identify candidate transcripts that distinguish AIEC from non-invasive E. coli (NIEC) strains and might be useful for rapid and accurate identification of AIEC by culture-independent technology. We performed comparative RNA-Sequence (RNASeq) analysis using AIEC strain LF82 and NIEC strain HS during exponential and stationary growth. Differential expression analysis of coding sequences (CDS) homologous to both strains demonstrated 224 and 241 genes with increased and decreased expression, respectively, in LF82 relative to HS. Transition metal transport and siderophore metabolism related pathway genes were up-regulated, while glycogen metabolic and oxidation-reduction related pathway genes were down-regulated, in LF82. Chemotaxis related transcripts were up-regulated in LF82 during the exponential phase, but flagellum-dependent motility pathway genes were down-regulated in LF82 during the stationary phase. CDS that mapped only to the LF82 genome accounted for 747 genes. We applied an in silico subtractive genomics approach to identify CDS specific to AIEC by incorporating the genomes of 10 other previously phenotyped NIEC. From this analysis, 166 CDS mapped to the LF82 genome and lacked homology to any of the 11 human NIEC strains. We compared these CDS across 13 AIEC, but none were homologous in each. Four LF82 gene loci belonging to clustered regularly interspaced short palindromic repeats region (CRISPR)—CRISPR-associated (Cas) genes were identified in 4 to 6 AIEC and absent from all non-pathogenic bacteria. As previously reported, AIEC strains were enriched for pdu operon genes. One CDS, encoding an excisionase, was shared by 9 AIEC strains. Reverse transcription quantitative polymerase chain reaction assays for 6 genes were conducted on fecal and ileal RNA samples from 22 inflammatory bowel disease (IBD), and 32 patients without IBD (non-IBD). The expression of Cas loci was detected in a higher proportion of CD than non-IBD fecal and ileal RNA samples (p <0.05). These results support a comparative genomic/transcriptomic approach towards identifying candidate AIEC signature transcripts.

Impaired granulocyte-macrophage colony-stimulating factor bioactivity accelerates surgical recurrence in ileal Crohn’s disease

AIM To examine the relationship between elevated granulocyte-macrophage colony-stimulating factor (GM-CSF) auto-antibodies (Ab) level and time to surgical recurrence after initial surgery for Crohn’s disease (CD). METHODS We reviewed 412 charts from a clinical database at tertiary academic hospital. Patients included in the study had ileal or ileocolonic CD and surgical resection of small bowel or ileocecal region for management of disease. Serum samples were analyzed for serological assays including GM-CSF cytokine, GM-CSF Ab, ASCA IgG and IgA, and genetic markers including SNPs rs2066843, rs2066844, rs2066845, rs2076756 and rs2066847 in NOD2, rs2241880 in ATG16L1, and rs13361189 in IRGM. Cox proportional-hazards models were used to assess the predictors of surgical recurrence. RESULTS Ninety six percent of patients underwent initial ileocecal resection (ICR) or ileal resection (IR) and subsequently 40% of patients required a second ICR/IR for CD. GM-CSF Ab level was elevated at a median of 3.81 mcg/mL. Factors predicting faster time to a second surgery included elevated GM-CSF Ab [hazard ratio (HR) 3.52, 95%CI: 1.45-8.53, P = 0.005] and elevated GM-CSF cytokine (HR = 2.48, 95%CI: 1.31-4.70, P = 0.005). Factors predicting longer duration between first and second surgery included use of Immunomodulators (HR = 0.49, 95%CI: 0.31-0.77, P = 0.002), the interaction effect of low GM-CSF Ab levels and smoking (HR = 0.60, 95%CI: 0.45-0.81, P = 0.001) and the interaction effect of low GM-CSF cytokine levels and ATG16L1 (HR = 0.65, 95%CI: 0.49-0.88, P = 0.006). CONCLUSION GM-CSF bioavailability plays a critical role in maintaining intestinal homeostasis. Decreased bioavailability coupled with the genetic risk markers and/or smoking results in aggressive CD behavior.

Correction: Aberrant DNA Methylation: Implications in Racial Health Disparity

[This corrects the article DOI: 10.1371/journal.pone.0153125.].

Tu1767 Detection of a LF82 Transcript in Fecal RNA Samples in Crohn's Disease Patients

G A A b st ra ct s to mesenteric lymph nodes and spleen. Increased TLR4 signaling resulted in significant down regulation of mucus protein Muc2 and epithelial barrier protein Zo-1. Although not significant, down regulation of polymeric immunoglobulin receptor (Pigr) and resistin-like molecule (Relmb) was observed in villin-TLR4 mice. Compared to WT mice, the villinTLR4 mice showed significant upregulation of antimicrobial peptides regenerating isletderived IIIγ (RegIIIg) possibly in response to increased mucosal-associated bacteria. Conclusion(s): Our results suggest that intestinal epithelial TLR4 signaling affects the abundance of the mucosa-associatedmicrobiota and regulates genes involved in host-microbial interactions. Dysregulation of TLR4 signaling may affect intestinal epithelial function and thus can influence the susceptibility to TLR4 dependent colitis-associated neoplasia.