Dataset: The Time-Series Transcriptomic Responses of THP-1 Cells to W-Beijing M. tuberculosis Strains of different sublineages

We report a whole-genome expression profiling of PMA-treated THP-1 human macrophages as a host system, infected by W-Beijing strains of different sublineages or the H37Rv reference strain. Surprisingly, we found that host transcriptional responses were irrespective of inter-sublineage variations. Based on this finding, we further showed that such common host transcriptional responses were probably under the control of a regulatory network involving STATs, IRF-1, IRF-7 and Oct-1 transcriptional factors. We also suggested that these putative regulators might act in a cooperative manner to induce the expression of host-responsive genes, particularly those involved in cytokine-cytokine receptor interactions. THP-1 cells were infected by 11 different sublineages plus one reference strain H37Rv (two replications) at three time points (i.e., 4h, 18h and 48h). For each infection as well as a non-infection control, RNA was extracted using Trizol (Life Technologies)TRIZOL® reagent (Invitrogen), and subsequently amplified and labeled with biotin according to the standard Affymetrix® protocol. Specifically, RNA was DNase-treated and purified with RNeasy columns (Qiagen). The purified RNA was quantified using the ND-1000 Spectrophotometer (NanoDrop Technologies) and Agilent 2100 Bioanalyzer (Agilent Technologies). The biotinylated cRNA was then fragmented and subjected to hybridization on Human Genome U133 Plus 2.0 Array (Affymetrix, USA). The scanned images were converted to cell intensity files (CEL) using GeneChip® Operating Software (GCOS) (Affymetrix). These CEL raw expression data were normalized using Robust Multi-array Averaging (RMA) with quantile normalization in R (Bioconductor). Detection call-based filter was applied to remove all the probe-sets whose expression values were consistently below an empirically-determined value of minimum sensitivity, which were evaluated according to the 95th percentile of all the ‘Absent’ call-flagged signals of the entire dataset. Following the normalization and filtering, a gene expression matrix (termed as TB expression matrix) was constructed. The TB expression matrix was then subjected to component plane presentation integrated self-organizing map (CPP-SOM) for visual comparisons, or subjected to Cluster3.0/TreeView-1.0.8 softwares with Euclidean distance algorithm for sample classification of 13 time-course transcriptome profiles. Also, Linear Models for Microarray Data (LIMMA) bioconductor library[18] was applied to identify those differentially expressed genes between any two successive time points. The criteria for identifying the top significant genes for the designed contrast was based on P-values (< 0.01) corrected using Benjamini and Hochberg false discovery rate (FDR) procedure. Sets of genes of interest such as differentially expressed genes between 4h and 18h infections were uploaded onto Database for Annotation, Visualization and Integrated Discovery (DAVID) v6.7 for identifying enriched gene ontology and KEGG pathways (FDR<0.01). For TFBS enrichment analysis, we applied PRIMA (PRomoter Integration in Microarray Analysis) program to identify TFs whose binding sites are significantly over-represented in a given gene set (i.e., differentially expressed genes) compared to background. Bonferroni-corrected P-values (< 0.01) were utilized to determine the significance of the binding site enrichments of TFs. Such computationally predicted TFs were considered as candidate regulators responsible for the putative co-regulations of differentially expressed genes.

Species:

human

Samples:

40

Source:

E-GEOD-29628

Updated:

Dec.12, 2014

Registered:

Sep.16, 2014