Host Transcriptome Analysis of Publicly Available Datasets reveal Major Contribution of Myeloid cells in TB specific Host Gene Signature

Anuradha Gautam ¹,  Saroj K. Mohapatra, MD ¹ , Bhaswati Pandit, PhD ¹

¹National Institute of Biomedical Genomics, Kalyani, West Bengal India

HYPOTHESIS: Development of tuberculosis is driven by the interaction of human host and M.tuberculosis. This interaction is known to dysregulate the expression of host genes and the study of these genes from whole blood can help identify active cellular pathways and compartments contributing to the dysregulation. 

METHODS: 

1. Selection of datasets: 10 Whole blood host transcriptomic datasets comprising of 594 subjects: 287 TB Cases without HIV and 307 healthy controls were identified from GEO, ArrayExpress and Pubmed.

2. Identification of Differentially Expressed Genes (DEGs): Quality control was performed for the normalized datasets.‘t’ test was performed on common genes to identify the DEGs. Up and downregulated genes have LFC greater than 0.26 or less than -0.26 at FDR corrected p value of 0.05.

3. Pathway analysis: Quality control was performed for the normalized datasets. Two-tailed, two-sample ‘t’ test was performed on common genes to identify the DEGs. Upregulated and down-regulated genes had a 20% difference in fold change of expression at FDR corrected p-value of 0.05.

4. Identification of miRNA targets from the DEGs: Identification of miRNAs interacting with the list of DEGs was found using a web-based tool Enrichr in the miRTarBase 2017 library.

5. Identification of active cellular compartments: The DEGs were identified with having a 2 fold change of expression. For these genes, ENCODE ChIP-seq experiment data was visualized using UCSC genome browser (hg19) to identify the presence of epigenetic activating marks (H3K4me3 and H3K27Ac) around regulatory elements characterized by Genehancer and EPD promoter database in haematopoietic cells

RESULTS: Pathway analysis was done for 239 upregulated genes and 83 downregulated genes with 20% difference in fold change of expression. respectively and an FDR corrected p value of 0.05. These DEGs showed enrichment of innate immune response, interferon response and response to virus pathways in GO, KEGG and WikiPathways. Non-genomic actions of 1,25 vitamin D3 and pathway for the human immune response to TB was enriched in WikiPathways. Among the down-regulated genes, adaptive immune pathways involving T cell receptor signalling and leukocyte differentiation were enriched for both GO and KEGG pathways. hsa-miR-146a-5p was found to have targets within the upregulated genes: STAT1, IFITM3, IFITM1, SAMDL9, S100A12, RSAD2, IFI44, EPSTI1 and TRIM22. Further increasing the LFC cutoff, 19 upregulated genes (LFC>1) and two downregulated genes (LFC<-1) were identified. Upregulated genes DHRS9, ANKRD22, TNFAIP6, CARD17 and KCNJ15 had activating epigenetic marks around their promoters exclusively in the myeloid lineage cells.13 genes had at least one of the two activation marks (H3K4me3 and H3K27Ac) in neutrophils, and 14 genes had either one or both activation marks in CD14⁺ monocytes accounting for two of the most numbers of genes with activation marks in any cell subset.

CONCLUSION: Greater number of innate immune pathways are enriched within the upregulated genes, while adaptive immune pathways are enriched within the downregulated genes.

miR146a-5p which targets STAT1 is known to be involved in chemoattraction of neutrophils by bronchial epithelial cells and upregulation of IL6 production by macrophages and monocyte migration. CD14+ Monocytes and Neutrophils could be recognized as the cellular source for this dysregulation evident from the presence of H3K4me3 and H3K27Ac peaks in the promoters of most top DEGs.

 Acknowledgements: The completion of this work has been supported by the National Institute of Biomedical Genomics and UGC, GOI.