Transcriptional Profiling of Early and Late Phases of Bovine Tuberculosis

Abstract

Bovine tuberculosis, caused by Mycobacterium tuberculosis var. bovis (M. bovis), is an important enzootic disease affecting mainly cattle, worldwide. Despite the implementation of national campaigns to eliminate the disease, bovine tuberculosis remains recalcitrant to eradication in several countries. Characterizing the host response to M. bovis infection is crucial for understanding the immunopathogenesis of the disease and for developing better control strategies. To profile the host responses to M. bovis infection, we analyzed the transcriptome of whole blood cells collected from experimentally infected calves with a virulent strain of M. bovis using RNA transcriptome sequencing (RNAseq). Comparative analysis of calf transcriptomes at early (8 weeks) versus late (20 weeks) aerosol infection with M. bovis revealed a divergent and unique profile for each stage of infection. Notably, at the early time point, transcriptional upregulation was observed among several of the top-ranking canonical pathways involved in T-cell chemotaxis. At the late time point, enrichment in the cell mediated cytotoxicity (e.g., Granzyme B) was the predominant host response. These results showed significant change in bovine transcriptional profiles and identified networks of chemokine receptors and monocyte chemoattractant protein (CCL) coregulated genes that underline the host-mycobacterial interactions during progression of bovine tuberculosis in cattle. Further analysis of the transcriptomic profiles identified potential biomarker targets for early and late phases of tuberculosis in cattle. Overall, the identified profiles better characterized identified novel immunomodulatory mechanisms and provided a list of targets for further development of potential diagnostics for tuberculosis in cattle.

Keywords: Mycobacterium tuberculosis var. bovis; RNAseq; bovine tuberculosis; host response; host-pathogen interactions; pathogenesis; transcriptom; transcriptome analysis.

FIG 1

Comparative transcriptomic analysis of infected and uninfected negative control groups. Venn diagram shows numbers of common differentially expressed (DE) genes for each comparison group when the transcriptomes of infected and uninfected control animals were compared. The upper panel represents upregulated DE genes between early infection, late infection, and age independent infection (A), early infection, late infection, and age-related changes (B). The lower panel represents common downregulated DE genes early infection, late infection, and age independent infection (C), early infection, late infection and age-related changes (D). Total analyzed genes: 24,597.

FIG 2

Differential expression analysis of infected and uninfected control animals. MA-plots represents the average normalized means versus the average log fold change between uninfected negative control as age-related changes (A) and infected and negative control age-independent infection (expression matrix) (B), at 8 wpi (C) and at 20 wpi (D). Red dots represent differentially expressed transcripts (fold change > 1.0 or <–1.0, P < 0.05).

FIG 3

Gene ontology analysis of infected and uninfected control animals. Significant terms in gene ontology analysis for the differentially expressed genes in infected animals compared to uninfected control animals. The PANTHER14.1 for automated identification of GO terms was used on the list of genes with significant differential expression when the transcriptomes of infected and uninfected control animals were compared. Pie chart shows significant GO in both comparisons: uninfected to uninfected groups at 8 wpi and 20 wpi (A). Infected to uninfected groups at 8 wpi and 20 wpi using an expression matrix (B). GO enrichment analysis of biological processes for DE genes in the age-related changes (C), age-independent infection (D), early infection (E), and late infection (F). To identify similar GO terms among the enriched terms, this set of GO terms was categorized using semantic clustering (REVIGO). The size and color of the dots represent the gene number and the range of P values, respectively. The cutoff P value for the GO enrichment analysis was set to 0.01. The background used for the GO enrichment analysis was all the annotated uni-genes of the assembly.

FIG 4

Gene network analysis of DE genes at different time points. Gene network analysis of DE genes in the uninfected groups showing age related changes (A), general M. bovis infection related changes (B), and the changes of infected group at 8 wpi (C) and 20 wpi (D), showing genes that were significantly upregulated. Light green lines represent connections between genes comentioned in an abstract in published studies, cyan lines represent putative pathway connections found in homologs in other species, black lines represent coexpression in Bos taurus or homologs in other species, and pink lines represent experimentally determined association.

FIG 5

Transcriptional profile of key genes after challenge with M. bovis. quantitative real-time PCR analysis of total RNA extracted from whole blood samples collected from calves’ groups at 8 wpi and 20 wpi in addition to the group selected for validating the performance of potential biomarkers for bTB. Expression levels were calculated with ΔΔCt relative quantitation method relative to the β-Actin gene expression in the naive uninfected animals. Target gene names are listed below each panel, and fold change for the infected animals relative to naive uninfected groups are listed on the y axis. At each time point, samples from three animals in each group were included and standard errors of the mean (SEM) of the three measurements were presented as error bars.

FIG 6

Validation of the performance of key genes as biomarker for bTB diagnosis. Statistical analysis of the gene expression of total RNA extracted from whole blood samples collected from calves’ groups at 8 wpi and 20 wpi compared to the additional group selected for validating the performance of potential biomarkers for bTB denoted as initial and validation on the x axis, respectively. Target gene names are listed above each panel, and mean log₂ relative normalized expression of infected animals relative to naive uninfected groups are listed on the y axis. Significance asterisks are shown above the bars in each panel and were calculated based on Z-score of first infection using the mean and standard deviation of second infection (see Materials and Methods). NS, not significant; *, P < 0.05; **, P <0.01; ***, P < 0.001.