Defining the stressome of Mycobacterium avium subsp. paratuberculosis in vitro and in naturally infected cows

Abstract

Mycobacterium avium subsp. paratuberculosis causes an enteric infection in cattle, with a great impact on the dairy industry in the United States and worldwide. Characterizing the gene expression profile of M. avium subsp. paratuberculosis exposed to different stress conditions, or shed in cow feces, could improve our understanding of the pathogenesis of M. avium subsp. paratuberculosis. In this report, the stress response of M. avium subsp. paratuberculosis on a genome-wide level (stressome) was defined for the first time using DNA microarrays. Expression data analysis revealed unique gene groups of M. avium subsp. paratuberculosis that were regulated under in vitro stressors while additional groups were regulated in the cow samples. Interestingly, acidic pH induced the regulation of a large number of genes (n=597), suggesting the high sensitivity of M. avium subsp. paratuberculosis to acidic environments. Generally, responses to heat shock, acidity, and oxidative stress were similar in M. avium subsp. paratuberculosis and Mycobacterium tuberculosis, suggesting common pathways for mycobacterial defense against stressors. Several sigma factors (e.g., sigH and sigE) were differentially coregulated with a large number of genes depending on the type of each stressor. Subsequently, we analyzed the virulence of six M. avium subsp. paratuberculosis mutants with inactivation of differentially regulated genes using a murine model of paratuberculosis. Both bacterial and histopathological examinations indicated the attenuation of all gene mutants, especially those selected based on their expression in the cow samples (e.g., lipN). Overall, the employed approach profiled mycobacterial genetic networks triggered by variable stressors and identified a novel set of putative virulence genes. A similar approach could be applied to analyze other intracellular pathogens.

FIG. 1.

Microarray analysis of M. avium subsp. paratuberculosis cultures exposed to variable stressors. (A) A scatter plot analysis displaying a high reproducibility between biological replicates (r > 0.9) of M. avium subsp. paratuberculosis cultures following exposure to oxidative stressors. (B) Venn diagram of genes that significantly changed expression levels under various in vitro stress conditions.

FIG. 2.

Hierarchical cluster analysis of the gene expression levels collected from M. avium subsp. paratuberculosis cultures exposed to variable stressors. (A) A heat map displaying the overall correlation among replicates of all examined stressors. (B) An example of cluster analysis showing genes activated only in the cow samples. Note the dendrogram displayed at the top of the image reflecting the overall relationship among examined samples. A color bar is presented at the top of each panel with a range from 0 to 1 (black to red) for panel A or from −3 to 3 (green to red) for panel B.

FIG. 3.

Gene deletion in M. avium subsp. paratuberculosis. (A) The design of lipN-knockout allelic-exchange substrates using the pYUB854 cloning vector (4). (B) PCR confirmation of the lipN-knockout mutant using genomic DNA (gDNA) from the wild type (wt) and the ΔlipN mutant and primer pairs designed for the hygromycin resistance gene, lipN, or the recombinant region after allelic exchange. A 2% agarose gel showed amplicons from the hygromycin resistance gene only when the mutant genomic DNA was used (lane 1), whereas the lipN sequence was amplified only from the wild-type genomic DNA (lane 4). (C) Southern blot analysis of the lipN-knockout mutant. Genomic DNA was digested with XhoI and Acc65I (for hygromycin detection) or XhoI and ScaI (for lipN detection) and detected with hygromycin or lipN probes. The lipN sequence was absent from the lipN-knockout mutant genomic DNA. (D) A genomic map showing the distribution of the 10 genes inactivated by transposon mutagenesis or homologous recombination examined in this and previous studies (36, 53).

FIG. 4.

Screening of M. avium subsp. paratuberculosis mutants in the murine model of paratuberculosis. Mouse groups were intraperitoneally inoculated with 10⁸ CFU/mouse for the M. avium subsp. paratuberculosis ATCC 19698 wild-type strain or its isogenic mutant while groups inoculated with M. avium subsp. paratuberculosis K-10 or its ΔlipN isogenic mutant were inoculated with 10⁷ CFU/mouse. Liver and intestine tissues were collected at 3, 6, and 12 WPI. Only data for the liver are shown here. (A) Colonization levels of three mutants with disruption of genes activated in cow samples compared to levels obtained from mouse groups infected with either M. avium subsp. paratuberculosis ATCC 19698 or M. avium subsp. paratuberculosis K-10. (B) Colonization levels of three mutants with disruption of genes activated in acidic pH compared to levels obtained from a mouse group infected with the wild-type M. avium subsp. paratuberculosis ATCC 19698. Error bars represent standard errors (±) of colony counts from different samples at each time (n = 3 to 6).