Predicting Toxin-Antitoxin Systems Responsible for Host-Induced Persistence of Burkholderia pseudomallei

Burkholderia pseudomallei (BPM) is the infectious agent that causes the disease. Melioidosis. The mortality rate for Melioidosis is up to 40% regardless of treatment. Despite the ability of certain antibiotics to reduce mortality, infection relapse occurs in 15-23% of treated patients. The inability of antibiotics to clear a pathogen is partly due to persistence, a mechanism employed by bacteria to enter a dormant state and evade the effects of antibiotics and host defenses. Toxin-antitoxin systems (TAS) have been identified as important modulators of persistence and targeting them provides an avenue for new drug development to reduce chronic infection. When investigating TAS, the predominant issue is the presence of a large number of TAS in an organism. Currently, there is no predictive model to determine which conditions toxins respond to. Because the major reason to investigate TAS is to define their role in chronic infection, this study aimed to identified and predict which toxins are associated with host survival. In the genome of BPM, 103 putative toxins were identified. Toxins were associated to a stress condition using unsupervised bipartite network analysis on an existing expression dataset that exposes BPM to 82 different conditions. Thirty-two toxins showed association to host-like conditions of which, six toxins were further examined for their role in host persistence. Toxins BPSS0390, BPSS1584, and BPSS0395 were upregulated in host-like conditions, but only moderately conserved among BPM strains. Loss of function moderately reduces persistence in macrophages and reduces chronic infection in a murine model of Melioidosis. Three novel toxins, BPSS0899, BPSSS1321, and BPSL1494 were found to be highly conserved among BPM strains and demonstrated to be functional but were constitutively expressed. These toxins played an active role in in vitro persistence and macrophage survival, but not macrophage or in vivo persistence, likely due to redundancy of toxin systems. These findings shift the paradigm of toxin-antitoxin (TA) research to use a data-driven approach and indicate the utility of investigating constitutively expressed toxins as highly conserved regulators of persistence.