Characterisation of two-component signal transduction systems in Burkholderia pseudomallei

Two-component signal transduction systems (TCSTSs) are abundant among prokaryotes. In pathogenic bacteria, TCSTSs regulate many different phenotypes, including drug resistance and virulence. The Burkholderia pseudomallei bacterium, which causes the severe infectious disease, melioidosis, is predicted to encode approximately 67 TCSTSs. However, only 4 of these TCSTSs have been characterised experimentally. This project aimed to shed light on additional TCSTSs. In order to identify candidate TCSTSs, in silico analyses were performed, using data from the Microbial Signal Transduction database. Amino acid and nucleotide sequences were compared to those of experimentally-characterised TCSTSs. Predicted sub-cellular localisation, conservation among related Burkholderia strains, and possible horizontal transfer of TCSTS genes from other bacteria were also examined. The BPSL1036-BPSL1037 OmpR-EnvZ orthologue, the BPSS2246 hybrid histidine kinase, and the BPSL1669 OmpR-like response regulator were selected for further characterisation. Either the BPSL1036-BPSL1037 operon [Δ(1036-1037)], BPSL1037 (Δ1037), BPSL1036 (Δ1036), BPSL1669 (Δ1669) or BPSS2246 (Δ2246) were deleted by allelic-exchange mutagenesis. A complementation vector for the Δ1036 and Δ1037 mutants was constructed by cloning the BPSL1036-BPSL1037 operon into pBHR1, a low-copy replicating plasmid, and successfully introduced into the B. pseudomallei wildtype, Δ1036 and Δ1037 strains. The OmpR-EnvZ system is an essential Escherichia coli osmoregulator. However, all 5 B. pseudomallei TCSTS mutants displayed similar survival to the wildtype strain on high-osmolarity media containing either sodium chloride or sucrose. The Δ(1036-1037), Δ1037 and Δ1036 strains also displayed similar swarming motility to the wildtype strain. When cultured on Ashdown’s agar for 21-28 days, Δ1037, Δ1036, Δ2246 and Δ1669 colonies developed a distinct, hypermucoid morphology, absent in similarly-cultured Δ(1036-1037) and wildtype colonies. Both the mucoid strains and non-mucoid Δ(1036-1037) and wildtype strains secreted similar levels of type I capsule. However, the hypermucoid strains released increased quantities of extracellular DNA (eDNA), even though increased cell lysis, a possible eDNA release mechanism, was not observed in these strains. In static liquid medium, the Δ1037 and Δ1036 mutants developed altered pellicle architecture compared to the wildtype, while Δ(1036-1037), Δ2246 and Δ1669 did not. The complemented Δ1037 and Δ1036 strains produced similar quantities of eDNA and similar levels of pellicle formation in static liquid culture as the wildtype strain. The virulence of each B. pseudomallei TCSTS mutant strain was assayed using the BALB/c acute melioidosis mouse model. All 5 TCSTS mutant strains caused disease at a similar rate to the wildtype strain. In vivo fitness of the Δ(1036-1037) and Δ1037 strains was determined by simultaneously infecting mice with both the wildtype and a mutant strain; neither the Δ(1036-1037) nor Δ1037 mutants displayed altered in vivo survival relative to the wildtype. Nevertheless, transcriptome profiling indicated that BPSL1037 influences the expression of several known virulence factors. In conclusion, while not essential for acute disease, BPSL1036-BPSL1037, BPSS2246 and BPSL1669 influence the release of eDNA, and the BPSL1036-BPSL1037 system is required for normal pellicle biofilm formation. Further investigations may characterise the protein-protein and protein-DNA interactions of each TCSTS that give rise to these phenotypes, contributing to an improved understanding of the B. pseudomallei regulatory mechanisms necessary for survival in a range of niches.