The BpsIR Quorum-Sensing System of Burkholderia pseudomallei

Autoinducer synthesis by BpsIR in B. pseudomallei KHW.

The bacterial strains and plasmids used in this study are described in Table 1. KHWbpsI::Km and KHWbpsR::Km insertion mutants were derived from a local virulent isolate, B. pseudomallei KHW, by gene replacement using the suicide vector pJQ200mp18 as described previously (2, 3). The bpsI and bpsR fragments were amplified from B. pseudomallei KHW genomic DNA using the primer pairs BpsIF (5′ATCTGCAGATGCGAACTTTCGTTCATGGC) and BpsIR (5′ATCTGCAGGAAATACCGTTGAATGGTCCA) and BpsRF (5′ATCTGCAGATGGAACTGCGCTGGCAAGA) and BpsRR (5′ATCTGCAGTTACGGCGCGTCGATGAGCC), respectively(Fig. 1). Located on chromosome 2 of B. pseudomallei, bpsIR is highly similar to pmlIR, which was recently described by Valade et al. (29), and the BpsI and BpsR proteins are 75 and 80% identical to B. cepacia CepI and CepR, respectively (29). bpsI and bpsR are divergently transcribed, and the intergenic 742-bp spacer region contained two lux box motifs composed of 20-bp palindromic sequences which matched the consensus lux box in 15 and 13 of 20 positions, respectively (Fig. 1, insert) (10).

• Open in new tab
• Download powerpoint

FIG. 1.

Genetic organization of the bpsI-bpsR locus of B. pseudomallei KHW (GenBank accession no. AY373337 ). Gray box arrows denote the divergently transcribed bpsI and bpsR genes; short thin arrows denote the PCR primer positions, while short white box arrows denote putative bpsI and bpsR promoters (P). Triangles indicate the locations where kanamycin resistance cassettes were inserted in the null mutants, KHWbpsI::Km and KHWbpsR::Km. The insertion sites were confirmed by PCR and DNA sequencing, while the null phenotypes were confirmed by Northern blotting (data not shown). Black rectangles denote the lux boxes, and their similarities with the palindromic lux box consensus sequence are shown in the insert (10). Consensus sequence abbreviations are as follows: N is A, T, C, or G; R is A or G; S is C or G; Y is T or C; and X is N or a gap in the sequence. The length of the bar represents a distance of 200 bp.

The promoter_bpsI-lacZ fusion (pSYI) and the promoter_bpsR-lacZ fusion (pSYR), obtained by ligating the putative bpsI and bpsR promoters, respectively, to lacZ on pCYY, were introduced into the wild type and bpsIR mutants to study transcriptional regulation of bpsI and bpsR. Exogenous addition of only 0.125 nM N-octanoyl-homoserine lactone (C8HSL) to KHWbpsI::Km (pSYI) restored the bpsI promoter activity to the wild-type level. The bpsI promoter was, in contrast, 10- and 1,600-fold less sensitive to C10HSL (N-decanoyl-l-homoserine lactone) and C6HSL (N-hexanoyl-l-homoserine lactone), respectively, and was insensitive to C4HSL (N-butyryl-l-homoserine lactone), 3OC6HSL [N-(3-oxohexanoyl)-l-homoserine lactone], 3OC8HSL [N-(3-oxooctanoyl)-l-homoserine lactone], 3OC10HSL [N-(3-oxodecanoyl)-l-homoserine lactone], and 3OC12HSL [N-(3-oxododecanoyl)-l-homoserine lactone] tested at concentrations up to 1 μM (data not shown).

Acyl-HSLs, extracted from the spent culture supernatant of Escherichia coli DH5α(pGEM-T-bpsIR) expressing bpsI, were analyzed by high-performance liquid chromatography (HPLC) using a C₁₈ reversed-phase column (Agilent Series 1100 Hypersil octyldecyl silane column; 200 by 4.6 mm; particle size, 5 μm). Upon elution at a flow rate of 1 ml/min with an isocratic profile of methanol-water (50:50, vol/vol) for 10 min, followed by a linear gradient of 50 to 90% methanol in water for 15 min, and an isocratic profile over 25 min, the amount of acyl-HSLs in each fraction was quantified by the β-galactosidase activity produced using KHWbpsI::Km (pSYI). BpsI synthesized mainly C8HSL (Fig. 2). The PmlIR system, in comparison, synthesized predominantly C10HSL, but this may be attributed to the different B. pseudomallei strains used (29).

• Open in new tab
• Download powerpoint

FIG. 2.

Detection of acyl-HSLs produced by BpsI using HPLC. The culture extract (filled circles) contained acyl-HSL extracted from the supernatant of a stationary-phase culture of E. coli DH5α (pGEM-T-bpsIR) in AB medium (4) supplemented with 0.1 μg of thiamine/ml, 0.3% Casamino Acids, and 20 mM glycerol. Concentrated extract was chromatographed on a C₁₈ reversed-phase HPLC column. Each 2-ml fraction collected was concentrated and assayed for β-galactosidase activity using KHWbpsI::Km (pSYI) as the reporter strain and according to the method described by Miller (15). For the profiles of the synthetic C8HSL (open squares) and C10HSL (open triangles) standards, only fractions 5 to 16 are represented, since the others did not yield any detectable activity.

Cell density-dependent expression and transcriptional regulation of bpsI and bpsR.

Cell density-dependent expression of bpsI was observed in the wild type, but not the bpsI or bpsR mutants, which is a characteristic of quorum-sensing genes (Fig. 3A and B). The addition of 0.125 nM C8HSL to the KHWbpsI::Km (pSYI) culture restored the cell density-dependent expression of bpsI previously absent in the bpsI mutant (Fig. 3A). The low-level expression of bpsI in KHWbpsI::Km suggests the possibility of residual autoinducers produced by other B. pseudomallei LuxIR homologs acting on the bpsI promoter (Fig. 3A and B). Like bpsI, transcription of bpsR was also cell density dependent (Fig. 3C) and was positively regulated by bpsI as well as its own gene product (Fig. 3C and D). Such findings are consistent with a model wherein the product of the BpsI synthase, an acyl-HSL, interacts with the BpsR, an acyl-HSL receptor protein, to activate the transcription of both bpsI and bpsR via interaction with the respective lux box motifs identified in the intergenic region between bpsI and bpsR (Fig. 1).

• Open in new tab
• Download powerpoint

FIG. 3.

Cell density-dependent expressions of bpsI and bpsR. Open symbols represent cell densities, while filled symbols represent β-galactosidase activities. Cell densities and β-galactosidase activities of KHW(pSYI) and KHW(pSYR) are represented as open and filled triangles, respectively. Cell density-dependent expression of bpsI was abolished in KHWbpsI::Km(pSYI) (circles, A), and KHWbpsR::Km(pSYI) (circles, B). Addition of exogenous 0.125 nM C8HSL to KHWbpsR::Km(pSYI) did not affect the growth curve (open squares, A) but restored bpsI expression to almost wild-type levels (filled squares, A). KHW did not express any endogenous β-galactosidase activity (data not shown). Cell density-dependent expression of bpsR was also abolished in KHWbpsI::Km(pSYR) (circles, C) and KHWbpsR::Km(pSYR) (circles, D). β-Galactosidase activities were expressed in Miller units. The experiments were performed in triplicate. OD₆₀₀, optical density at 600 nm.

bpsI and bpsR mutants are partially attenuated in virulence in the Caenorhabditis elegans model.

In the B. pseudomallei-C. elegans coculture assay using synchronized L2-stage worms, twice as many worms survived after 48 h of coculture with KHWbpsI::Km as with KHW and the complemented KHWbpsI::Km mutant (Fig. 4A) (8). Similarly, twice as many worms survived when fed on KHWbpsR::Km as when fed on KHW and the complemented KHWbpsR::Km mutant (Fig. 4B). The partial attenuation in C. elegans killing observed with the bpsI and bpsR mutants may be attributed to the presence of other luxIR homologs which may interact to control virulence in B. pseudomallei as in the cases of Vibrio cholera and Pseudomonas aeruginosa (16, 19). Two other luxIR homologs have been identified in the recently sequenced B. pseudomallei K96243 genome. In P. aeruginosa, the quorum-sensing mutants were also not completely avirulent in both mammalian models of infection and pathogen-C. elegans coculture assay (25). The expression of B. pseudomallei virulence is probably multifactorial, and although quorum-sensing genes have significant effects on virulence, many other factors also play important roles in regulating pathogenesis.

• Open in new tab
• Download powerpoint

FIG. 4.

Effects of bpsI and bpsR mutations on B. pseudomallei virulence. Both KHWbpsI::Km and KHWbspR::Km were attenuated in virulence in the B. pseudomallei-C. elegans coculture assays. After 48 h of coculture, 79% of the worms fed on KHWbpsI::Km survived, compared to 39 and 46% survival rates for those worms fed on KHW and the complemented KHWbpsI::Km strains, respectively (A). KHWbpsR::Km was also attenuated in virulence with 62% of the worms surviving after 48 h of coculture, compared to 31 and 37% survival rates in the wild-type KHW and complemented KHWbpsR::Km strains, respectively (B). E. coli OP50 was used as a negative control in this assay.

BpsIR is involved in the secretion of some exoproducts.

Previous studies have demonstrated that B. pseudomallei secreted protease, lipase, and phospholipase C (PLC) into the extracellular milieu, but the roles of these in pathogenesis have not been elucidated. Siderophore production in B. pseudomallei KHW was growth phase dependent, with maximal siderophore production occurring in the stationary-phase culture supernatants (data not shown). Both the bpsI and bpsR mutants yielded two- to threefold more siderophores than the wild type (Fig. 5A). Complementation of the mutants by use of the plasmids pUCP28T-bpsI and pUCP28T-bpsR restored the siderophores' levels to those of the wild type (Fig. 5A). Siderophores, which function in the sequestration of iron, are implicated in the virulence of several pathogenic bacteria, including B. cepacia, where the CepIR quorum-sensing system also negatively regulates ornibactin synthesis, but the significance of such a mechanism is also not established (14).

• Open in new tab
• Download powerpoint

FIG. 5.

(A) Siderophore secretion in B. pseudomallei KHW was negatively regulated by the BpsIR quorum-sensing system. Siderophore activities were assayed in the 24-h-old culture supernatants of B. pseudomallei of KHW, KHWbpsI::Km, KHWbpsI::Km(pUCP28T-bpsI), KHWbpsR::Km, and KHWbpsR::Km(pUCP28T-bpsR) by measuring the differential in readings of optical density at 630 nm (OD₆₃₀) between the test and the sample blank using the chrome azurol S assay as described in Yang et al. (31). The values shown have been normalized for cell density by expressing them as a ratio of ΔOD₆₃₀/OD₆₀₀. Each bar represents the average (± standard deviation) of readings from three independent experiments. (B) PLC secretion by B. pseudomallei KHW is positively regulated by the BpsIR quorum-sensing system. PLC activities were determined in the supernatants of 24-h cultures of B. pseudomallei KHW, KHWbpsI::Km, KHWbpsI::Km(pUCP28T-bpsI), KHWbpsR::Km, and KHWbpsR::Km(pUCP28T-bpsR) by the method described by Kurioka et al. (12). The data presented are the averages (± standard deviations) of the results from three independent experiments.

PLC production in the culture supernatants of B. pseudomallei was also dependent on the growth phase, with maximal production of PLC occurring at the late log phase (data not shown). Unlike siderophore production, PLC production in B. pseudomallei is positively regulated by the BpsIR quorum-sensing system, and the production of PLC in the supernatants of 24-h-old cultures was reduced to half in the bpsI and bpsR mutants compared to that in the wild type (Fig. 5B). The production of PLC was likewise restored to wild-type levels in the trans-complemented KHWbpsI::Km and KHWbpsR::Km mutants (Fig. 5B). The twofold difference between the wild type and the mutants suggests either that the PLC promoter may be indirectly controlled by the BpsIR quorum-sensing system or, alternatively, that the high basal level of PLC expression in the wild type and mutants might be attributable to a second acyl-HSL system which bears upon the PLC promoter. Since PLC is believed to be important for interaction with the phospholipids in eukaryotic cell membranes during infection, the former suggestion would explain how its production is positively regulated by quorum sensing (27, 28).

The secretion of lipase by B. pseudomallei KHW, which was also growth phase dependent, was unaffected in the bpsIR mutants (data not shown). Likewise, protease secretion in B. pseudomallei KHW, as detected by a zone of clearance around the colonies on dialyzed brain heart infusion agar supplemented with 1.5% skim milk (26), was also unaffected by the bpsIR mutations (data not shown). It is also unclear if protease is a virulence determinant in B. pseudomallei, since there was no correlation between virulence and the level of exoproteolytic activity when B. pseudomallei was injected into mice via the intraperitoneal route (9). Further studies are needed to determine the mechanisms by which BpsIR regulates the production of siderophores and phospholipase C in B. pseudomallei.