Burkholderia thailandensis Methylated Hydroxyalkylquinolines: Biosynthesis and Antimicrobial Activity in Cocultures

Abstract

The bacterium Burkholderia thailandensis produces an arsenal of secondary metabolites that have diverse structures and roles in the ecology of this soil-dwelling bacterium. In coculture experiments, B. thailandensis strain E264 secretes an antimicrobial that nearly eliminates another soil bacterium, Bacillus subtilis strain 168. To identify the antimicrobial, we used a transposon mutagenesis approach. This screen identified antimicrobial-defective mutants with insertions in the hmqA, hmqC, and hmqF genes involved in biosynthesis of a family of 2-alkyl-4(1H)-quinolones called 4-hydroxy-3-methyl-2-alkenylquinolines (HMAQs), which are closely related to the Pseudomonas aeruginosa 4-hydroxy-2-alkylquinolines (HAQs). Insertions also occurred in the previously uncharacterized gene BTH_II1576 ("hmqL"). The results confirm that BTH_II1576 is involved in generating N-oxide derivatives of HMAQs (HMAQ-NOs). Synthetic HMAQ-NO is active against B. subtilis 168, showing ∼50-fold more activity than HMAQ. Both the methyl group and the length of the carbon side chain account for the high activity of HMAQ-NO. The results provide new information on the biosynthesis and activities of HMAQs and reveal new insight into how these molecules might be important for the ecology of B. thailandensisIMPORTANCE The soil bacterium Burkholderia thailandensis produces 2-alkyl-4(1H)-quinolones that are mostly methylated 4-hydroxyalkenylquinolines, a family of relatively unstudied metabolites similar to molecules also synthesized by Pseudomonas aeruginosa Several of the methylated 4-hydroxyalkenylquinolines have antimicrobial activity against other species. We show that Bacillus subtilis strain 168 is particularly susceptible to N-oxidated methylalkenylquinolines (HMAQ-NOs). We confirmed that HMAQ-NO biosynthesis requires the previously unstudied protein HmqL. These results provide new information about the biology of 2-alkyl-4(1H)-quinolones, particularly the methylated 4-hydroxyalkenylquinolines, which are unique to B. thailandensis This study also has importance for understanding B. thailandensis secondary metabolites and has implications for potential therapeutic development.

Keywords: Burkholderia; cell-cell interaction; natural antimicrobial products; quinolones.

FIG 1

Biosynthesis of 4-hydroxy-2-alkylquinoline congeners. Burkholderia thailandensis uses the hmq gene products to synthesize HAQs including HMAQ and HMAQ-NO. In Pseudomonas aeruginosa, the analogous pqs gene products synthesize the related compounds HAQ, HAQ-NO, and PQS. Shown are the N-oxidated species referred to in the text, HQNO and HMAQ-NO-C₉ with a double bond at the 1′-2′ position added by HmqF. The B. thailandensis compounds are methylated by HmqG, which does not have a homolog in P. aeruginosa. PqsH is needed for production of PQS, which is specific to P. aeruginosa.

FIG 2

Sensitivity of Bacillus subtilis strain 168 to a substance produced by Burkholderia thailandensis. (A) For liquid coculture growth, B. subtilis strain 168 was combined in a 1:1 ratio with either Burkholderia thailandensis E264 (WT) or bactobolin-deficient B. thailandensis (Bacto⁻, strain BD20) in LB broth and grown for 24 h at 37°C prior to plating to determine surviving CFU as described in Materials and Methods. Data are representative of three biological replicates. Error bars show standard deviations. (B) Growth inhibition of B. subtilis strain 168 following treatment with cultures or culture fluid from B. thailandensis after 18 h of growth on plates. B. thailandensis wild type (E264) or the bactobolin-defective mutant (Bacto⁻, strain BD20) was applied to a lawn of freshly plated B. subtilis 168, and plates were incubated at 30°C prior to imaging. Top, B. thailandensis culture fluid was filtered and used to saturate paper diffusion discs applied to the lawn of B. subtilis 168. A zone of clearing around a diffusion disc indicates the region where B. subtilis growth was inhibited. Results are similar to those previously reported (11). Middle, unfiltered B. thailandensis fluid (10 μl) was spotted directly onto B. subtilis 168. Bottom, unfiltered B. thailandensis fluid was spotted onto a lawn of B. subtilis 168, as in the experiment whose results are shown in the middle panel, but on medium containing 5% NaCl, which inhibits B. thailandensis growth.

FIG 3

B. thailandensis transposon mutants with reduced effects on Bacillus subtilis strain 168. (A) Amounts of 5 μl of unfiltered fluid from B. thailandensis stationary-phase cultures were spotted onto lawns of freshly plated B. subtilis strain 168 and incubated overnight at 30°C. Results are shown as the diameters of the zones of inhibition. The black dashed line indicates the diameter of the spot of B. thailandensis culture. Transposon mutant numbers correspond with those in Table 1 and are shaded according to the gene mutation, as follows: dark gray, hmqA disruptions; light gray, hmqC disruption; white, hmqF disruptions; hatched, BTH_II1576 disruptions. P (parent), B. thailandensis bactobolin-deficient mutant BD20 used for transposon mutagenesis. Data are the averages of two biological replicates. Error bars show standard deviations. (B) Images of B. subtilis 168 lawns spotted with 5 μl unfiltered fluid from cultures of the B. thailandensis bactobolin-deficient strain BD20 (Parent) or BD20 with a disruption in hmqA (hmqA::cm) or BTH_II1576 (BTH_II1576::FRT-tmp) introduced by homologous recombination.

FIG 4

BTH_II1576 (hmqL) involvement in HMAQ-NO production and activity against B. subtilis strain 168. (A) HMAQ-NO (C₉) was quantified in stationary-phase B. thailandensis strains using LC-MS/MS and methods described previously (21). Error bars show standard deviations. (B) Antimicrobial activity of unfiltered B. thailandensis fluid (5 μl) on a lawn of freshly plated B. subtilis on plates containing 1 mM IPTG. Strains tested were the B. thailandensis bactobolin-deficient BD20 (Parent), BD20 with the IPTG-inducible Plac expression cassette inserted into the neutral glmS1 site in the genome (BD20 Plac), the constructed BD20 BTH_II1576 (hmqL) mutant with the Plac cassette in glmS1 (hmqL::tmp/Plac), or the BD20 hmqL mutant with Plac-hmqL in glmS1 (hmqL::tmp/Plac-hmqL).

FIG 5

Involvement of BTH_II1576 (hmqL) in liquid cocultures and cell pellet fraction localization of HMAQ and HMAQ-NO molecules. (A) Results of cocultures of B. subtilis strain 168 combined in a 1:1 ratio with the bactobolin-deficient B. thailandensis (BD20) parent strain or the parent strain bearing a constructed disruption of hmqA, hmqL, or both in LB broth and grown for 24 h at 37°C. Surviving CFU were enumerated by serial dilution and plating on LB agar containing, for B. subtilis, 5% NaCl (nonpermissive for B. thailandensis growth), and for B. thailandensis, 100 μg/ml gentamicin (nonpermissive for B. subtilis). Data are representative of three biological replicates, and the error bars represent standard deviations. Statistical significance was determined by repeated measures one-way analysis of variance (ANOVA), followed by Tukey’s multiple-comparison tests, and the letters A and B represent statistically significantly different groups (P < 0.05). (B) C₉ congeners of HMAQ and HMAQ-NO were quantified in unfiltered and filtered fluid from cell-free B. thailandensis cultures, as well as in pelleted cells, using LC-MS/MS and methods described previously (21). Error bars show standard deviations.

FIG 6

Heterologous expression of hmqL in Burkholderia ambifaria. (A) HMAQ and HMAQ-NO in cultures of B. ambifaria HSJ1 cells containing either pME6010 or pMP6010-hmqL. Results are the average of three biological replicates and represent the sum of the C₇, C₈, and C₉ congeners of each molecule. Error bars show standard deviations. (B) Antimicrobial activities of unfiltered fluid (5 μl) from cultures of Burkholderia ambifaria HSJ1 containing pME6010 or pME6010-hmqL spotted onto freshly spread lawns of B. subtilis on plates. Plates were imaged after 24 h of incubation at 37°C.