Genetic Determinants Associated With in Vivo Survival of Burkholderia cenocepacia in the Caenorhabditis elegans Model

Abstract

A Burkholderia cenocepacia infection usually leads to reduced survival and fatal cepacia syndrome in cystic fibrosis patients. The identification of B. cenocepacia essential genes for in vivo survival is key to designing new anti-infectives therapies. We used the Transposon-Directed Insertion Sequencing (TraDIS) approach to identify genes required for B. cenocepacia survival in the model infection host, Caenorhabditis elegans. A B. cenocepacia J2315 transposon pool of ∼500,000 mutants was used to infect C. elegans. We identified 178 genes as crucial for B. cenocepacia survival in the infected nematode. The majority of these genes code for proteins of unknown function, many of which are encoded by the genomic island BcenGI13, while other gene products are involved in nutrient acquisition, general stress responses and LPS O-antigen biosynthesis. Deletion of the glycosyltransferase gene wbxB and a histone-like nucleoid structuring (H-NS) protein-encoding gene (BCAL0154) reduced bacterial accumulation and attenuated virulence in C. elegans. Further analysis using quantitative RT-PCR indicated that BCAL0154 modulates B. cenocepacia pathogenesis via transcriptional regulation of motility-associated genes including fliC, fliG, flhD, and cheB1. This screen has successfully identified genes required for B. cenocepacia survival within the host-associated environment, many of which are potential targets for developing new antimicrobials.

Keywords: Burkholderia cenocepacia; Caenorhabditis elegans; TraDIS; essential genes; in vivo survival.

FIGURE 1

Genes required for B. cenocepacia for in vivo survival. A Circos-derived (Krzywinski et al., 2009; Cheong et al., 2015) atlas representation of Burkholderia cenocepacia J2315 genome is shown (each chromosome is not drawn in scale). Outer ring (track 1; arranged clockwise) shows the chromosomal locations of the genes predicted as essential for in vivo survival at selected time points (locus tag labeled blue: essential for 24 hpi only; black: essential for both 6 and 24 hpi). The log₂ fold-change level (relative reads abundance in the in vivo output pools compared to that in the input pool) is depicted by the histogram in the inner-most two rings of the circular maps (track 2 and 3). Red bars (track 2) represent data obtained for output-6 hpi; blue bars (track 3) represent data obtained for output-24 hpi. The height of the histogram bars corresponds to the degree of fold-change level (log₂-transformed).

FIGURE 2

Functional distribution of genes required for B. cenocepacia in vivo survival in C. elegans. The number of genes for which Tn5 insertions resulted in decreased read abundance in the in vivo output pools relative to the input pool (log₂ fold change ≤2; adjusted p-value < 0.05) in each COG category is shown.

FIGURE 3

Bacterial counts for B. cenocepacia J2315 wild type and knockout mutants in C. elegans. The CFU number of B. cenocepacia knockout mutants ΔBCAL0154, ΔBCAL3116, and ΔBCAL3135 and their isogenic wild type (WT) strain recovered from the infected worms at different time points (6, 24, 48, and 72 hpi) are illustrated. The bars correspond to mean ± SEM of the number of CFU (log₁₀) per worm from two independent experiments of three replicates each; in each replicate, ingested bacteria were extracted from 10 infected worms. The asterisks represent P-values of the WT counts versus the mutants (unpaired, two-tailed Student’s t-test; ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001). Throughout the infection period, the number of ΔBCAL0154 mutants recovered is significantly lower than that of the wild type strain. Disruption of BCAL3116 affected bacterial survival after 24 h of infection. At all-time points tested, ΔBCAL3135 accumulated in C. elegans to a level similar to wild type J2315.

FIGURE 4

Phenotypic characterization of B. cenocepacia J2315 wild type (WT) and the knockout mutants ΔBCAL0154, ΔBCAL3116, and ΔBCAL3135. (A) Swimming and swarming at 25°C and 37°C. The images shown are representative of two independent experiments (for each experiment, three technical replicates were performed). The motility of ΔBCAL0154 and ΔBCAL3116 decreased, whereas no defect in motility was observed for ΔBCAL3135 (unpaired, two-tailed Student’s t-test). (B) Expression levels of motility-associated genes in ΔBCAL0154 relative to wild type. Data are mean ± SEM from two independent experiments with replicate measurements. The asterisks represent P-values of the WT expression levels versus ΔBCAL0154 (unpaired, two-tailed Student’s t-test; ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗∗p < 0.0001). (C) Biofilm formation at 25°C and 37°C. Each bar represents the mean OD_570nm value ± SEM obtained from two independent experiments; in each experiment, measurements of eight technical replicates were taken. At both temperatures tested, levels of biofilm produced by ΔBCAL0154 and ΔBCAL3116 were significantly higher (unpaired, two-tailed Student’s t-test; ^∗∗p < 0.01, ^∗∗∗∗p < 0.0001). (D) Survival of C. elegans infected with the deletion mutants and wild type J2315. Uninfected worms fed with E. coli OP50 served as a control. Graph shows the mean ± SD from a representative of three independent assays. In a pair-wise analysis using the log-rank test, ΔBCAL0154, ΔBCAL3116, and ΔBCAL3135 were significantly attenuated compared to wild type (Kaplan-Meier; p < 0.0001).