Identification of Burkholderia cenocepacia non-coding RNAs expressed during Caenorhabditis elegans infection

Abstract

Small non-coding RNAs (sRNAs) are key regulators of post-transcriptional gene expression in bacteria. Despite the identification of hundreds of bacterial sRNAs, their roles on bacterial physiology and virulence remain largely unknown, as is the case of bacteria of the Burkholderia cepacia complex (Bcc). Bcc is a group of opportunistic pathogens with relatively large genomes that can cause lethal lung infections amongst cystic fibrosis (CF) patients. To characterise sRNAs expressed by Bcc bacteria when infecting a host, the nematode Caenorhabditis elegans was used as an infection model by the epidemic CF strain B. cenocepacia J2315. A total of 108 new and 31 previously described sRNAs with a predicted Rho independent terminator were identified, most of them located on chromosome 1. RIT11b, a sRNA downregulated under C. elegans infection conditions, was shown to directly affect B. cenocepacia virulence, biofilm formation, and swimming motility. RIT11b overexpression reduced the expression of the direct targets dusA and pyrC, involved in biofilm formation, epithelial cell adherence, and chronic infections in other organisms. The in vitro direct interaction of RIT11b with the dusA and pyrC messengers was demonstrated by electrophoretic mobility shift assays. To the best of our knowledge this is the first report on the functional characterization of a sRNA directly involved in B. cenocepacia virulence. KEY POINTS: • 139 sRNAs expressed by B. cenocepacia during C. elegans infection were identified • The sRNA RIT11b affects B. cenocepacia virulence, biofilm formation, and motility • RIT11b directly binds to and regulates dusA and pyrC mRNAs.

Keywords: Burkholderia cenocepacia; C. elegans infection model; Non-coding RNAs; RNA-seq.

Fig. 1

Schematic representation of the methodology implemented to infect C. elegans nematodes with B. cenocepacia J2315 and recover total RNA to study the bacterial transcriptome in conditions of infection. Abbreviations: CDS, coding sequences; 5′ UTR, 5′ untranslated region

Fig. 2

Quality control and characterization of B. cenocepacia J2315 Cappable-seq dataset. A Principal component analysis (PCA) plots of control and infection samples. B Read counting: total, trimmed, and reads mapped to the reference genome of B. cenocepacia J2315. C Number of TSSs obtained per replicon: gene – TSS located up to 100 nt upstream of a gene, 5′ UTR – TSS located from 100 to 300 nt upstream of a gene, intergenic – TSS located over 300 nts upstream of a gene. D Percentage of gene-related TSSs upregulated (twofold), unchanged (any difference in expression level or no expression), and downregulated (twofold) on infection conditions in each replicon. E KEGG pathways overrepresented by the altered genes of B. cenocepacia J2315 infecting C. elegans N2. S1, S2, S3 – B. cenocepacia J2315 samples collected from C. elegans after 48 h of infection. C1, C2, C3 – B. cenocepacia J2315 control samples grown on NGMII agar plates. The Fisher’s exact test was used to determine the statistical significance of the results and the calculated p-value is represented by * when p-value < 0.05, ** when p-value < 0.01, and *** when p-value < 0.001

Fig. 3

Genomic distribution and characteristics of the B. cenocepacia J2315 sRNAs identified in this work. A TSS and sRNA distribution by replicon. Plus: positive or sense strand; minus: negative or antisense strand. B sRNA length distribution. The average sRNA length is 200 nts, which is represented by the Gaussian distribution. C sRNA conservation. The majority of the identified sRNAs are conserved in B. cenocepacia, conserved or semi-conserved in Bcc, and non-conserved in B. pseudomallei group (classification criteria are detailed in Supplemental Table S9)

Fig. 4

Effect of RIT11b overexpression and RIT43 silencing in B. cenocepacia K56-2. A and B Kaplan–Meier survival plots of C. elegans infected with B. cenocepacia K56-2 carrying the empty vector pIN29 (control), the pTAP3 plasmid overexpressing RIT11b (A), or the pTAP11 plasmid silencing RIT43 using an antisense strategy (B). Survival curves show the results of three representative independent assays. Log-rank (Mantel-Cox) test was used to determine significance of survival differences. C Relative expression levels of the sRNAs RIT11b and RIT43 quantified by qRT-PCR. Multiple t tests (one per row) were used to determine statistical significance. D Bacterial growth curves in rich medium (LB), monitored by optical density at 640 nm. C and D graphs show the mean ± SD from a representative of three independent assays. Statistical significance is represented in accordance with p-value, *p-value < 0.05, **p-value < 0.01, and ****p-value < 0.0001

Fig. 5

Effect of RIT11b overexpression in B. cenocepacia K56-2. A Regulatory effect of RIT11b in two predicted targets. Relative expression levels of the mRNAs dusA and pyrC, when RIT11b is overexpressed in B. cenocepacia K56-2 (pTAP3 plasmid). B and C Electrophoretic mobility shift assay (EMSA). Interactions between the unlabelled sRNA RIT11b and the labelled BCAL1625 (B) or BCAL3351 (C) RNAs. D RIT11b impairs biofilm formation. Comparison of the amount of biofilm formed by B. cenocepacia K56-2 with pIN29 or pTAP3 plasmids in polystyrene microtitre plates, after 24 h or 48 h of growth in LB medium. E RIT11b increases the B. cenocepacia motility. Quantification of the swimming motility diameter of bacterial cells at 24, 48, and 72 h after inoculation onto swimming agar plates. B. cenocepacia K56-2 carrying the pIN29 plasmid was used as control. All graphs show the mean ± SD from a representative of three independent assays. Two-way ANOVA was used to assess the statistical significance of biofilm formation and swimming motility, represented in accordance with the p-value, *p-value < 0.05, **p-value < 0.01, ***p-value < 0.001, and ****p-value < 0.0001