Tn-Seq analysis of Vibrio cholerae intestinal colonization reveals a role for T6SS-mediated antibacterial activity in the host

Abstract

Analysis of genes required for host infection will provide clues to the drivers of evolutionary fitness of pathogens like Vibrio cholerae, a mounting threat to global heath. We used transposon insertion site sequencing (Tn-seq) to comprehensively assess the contribution of nearly all V. cholerae genes toward growth in the infant rabbit intestine. Four hundred genes were identified as critical to V. cholerae in vivo fitness. These included most known colonization factors and several new genes affecting the bacterium's metabolic properties, resistance to bile, and ability to synthesize cyclic AMP-GMP. Notably, a mutant carrying an insertion in tsiV3, encoding immunity to a bacteriocidal type VI secretion system (T6SS) effector VgrG3, exhibited a colonization defect. The reduced in vivo fitness of tsiV3 mutants depends on their cocolonization with bacterial cells carrying an intact T6SS locus and VgrG3 gene, suggesting that the V. cholerae T6SS is functional and mediates antagonistic interbacterial interactions during infection.

Figure 1. Preparation and Characterization of Infant Rabbit STAT-Tn-Seq Library

(A) Schematic depicting changes in mutant abundance during colonization of a pool of mutants in the infant rabbit host. Mutants that exhibit no defect in intestinal replication and survival are depicted in black, colonization defective mutants (CD) in red, reduced colonization mutants (RC) in green, and mutants displaying hypercolonization in gray. After orogastric inoculation of infant rabbits, the population of each mutant changes according to its in vivo fitness difference; these changes in abundance can be deduced by STAT-Tn-seq analysis of the inoculum input pool and rabbit-derived output pool. (B) Library building protocol. Genomic DNA pool was sheared by sonication, and frayed ends were repaired and modified by A tailing. An adaptor composed of short 3′ blocked (YM1.0) and long (YM2.0) oligonucleotides was ligated onto the repaired genomic DNA. The 3′ end of the short adaptor oligo (YM1.0) was blocked (*) so that it could not be extended during amplification. PCR amplification of the transposon junction site in the interrupted gene was performed using primers that recognize the end of the transposon (YM3.0) and the 5′ end of the long adaptor oligo (YM4.0). These primers contain all requisite sequences to permit direct multiplex sequencing of amplicons on an Illumina HighSeq instrument and the sequencing primer Tn-FY-seq (YM5.0). (C) STAT-Tn-seq analysis of input library. As a quality control step, two STAT-Tn-seq libraries were prepared from the same input subpool. Each point represents the number of mapped reads of one mutant in each of these two technical replicates of the input pool libraries. The correlation coefficient R² on log-transformed abundance values is shown. (D) STAT-Tn-seq analysis of two output libraries. Two STAT-Tn-seq libraries were prepared from bacterial cells recovered from two rabbit pups that had been infected with the same input subpool. Each point represents the number of mapped reads of one mutant in each of the two biological replicates in the output libraries. The correlation coefficient R² on log-transformed abundance values is shown. Figure 1 is related to Table S1.

Figure 2. STAT-Tn-Seq Data Analysis

(A) Example data for three mutants showing normal, reduced colonization (RC), or colonization defective (CD) phenotypes. The number of STAT-Tn-seq mapped reads to genes VC1403, VC1545, and VC0239 are shown for three technical replicates of the input inoculum library (Input R1-R3) and four output libraries (Out R1-R4) derived from four separate infected rabbits. Normalized reads are shown. The t test was used to compare normalized reads to derive p values for the differences between the average number of mapped reads for the input (In AVG) versus the output (Out AVG) libraries as well as the calculated input/output fold change in the ratio of the mapped reads (Fold I/O). (B) Normalized total mapped read count of 2,951 defined Tn mutants. Each spot represents the mapped reads in the input and output populations for an identified TnFGL3 mutant. CD mutants are shown in red, RC mutants in green, hypercolonizing mutants in gray, and all other mutants are shown in black. The location of the dot corresponding to the mapped reads for the mutant carrying a transposon insertion in VCA0124 (tsiV3) is circled in blue, and in VCA0177 (vspR) is circled in black. (C) Distribution of the number of mutants displaying different output/input ratios. Each bar reflects the total number of the different transposon mutants that displayed various output/input ratios in the particular interval on the scale. The cut-off for CD mutants is 0.067 (1/15), RC mutants 0.2 (1/5), and IP mutants 0.5. (D) Fold defect in intestinal colonization compared to p value determined for normalized output versus input mapped STAT-Tn-seq reads. Each point represents a different TnFGL3 mutant. Fold change in the input/output ratio is expressed in log2 fold and the p values in log10. The location of the mutant carrying an insertion in vspR is indicated by the red triangle, and the tsiV3 mutant is denoted by a black rhombus. Figure 2 is related to Table S2.

Figure 3. Verification of Colonization Defective Mutants Identified with STAT-Tn-Seq

(A) Venn diagram of genes found to be defective in colonization of mice and rabbits. Of 230 CD mutants found in this infant rabbit study, 38 mutants were also present in the group of 47 genes important in the colonization of infant mice in previous studies. (B) Competition assays with multi-mutant subpools of CD mutants. Seven groups of seven LacZ⁺ TnFGL3 mutant strains were each mixed with the wild-type LacZ^– C6706 parental strain in an input ratio of 1:1 for all eight strains. Each group of eight strains was used to infect groups of infant rabbits, and after infection the ratio of LacZ^– to LacZ⁺ was determined on indicator medium. Each point reveals the average C.I. of the entire LacZ⁺ mutant pool in a given subgroup (M1-M7) compared to the LacZ^– WT in the corresponding subgroup. The unusually higher competitive index in group M4 was determined to be caused by one mutant (VC0126) that displayed only about a 3-fold defect in colonization. Statistical significance was determined by t test comparing the C.I. for various pairs to the C.I. determined for a competition between WT C6706 versus C6706 lacZ::Tn, with **** indicating a p value < 0.0001. Mean with SEM was shown. Figure 3 is related to Figure S2 and Table S3.

Figure 4. Predicted Function and Subcellular Location of CD Gene Products

(A) Pie chart for functional categories. (B) Pie chart for metabolic functional categories for the 94 CD mutants that were predicted to affect metabolic functions. (C) Pie chart for the predicted subcellular localization for the protein products encoded by CD mutants.

Figure 5. Colonization Assays on vspR and dncV Mutants

Left: The in vivo competitive index was determined for a phenotypically LacZ⁺ TnFGL3 mutant carrying an insertion in lacZ (lacZ::Tn) competed against its parental ΔlacZ WT strain (WT). The in vivo competitive index was determined for a phenotypically LacZ⁺ TnFGL3 insertion in vspR (vspR::Tn) when competed against its ΔlacZ parental strain (WT) at an input inoculum dose of either 10⁹ (middle left) or 10⁶ (middle right). Right: The in vivo competitive index was determined for dncV::Tn competed with WT strain. Mean with SEM is shown.

Figure 6. Immunity to and Induction of the T6SS of C6706 Contributes to Intestinal Colonization in the Infant Rabbit Model

(A and B) In vivo competition experiments were performed comparing the indicated tsiV3 related mutants (A) and tsiV2 related mutants (B) to parental strain C6706. Statistical significance was determined by t test relative to the C.I. determined for a competition between WT C6706 versus C6706 lacZ::TnFGL3, with **** indicating a p value < 0.0001. (C) Bacterial killing of T6SS immunity mutant defective in TsiV3 under in vitro culture conditions. The tsiV3::TnFGL3 mutant was used as a prey strain for killing by V. cholerae predator strains V52 (expresses functional T6SS killing activity in vitro), V52 ΔvasK (defective in T6SS killing), V52 ΔvgrG3 (has functional T6SS but does not secrete the effector protein of TsiV3), and C6706 WT (inactive in T6SS killingunder invitro growthconditions). The predator strains were individually mixed with the tsiV3 mutant on L agar at 37°C for 3 hr prior to determine the viable counts of the TsiV3 mutant on L agar containing kanamycin. Only the V52 predator noticeably kills the TsiV3 under these in vitro conditions. (D) Transcript level determination for T6SS genes of C6706 WT when grown in vitro in LB versus in vivo in the infant rabbit host. Transcript level was determined by quantitative RT-PCR and is expressed as fold enrichment of transcripts in vivo compared to in vitro relative to 16S rRNA. Transcripts for two housekeeping genes (gyrB and rpoB) that are not directly involved in T6SS function are shown as controls. Three rabbit biological replicates were sampled for each gene. Significance was determined by t test: **p < 0.01, ***p < 0.001, ****p < 0.0001. Mean with SEM is shown.