Single nucleotide polymorphism determines constitutive versus inducible type VI secretion in Vibrio cholerae

Vibrio cholerae is a well-studied human pathogen that is also a common inhabitant of marine habitats. In both environments, the bacterium is subject to interbacterial competition. A molecular nanomachine that is often involved in such competitive behavior is the type VI secretion system (T6SS). Interestingly and in contrast to non-pandemic or environmental isolates, the T6SS of the O1 El Tor clade of V. cholerae, which is responsible for the ongoing 7th cholera pandemic, is largely silent under standard laboratory culture conditions. Instead, these strains induce their full T6SS capacity only under specific conditions such as growth on chitinous surfaces (signaled through TfoX and QstR) or when the cells encounter low intracellular c-di-GMP levels (TfoY-driven). In this study, we identified a single nucleotide polymorphism (SNP) within an intergenic region of the major T6SS gene cluster of V. cholerae that determines the T6SS status of the cell. We show that SNP conversion is sufficient to induce T6SS production in numerous pandemic strains, while the converse approach renders non-pandemic/environmental V. cholerae strains T6SS-silent. We further demonstrate that SNP-dependent T6SS production occurs independently of the known T6SS regulators TfoX, QstR, and TfoY. Finally, we identify a putative promoter region adjacent to the identified SNP that is required for all forms of T6SS regulation in V. cholerae.


Bacterial strains and growth conditions
Bacterial strains used in this study are listed in Supplementary Table S3. Unless otherwise stated, strains were grown aerobically in Lysogeny broth (LB; 10 g/L of tryptone, 5 g/L of yeast extract, 10 g/L of sodium chloride; Carl Roth) or on LB agar plates at 30°C or 37°C.

Recombinant DNA techniques and genetic engineering
DNA manipulations/cloning was carried out using standard methods. PCR amplifications were performed using GoTaq (Promega), Pwo (Roche), or Expand High Fidelity (Roche) polymerases according to the suppliers' recommendations. Genetically modified loci were checked by colony PCR and Sanger sequencing (Microsynth, Switzerland).
V. cholerae were genetically engineered using a set of different methods. Natural transformation on chitin flakes and FLP-recombination (TransFLP; [2,3]) were used to integrate the vipA-sfGFP translational fusion construct allele into the chromosome of diverse strains as reported [4]. Natural transformation on chitin flakes using PCR fragments was also used for the hybrid strain library construction (see details below). All remaining mutations and deletions were done by allelic exchange using the counter-selectable plasmids pGP704-Sac28 and pGP704-Sac-Kan [5,6].

Construction of a pandemic/non-pandemic hybrid strain library
The kanamycin-resistance marker aph was integrated by natural transformation into the genome of pandemic V. cholerae strain A1552. To do so, amplified PCR fragments served as transforming material, whereby each fragment combined an upstream region, aph preceded by its promoter, and a downstream region (fused using overlapping PCR). The PCR fragments were prepared to generate 40 transformants (aph#1 to aph#40) whereby each strain contains the aph cassette at a different location (roughly every 100 kb). Genomic DNA (gDNA) of these 40 donor strains was isolated from 2 ml of an overnight culture using 100/G Genomic-tips together with a Genomic DNA buffer set (Qiagen) as described in the manufacturer's instructions.
To generate the pandemic/non-pandemic hybrid strain library, the acceptor strain ATCC25872 was grown in 40 parallel tubes on chitin flakes and each donor strain-derived gDNA was added to one of those tubes. After incubation at 30°C for 30 h, bacteria were selected on LB plates containing kanamycin and 20 transformants were isolated and stocked from each independent reaction, resulting in 800 strains in total.
After the first screen of the 800 strains library, the reverse experiment from what was described above was performed, generating a set of ATCC25872 strains carrying aph at 40 different genomic locations. However, transformation of A1552 as acceptor strain was only performed using gDNA of strain ATCC25872-aph#32 as transforming material. 20 transformants of this reaction were isolated and stocked and further explored for their T6SS activity. Based on the screening result and the proximity of aph#32 to the large T6SS cluster, an additional construct was generated in both strains (A1552-aph#42 and ATCC25872-aph#42) and their gDNA used to transform in each case the opposite strain to obtain 2 x 20 additional hybrid clones.

Phenotypic screening of the hybrid strain library
The initial library was screened for the strains' T6SS activities using a fluorescence-based E. coli killing experiment, which was adapted from [7]. Briefly, 200 µL of the predator hybrid strain were mixed with 40 µL of GFP-labelled E. coli prey (strain MC4100-TnGFP) in 96-well plates and 5 µL of the mixtures were spotted onto LB plates. The donor (A1552; T6SS OFF) and acceptor (ATCC25872; T6SS ON) strains served as controls. After 4 h incubation at 37°C, the plates were observed under a stereo microscope (Leica EL6000) equipped with a green fluorescence (FITC) filter cube. Prey survival was scored based on their maintained GFP signal. To properly quantify T6SS activity, all aph#32 and aph#42 transformants were rechecked using the standard interbacterial killing assay described below with E. coli TOP10-TnGFP (Cm R ) as prey.

Interbacterial killing assay
Bacterial killing was assessed following a previously established assay [4]. Briefly, the respective predator and the E. coli prey were mixed at a ratio of 10:1 and spotted onto filters on prewarmed LB agar plates (containing 0.2% arabinose where indicated in the figure legend).
After 4 h of incubation at 37°C, the bacteria were resuspended, serially diluted, and spotted onto antibiotic-containing (streptomycin, kanamycin, or chloramphenicol) LB agar plates to enumerate colony-forming units (shown as log-transformed CFU/mL in the graphs).
Experiments were performed at least three independent times. Statistically significance was determined using using GraphPad Prism 9.1.1 (for macOS) on log-transformed data [8] using a one-or two-way ANOVA followed by a Šídák's multiple comparisons test, as indicated in the figure legends. If no prey bacteria were recovered, the detection limit was used to calculate the mean of the independent experiments and to perform the statistical analysis.

Imaging of T6SS sheath structure
To image T6SS sheath structures in strains carrying the translational fusion vipA-sfGFP, the bacteria were mounted on microscope slides coated with a thin agarose pad (1.2% in 0.5X PBS), covered with a coverslip, and observed in the phase contrast and epifluorescence mode (green channel) using a Zeiss LSM 700 inverted confocal laser scanning microscope with an attached HXP 120 light unit (Zeiss, Feldbach, Switzerland). Images were adjusted for contrast and brightness and cropped using the Fiji software [9]. The images are overlays of the Ph and GFP channels and representative of at least three biologically independent replicates.

Whole-genome sequencing and de novo assembly of strain ATCC25872
Bacterial growth and gDNA extraction were done as described [10]. DNA sample preparation and genome sequencing was performed by the Genomic Technology Facility of the University of Lausanne (Switzerland). Briefly, the DNA sample was sheared in Covaris g-TUBEs resulting in fragments with a mean length of 20 kb. The DNA library was prepared using the PacBio SMRTbell template prep kit 1 (Pacific Biosciences) according to the manufacturer's recommendations and size selected on a BluePippin system (Sage Science, Inc.) for molecules larger than 15 kb. The library was sequenced on a PacBio system within one single-molecule real-time (SMRT) cell with P6/C4 chemistry and MagBeads at a movie length of 360 min. The genome was de novo assembled using the protocol RS_HGAP_Assembly.3 in SMRT Pipe 2.3.0 and circularized using the Minimus assembler of the AMOS software package 3.1.0 using default parameters [11]. The assembled genome was annotated using Prokka 1.12 [12] but due to incompatibly with the NCBI database, reannotated using their Prokaryotic Genome Annotation Pipeline (PGAP) during data submission. The data are available under BioSample SAMN13736322 and BioProject PRJNA599000 and the NCBI accession numbers CP047305 (chromosome 1) and CP047306 (chromosome 2). Details on the genome sequencing and assembly are provided in Supplementary Table S4.

Expression profiling by RNA sequencing
Overnight cultures of strains A1552 and ATCC25872 and their SNP45-converted derivatives were back-diluted 1:100 in LB medium and grown for 3 h at 30°C with agitation. Cells were harvested by centrifugation at 4°C and washed with PBS buffer, followed by lysis with Tri Reagent (Sigma-Aldrich) and shock freezing in a dry-ice ethanol bath. The samples were stored at -80 °C prior to processing. RNA preparation and DNase treatment were performed as previously described [13]. After DNase treatment, an additional purification step was performed using the GenElute Mammalian Total RNA miniprep kit (Sigma-Aldrich).  [14] in local mapping mode with very sensitive pre-settings. To count the uniquely mapped reads to annotated genes, the software htseq-count (HTSeq version 0.11.2) [15] was used.
Normalization of the raw counts and differential gene expression analysis was carried out with help of the R software package DESeq2 (version 1.22.2) [16]. Data were visualized using the Integrative Genomics Viewer [17].

Quantitative Reverse Transcription PCR (qRT-PCR)
To analyze gene expression using quantitative reverse transcription PCR (qRT-PCR), overnight cultures were back-diluted 1:100 in LB medium and grown for 3 h at 30°C with agitation. RNA purification, DNase treatment, cDNA synthesis and qPCR followed a previously established protocol [13]. Samples were analyzed on a LightCycler Real-Time PCR System (LightCycler Nano or LightCycler 96; Roche) using the standard curve method.
Expression values are presented relative to the transcript levels of the reference gene gyrA.
Each experiment was performed two independent times.

Western blotting
To check the production of the inner tube T6SS protein Hcp, cell lysates were prepared as described previously [18]. In brief, overnight cultures were back-diluted 1:100 in LB medium and grown with agitation at 30°C for 3 h. Cells were harvested by centrifugation and the bacterial pellet was resuspended in 2× Laemmli buffer (Sigma-Aldrich), adjusting for the total number of bacteria according to the cultures' optical density at 600nm (OD600) values. To check for T6SS-secreted Hcp, 1.5 ml of the culture supernatant was filter sterilized (0.2 μm filter; VWR) and the proteins were precipitated using trichloroacetic acid (TCA). The precipitated proteins were washed with acetone before resuspension in 30 μl of 2× Laemmli buffer (Sigma-Aldrich). All samples were heated at 95°C for 15 min.

Data availability
The RNAseq dataset (for strains A1552, SNP45-converted A1552, ATCC25872, and SNP45converted ATCC25872) is accessible through the GEO Series accession number GSE196165.  Tables S1 and S2 for details.  Figure S2. SNP45-conversion changes T6SS status in pandemic and environmental strains. SNP45-conversion inverts the ability of strains to kill prey bacteria, as assessed in an E. coli killing assay. Results for WT and SNP45-converted derivatives of six well-studied pandemic strains (N16961 derivative is QS-repaired; [19]) (A) or five environmental isolates (with ATCC25872 as control) (C) are shown. d.l., detection limit. Bar plots represent the average of three independent biological replicates as shown by the individual dots (±SD). Statistically analysis was done using one-way ANOVA followed by a Šídák's multiple comparisons test whereby WT and SNP-converted derivatives were compared. **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.     pGP704-28-SacB-∆qstR pGP704-Sac28 carrying a gene fragment resulting in a deletion within qstR (VC0396). Amp R .

MB_1038
[1]  pGP704-Sac-Kan carrying a genome fragment resulting in a site-directed mutation in the -10 element (AA to GC) located in the intergenic region between VCA0106 and VCA0107 (vipA) with a SNP45 as "G"; Kan R