Identification of V. parahaemolyticus biologically essential genes for glycolytic or chitinolytic growth

We developed an efficient genetic transposition system to initially identify biologically essential genes within V. parahaemolyticus RIMD2210633. The bacterial strain was selected for stable streptomycin resistance and its genome sequenced to provide a baseline reference (see methods and S1 Data). Next, we used an efficient biparental conjugation approach to introduce a suicide plasmid containing a mariner-type transposon into V. parahaemolyticus. Cells undergoing transposition were selected on minimal medium with glucose (see methods) supplemented with chloramphenicol and streptomycin to select against donor bacteria.

Next, isolated transposon mutant populations were then grown in liquid minimal media containing either glucose or colloidal chitin as their sole carbon source and incubated until mid-log growth. Genomic DNA was isolated from the population of transposon mutants and served as the reference data set of established (viable) mutants with transposon insertions. The TRANSIT insertion sequencing analysis software was used to determine essential genes in the glucose condition (minimal medium + glucose) as well as the chitin condition (minimal medium + 0.4% colloidal chitin) and compare the two conditions to find conditionally essential genes [23] (S1 Data). This was done by first using a hidden markov model (HMM) approach to identify essential regions of the genome in each condition, followed by a zero-inflated negative binomial (ZINB) approach to compare the two conditions. First, each condition was subjected to an HMM analysis which locally determines differences in transposon insertion density and provides essentiality calls for genes: growth advantage, non-essential, growth disadvantage, and essential. HMM analysis of the glucose condition determined 9 genes whose mutations conferred a growth advantage, 3671 genes that were non-essential, 223 genes whose mutations generated a growth defect, and 663 genes that were essential for growth (Fig 1A and 1B). In the chitin condition, mutations in 82 genes conferred a growth advantage, 3725 genes were non-essential, mutations in 148 genes conferred a growth disadvantage, and 611 genes were labelled essential for growth (Fig 1A and 1B). 3547 genes were non-essential and 539 were essential in both conditions.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Summary of Tn-seq analysis.

(A) Circos plot showing transposon insertion data across both chromosome 1 and chromosome 2; (1.) the loci indicated as essential (combined growth disadvantage and essential from HMM analysis) and non-essential (combined growth advantage and non-essential from HMM analysis) (2.). Red loci are essential, while grey loci are non-essential. Both glucose (Glc and Chitin (Chi) growth condition datasets are shown. (B) Genes indicated as non-essential (NE), essential (ES), growth defect (GD), or growth advantage (GA) by HMM analysis for both chromosomes.

https://doi.org/10.1371/journal.pgen.1011370.g001

Tn-seq identifies fitness determinants linked to chitinolytic growth and validates known Vibrio genes in chitin metabolism

To assess if the chitin selection media was performing as expected, we compared our data to modeled Vibrio chitin catabolic pathways [4,14] with an expectation to identify a subset of genes previously linked to chitin metabolism. As expected, conditionally essential genes for the chitin condition included nagA (vpa0038), nagB (vp0829), chiS (vp2478), cbp (vp2479), almost the entire chb operon (vp2479-vp2488, except vp2485), one of three encoded β-N-acetylhexosaminidases (vp2486), and a chitodextrinase (vp0832) (Fig 2A, 2B, and 2C). Importantly, these genes comprise the two major chitin catabolism pathways in Vibrio spp. and all had significant p-values by ZINB (S1 and S2 Data). Furthermore, most (but not all) of the type II secretion system (T2SS) associated genes were identified as essential (Fig 2D), consistent with the established role of this system in secreting chitinase enzymes that degrade chitin polymers. These data suggest that our applied Tn-seq approach for V. parahaemolyticus was successfully interrogating the biological system related to chitin metabolism. Critically, the data identifies each sequential enzymatic reaction required for intracellular oligomeric GlcNAc catabolism (Fig 3). Interestingly, nagE (vp0831), the glucosamine kinase (vp2485), the cellobiose PTS (vp2635-vp2637), ChiP (a chitoporin, vp0760) (Fig 2E), two of the three encoded β-N-acetylhexosaminidases (vp0755; p=0.283 and vp0545; p=0.407), as well as all the encoded chitinases (vp2338, vpa0055, vpa1177, vp0619) were not conditionally essential in the chitin condition (Fig 3). Some of these observations can potentially be explained by known functional genetic redundancy within the genome (e.g., multiple chitinase and hexosaminidase genes) or bifunctional enzymatic activities for some proteins. In contrast, the conditional non-essential finding for the chitoporin (Vp0760) raised the perplexing question as to how large chitooligosaccharides, as a sole carbon source, would enter the bacterial cell to support growth. Interestingly, an unclassified OprD-like porin, encoded by vp0802, was identified as essential for chitinolytic growth (Fig 2F), a finding that is novel since these porins typically transport amino acids and other carboxylic acid compounds, and not N-acetylated compounds like chitooligosaccharides (see detailed section below). Furthermore, the striking fitness defect for vp1236 (p=0.000016) (S2 Data) revealed a novel discovery for a helix-turn-helix transcriptional regulator with respect to specific chitinolytic growth (see detailed sections below).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Tn-seq insertion data indicating specific gene regions.

(A) nagE-nagAC operon, (B) nagB and surrounding genomic region, (C) chiS and the chb operon, (D) gsp operon for T2SS, (E) chiP, (F) vp0802. (G) vp1236 and surrounding genomic region. Chromosome numbers are indicated with approximate kilobase markers for location. The gene annotation for each region is shown, with genes demonstrating statistically underrepresented transposon insertions by ZINB analysis shown in red. Insertion counts for each genomic region are shown as peaks for both the chitin and glucose growth conditions. vp1237 is an essential gene (glutamate decarboxylase) under both growth conditions.

https://doi.org/10.1371/journal.pgen.1011370.g002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Genes essential for chitin metabolism as identified by Tn-seq.

A pathway diagram for catabolism of chitin to fructose-6-phosphate is shown, with component enzymes and their associated locus tags indicated. Chemical conversions which are known to occur, but do not have associated enzymes are shown with a question mark and dotted arrow. Conversions and locus tags shown in red are indicated as conditionally essential (p < 0.05) for growth on chitin by ZINB analysis of the Tn-seq dataset.

https://doi.org/10.1371/journal.pgen.1011370.g003

The V. parahaemolyticus Type II Secretion System (T2SS) is required to support efficient chitinolytic growth

T2SS are multicomponent protein assemblies that secrete proteins into the extracellular environment [24]. Vibrio species are known to secrete their cognate chitinases via the T2SS [25]. Strikingly, while none of the chitinase genes were essential on their own (S1 Data), many genes identified in the gspCDEFGHIJKL type II secretion system operon were essential for growth on chitin (Fig 2D). While some of the genes in this operon did not meet the p-value cutoff of 0.05, many approached significance and were essential by HMM analysis (S2 Data). gspD, gspE, gspF, gspG, and gspK were statistically underrepresented in the chitin growth condition, while gspC and gspI approached significance. gspH, gspJ, and gspL were not underrepresented in the chitin growth condition (Fig 2D). Importantly, gspH, gspJ, and gspI encode minor pseudopilin subunits, and this data would suggest that mutations in these genes do not affect the function of the T2SS with respect to supporting growth on colloidal chitin. Taken together, these data suggest that a functional T2SS is a critical fitness determinant for V. parahaemolyticus chitinolytic growth.

An unclassified OprD-like porin is required for efficient growth on colloidal chitin

Vibrio species are known to utilize a trimeric ChiP porin complex to selectively import chitooligosaccharides into the periplasm (Figs 4A and S1) [26–29]. The porin complex is in the bacterial outer membrane where it supports facilitated diffusion of chitooligosaccharides driven by binding affinities within the hydrophilic porin channels [27,28,30]. ChiP is often considered essential for growth on chitin as this chitoporin is found in most Vibrio species [13]. In contrast, a study in V. parahaemolyticus RIMD2210633 (the same pandemic strain in this study) found that ChiP was not essential for growth on chitin flakes [31]. This latter observation is concordant with our Tn-seq data for ChiP being non-essential for chitin uptake. Instead, our Tn-seq data revealed a significant growth defect for mutants with insertions within vp0802 (p<0.0001, Fig 2F), an open reading frame that putatively encodes a protein with approximately 35-40% sequence similarity to OprD-like single channel porins in various Proteobacteria [32,33] and is broadly conserved in various Vibrio DNA sequence database entries (S1 Fig). This was unexpected, especially since OprD family porins are considered selective to substrates with carboxylic acid (-COOH) groups, including amino acids and some antibiotics (e.g., carbapenems) [34]. AlphaFold2 structural modelling analyses [35] of Vp0802 predicts a single channel beta-barrel presumably located in the outer membrane of bacterial cells, with remarkable structural similarity to OprD family porins [34] (Figs 4A and S1). Moreover, SignalP analysis identified a high confidence putative signal peptide sequence, consistent with protein export beyond the cytoplasmic membrane, presumably with an outer membrane localization for the barrel structure (S2 Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Vp0802 is an outer membrane OprD-like

β-barrel porin that supports efficient chitinolytic growth. (A) Alphafold2 structural prediction of Vp0802 compared with ChiP trimer, coloured by predicted local distance difference test (pLDDT) confidence values. (B) Chitinolytic growth of indicated bacterial strains as determined by viable plate counts. Note that wild type cells with high growth (t=48 to 54 hours) enter a putative viable but not culturable (VBNC) condition after depleting the media of colloidal chitin. Statistical significance was determined with an unpaired t-test between wild type and each of the strains at the indicated time points (n=3, *:p < 0.05). (C) Coomassie stain of bacterial outer membrane fractions isolated from the indicated strains at early stationary phase of growth for the respective strains. The black arrowhead points to Vp0802 and the white arrowhead points to ChiP with mass spectrometry validation of the protein species. M, molecular mass standards.

https://doi.org/10.1371/journal.pgen.1011370.g004

To investigate the contribution of Vp0802 relating to chitinolytic growth, we constructed a null mutant strain deleted for the vp0802 DNA coding sequence (Δvp0802) for comparison to wild type V. parahaemolyticus (RIMD2210633, parent strain). After an initial lag phase of approximately 24 hours, wild type bacteria exhibited efficient logarithmic growth and entered a stationary phase at 36 hours (Fig 4B) (S3 Data, for statistical analyses between specific time points for a given strain). Viable cell counts (achieved by plating on LBS agar) appeared to decline however there was not a decrease in the colloidal cell culture density or any visible aggregation. This may represent a temporary state of a viable but not culturable condition (see discussion) as a sporadic increase in viable bacteria at 66 hours was observed. When compared to the wild type, the Δvp0802 strain had a significant growth defect when colloidal chitin was the sole carbon source and exhibited an extended lag phase until 66 hours (Fig 4B). After 66 hours, the Δvp0802 strain did eventually exhibit growth as revealed by an increase in cell number between 66 and 72 hours post inoculation. Genetic complementation with a plasmid encoding vp0802 restored chitinolytic growth, albeit delayed compared to the parental strain (S3 Fig). These data suggest that vp0802 contributes to efficient chitinolytic growth. Furthermore, bacteria deficient for vp0802 expression can utilize an alternate and temporally separate mechanism to produce chitinolytic growth.

Next, we set out to determine the cellular localization of Vp0802. After 80 hours of colloidal chitin growth, bacterial cells were harvested, followed by sonication and lauryl sarcosine sulfate treatment to enrich an outer membrane fraction [26,36] for SDS-PAGE analyses. A dominant protein species was visible at approximately 50 kDa for the parental strain, consistent with the predicted molecular mass of 49.25 kDa for processed (mature) Vp0802 (Fig 4C). Notably, this protein species was absent from the outer membrane preparation derived from the Δvp0802 strain that exhibited delayed growth and was clearly restored in the genetically complemented strain (Fig 4C). Interestingly, a dominant protein species at approximately 40 kDa was apparent in the Δvp0802 strain but missing in the parental and genetically complemented strains (Fig 4C). We hypothesized that this apparent 40 kDa protein species was the ChiP chitoporin (Vp0760), as its predicted mature molecular mass is 38.33 kDa. Therefore, we subjected the respective 50 and 40 kDa protein species to mass spectrometry analyses to determine their identities. As expected, the 50 kDa protein outer membrane protein species was determined to be Vp0802 by peptide matching (S4 Fig). Furthermore, the 40 kDa protein species was conclusively identified as ChiP. Moreover, we generated a vp0802, chiP double mutant strain which was subsequently demonstrated to be severely compromised for growth on colloidal chitin (Fig 4B). Genetic complementation of the vp0802, chiP double mutant with a plasmid expressing Vp0802 or ChiP partially restored chitinolytic growth (S3 Fig), suggesting that these two chitoporins are the main outer membrane uptake mechanisms for large chitooligosaccharides. Collectively these data indicate that Vp0802, encoding an OprD-like porin, is localized to the outer membrane of V. parahaemolyticus cells grown in colloidal chitin as a sole source and that ChiP outer membrane protein levels are minimal in that context. In bacterial cells without Vp0802, the ChiP chitoporin was expressed and accumulated in the outer membrane to support chitinolytic growth.

Vp1236 is an important regulator for growth on diverse nutrient sources

The genomic location of Vp1236 exhibits conserved gene synteny in various Vibrio species (S5 Fig). These proteins are believed to be linked to central carbon metabolism based on sequence similarity to the HexR/MurR/RpiR family of transcriptional regulators. Notably, Vp1236 exhibits 59% amino acid identity (77% similarity) with Pseudomonas aeruginosa (PAO1) HexR where its role in carbohydrate utilization was initially characterized [37]. We opted to use HexR nomenclature for Vp1236 from here on, based on HexR being linked to six carbon sugar metabolism in other bacteria (such as glucose and other hexoses) and other known roles in central carbon metabolism [38–40]. Vp1236 shares 95% sequence identity with Vc1148 of V. cholerae and neither of these proteins have been functionally implicated in chitinolytic growth [13,31]. To investigate the contribution of HexR in V. parahaemolyticus fitness, a chromosomal hexR deletion (ΔhexR) was generated and compared to the wild type strain for growth in both rich and minimal nutrient sources. Under static conditions in rich LB media, wild type bacterial growth produced a sigmoidal growth curve, entering an initial stationary phase at 300 minutes and then showed a steady rate of ‘secondary’ growth that reached a plateau around 800 minutes with a maximum 0.196 (Fig 5A). The ΔhexR strain produced a similar initial sigmoidal growth trend to wild type bacteria but its secondary growth rate began to exhibit a statistically significant reduction in growth with a maximum of 0.157. Genetic complementation with a plasmid expressing HexR (pVSV105-hexR) restored the growth pattern to that of wild type bacteria (Fig 5A). These data suggest that the ΔhexR strain growth defect was due to the functional absence of HexR.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. The role of HexR in supporting V. parahaemolyticus growth under different nutrient conditions.

(A) Growth of indicated strains in LB medium. Strains were cultured overnight in LBS and normalized to starting OD600nm of 0.025. Cell density readings were taken every 20 minutes for 14 hours at 37°C in a VictorX5 Multi Label plate reader. (B) Growth of indicated strains in MM9 media supplemented with glycerol. (C) Growth of indicated strains in LB Spent Media. Strains were inoculated into spent media following growth in LBS at a starting OD_600nm = 0.025. Optical density (OD_600nm) measurements were taken in duplicate for each strain every 6 hours for 36 hours. Statistical significance was calculated at the indicated time points using an unpaired t-test against wild type (n=3, *: p<0.05, **: p<0.01).

https://doi.org/10.1371/journal.pgen.1011370.g005

In the case of nutrient limiting Marine M9 media (MM9) containing glycerol as the sole carbon source, the hexR mutant growth defect was further exacerbated (Fig 5B). Glycerol is metabolized into dihydroxyacetone-phosphate (DHAP) which can enter either the glycolytic or gluconeogenic pathways [41]. While the wild type V. parahaemolyticus strain did have an extended lag phase compared to rich (LB) conditions, the sigmoidal growth trend remained the same with a maximum of 0.054. In contrast, ΔhexR exhibited an extended lag phase in the first 750 minutes followed by statistically significant reduction in growth where cultures reached a maximum of 0.025 (Fig 5B). Genetic complementation of ΔhexR restored the growth phenotype albeit delayed, back to the pattern observed for the wild type strain.

Entrance into stationary phase prompts significant metabolic changes and shifts the flux of carbon [42]. Likewise, adaptations of bacteria into stationary phase or nutrient depleted conditions include significant changes to cell morphology and physiology to allow the cell to persist during environmental stressors (e.g., changes in pH or nutrient depletion) [43]. To further investigate the growth of ΔhexR in nutrient depleted conditions, we conducted cell growth experiments using filtered LB spent media derived from 18 hour wild type cultures (shaking, aerobic, late stationary phase). Under aerobic shaking spent media conditions, wild type bacteria displayed a long lag phase of approximately 16 to 18 hours before entering logarithmic growth. The ΔhexR strain lag phase was even longer at 24 hours, further suggesting that these bacteria have a defect in nutrient assimilation. Strikingly, the ΔhexR genetically complemented strain did not exhibit a lag phase and grew faster than wild type bacteria between 6 to 18 hours. After 18 hours, HexR overexpression appeared to limit additional growth as cell density reached a plateau (between 18-24 hours) and declined after 24 hours (Fig 5C). In contrast, between 18-30 hours wild type bacteria were still growing as revealed by a continued increase in cell density. Moreover, ΔhexR bacteria exhibited a delayed increase in cell density between 24-36 hours. This suggested that hexR expression provides an initial growth advantage in nutrient limited environments, but its dysregulated expression or absence is inhibitory in extremely low nutrient conditions. Indeed, wild type V. parahaemolyticus overexpressing HexR did not exhibit a lag phase and had an initial increased growth rate over the normal wild type strain (Fig 5C). Furthermore, the growth trends of the overexpression HexR strain and the ΔhexR genetically complemented were similar and both strains entered stationary phase at 18 hours. Taken together, this suggests that while the overexpression of HexR provides an initial growth advantage in nutrient-limited environments, ultimately wild type bacteria with inherent coordinated central carbon metabolism mediated by HexR exhibit efficient growth.

HexR is an important regulator for growth on chitin as a sole carbon source

To validate the Tn-seq hexR observations relating to chitinolytic growth, we performed growth assays using MM9 supplemented with 0.4% colloidal chitin as the sole carbon source. Unexpectedly, while the ΔhexR strain had a statistically significant growth defect in the first 16 hours, it consistently outgrew the wild type strain by 20 hours and continued this faster rate of growth until reaching stationary phase (Fig 6A). The wild type strain showed slower growth kinetics and reached a similar final cell density by 54 hours. Notably, when ΔhexR was genetically complemented, the strain returned to wild type pattern of growth (Fig 6A). Therefore, the ΔhexR strain displayed dysregulated and an altered growth rate on chitin as a sole carbon source.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. HexR is a fitness determinant for V. parahaemolyticus on chitin as a sole carbon source.

(A) Overnight LBS cultures of strains were inoculated into M9 Minimal Media containing colloidal chitin as a sole carbon source. Samples were collected every 6-12 hours for 54 hours. Serial dilutions were plated to determine the number of cells present. Statistical significance was determined with an unpaired t-test between wild type and the indicated strain at indicated time points (n=3, *:p < 0.05). (B) Competition assay to assess fitness defect of ΔhexR compared to wild type V. parahaemolyticus. Overnight LBS cultures of wild type GFP (+) transposant and ΔhexR GFP (-) transposant (strain A) were inoculated together into MM9 (0.4% colloidal chitin) at a starting OD600nm of 0.025. Samples were collected every 6-12 hours for 48 hours. Serial dilutions were plated to determine the number of fluorescent cells and total cells present. Statistical significance was measured with an unpaired t-test (n=3, *: p<0.05). (C) Competition Indices (CI) of three independent GFP (-) hexR mutant transposant strains (A, B, C). The transposon insertion site for each strain was determined and the respective strains did not differ for growth when compared to the ΔhexR parent (S6 Fig). The individual GFP (-) hexR mutant transposants were subjected to a head-to-head competition assay with the WT GFP (+) strain in MM9 (0.4% Colloidal Chitin). The CI for three experiments (n=3) were plotted, CI < 1 indicates a fitness defect.

https://doi.org/10.1371/journal.pgen.1011370.g006

The observation of faster dysregulated chitinolytic growth for the ΔhexR strain was paradoxical to the Tn-seq gene fitness data where a gene fitness defect was observed for hexR transposants (insertional mutants). A key difference in the experiments is that Tn-seq is a mixed population-based competitive growth context, whereas ΔhexR bacteria grown in isolation (i.e., pure culture) represents an isogenic condition. Therefore, to investigate the competitive fitness of wild type and ΔhexR bacteria, we generated a fluorescently tagged wild type strain with a transposon system capable of expressing stable green fluorescent protein (GFP +). A similar approach was performed to generate three ΔhexR transposant strains that were non-fluorescent (GFP-) (see methods). This allowed for an experimental setup with a tractable mixed culture competition assay. Critically, we initially confirmed that all the transposant strains were unchanged for LB growth compared to their parental strains (original sources) (S6 Fig). Furthermore, we mapped the chromosomal transposon insertion site for each strain and confirmed that the disrupted genes do not contribute to chitinolytic genetic fitness based on our Tn-seq dataset (S6 Fig and S1 Data). Next, following co-inoculation of the transformant strains at the same starting OD_600nm into MM9 (0.4% colloidal chitin), serial dilutions of the cultures were plated over a time course of 48 hours. There was no significant change in growth between wild type or a hexR mutant transposant strain during the first 24 hours post-inoculation, however, after 48 hours the wild type transposant (GFP+) had significantly outgrown ΔhexR bacteria (GFP-)(Fig 6B). The average competition index (CI) derived from three independent competition experiments between wild type (GFP+) and each of the three hexR transposants (GFP-) at 48 hours was 0.4 (Fig 6C). Incorporating GFP expression in the hexR mutant (GFP marker swap) also demonstrated a competitive growth defect (S6 Fig). Given that a CI < 1 indicates a fitness defect, these results support the interpretation that cells lacking functional hexR were at a significant competitive disadvantage in mixed culture chitinolytic growth conditions. Importantly, this competition assay data validates the observations from the Tn-seq data and implicate contextual hexR expression as a critical fitness determinant, especially under nutrient limiting conditions.

HexR is a regulator for biofilm formation and cell differentiation under nutrient limiting conditions

The data implicating HexR as a fitness determinant in nutrient limiting conditions suggests that HexR functionality would contribute to physiological factors that are important for V. parahaemolyticus environmental persistence. Notably, biofilms contribute significantly to the aquatic survival of V. parahaemolyticus and require considerable investment in the synthesis and export of polysaccharides [44]. Biofilm formation is additionally important for efficient chitinolytic growth which occurs post-cellular attachment to chitin polymers. As ΔhexR bacteria displayed growth defects under nutrient limiting conditions, this led us to investigate its biofilm phenotype. As expected, in MM9 media (with glycerol and casamino acids) wild type V. parahaemolyticus displayed biofilm formation. In contrast, ΔhexR exhibited a statistically significant reduction in biofilm formation even when cell number was normalized (Fig 7). Genetic complementation restored biofilm formation to wild type levels indicating that HexR contributes to proper biofilm formation under nutrient limited conditions. Critically, we tested all the strains for biofilm formation in rich LB media and observed no difference, suggesting that the role of HexR in biofilm formation is likely nutrient or context dependent (Fig 7).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. HexR contributes to V. parahaemolyticus biofilm formation under nutrient limiting conditions.

Biofilm formation of strains grown in MM9 (0.4% Glycerol, 0.2% Casamino Acids) or LB for 18 hours at 37°C. Each dot represents an individual sample. Statistical significance was calculated using one-way ANOVA (n=9, *:p < 0.05, **: p < 0.01).

https://doi.org/10.1371/journal.pgen.1011370.g007

Our data implicating HexR for biofilm and chitin association prompted us to evaluate cell shape and cell differentiation as previous studies have reported that dysregulated central carbon metabolism often leads to defects in cell morphology [45, 46]. Moreover, for growth on chitinaceous sources, V. parahaemolyticus is known to exhibit elongated cell morphologies as a response to environmental cues [21]. Observations of all strains cultured in LB medium revealed no apparent differences in cell shape, with slightly curved rods being uniformly apparent (Fig 8A). In contrast, wild type bacteria cultured in nutrient limiting MM9 (glycerol) or MM9 with colloidal chitin developed a filamentous subpopulation of cells. Strikingly, ΔhexR cells remained homogenously and strictly curved rod-shaped and were devoid of filamentous cells (Fig 8A). Genetic complementation of ΔhexR restored the filamentous cell subpopulation. This phenotype was extremely robust and was quantified across multiple fields of view in microscopy observations. Across 50 fields of view (FOV) with a minimum of 30 cells, the filamentous cells made up approximately 20-32% of the total cells observed in both wild type and genetically complemented strain backgrounds but were completely absent for the ΔhexR strain (Fig 8B). Under the MM9 glycerol condition, we measured cell lengths from the respective cultures and observed a significant reduction in cell length for the ΔhexR cells compared to wild type bacteria (Fig 8C). When the same strains were cultured in MM9 supplemented with colloidal chitin, the ΔhexR cells were predominantly observed to be short rods in contrast to the wild type which exhibited a mix of filamented cells and short rods (Fig 8D). Based on these observations, we conclude that HexR contributes to V. parahaemolyticus cell filamentation under nutrient limiting conditions.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. HexR contributes to cell filamentation in nutrient limiting conditions.

(A) Representative microscopy of wild type (WT), hexR mutant, and hexR complement strains (top to bottom) in LB and MM9 (0.4% glycerol). Cells were observed at 100x. Scale bar = 5μm. Insets (I1) identify cell morphologies for each genotype. (B) Quantification of filamentous cell subpopulation for the indicated strains. Cells were imaged with microscopy after 16 hours of growth in MM9 (0.4% Glycerol). All cells were counted in a field of view (FOV), and filamentous cells further enumerated to determine the % of filamentous cells per FOV. (C) Measurement of cell length of for bacteria cultured in MM9 (0.4% glycerol) conditions. A minimum of 50 cells across 5 independent observations were measured using Image J. Data were plotted with standard deviation and statistical significance was calculated using an unpaired t-test (n=50, ****: p<0.0001). Mean wild type cell length was 21μm, mean hexR mutant cell length was 3μm, and mean hexR complement cell length was 17μm. (D) Representative microscopy of wild type, hexR mutant, and hexR complement strains cultured in MM9 (0.4% colloidal chitin). Cells were observed at 100x. Scale bar = 5μm.

https://doi.org/10.1371/journal.pgen.1011370.g008

HexR activity is important for regulated swimming motility but not surface swarming

V. parahaemolyticus cells are highly motile in the aquatic environment where the different cell morphologies (swimming and swarming cells) utilize different flagellar systems to promote dissemination and colonization [47]. Therefore, we initially set out to assess motility by stab inoculating each strain in soft agar allowing for quantifiable motility measurements via radial growth distance. As expected, measurements of radial growth confirmed wild type V. parahaemolyticus motility (Fig 9A and 9B). In contrast, hexR mutants had a statistically significant increase in radial growth indicative of increased cell motility. This increase in swimming motility was lost following genetic complementation with hexR. This data reveals that HexR impacts on regulated motility, and more specifically that its absence leads to increased motility.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 9. HexR contributes to regulated swimming motility but not swarming.

(A) Evaluation of swimming motility for the indicated strains using motility agar. The bacterial strains were cultured overnight in LBS, normalized to OD600 of 0.05 in 1x MM9 salts and stab inoculated into 0.2% (w/v) agar LBS plates. (B) Quantification of swimming phenotype by measuring the diameter of radial growth. Measurements were recorded after 5 hour incubation at 37°C. Statistical significance was calculated using one-way ANOVA (n=4, ****:p < 0.0001). Each dot represents a biological replicate. (C) Swarmer cell differentiation for the indicated strains was induced (see methods) and 1μL of cultures were spotted onto fresh BHI agar (4mM CaCl₂, 50μM 2,2-Bipyridil). Plates were sealed and incubated at 24°C overnight. Swarmer cell flares were imaged using CannoScan 5600F (Cannon) scanner. Images were processed on ImageJ. (D) Quantification of swarming phenotype by measuring the diameter of radial growth. Measurements were recorded after overnight growth. Statistical significance was calculated using one-way ANOVA (n=5). Each dot represents a biological replicate. (E) Visualization of swarmer cells obtained directly from the peripheral flare region of growth. Peripheral cells from wild type and hexR mutant strains each expressing a PlafK-GFP reporter were stamped onto 1% agarose pads and imaged with DIC and GFP fluorescence on Z2 Axio Imager (Zeiss). All samples were viewed at 63x. Scale bar 5μm.

https://doi.org/10.1371/journal.pgen.1011370.g009

V. parahaemolyticus surface swarming is associated with cell elongation and expression of lateral flagella in a lafK gene dependent manner [21,48]. Cells typically exhibit swarming when cultured using rich (complex) nutrients, low iron conditions, and in the presence of calcium [48]. We observed equivalent swarming (no significant difference) for both wild type and ΔhexR bacteria with characteristic growth ‘flares’ at the periphery of swarmer colony growth (Fig 9C and 9D). It should be noted that there was considerable variability in swarming radius for the 6 technical replicates for each strain ranging from 45-80 millimetres, therefore subtle differences could be masked. Nonetheless, in each case, the localized flare bacteria were expressing lafK as demonstrated by the strains harboring a transcriptional fusion to the lafK promoter (PlafK-gfp), thus producing green fluorescent cells (Fig 9E). Notably, ΔhexR bacteria were elongated similar to wild type cells under these nutrient replete conditions. For all strains tested, we were unable to stimulate a swarming phenotype under nutrient limiting conditions of growth (e.g., minimal or colloidal chitin solid media) for unknown reasons. Nonetheless, under nutrient rich conditions ΔhexR bacteria can produce swarmer cells based on elongated cell morphology, lafK promoter activity, and a swarming phenotype on the agar surface.

HexR regulates central carbon metabolism-associated genes important for cell fitness

HexR is a helix-turn-helix (HTH) global transcriptional regulator of central carbon metabolism (CCM) genes in many Proteobacteria [38, 39]. In the case of V. parahaemolyticus, bioinformatic analyses predict HexR to bind at a conserved pseudo-palindromic motif (TGTAATTAAATTACA) within promoter regions of various genes [49](S4 Data). Depending on the genomic context, HexR may act as a negative or positive regulator of gene expression [39]. HexR binding downstream of a promoter is thought to repress gene expression, whereas HexR binding upstream of a promoter is viewed to positively regulate gene expression. Therefore, we hypothesized that disruptions to CCM via the chromosomal deletion of hexR would likely contribute to gene expression differences with an outcome that impacts genetic fitness. To investigate this, we assessed gene promoter activities of two predicted V. parahaemolyticus HexR regulon members: Glucose-6-P isomerase (pgi), and Glycogen Debranching Enzyme (glgX), in both wild type and ΔhexR strains. These putative HexR regulon members were selected based on their respective high score for HexR consensus binding sequences identified by RegPrecise [49] and their linkage to central carbon metabolism Proteobacteria [39]. Real-time quantitative cell-based luciferase reporter assays were performed to measure promoter activities.

In the cases of pgi, ΔhexR bacteria demonstrated significantly higher promoter activity than wild type bacteria at the 4 and 4.5 hour time points over a time course experiment (Fig 10A). Genetic complementation using a plasmid expressing HexR repressed pgi promoter activity below that of wild type levels, suggestive of HexR being a negative regulator of pgi expression. Consistent with this view, a putative HexR binding site (TGAAAAAAAATTACA) was localized downstream of the predicted pgi promoter element (S7 Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 10. HexR directly regulates pgi and glgX expression in V. parahaemolyticus.

(A) Luciferase Reporter assay of pgi promoter region in wild type (WT), ΔhexR, and hexR genetically complemented strain. (B) Luciferase reporter assay of glgX promoter region in WT, ΔhexR and a hexR genetically complemented strain. The strains were grown overnight in LBS and inoculated at a starting OD600nm of 0.05 into LB. CPS (photon counts per second) and OD600 readings were recorded at the indicated time points. Complementation of hexR was performed using plasmid pDSK519-hexR. Statistical significance was calculated with an unpaired t-test compared to wild type (n=3, *: p<0.04,**: p<0.01, ****: p<0.0001).

https://doi.org/10.1371/journal.pgen.1011370.g010

Comparison of glgX promoter activity between wild type and ΔhexR bacteria revealed significantly reduced transcriptional activity for ΔhexR bacteria (Fig 10B). Genetic complementation with a plasmid expressing HexR resulted in glgX promoter activity at levels above that observed for wild type bacteria. These data suggest that glgX expression is positively regulated by HexR, a finding consistent with a putative HexR binding site (AGTAATTAAATTACA) being identified upstream of a the predicted glgX promoter (S7 Fig).

Collectively, the promoter activity data for the HexR regulon members pgi, and glgX, suggest that HexR is involved in regulating central carbon metabolism associated genes. This data further suggests a pivotal role for HexR mediated gene regulation in chitinolytic growth, as derived N-acetylglucosamine (GlcNAc) import and processing requires coordinately regulated enzymatic activities (Glucose-6-P isomerase, glycogen debranching enzyme) for efficient carbon assimilation leading to essential amino acid and nucleotide biogenesis.

Bacterial strains & growth conditions

Vibrio parahaemolyticus RIMD2210633 was grown in Luria Miller Broth (LB; Bioshop; LBL417.1), Luria Broth Salt (LBS; 10g tryptone, 5g yeast extract, 20g NaCl, 20mM Tris-HCl pH 8.0), or M9 Minimal Media (MM9; 420mM , 220mM , 86mM NaCl, 187mM , , , 0.4% w/v glycerol). Cultures were grown at 37°C and 200rpm unless otherwise stated. Antibiotics were used in the growth medium as required: chloramphenicol (Sigma) at 2.5 μ g/ml or 30 μ g/ml (for V. parahaemolyticus and Escherichia coli respectively), ampicillin at 100 μ g/ml (for E. coli), erythromycin at 10 μ g/ml (for V. parahaemolyticus) and neomycin at 25 μ g/ml (for E. coli). For growth on colloidal chitin, 0.4% colloidal chitin was substituted for glycerol and cultures were grown at 30°C and 250 rpm. For biofilm assays, 0.2% casamino acids (Fluka, 70171) was added to MM9. Agar (Bioshop; AGR003.1) was added to a final concentration of 1.5% (w/v) for solid media preparations unless otherwise stated. Spent Media was prepared by growing V. parahaemolyticus in LB, subculturing into LB at a starting of 0.025 for 16 hours at 37°C and 200 rpm. Cells were pelleted at 5000rpm and 4°C until supernatant was clear (30 minutes) and then liquid media was filter sterilized using Nalgene Rapid-Flow 0.22 μ m filter (Thermofisher; 564-0020). Table 1 contains information relating to strains and plasmids used in this study.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Strains and plasmids used in this study.

https://doi.org/10.1371/journal.pgen.1011370.t001

Generation of a streptomycin resistant strain of V. parahaemolyticus

Wild type RIMD2210633 was cultured overnight in LBS media without antibiotics overnight at 37°C/200RPM. The culture was sub-cultured into LBS containing 25 μg/mL of streptomycin and incubated at 37°C/200RPM until cell growth was visible, around 48 hrs. Once cell density was visible in the culture tubes, the cells were sub-cultured into LBS containing increasing concentrations of streptomycin until they reliably reached stationary phase after overnight growth at 200 μg/mL streptomycin. This culture was streak-purified and single colonies were screened for their ability to grow in LBS containing 200 μg/mL streptomycin. A colony which reliably reached stationary phase after overnight growth in media containing 200 μg/mL streptomycin was stored and named VpSm.

Genome sequencing, quality analysis, and mutation analysis

Genome sequencing of the wild type and VpSm strain were completed following phenol-chloroform extraction of genomic DNA from both strains. Following, sequencing libraries were prepared using the Illumina XT DNA Library Preparation Kit (Illumina, FC-131-1024) and sequenced on the Illumina MiSeq platform. Following collection of the raw sequencing data, sequence files were analyzed for quality by FastQC, followed by assembly of the genomes of both strains using SPAdes, with default settings. Following assembly of the genomes, the genome assembly quality was determined using QUAST, with default settings except the minimum contig length for inclusion, which was increased to 700 bp based on the genome assemblies. Finally, Snippy was used to identify mutations present in both strains compared to the V. parahaemolyticus RIMD2210633 reference genome, using default settings.

Transposition experiments

Transposition of the VpSm strain was performed using conjugation between VpSm recipients and SM10λpir E. coli pSC189-CmR donors. VpSm and SM10λpir were mixed at a 1:1 ratio normalized by OD600 and washed twice in PBS to remove overnight culture antibiotics. Cells were resuspended in LBS, plated on LBS media on an 0.45μm MLE filter (Millipore), and incubated at 37°C for 4 hours (or various times for the optimization experiments). Cells were collected in 1x MM9 salts by vortexing the filter in a 15 mL conical tube containing 1x MM9 salts, and the cells were then plated on MM9 glucose agar plates containing streptomycin (100 μg/mL) and chloramphenicol (5 μg/mL) to select only VpSm::Himar1 transposon mutants. Cells to be counted for colony-forming units were serially diluted, while transpositions to be used in a selective growth experiment for transposon sequencing were diluted to a total volume of 15 mL in 1x MM9 salts and plated over 150 MM9 glucose plates (FisherBrand, 100mm x 15mm), and incubated at 30°C for 36 hrs prior to collection.

Colloidal chitin preparation

Colloidal chitin was prepared using shrimp chitin flakes (Sigma). Briefly, 20g of shrimp chitin flakes were measured and blended into a fine powder using a blender. In a glass beaker, 100mL of concentrated HCl was added slowly to the powdered chitin and stirred gently with a glass rod every 15 min until the mixture was homogenous. The chitin-HCl mixture was added to 2L of distilled H2O, and allowed to precipitate overnight at 4°C. Following, filtration through a buchner funnel and Whatman filter paper was performed using vacuum suction. The moist chitin cake was rinsed with distilled water in the Buchner funnel until the pH was neutral. The chitin cake was aliquoted into 50 mL conical tubes and sterilized by autoclave.

Generation of insertion sequencing libraries using HTML-PCR

Following transposon mutagenesis and selection in either chitin or glucose containing minimal media, DNA was extracted from the selected population using phenol-chloroform extraction (see Genome Sequencing, Quality Analysis, and Mutation Analysis). HTML-PCR was performed as previously described with a few modifications [59]. Mainly, HTML-PCR does not require blunting of fragmented DNA ends, as previously published, prior to poly-C tailing. 10μg of DNA was fragmented using 182 Fragmentase (NEB) for 30 min followed by DNA purification using AMPure XP Beads for DNA purification (1X volume, Beckman Coulter). Poly-C tailing was performed followed by DNA cleanup as before, and PCR amplification using LG17 and LG18 oligos as previously described. The second PCR was performed using variable primers – depending on the indices required for illumina sequencing (LG503 and LG710 for the glucose sample, along with LG510, LG706 for the chitin sample) – and 1 μL of the product from PCR1 as template. Final PCR amplifications were cleaned up, visualized using agarose gel electrophoresis, and quantified using the Illumina Quant Kit (NEB). Sequencing of these libraries was performed on the Illumina NextSeq platform, using single end reads, 150 bp in length. Table 2 contains a list of primers used in this study.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Primers used in this study.

https://doi.org/10.1371/journal.pgen.1011370.t002

TRANSIT analysis of sequencing data

Following sequencing, sequencing output was analyzed for overall quality using FastQC, followed by trimming and genome mapping using TRANSIT’s tpp tool against the RIMD2210633 reference genom (NCBI Accession: NC_004603.1 and NC_004605.1), using default settings except the transposon primer window, which was increased to 40 bp from the start of the read due to differences in the HTML-PCR approach. Output from tpp was used to determine success of the transposon sequencing experiments, particularly considering Tn saturation and density of the libraries. TRANSIT was then used to perform two analyses on the data for both chromosomes: a hidden markov model analysis to identify regions of the genome that are essential for growth on either condition, and a zero-inflated negative binomial approach which compared the glucose and chitin condition data following normalization. Both of these analyses were performed using default TRANSIT settings and protein tables generated from gff3 files from the RIMD2210633 reference genome (NCBI Accession: NC_004603.1 and 183 NC_004605.1). Data from these analyses were explored using LibreOffice Calc to sort and identify genes that carried a ZINB adjusted p-value of less than 0.05, and visualized using R and the GViz package, along with modified R scripts. Circos plots were generated from TRANSIT outputs for transposon insertion counts, and HMM essential gene analysis.

Recombinant DNA and cloning techniques

DNA fragments of interest were amplified by PCR using primers (Table 2) synthesized by Integrated DNA Technologies (IDT) and V. parahaemolyticus genomic DNA as template. All recombinant plasmid DNA constructs were generated using standard techniques with restriction enzymes from New England Biolabs. Plasmid DNA constructs were verified by DNA sequencing.

Generation of Δ hexR, Δvp0802, ΔchiP chromosomal mutant strains

Gene deletion fragments for the respective alleles were generated by PCR and cloned into pRE112 following delivery into V. parahaemolyticus through tri-parental mating. Chromosomal integrants were selected on LBS containing chloramphenicol and then were subjected to sacB-mediated sucrose selection to drive allelic exchange and chromosomal deletions. Mutants with targeted gene deletions were identified by PCR. For the vp802, chiP double mutant, the vp0802 null strain was used as the parent and was then constructed with a chiP gene deletion.

Genetic complementation constructs were generated by PCR amplifying chromosomal DNA from wild type V. parahaemolyticus with gene specific primers (Table 2). In each case, the gene specific promoter region (located in the intergenic region upstream from the gene coding sequence) was included in the DNA amplicon that was eventually cloned into plasmid pVSV105 or pDSK519. The resulting plasmids express the respective cloned gene (hexR, vp0802, or chiP) under the control of its native promoter.

Generation of lafK promoter GFP fusion constructs

Primers NT540 and NT541 were used in a PCR reaction to amplify the upstream intergenic region of lafK from wild type V. parahaemolyticus RIMD2210633 to capture the lafK promoter and regulatory elements. Separately, primers NT542 and NT543 were used to amplify gfp from pEVS168. The resulting amplicons were subjected to restriction enzyme digests, lafK (Sac1/EagI) and gfp (EagI/SphI) respectively. Likewise, pVSV105 was digested with SacI and SphI. Following digestions, a triple ligation was performed with the respective digested DNA fragments, prior to transformation into DH5αλpir. The resulting pVSV105-lafK-gfp plasmid construct was delivered via tri-parental mating with E. coli EVS104 into wild type and ΔhexR V. parahaemolyticus. Transconjugants were selected on LBS supplemented with chloramphenicol and were streak purified.

Recombinant DNA constructs for luciferase reporter assays

Intergenic regions upstream of pgi and glgX were PCR amplified. The subsequent amplicons were digested with SacI and KpnI and directionally cloned into digested pVSVlux. The resulting plasmid constructs were delivered via conjugation into wild type and ΔhexR V. parahaemolyticus strains.

Transcriptional luciferase reporter assays to assess promoter activity

Wild type, ΔhexR, and genetically complemented V. parahaemolyticus strains containing pVSVlux bearing an upstream intergenic region of regulon members (namely glgX and pgi) were grown overnight in LBS with chloramphenicol. Cells were inoculated into LB at a starting OD_600nm of 0.05 at 30°C, 250rpm. Light emission (counts per second (CPS)) and OD_600nm measurements were taken in triplicate in a VictorX5 Multi Label Plate Reader.

Isolation of mini-Tn5 mutant transposants

Briefly, a conjugal tri-parental mating occurred on LBS agar at 28°C and allowed for the delivery of the plasposon pEVS168 [60] into either ΔhexR or wild type V. parahaemolyticus RIMD 2210633. The mixture was serially diluted and plated on LBS containing 10 μ g/mL erythromycin and incubated at 22°C for 36 hours to select for transposants. GFP and non-GFP producing strains for both genetic backgrounds were selected for further characterization. Marker retrieval experiments to map transposition insertion sites (disrupted genes) were conducted as previously described [61, 62].

Preparation of outer membrane fractions from V. parahaemolyticus strains

Outer membrane preparations were derived from colloidal chitin grown stationary phase bacterial cultures using a previously described method [26]. Briefly, bacterial cells were sonicated in 50 mM Tris-HCl buffer followed by the addition of N-Lauryl sarcosine to a final concentration of 0.5% (w/v) which selectively solubilizes the cytoplasmic membrane and maintains outer membrane composition [36]. The mixture was incubated at 25 °C for 30 min. Samples were then centrifuged at 5000 rpm to remove unlysed cells. The supernatant fraction was then centrifuged at 150,000 X g for 30 min. The resulting supernatant and membranous pellet were resuspended in electrophoretic sample buffer for SDS-PAGE analysis to assess protein profiles and efficiency of fractionation. Protein gels were prepared and run as previously described [63] using a Bio-Rad system.

Sample preparation for mass spectrometry and proteomics

Excised gel bands were cut into 1mm cubes and incubated in destain solution (100mM ammonium bicarbonate, 20% (v/v) acetonitrile) until clear. Destained gel pieces were dehydrated using acetonitrile and subsequently rehydrated using a solution of 1mM tris(2-carboxyethyl)phosphine and 4mM chloroacetamide and incubate for 30-minutes at room temperature (+21C). Excess solution was discarded, and gel pieces were dehydrated using acetonitrile, rehydrated with 100mM ammonium bicarbonate, and dehydrated again with acetonitrile. For digestion, gel pieces were rehydrated with a solution of 100mM ammonium bicarbonate containing 100ng of trypsin (Promega) and incubated at +37C for 16-hours in a Thermomixer (1000rpm). After digestion, peptides were extracted from the gel pieces using two incubations in 1% formic acid, and one in acetonitrile. Each extraction step was carried out in a bath sonicator for 5-minutes. Recovered peptide solutions were concentrated using a SpeedVac and reconstituted in 0.1% (v/v) trifluoroacetic acid. Reconstituted peptides were desalted prior to mass spectrometry analysis using StageTips. Packed StageTips (2-discs, Empore C18) wells were initially conditioned with 100% (v/v) methanol (100 μL) followed by equilibration with 0.1% (v/v) trifluoroacetic acid (100 μL). Loaded peptides were rinsed using 4% (v/v) methanol in 0.1% (v/v) formic acid (200 μL) and eluted using 60% (v/v) methanol in 0.1% (v/v) formic acid (100 μL). All StageTip steps were carried out using a centrifuge with spinning at 250 x g. Eluted peptides were evaporated to dryness in a SpeedVac centrifuge and reconstituted in a solution of 1% dimethylsulfoxide (v/v) in 1% (v/v) formic acid.

Data dependent mass spectrometry analysis was carried out on an Orbitrap Lumos instrument interfaced to an UltiMate 3000 liquid chromatography (LC) system. For analysis, peptide samples were directly separated by an analytical column (75µm x 15 cm 1.8µm bioZen Peptide XB-C18 beads). The gradient ramped from an initial starting condition of 5% mobile phase B (80% acetonitrile, 0.1% formic acid) to 32% B over 39.5-minutes, with sample pickup and column equilibration using an additional 15.5-minutes (45-minute total run time, mobile phase A = 0.1% formic acid in water). Data acquisition on the Orbitrap Lumos utilized a standard data-dependent acquisition scheme. Specifically, the Orbitrap Lumos was globally set to use a positive ion spray voltage of 2200V, an ion transfer tube temperature of 275°C, a default charge state of 2, advanced precursor determination activated, and an RF Lens setting of 45%. MS1 survey scans covered a mass range of 375-1400m/z at a resolution of 60,000 with an automatic gain control (AGC) target of 4e5 (100%) and the injection time set to ‘Auto’. Precursors for tandem MS/MS (MS2) analysis were selected using monoisotopic precursor selection (‘Peptide’ mode), charge state filtering (2-4z), dynamic exclusion (20 ppm low and high, 60s duration, exclude isotopes = TRUE, exclude within cycle = TRUE), and an intensity threshold of 2.5e4. MS2 scans were carried out in the Orbitrap at a resolution of 30,000 using a higher-energy collision dissociation setting of 30%, an AGC target of 2e5 (400%), injection time set to ‘Auto’, and an isolation window setting of 1.6m/z. The total allowable MS2 cycle time was set to 3s. MS1 data were acquired in profile mode, and MS2 in centroid.

DDA-MS raw data were analyzed using FragPipe software (version 22.0) with MSFragger (version 4.1) using the default workflow settings (PMID: 28394336). Specifically, raw data files were searched using MSFragger against a representative Vibrio parahaemolyticus proteome fasta database (NCBI taxonomy 223926, version 06/2024, 13,642 entries, includes contaminants) using precursor and fragment tolerances of 20 parts-per million (PPM), a fixed modification of 57.02146 at C, and variable modifications of 15.9949 at M and 42.0106 at protein N-termini. Matching spectra were rescored using MSBooster and peptide spectral match (PSM) validation utilized Percolator prior to ProteinProphet scoring to support filtering at <1% FDR at the PSM, peptide, and protein levels. Peptides were quantified using the IonQuant package built into FragPipe with the default settings. Resulting data were parsed to facilitate comparisons and final reporting using R.

Growth assays

Strains were grown overnight in LBS supplemented with chloramphenicol. Aliquots of cultures were collected, and cells were washed twice with 1x MM9 salts. Strains were normalized to a starting of 0.025 in either LB or MM9 (0.4% Glycerol) and inoculated in triplicate into a 96 well plate. measurements were taken at 37°C every 20 minutes for 18 hours in a VictorX5 Multi Label Plate Reader. Each experiment was performed with three technical and biological triplicates. For colloidal chitin growth assays, strains were grown overnight in LBS supplemented with chloramphenicol and inoculated at a starting CFU/mL of approximately cells into MM9 (0.4% colloidal chitin). Sampling was done every 6-12 hours post-inoculation and serial dilutions were plated in duplicate on LBS Agar to determine the number of cells present. For some of the colloidal chitin growth experiments, a viable but not culturable (VBNC) state may have been induced at later sampling time points (beyond 48 hours) and after a logarithmic phase of growth and when cells might experience starvation due to rapid consumption of available GlcNAc derivatives. Specifically, we observed a decrease in viable plate counts on LBS media in the absence of a decrease in cell culture density or any appearance of cell aggregates.

Spent media assay

A wild type V. parahaemolyticus strain containing pVSV105-hexR was generated via tri-parental mating using helper strain pEVS104 and then subjected to growth in Spent Media along with wild type, hexR mutant, and complement strains. Growth assays were performed as previously described with minor adjustments. Flasks of 10mL of each strain were grown for 36 hours at 37°C and 200rpm in Spent Media. OD₆₀₀ measurements were taken at every 6 hours in duplicate.

Biofilm formation assay

Biofilm assays were performed as previously described [64], with minor adjustments. Biofilms were grown in glass tubes by inoculating cells in LBS containing chloramphenicol overnight followed by subculturing overnight in MM9 at a starting OD₆₀₀ of 0.025. Liquid media was removed, and cells were washed twice in sterile phosphate buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 8.1 mM , 1.46mM ) before staining in 0.5% crystal violet for 30 minutes. Biofilms were washed twice with PBS to remove excess crystal violet from the tubes. The biofilms were de-stained with 3mL of 95% ethanol. Aliquots were collected and measurements were taken at an absorbance of 565nm in a 1 cm cuvette on a Biophotometer plus. For quantification of biofilms in LB, strains were grown in LBS in glass tubes overnight followed by a subculture into LB at a starting OD₆₀₀ of 0.025. Following consecutive washing with PBS, staining with 0.5% crystal violet, and destaining with 3mL of 95% ethanol, aliquots were diluted 1/10 in 95% ethanol prior to measuring.

Microscopy

Microscopy was performed as previously described [45,65], with minor adjustments. Overnight cultures were inoculated into LB and MM9 at a starting of 0.025 for 16 hours at 37°C, 200rpm. Strains were fixed in 2 volumes on 1% agarose pads containing 0.5% PBS. A cover slip was added, and slides were immediately imaged using a Zeiss Axio Imager Z1 microscope at 100x via Differential Interference Contrast (DIC) settings. Cell lengths were analyzed and measured using ImageJ Fiji.

For the observation of swarmer cells, swarmer cell differentiation was induced overnight and 5-10mm cubes of agar were cut out to isolate swarmer cell flares and placed cell-side down onto a 1% agarose pad on a glass slide for 30 seconds. A coverslip was added immediately, and slides were imaged at 63X-100X via DIC and fluorescence on a Z2 Axio Imager (Zeiss).

Motility assays

Swimming motility assays were performed as previously described [66]. Strains were cultured overnight in LBS, normalized to of 0.05 in 1x MM9 salt and stab inoculated into 0.2% (w/v) agar LBS plates. Measurements were recorded after 5 hours of incubation at 37°C. Swimming of each strain was documented via photograph.

Induction of swarming motility

Swarming agar was prepared as described previously [65]. Briefly, plates containing 40g/L Brain Heart Infusion (EMD) supplemented with 15g/L agar (Bioshop), 4mM CaCl₂, and 50mM 2,2’-Bipyridyl (Sigma) were prepared and dried for 10 minutes at room temperature with lids off. The same day, strains were sub-cultured in LB with appropriate antibiotics to an OD_600nm = 0.8 before being plated in 1μL spots. Plates were sealed with parafilm and incubated overnight at 24°C. Swarming of each strain was documented with an image captured using a scanner.

Competition assay

Overnight cultures of both GFP and non-GFP producing strains were grown in LBS supplemented with erythromycin at 37°C, 200rpm. The following day, 1 mL of culture was washed twice in 1x MM9 Salts and inoculated into 50 mL MM9 (0.4% Colloidal Chitin) at a starting OD_600nm of 0.025. Cultures grew at 30°C, 250rpm. Aliquots of 250uL were collected (at 0 hours, 6 hours, 24 hours, and 48 hours), immediately serially diluted, and plated onto LB. Single colonies were counted (both the total and fluorescent sub-population) to determine the proportion of fluorescent and non-fluorescent strains using Quantity One software (Bio Rad). A Competition Index (CI) was calculated using the equation:

Where subscript (f) denotes colony counts (CFU/mL) at final time point (t=48) and subscript (i) denotes colony counts of starting inoculum (t=0). Statistical significance was calculated using un-paired t-test against baseline of CI=1, where a value of 1 indicates no fitness defect.