Quorum sensing in Aliivibrio wodanis 06/09/139 and its role in controlling various phenotypic traits

Bacterial strains, plasmids and growth conditions

Bacterial cells and plasmids used in this study are listed in Table 1. The wild type A. wodanis 06/09/139 and the constructed A. wodanis mutants were grown from glycerol stocks (−80 °C) on Luria-Bertani Agar (Difco BD Diagnostics Sparks, MD, USA) plates containing 1.5% agar (Sigma-Aldrich, St. Louis, MO, USA) and supplemented with a final concentration of 2.5% NaCl (wt/vol) (LA2.5) for 3 days at 12 °C. The Pre-culture of A. wodanis was grown overnight in 2 ml of Luria-Bertani broth (LB2.5) at 12 °C and 220 rpm.

The Escherichia coli (E. coli) strains JM109 and S.17-1lambdapir were grown on LA or LB supplemented with 1% (wt/vol) NaCl (LA1 and LB1, respectively) at 37 °C and 220 rpm overnight. The TA plasmid pGEM-T was propagated in E. coli JM109 (Promega). The suicide plasmid pDM4 (GenBank: KC795686.1) was propagated in S.17-1lambdapir cells. The E.coli JM109 and S.17-1lambdapir transformants were selected on LA1 with 100 µg/ml ampicillin and 25 µg/ml chloramphenicol respectively. The potential transconjugants of A. wodanis were selected on LA2.5 supplemented with 2 µg/ml chloramphenicol at 12 °C for 5 days.

A sea water-based medium (SWT) was used in the biofilm and colony morphology assays consists of 5 g/L of Bacto peptone (BD), 3 g/L of yeast extract (Sigma-Aldrich, St. Louis, MO, USA) supplemented with different sea salt (Instant Ocean, Aquarium systems, Sarrebourg, France) concentrations (1.0, 2.8 and 4.0% per litre). For solid SWT plates, 1.5% agar (Sigma-Aldrich, St. Louis, MO, USA) was added. The hemolysin assay was carried on blood agar plates (BA) base no.2 (OXOID, Thermo Scientific) supplemented with 5% blood and 2.5% NaCl. Leibowitch-15 (L-15) medium (Thermo Fisher Scientific, USA) was used for cytotoxicity assay and supplemented with 200 mM L-glutamine, antibiotic-antimycotic solution 100 units/ml penicillin, 100 µg/ml Streptomycin, 250 ng/ml amphotericin B (P/S/A) (Sigma-Aldrich, St. Louis, MO, USA) and fetal bovine serum (FBS) (8%).

DNA extraction, PCR and sequencing

Genomic DNA was purified using Masterpure™ complete DNA/RNA purification kit (Epicentre, Cambio Ltd., Cambridge) and plasmids were purified using E.Z.N.A^® plasmid mini kit (Omega Bio-tek, Inc., Norcross, GA). The DNA concentration was measured using NanoDrop™ 2000c spectrophotometer (Thermo Scientific, DE, USA). The PCR amplification was performed using Phusion^® polymerase (Thermo Fisher Scientific, Waltham, MA, USA) or Taq polymerase (Sigma, St. Louis, MO, USA) in a Arktik™ thermal cycler (Thermo Fisher Scientific, USA). Restriction digestion using XhoI and SpeI restriction enzymes and DNA ligation using T4 DNA ligase were performed as recommended by the manufactures and were obtained from New England Biolabs (Ipswich, MA, USA). The PCR products and the digested DNA fragments were separated using agarose gel electrophoresis and extracted using Montage^® gel extraction kit (Millipore, MA, USA). DNA sequencing was performed using Big Dye (Applied biosystems, CA, USA). The primers used in the PCR and sequencing reactions were synthesized by Sigma-Aldrich and are listed in Table 1.

Construction of ΔlitR and ΔainS mutants

The litR (AWOD_I_0419) and ainS (AWOD_I_1040) genes were in-frame deleted in A. wodanis using allelic exchange as described by others (Milton et al., 1996; Bjelland et al., 2012; Hansen et al., 2015; Khider, Willassen & Hansen, 2018). Briefly, litR and ainS genes were deleted by amplifying and fusing regions flanking these genes. The upstream (280 bp) and downstream (263 bp) flanking regions of litR gene were amplified by primer pairs LitRA/LitRB and LitRC/LitRD respectively. The upstream region contains the start codon (ATG), and the downstream region contains the last three codons (TAA) at the C-terminal end of litR gene. The upstream (253 bp) and downstream (271 bp) flanking regions of ainS gene were amplified using primer pairs AinSA/AinSB, and AinSC/AinSD, respectively. The upstream PCR product ends before the start codon (ATG), and the downstream PCR product contained the last 149 bp (50 codons) of the ainS gene. PCR amplification was performed with an initial denaturation at 98 °C for 30 s (s), followed by 30 cycles of 98 °C for 10 s, 60 °C for 20 s, 72 °C for 30 s, finishing with a final extension at 72 °C for 5 min and storage at 4 °C thereafter. The upstream and downstream PCR products of either litR or ainS were fused by overlap extension PCR. This overlap PCR was performed by mixing the upstream and downstream PCR products with DNA Phusion polymerase, deoxynucleoside triphosphates (dNTPs) and buffer and cycling for seven times. Then the primer pairs LitRA/D or AinSA/D were added, and 25 more cycles were run (PCR amplification was performed similarly to the stated above). A’overhangs were added to the fused PCR products and ligated into pGEM-T Easy vector. The ligated constructs were transformed into E. coli JM109 cells. The inserts (PCR overlap products) were digested from the pGEM plasmid using SpeI and XhoI restriction enzymes as the primer pairs LitRA/D and AinSA/D contain restriction sites (SpeI and XhoI) to enable further ligation. The digested fused PCR products were further ligated into corresponding restriction sites of the suicide vector pDM4. The pDM4 plasmid with either litR or ainS fused PCR product was transformed to E.coli S.17-1lambdapir cells. The resulting plasmids were designated as pDM4ΔlitR and pDM4ΔainS.

The pDM4ΔlitR and pDM4ΔainS constructs were transferred into wild type A. wodanis by bacterial conjugation as described previously (Milton et al., 1996; Bjelland et al., 2012; Khider, Willassen & Hansen, 2018). Briefly, the donor cells E.coli S.17-1lambdapir harboring the pDM4ΔlitR or pDM4ΔainS were grown until mid-exponential phase to OD₆₀₀ (optical density at 600 nm) of 1.0 and the recipient strain (A. wodanis) to an early-exponential phase OD₆₀₀ of 2.0. One ml from each culture was centrifuged separately at room temperature and the pellets were separately washed twice with LB1 medium. The washed bacterial pellets were mixed and resuspended in 10 µl of LB1. The resuspended pellet was spotted onto LA1 plates and incubated at 19 °C for 6 h. The plates were further incubated at 12 °C for 48 h. The spotted cells were then resuspended in 2 ml LB2.5 and incubated overnight at 12 °C before plating (20, 40, 60, 80, and 100 µl) on LA2.5 plates with 2 µg/ml chloramphenicol. Potential transconjugants were selected after 3 to 5 days and confirmed using colony PCR. To complete the allelic exchange needed to generate the complete deletion mutants (ΔainS and ΔlitR) potential transconjugants (A.wodanis-pDM4-ΔainS or A.wodanis-pDM4-ΔlitR) were plated on LA2.5 plates with 5% sucrose to allow the second cross over. The cells that were able to grow were selected based on their sensitivity to chloramphenicol. The antibiotics sensitive cells were confirmed by colony PCR and further verified by sequencing.

Construction of ΔlitR complementary strain

The complementary strain (litR⁺) of the ΔlitR mutant was constructed by amplifying the full-length wild type litR gene using primers LitRA and LitRD. Briefly, the length of both the parental litR gene and its flanking region used to generate litR⁺ was 1,145 bp, which was amplified from wild type using primers LitRA and LitRD. The PCR product was then cloned into pDM4 using restriction digestion and ligation as described above. The resulting plasmid pDM4litR⁺ was transformed into E.coli S.17-1lambdapir cells and further transferred to the A. wodanis ΔlitR mutant by bacterial conjugation (described above). The selection and verification of the potential complementary strain were performed as described above. The region that flanks the complemented region (litR⁺) after allelic exchange was confirmed using primers LitRG and LitRH. The length of products amplified using LitRG and LitRH was 1,365 bp.

Preparation of bacterial supernatants for AHL measurements

The wild type A. wodanis, ΔlitR and litR⁺ were cultivated in parallel at 6 and 12 °C. The cultures were diluted to a start OD₆₀₀ of 0.001 in a total volume of 60 ml LB2.5 in a 250 ml baffled flasks. The cultures were grown further at the selected temperatures and 220 rpm. Cultures of 1 ml (wild type, ΔlitR and litR⁺) were collected at seven different cell densities in total, six at the log phase (OD₆₀₀ of 0.5, 1.0, 2.0, 3.0, 4.0, 5.0) and one at the stationary phase (8.0). For ΔainS AHL measurements, samples were only harvested at the early stationary phase (OD₆₀₀ of 6.0). The cultures were centrifuged at 13,000× g for 2 min at 4 °C (Heraeus fresco 21; Thermo Scientific, Waltham, MA, USA). Seventy-five microliters of each supernatant were acidified with 4 µl of 1M HCl and stored in three technical replicates at −20 °C before measuring the AHLs. A commercial 3-OH-C10-HSL was used as a standard (Sigma-Aldrich, St. Louis, MO, USA). The sample preparations for High-Performance Liquid Chromatography-Tandem Mass Spectrometry (HPLC-MS/MS) were done as described by others (Purohit et al., 2013; Hansen et al., 2015). Briefly, the acidified supernatants were mixed with three volumes of ethyl acetate (225 µl) and vortexed. The ethyl acetate phase of the three technical replicates was pooled together into a 1 ml 96 well plate and dried in a rotary vacuum centrifuge at −90 °C for 2 h (SpeedVac Savant™ concentrator; Thermo Scientific). The dried samples were dissolved in 150 µl of 20% acetonitrile containing 0.1% formic acid and 660 ng/ml of internal standard 3-oxo-C12-HSL (Sigma-Aldrich, St. Louis, MO, USA).

HPLC-MS/MS analysis

The HPLC-MS/MS analysis was performed as described in Purohit et al. (2013) and Hansen et al. (2015). Briefly HPLC-MS/MS was performed using an Ascentis Express C18 reversed-phase column (50 × 2.1 mm, 2.7 µm particle size; Sigma). A sample of 20 µl was injected into the column and eluted using 0.1% formic acid in water and 0.1% formic acid in acetonitrile at a flow rate of 200 µl/min. The elution profile obtained was 5% acetonitrile in 30 s, 90% in 300 s and 5% in the next 60 s. The separated compounds were detected by Linear Ion Trap Quadrapole (LTQ) part of the LTQ-Orbitrap (Thermo Fisher Scientific). The LTQ was used in selected reaction monitoring (SRM) mode, and the SRM was divided into two segments. Segment 1 scanned 3OHC10-HSL and segment 2 scanned the internal standard 3O-C12-HSL with a retention time of 0–3.15 min and 3.15–6.00 min, respectively. The ion trap parameters chosen for MS/MS were maximum injection time 50 ms, isowidth 1.0 m/z, collision energy 35, act Q 0.25 and act time 30 ms. The measured AHLs are presented in ng/ml/OD₆₀₀. The AHL measurements at different temperatures were performed twice.

Motility assay

Motility assay was performed in LA2.5 soft agar plates with 0.25% agar (Bjelland et al., 2012; Khider, Willassen & Hansen, 2018). Pre-cultures of A. wodanis wild type, ΔainS, ΔlitR and litR⁺ were diluted 1:100 in LB2.5 and grown overnight at 12 °C to an OD₆₀₀ corresponding to 1.0. Then 2 µl of each culture was spotted onto the LA2.5 soft agar plates and incubated at 6 °C and 12 °C. The diameter of motility zones were measured in millimeters every 24 h for 5 days.

Siderophore, hemolysis, chitinase and protease assays

Siderophore production was screened using Chromeazurol S (CAS) agar plates as described by others (Lauzon et al., 2008), with the exception that 2.5% of NaCl was used in this study. Culture supernatants harvested at OD₆₀₀ 6.0 (100 µl) from strains grown at 6 and 12 °C were added to six mm wells casted in CAS agar. The CAS agar plates were incubated at 20 °C for 2 days. Hemolysin production was estimated by spotting 2 µl cultures of each strain on BA plates. The protease production was estimated by spotting 2 µl cultures of each strain on LA2.5 agar plates supplemented with 2% skim milk (Sigma-Aldrich, St. Louis, MO, USA). The chitinase assay was performed by spotting 2 µl cultures of each strain on LA2.5 supplemented with 2% colloidal chitin (Sigma-Aldrich, St. Louis, MO, USA) and the plates were stained with 0.5% congo red (Sigma-Aldrich, St. Louis, MO, USA) for 30 min before destaining with 1 M NaCl for 20 min for chitinase zones measurement. All the assays were performed with at least three biological replicates of strains A. wodanis wild type, ΔainS, ΔlitR and litR⁺ and were incubated at 6 and 12 °C. The clear zones ratio values for hemolysis, protease and chitinase assays were calculated as clear zone diameter/colony diameter.

Biofilm and colony morphology assays

The biofilm and colony morphology assays were performed as described previously (Hansen et al., 2014) using SWT media and plates, respectively. Pre-cultures of A. wodanis wild type, ΔainS, and ΔlitR were diluted 1:100 in LB2.5 and grown overnight at 12 °C to an OD₆₀₀ of 1.0. For the biofilm assay, the cultures were further diluted 1:10 in SWT media, and a total volume of 300 µl was added to Falcon 24 well plates (BD Biosciences) and incubated statically at 6 °C. The plates were monitored every 24 h. For the colony morphology assay, a 250 µl of each bacterial culture was harvested by centrifugation, and the pellet was resuspended in 250 µl SWT. Then, 2 µl of each culture was spotted onto SWT plates and incubated at 6 °C for up to 4 weeks. The biofilm formation and colony morphology was visualized using Ziess Primovert microscope at 10x and 4x magnification, respectively and were photographed with AxioCam ERc5s.

Cytotoxicity assay and crystal violet staining

Chinook salmon embryo (CHSE-214) cells were purchased from American Type Culture Collection (Nicholson & Byrne, 1973). Chinook salmon embryonic (CHSE) cells (passage 55) were grown in L-15. The CHSE cells were seeded 1 × 10⁵ cells/ml in a flat-bottom tissue culture 24-well plates (Falcon; BD Biosciences) and incubated for 48 h at 20 °C. Supernatants of A. wodanis, ΔainS, ΔlitR and litR⁺ grown at 6 °C and 12 °C were harvested at OD₆₀₀ of 6.0, 7.0 and 8.0 and filter sterilized through 0.22 µm filter. The 100% confluent fish cells were washed with L-15 without supplements and treated with bacterial supernatants (1:10 to L-15 with antibiotics) before incubating at 12 °C. LB2.5 was used as a negative control. The plates were monitored after 4 and 24 h. The treated CHSE cells were quantified using crystal violet staining. Briefly, the wells with treated CHSE cells were stained with 500 µL of 0.1% crystal violet for 20 min before washing with water. The plates were air-dried for 1 or 2 days, and the cells were dissolved in 96% ethanol before measuring the absorbance at 590 nm (100 µL) using a spectrophotometer (Spectromax, Molecular devices).

All assays were carried out in biological triplicates, unless otherwise indicated. The assays were also performed in two to three independent experiments to validate the results.

The AinS autoinducer synthase is responsible for the production of 3OHC10-HSL

In previous work, we mapped AHL profiles among members of the Vibrionaceae family. Only one single AHL, the 3OHC10-HSL, was identified in A. wodanis 06/09/139 (Purohit et al., 2013). In A. salmonicida, the autoinducer synthase AinS was responsible for 3OHC10-HSL synthesis (Hansen et al., 2015). Thus our first aim was to verify if 3OHC10-HSL is produced by AinS. To this end, supernatants harvested from the wild type and mutants (ΔainS, ΔlitR and litR⁺) were analyzed by HPLC-MS/MS as previously described (Purohit et al., 2013; Hansen et al., 2015). A peak corresponding to 3OHC10-HSL was present only in supernatants harvested from the wild type, ΔlitR mutant and the litR⁺ (Fig. 1). This suggests that AinS is the autoinducer synthase responsible for 3OHC10-HSL production in A. wodanis 06/09/139. Unfortunately, we were not able to complement the ainS mutant which could have given absolute proof for AinS being the 3OHC10-HSL synthase in A. wodanis. Since ainS is the only AHL-linked gene annotated in A. wodanis we find other explanations unlikely.

LitR represses growth of A. wodanis at 20 °C

Mutations are known to affect the growth of bacteria. A. wodanis strains grow in a range of 4–25 °C and in a recent study in our lab, the optimal growth temperatures were found to be 12–18 °C (Lunder et al., 1995; Lunder et al., 2000; Soderberg et al., 2019). We, therefore tested the strains ability to grow at three different temperatures (6 °C, 12 °C and 20  °C) within the reported temperature range for A. wodanis. As shown in Fig. 2, the litR or ainS mutations did not alter the growth rate of A. wodanis at 6 °C and 12 °C, and all strains reached a maximum OD₆₀₀ of ∼8.0. The strains grew considerably faster at 12 °C than at 6  °C, and the duration of the log/exponential phase (OD₆₀₀ 0.5 to 8.0) for the wild type lasted for 22 h when grown at 12 °C compared to 45 h at 6 °C (Fig. 2). At 20  °C, which is a non-optimal temperature for growth of A. wodanis in the laboratory (Soderberg et al., 2019), the wild type and ΔainS mutant showed a growth deficiency, and the growth halted after reaching an OD₆₀₀  of 2.0–3.0 before it finally reached maximum OD₆₀₀ of 5.0–6.0. On the other hand, the ΔlitR mutant grew steadily and was able to reach an OD₆₀₀ of 8.0. Neither of the strains grew in liquid media (LB2.5) at 25  °C, and after streaking single colonies of the different strains onto blood agar plates (BA2.5) only the ΔlitR mutant was able to form small colonies when incubated at 25 °C (Fig. S2). Thus, LitR represses the ability of A. wodanis to grow at 20 °C and 25 °C.

3OHC10-HSL production in A. wodanis is cell density and temperature-dependent, and weakly regulated by LitR

To explore the role of temperature on the production of 3OHC10-HSL in A. wodanis, we analyzed supernatants harvested from the wild type 06/09/139 at different cell densities after growth at 6 and 12 °C. The HPLC-MS/MS analyses showed that the 3OHC10-HSL production was detectable from the measurements started at OD₆₀₀ of 0.5 and increased along the growth curve in a cell density-dependent manner. The bacterium produced higher concentrations of 3OHC10-HSL when it was grown at 6 °C compared to at 12 °C (P < 0.05, by Students t test). Highest 3OHC10-HSL concentrations were measured in the stationary phase (OD₆₀₀ of 8.0) where the wild type reached concentrations of 21.06 ± 0.43 ng/ml/OD₆₀₀ and 15.12 ± 0.94 ng/ml/OD₆₀₀ after growth at 6  °C and 12 °C, respectively (Table S1, Fig. 3).

We also analyzed supernatants harvested from the ΔlitR mutant to examine if LitR is a regulator of AHL production in A. wodanis, similar to what has been shown for other aliivibrios (Lupp & Ruby, 2004; Hansen et al., 2015). As shown in Fig. 3, the ΔlitR mutant produced lower concentrations of 3OHC10-HSL than the wild type did in the stationary phase, and the deletion of litR led to an 24% reduction when the maximum concentrations (measured at OD₆₀₀ = 8.0) of the wild type and ΔlitR were compared at 6 °C (WT = 21.06 ± 0.43 ng/ml/OD₆₀₀ and ΔlitR = 16.01 ± 0.96 ng/ml/OD₆₀₀; P < 0.05 by Students t test) and 22% reduction after growth at 12 °C (WT = 15.12 ± 0.94 ng/ml/OD₆₀₀ and ΔlitR = 11.78 ± 0.94 ng/ml/OD₆₀₀; P < 0.05 by Students t test). The complementary mutant litR⁺ behaved as the wild type with regard to 3OHC10-HSL production.

Phenotypic traits regulated by QS in A. wodanis06/09/139

QS is known to regulate several activities or phenotypic traits in vibrios and allivibrios (Croxatto et al., 2002; Zhu et al., 2002; Lee et al., 2004; Tsou & Zhu, 2010; Bjelland et al., 2012; Khider et al., 2019). Therefore, we analyzed the wild type A. wodanis 06/09/139 and QS mutants (ΔainS and ΔlitR) with regard to motility, protease and siderophore production, hemolysis, chitinase activity, biofilm formation and colony morphology. The experiments were performed at 6  °C and 12 °C to determine if the temperature has an influence on the phenotypic traits exhibited by the wild type, ΔainS and ΔlitR mutants.

Motility

The motility assay showed that the wild type A. wodanis 06/09/139 was motile at both 6 °C and 12  °C. The motility of wild type A. wodanis was 57% higher at 12 °C compared to at 6  °C (12 °C =42.17 ± 3.19 mm and 6 °C =18.00 ± 0.89 mm; P < 0.05 by Students t test) (Fig. 4A). The ΔainS and ΔlitR mutants showed significantly higher motility than the wild type at both temperatures. Compared to wild type, the ΔlitR mutant showed 27% larger motility zones both at 6 °C (WT = 18.00 ± 0.89 mm and ΔlitR = 24.58 ± 1.74 mm; P < 0.05 by Students t test) and at 12 °C (WT = 42.17 ± 3.19 mm and ΔlitR = 57.67 ± 1.97 mm; P < 0.05 by Students t test). Similarly, the ΔainS mutant showed 17% larger motility zones at 6  °C (WT = 18.00 ± 0.89 mm and ΔainS = 21.67 ± 1.51 mm; P < 0.05 by Students t test) and 26% larger zones at 12  °C (WT = 42.17 ± 3.19 mm and ΔainS = 57.17 ± 3.87 mm; P < 0.05 by Students t test) (Fig. 4A, Table S2).

Cytopathogenic effect on Chinook salmon embryonic cells (CHSE)

To test whether QS affects the cytopathogenic effect (CPE) of A. wodanis 06/09/139, CHSE cells were treated with supernatants harvested at different cell densities (OD₆₀₀ of 6.0, 7.0 and 8.0) from the wild type and mutants (ΔainS and ΔlitR) after growth at 6 and 12 °C. The CPE was observed microscopically (Fig. 5A), and cells that survived and remained attached to the substratum were thereafter quantified using a crystal violet staining method (Figs. 5B and 5C). Supernatants harvested from the wild type had a CPE on the CHSE cells similar to what has been described earlier (Karlsen et al., 2014). The temperature at which the bacterium was grown and the time of harvest (cell density) determined the severity of CPE. After growth at 6 °C, wild type supernatants harvested at OD₆₀₀ of 6.0 showed highest CPE with complete lysis of the cells (Figs. 5A and 5B). Wild type supernatants harvested at OD₆₀₀ of 6.0 after growth at 12 °C were less cytotoxic to the cells (Fig. 5A), but the CPE increased with increasing cell density, and after treatment with supernatants harvested at OD₆₀₀ of 8.0 the cells suffered from severe CPE and few cells remained viable and attached to the substratum (Fig. 5C).

Compared to the negative control, some CPE was observed for cells treated with supernatants harvested from the ΔlitR mutant, but most cells remained intact without losing the cell to cell contact (Fig. 5A). On the other hand, supernatants harvested at OD ₆₀₀ of 6.0 from the ΔainS mutant grown at 6 °C induced severe CPE and few cells survived (Figs. 5A and 5B). However, after growth at 12 °C the supernatants harvested from the ΔainS and ΔlitR induced similar CPE and were less cytotoxic than the corresponding supernatants harvested from the wild type. Hence, in particular, QS and LitR play a role in regulation of CPE towards CHSE cells in A. wodanis, and this regulatory role is somewhat stronger at 6 °C compared to at 12 °C (Fig. 5, Table S3).