Characterization of a biofilm-forming Shigella flexneri phenotype due to deficiency in Hep biosynthesis

Bacterial strains and plasmids

Bacterial strains and plasmids used in this study are listed in Table 1. Shigella strains were cultured aerobically at 37°C in Tryptic Soy (TS) broth (Aoboxing, Beijing, China) or on TS agar plates with 0.1% Congo Red. Antibiotics (Sigma) were used as follows: ampicillin 100 µg/ml; kanamycin 100 µg/ml.

Strain construction

Bacterial gene knockout was performed using the λ Red recombination system (Datsenko & Wanner, 2000). Briefly, bacterial cells transformed with pKD46 were grown in the presence of L-arabinose to induce the expression of the lambda Red recombinase. A linear PCR product, amplified using the primers listed in Table 2, containing a kanamycin-resistance cassette (KRC) flanked by FLP and 50 bp of the 5′- and 3′-end homologous sequences of the target gene was electroporated into the bacterial cells and kanamycin was used to select the transformants. The plasmid pKD46 was eliminated by incubation at 37°C. To cure the kanamycin maker, pCP20 was introduced into kanamycin-resistant cells to elicit the recombination of flanking FLP sequences at both ends of the kanamycin cassettes. PCR screening for cured colonies were performed using specific primers listed in Table 2.

LPS preparation and electrophoresis

LPS was prepared from Shigella strains as described by Hitchcock & Brown (1983). Bacteria grown overnight on TSA were resuspended in PBS to an OD600 of 0.8. The bacterial pellet from 1.5 ml of the suspension was then resuspended in 125 ml lysis buffer (0.1 M Tris/HCl pH 6.8, 2% SDS, 4% β-mercaptoethanol and 10%, v/v, glycerol) and boiled for 10 min. Proteinase K (50 mg, Sigma) was added and the mixture incubated for 1 h at 60°C. Samples were run on a Tris-Tricine gel) and visualized by silver staining (Hitchcock & Brown, 1983).

Autoaggregation assay

Overnight-cultured bacteria were harvested by centrifugation (10,000× g for 2 min) and two ml of whole cells standardized at OD₆₀₀ = 1 after suspension in PBS were placed in a 14-ml polyethylene tube and incubated at 4°C under a static condition. The OD₆₀₀ of the phase above the sediment by aggregation was recorded at different time points.

Biofilm formation assays

Biofilm formation by Shigella strains was assayed as previously described with some modifications. For biofilm analysis on polystyrene surface, 10⁷ CFU of Shigella in 100 µl of TSB broth was inoculated into the wells of a 96-well flat-bottom polystyrene microtiter plate. The bacterial strains were grown at 37°C for 48 h under a static condition and the planktonic cells in liquid medium were discarded. The plate or tube was washed twice with distilled water and air-dried. Attached biofilms were stained with 0.1% crystal violet for 20 min. Then, the plates were rinsed twice with distilled water to remove excess stain and air-dried. In order to quantify the amount of biofilm on a 96-well plate, all stain associated with the attached biofilms was dissolved with 95% ethanol, then OD₅₉₅ absorbance was measured using a microplate reader.

To prepare the biofilm sample for Confocal laser scanning microscopy (CLSM) analysis, GFP-expressing plasmid was electroporated into Shigella strains. 10⁸ CFU of GFP-Shigella was added in 1 ml of TSB broth per well and incubated at 37°C sunder a static condition. Biofilms were grown on cover glass (φ = 14 mm) placed in 24-well polystyrene cell culture plates. After 48 h, the cover glass was rinsed twice with PBS to remove any planktonic cells. After washing, the cells were fixed in 3% paraformaldehyde/PBS and mounted with Anti-Fade solution (Invitrogen) containing DAPI onto glass slides. After the preparation, the samples were examined under the confocal laser scanning microscope ZEISS LSM 710 (Carl-Zeiss), and images were processed by LSM software ZEN (Carl-Zeiss, Oberkochen, Germany).

To prepare the biofilm sample for scanning electron microscopy (SEM), the biofilms grown in the cover glass dehydrated in graded ethanol, critical point dried with CO2 and coated with gold-palladium beads with a diameter of 15 nm. Samples were photographed using a Philips XL-30 scanning electron microscope at 20 kV.

In vitro adhesion and invasion assays and microscopy

One day before the assays, HeLa cells (ATCC CCL-2) were seeded into 24-well plates at a density of ∼10⁵ cells per well. One hour before the infection, cell culture medium were changed into serum-free medium and ∼10⁶ CFU Sf301 or its LPS mutants from mid-exponential phase was added to the cells together. Bacteria were centrifuged (2,000 rpm, 10 min, RT) onto HeLa cells (moi 10:1, or indicated moi) to synchronize the infection. For adhesion assay, after washing, the cells were lysed with distilled water and the CFU was enumerated after plating. For invasion and proliferation assays, bacteria/HeLa mixtures were incubated for 40 min after centrifugation and then washed, treated with gentamycin-containing (25 µg/ml) medium for another 1 h (invasion) or 4 h (proliferation) before lysed for plating. Adhesion was defined as the total number of HeLa cell-associated bacteria and is shown as the percentage of input. Invasion and proliferation was defined as the total number of intracellular bacteria in cells (extracellular bacteria was killed by gentamycin, a cell-impermeable antibiotic). Average results of three independent experiments are shown as mean ± SD.

For fluorescent microscopy of bacterial adhesion to HeLa cells, cells were plated onto glass coverslips and adhesion assays were performed as described using Sf301 harboring the GFP_UV-expressing plasmid (moi 100:1). Cells were fixed in 3% paraformaldehyde/PBS at room temperature for 15 min, washed in PBS and mounted with Anti-Fade solution (Invitrogen) containing DAPI onto glass slides and visualized under Zeiss confocal microscope. For scanning electron microscopy (SEM), the cover glass was dehydrated in graded ethanol, critical point dried with CO2 and coated with gold-palladium beads with a diameter of 15 nm. Samples were photographed using a Philips XL-30 scanning electron microscope at 20 kV.

Lactate dehydrogenase (LDH) activity assay

The LDH assay was performed using the LDH cytotoxicity assay detection kit (Beyotime, China), according to the manufacturer’s instructions. The assay measures the conversion of a tetrazolium salt to a red formazan product, detectable by absorbance measurement at 490 nm. For this, a SpectraMax 190 Microplate Reader (Molecular Devices) was used.

Phalloidin staining assay

For phalloidin staining of F-actin in Shigella-infected HeLa cells, cells were plated onto glass coverslips and invasion assay was performed as described using Sf301 harboring the GFP_UV-expressing plasmid (moi 100:1) for two hours. Cells were fixed in 3% paraformaldehyde/PBS at room temperature for 15 min, washed in PBS and permeabilized in 0.1% Triton-X100 in PBS for 1 min. F-actin were stained with 80nM TRITC-phalloidin (Yeasen, Shanghai, China) for 30 min. The coverslips was mounted with Anti-Fade solution (Invitrogen) containing DAPI onto glass slides and visualized under Zeiss confocal microscope.

Plaque assay

This was carried out according to Oaks, Wingfield & Formal (1985) using confluent HeLa cell monolayers. Briefly, HeLa cells to be used in the plaque assay were grown to 100% confluency in 6-well polystyrene plates (Corning) in DMEM supplemented with 10% FBS. One hour before the infection, cell culture medium were changed into fresh medium and Sf301 or its LPS mutants from mid-exponential phase (m.o.i 100:1) was added to the cells and subsequently incubated at 37°C for 90 min. During this adsorption or attachment phase, the plates were rocked every 30 min to assure uniform distribution of the plaque-forming bacteria. Next, an agarose overlay (5 ml) consisting of DMEM, 5%FBS, 25 µg of gentamicin per ml, and 0.5% agarose was added to each plate. The plates were incubated at 37°C in a humidified 5% CO₂ and examined daily for up to 3 days for plaque formation. The agar was carefully removed three days later and the HeLa monolayers were stained by Giemsa staining kit (Yeasen, Shanghai, China) in order to visualize the plaques.

Sereny test

Female Hartley guinea pigs, aged 6–8 weeks, weighing 200–300 g, were inoculated with 10⁷ CFU/eye of mid-log phase Sf301 and its LPS mutants via conjunctival route as described, with 3 animals in each group. The protocol has been approved by the Animal Research Ethical Committee of School of Life Science and Technology the Xi’an Jiaotong University (Approval No. 201411). Inoculated animals were observed and scored for 7 consecutive days for development of the conjunctivitis. Eyes were blindly scored by three individuals (DX, YPS, YC) on a scale of 0–3 defined as follows: grade 0 (no disease or mild irritation), grade 1 (mild conjunctivitis or late development and/or rapid clearing of symptoms), grade 2 (keratoconjunctivitis without purulence), grade 3 (fully developed keratoconjunctivitis with purulence).

Guinea pigs were anesthetized using chloral hydrate (10 mg/kg of body weight) before being inoculated via an intrarectal (i.r.) route with 10⁸ CFU of Sf301 and its LPS mutants in 100 µl PBS, with 3 animals in each group. The protocol has been approved by the Animal Research Ethical Committee of the Xi’an Jiaotong University. Three pieces of feces of each inoculated animal were collected at the same time of the day for 18 days. Each piece of the feces were dispersed and 1 ml PBS. After centrifuge, 100 µl of the supernatant was plated onto MaConkey plates. Shigella bacteria were recognized as smooth, white colonies on MaConkey plates and further confirmed by PCR of spa33 gene.

Statistical analysis

All data are presented as means ± SEMs. Statistical analysis was performed within the subgroups by using Student’s t-test for autoaggregation assays, biofilm formation assays, in vitro adhesion/invasion assays and LDH assays in this study. p values of less than 0.05 were regarded as statistically significant. Statistical analysis was performed with GraphPad Instat 3 software (GraphPad Software, Inc, La Jolla, Calif).

Generation and characterization of the LPS mutants of Shigella flexneri

To generate Shigella LPS mutants of different chain length, we knocked out wzy, waaL and rfaC individually using the lambda red system (Fig. S1). As shown in Fig. 1A, the wzy mutant contains a complete core and one copy of the O-antigen; the waaL mutant has a complete core but no O-antigen; the rfaC mutant is deficient of Hep but retains the Kdo. SDS-PAGE analysis confirmed that the all mutant strains produced proper LPS variants as expected (Fig. 1A). The Δwzy and ΔwaaL strains formed WT-like colonies on Tryptone Soya Agar, while the ΔrfaC strain grew round and smooth colonies with smaller size. In accordance with this, the ΔrfaC mutant grew slightly slower in liquid tryptone soya broth medium than other strains (Fig. 1B). LPS truncation had no apparent impact on the morphology of individual bacteria cells as illustrated by transmission electron microscopy (TEM) (Fig. 1C).

Biofilm formation is enhanced in the rfaC-deleted Shigella strain

The aggregation tendency of LPS mutants was initially evaluated in PBS and the results showed that LPS truncation led to enhanced bacterial aggregation (Fig. 2A). Next, we examined the biofilm formation of the LPS mutants on the polystyrene surface under a static culture condition. While the WT strain had a poor ability to form biofilm, the LPS mutants showed improved potential to do so (Fig. 2B). Moreover, the ability to form biofilm appeared to be negatively correlated with the LPS length with the ΔrfaC strain (shortest LPS) being the most effective one (Fig. 2B). Fluorescence microscopy and scanning electron microscopy (SEM) further confirmed the enhanced biofilm formation by the ΔrfaC strain on glass surface (Figs. 2C–2D). To test the involvement of extracellular DNA in the biofilm formation by the ΔrfaC strain, DNase I was added when the bacteria were seeded into the polystyrene plates. Disruption of extracellular DNA significantly abolished the formation of biofilm without affecting the viability of the bacteria (Fig. 2E and Fig. S2). In addition, SEM analysis demonstrated that there were fibrous connections among bacterial cells in the biofilm (Fig. 2F).

In vitro adhesiveness and thus invasiveness is promoted in biofilm-forming rfaC-deleted Shigella flexneri strain independent of activation of type III secretion system

Since biofilm formation is often associated with bacterial virulence, we set out to evaluate the pathogenic activities of the biofilm-forming LPS mutants using the gentamicin protection assay. The Δwzy and ΔwaaL strains showed almost identical host invasion efficiency to the WT strain. By contrast, the ΔrfaC strain showed a much stronger invasiveness and intracellular proliferation than wild type (Fig. 3A). Further dissection of time-dependent infection process revealed that the improved invasiveness came from enhanced adhesion during the initial host cell-Shigella contact (within 10 min), which was further confirmed by fluorescence microscopy and scanning electron microscopy (Figs. 3B–3D). To be specific, ΔrfaC mutant exhibited pronounced adhesiveness even when the type III secretion system (T3SS), responsible for bacterial invasion and virulence, was inactivated transcriptionally at low growth temperature (28–30°C). Of note, the invasion of low temperature-cultured bacteria was totally abolished as expected since no intracellular bacteria were recovered. To verify the cell-lysing ability of ΔrfaC mutant, we assayed the level of LDH in the cell culture medium 90 min post-infection. LDH is a stable cytoplasmic enzyme that is only released on loss of membrane integrity. The LDH level in the supernatant of ΔrfaC-infected HeLa cells was more than twice that of the M90TS-infected cell supernatant (Fig. 3E).

rfaC-deleted strain showed different actin-based motility from other LPS-truncated strains

To examine the distribution of IcsA in the LPS-truncated mutants, intracellular F-actin of the infected host cells were visualized by fluorescent phalloidin, which binds to F-actin of mammalian cells. Although F-actins could be recruited by all the Shigella strains except the icsA-deleted strain (Fig. 4A and Fig. S3), their cellular distribution varies between WT and LPS mutants. Typical “actin comets” and long protrusion were observed with the wild type strain (Fig. 4A), indicating a normal polar distribution of IcsA, which was further confirmed by the competence to form regular plaques on HeLa cell monolayer in a plaque assay (Fig. 4B). By contrast, the Δwzy and ΔwaaL mutants assembled F-actins all around the bacteria cell (Fig. S3), implying a circumferential distribution of IcsA, which caused aberrant actin-based motility (ABM) and thus failure to form plaques on HeLa cells monolayer (Fig. 4B). The ΔrfaC bacteria also recruited F-actin in a circumferential pattern as other LPS-truncated strains do. Nevertheless, a small amount of bacteria could assemble short and curly “actin-tail” behind the bacteria (as arrows indicated in Fig. 4A). Although ΔrfaC mutant could not form classical plaques on HeLa cell monolayer in 3 days, the integrity of the monolayer were destroyed due to massive cell detachment (Fig. 4B).

Truncation of LPS rendered bacteria susceptible in vitro and in vivo

LPS, especially long-chain LPS of gram-negative bacteria, provides protection for the bacteria against unfavorable environment. While loss of LPS may contribute to biofilm formation and pathogenic activities such as adhesion and invasion, the adverse effects associated with LPS deletion must also be taken into consideration to accurately gauge the overall influences on bacterial fitness. To do this, we measured the susceptibility of the LPS mutants to distilled water (to mimic the low osmotic shock in environment) and 10% pooled human serum (to mimic the adverse environment in vivo). While WT bacteria well survived the low osmotic shock (water) during the experiment, all three LPS mutants showed severely compromised viabilities after two hours in water. The ΔrfaC mutant in particular, even begun to exhibit significantly reduced viability within 30 min (Fig. 5A). Human serum showed much stronger killing activity than water toward Δwzy, ΔwaaL and ΔrfaC strains (Table 3). It is interesting to note that although most of the bacteria were killed by the serum, a small number of colonies were recovered for the ΔrfaC strain versus no colony for Δwzy and ΔwaaL strains (Fig. 5B). Whether this improved survivability of the ΔrfaC strain against serum is related to its biofilm-forming ability warrants further investigation.

Given the dramatic virulence-boosting phenotype of the ΔrfaC mutant in vitro, we next examined the in vivo effects of LPS mutants on virulence using two independent animal models. The Sereny test (guinea pig keratoconjunctivitis) indicated that Δwzy, ΔwaaL and Δrfac strains failed to cause any manifestation of conjunctivitis during a period of 2 weeks’ infection. Wild type, on the other hand, provoked a typical keratoconjunctivitis 72 h post-infection (Table 4). To test the colonizing ability of the “super-adhesive” ΔrfaC mutant in the intestine, guniea pigs were inoculated with 10⁸ CFU of Shigella via intrarectal route. Feces analysis revealed that Δwzy, ΔwaaL and ΔrfaC failed to colonize the intestines of guinea pigs, and bacteria were evacuated in three days without any exhibited symptoms. Nevertheless, animals inoculated with wild type stably excreted bacteria in 18 tested days, albeit without observed symptoms either (Fig. 5B).