Deletion of two-component system QseBC weakened virulence of Glaesserella parasuis in a murine acute infection model and adhesion to host cells

Animals and ethics statement

Female BALB/c mice (6-week-old) were purchased from Chengdu Dossy Experimental Animal Co., Ltd (Sichuan, China). The animal experiments were conducted in strict accordance with the recommendations in the China Regulations for the Administration of Affairs Concerning Experimental Animals (1988) and had been approved by the Institutional Animal Care and Use Committee of Southwest Medical University (approval number 20211017-001), Luzhou, China. We did our best to provide maximum comfort and minimal stress for the mice.

Bacteria strains, plasmids and culture conditions

Bacteria strains and plasmids employed in this study are listed in Table 1. Escherichia coli DH5α (Biomed, Beijing, China) was cultured in liquid Lysogeny Broth (LB, Difco, Thermo Fisher, USA) medium or on LB agar (Invitrogen, Shanghai, China) plates. Glaesserella parasuis strain SC1401 and its derivatives were grown in tryptic soy broth (TSB; Difco, NJ, USA) or on tryptic soy agar (TSA; Difco, Thermo Fisher, USA) supplemented with 0.001% nicotinamide adenine dinucleotide (NAD; Sigma-Aldrich, St Louis, MO) and 5% inactivated bovine serum (Solarbio, Beijing, China). When required, the medium was supplemented with kanamycin (kan; 100 µg/mL) or gentamicin (50 µg/mL).

Construction and complementation of the qseBC mutant

The ΔqseBC strain and genetically complemented strains were constructed by natural transformation. Naturally competent bacteria are able to take up DNA fragments from the environment into their cytoplasm, and these foreign DNA fragments provide the bacteria with a source of nucleic acid (Patrick & Melanie, 2013). This transformation method is the third parallel gene transfer (HGT) mode, which is different from phage transduction and combined transfer, and is called natural transformation.

To construct an in-frame nonpolar qseBC mutant strain of G. parasuis, the target vector pLQ4 was built as follows: the upstream (958 bp) and downstream (947 bp) fragments of the in-situ qseB and qseC genes were amplified using primers P1+P2 (qseBC-Up-F/R, Table 2) and P3+P4 from the genomic DNA of SC1401. The qseB and qseC genes have a 14 bp overlap in G. parasuis. The kanamycin resistance cassette (935 bp) was amplified from pKD4 using primers P7+P8 (Kan-F/R). The three PCR fragments were integrated by overlap PCR with primers P1+P4 (qseBC-Up-F/qseBC-Down-R). The fusion segment was inserted into pk18mobsacB at the BamHI and HindIII sites to generate the recombinant plasmid pLQ4. Finally, the resulting plasmid pLQ4 was transformed into SC1401 using a natural transformation technique. Transformant bacteria were incubated at 37 °C for 36 h. The kanamycin-resistant transformants were identified by PCR (using P7+P8 for the presence of Kan cassette, and P9+P10 for qseBC deletion). The target vector pLQ4 was built as follows: the fragment was ligated into a linearized vector PSF116, which was digested by Kpn I and BamHI using a BM seamless cloning kit (Biomed, Beijing, China). The pLQ5 was introduced into ΔqseBC by natural transformation. The complementary strain C-ΔqseBC was selected on TSA supplemented with 20 µg/mL gentamicin. The gentamicin-resistant transformants were identified by PCR. The flow chart for construction of the ΔqseBC mutant is shown in the attachment (Fig. S1).

To rule out the possible polar effects resulting from the deletion of qseB and qseC genes, relative quantification of 2 (−ΔΔC(T)) method was adopted to analyze the transcriptional levels of the flanking genes by quantitative real-time PCR (qRT-PCR) using primer sets P19/P20 for A4U84_RS03675 (hypothetical protein) and P21/22 for groES (co-chaperone), respectively. The stably transcriptional 16S RNA of G. parasuis was used as an internal reference gene. qRT-PCR was performed by using an iTaqTM universal SYBR Green Supermix (Bio-Rad, Hercules, CA) in a Lightcycler 96 (Roche, Switzerland) system.

Preparation of 3D4/21 cell culture and adhesion and invasion assays

The study was performed as described in previous papers with some modifications (Ji et al., 2017; Xu et al., 2015; Zhou et al., 2016). The colony count method was used to study the adhesion and invasion ability of the ΔqseBC mutant strain to host cells. Porcine alveolar macrophages (PAM) 3D4/21 were purchased from Shanghai Zeye Biotechnology Co., Ltd. The 3D4/21 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS), 100 U/mL of penicillin G (Invitrogen, Waltham, MA), 100 mg/mL of streptomycin (Invitrogen, Waltham, MA) and MEM non-essential amino acids (Invitrogen, Waltham, MA). Cells were cultured at 37 °C in a humidified incubator with 5% CO₂ for 12 h before use. Then, cells at more than 80% confluence in 6-well plates were used for subsequent assays.

For the adhesion assay, the wild-type SC1401, derivatives of ΔqseBC and C-ΔqseBC were grown to logarithmic growth. These bacteria were washed in sterile PBS by centrifugation, resuspended in DMEM and adjusted by dilution to a multiplicity of infection (MOI) of 100:1 (1  × 10⁷ bacteria vs. 1  × 10⁵ host cells). Plates were incubated at 37 ° C in 5% CO₂. The cells were incubated for 2 h for bacterial adhesion, and then vigorously washed three times with PBS to remove nonspecifically attached bacteria. Then the cells were incubated for 10 min at 37 °C with 100 µL 0.25% trypsin/0.02% EDTA. After incubation, 900 µL TSB was added, the cells were diluted with 10-fold serial for 4 times and 100 µL cultures were spread on TSA++ plates. The visible colonies were counted manually to calculate the colony forming units (CFU).

For the invasion assay, complete growth medium (including 150 µg/mL of gentamicin) was added to each well, and the plates were incubated for additional 2 h to kill extracellular G. parasuis and washed three times with PBS. The above cells were lysed with 1% Triton X-100 to determine the intracellular bacteria by serial dilutions. The visible colonies were counted manually to calculate the CFU. The assay was performed in triplicates.

Indirect immunofluorescent assay of adherent bacteria on MLE-12 cells

Mouse lung epithelial cells MLE-12 were purchased from the Beijing Institute of BeNa Biotechnology (ATCC-sourced strain, the 4th passage). In the indirect immunofluorescent (IIF) assay, MLE-12 cells were grown on coverslips in a six-well plate to reach a cell concentration of 90% confluence. The bacteria and cells were co-incubated following the adhesion assay. In this study, the G. parasuis HpGbpA polyclonal antibody (1:500 dil.) prepared and stored in our laboratory was used as the primary antibody, and the commercial CY3-labeled goat anti-mouse IgG (Servicebio, Gent, Belgium) was used as the secondary antibody to detect the distribution of bacteria. Cy3-labeled fluorescent secondary antibodies (EX WL: 510–560 nm, EM WL: 590 nm) were detected by a fluorescence microscope Nikon Eclipse ci&NIKON DS-U3, and the fluorescence intensity was analyzed by CaseViewer software. Fluorescence signal intensity was analyzed using ImageProPlus software and expressed as integrated optical density (IOD). The relative IOD value (IOD_R) is the ratio of the fluorescence intensity of CY3 (IOD_C) to the fluorescence intensity of DAPI (IOD_D).

Animal infection studies

Animal infection studies were performed as described previously (Dai et al., 2019a; Dai et al., 2021; Ding et al., 2016; Zhang et al., 2016). All efforts were made to provide for maximum comfort and minimal suffering of the animals. The study was conducted in accordance with ethic guidelines of the recommendations in the China Regulations for the Administration of Affairs Concerning Experimental Animals (1988) and the study was approved by the Institutional Animal Care and Use Committee of Southwest Medical University (20211017-001), Sichuan, China.

Thirty-two mice were allocated randomly to four groups of eight. Three groups were intraperitoneally injected with the G. parasuis SC1401, △qseBC or C-△qseBC strain at a dose of 1. 3  ×  10⁸ CFU per mouse. The remaining group was allocated to the PBS control group and injected intraperitoneally with 0.5 mL PBS. Dying animals that exhibited lethargy, hunched backs, rough coats, swollen abdomens, or inability to eat or drink were euthanized and considered dead.

Random block grouping method: mice were allocated randomly to SC1401, △qseBC, C-△qseBC and mock groups, with each group containing eight randomized mice. The random number table was then used to randomly assign the animals to the treatment group and the control group for experimental observation. In order to reduce the differences in experiments for non-research purposes, when the animals were injected intraperitoneally, groups were randomly selected, and a new set of syringes and needles was used for each individual animal. The animal cages were placed on the same level in the cage rack. In this blinded experiment, animal management personnel were not informed about the animal groupings, and the observations were done without information on the grouping of animals, so as to reduce the artificial selection bias in the observation of the results.

All mice were intraperitoneally injected with bacteria at a dose of 1.3 ×10⁸ CFU/animal (0.5 mL) or with 0.5 mL PBS. Clinical symptoms were carefully monitored for 7 days. Euthanasia was carried out on the animals with sleepiness, hunchback posture, rough hair, abdominal distension, or inability to eat or drink. The method of euthanasia was intraperitoneal injection of sodium pentobarbital (150 mg/kg). After 72 h of challenge, three animals in each group were euthanized humanely. The spleens and lungs were autopsied and treated with formalin/paraformaldehyde-fixed, paraffin-embedded (FFPE), and used for hematoxylin eosin (H&E) analysis to determine the pathological damage before and after immunization. The surviving mice were humanely euthanized (7 dpi).

Histological examination (H&E) and immunofluorescence assay

Mice were euthanized using sodium pentobarbitone (150 mg/kg). Lung and spleen tissues were harvested from mice in different groups (SC1401, △qseBC, C-△qseBC and PBS) after 4 dpi and fixed with 4% paraformaldehyde for histopathological and immunofluorescence (IF) assay. The paraffin sections were deparaffinized to water, and the sections were stained with hematoxylin and eosin (H&E) in turn, dehydrated, sealed with neutral gum and examined under a microscope (Dai et al., 2021).

Myeloperoxidase (MPO) levels were measured through the IF assay to quantitatively evaluate the levels of inflammation in the lungs of mice infected with SC1401, △qseBC and △qseC. The experimental procedure is described in Dai et al. (2019a). Mouse tissue and partial preparations were similar to the preparation of H&E materials. The fluorescent secondary antibody labeled with CY3 (EX WL: 510–560 nm, EM WL: 590 nm) was detected by a fluorescent microscope Nikon Eclipse ci&NIKON DS-U3, and the fluorescence intensity was analyzed by CaseViewer software. Fluorescence signal intensity was analyzed using ImageProPlus software and expressed as integrated optical density (IOD). The relative IOD value (IOD_R) is the ratio of the fluorescence intensity of CY3 (IOD_C) to the fluorescence intensity of DAPI (IOD_D).

Determination of the level of apoptosis in lung tissue (TUNEL staining)

In order to study cell apoptosis in situ in mice, the lungs of mice in the SC1401, △qseBC and △qseC challenge groups were collected, and then the TUNEL (TdT-mediated dUTP nick end labeling) kit was used to detect the level of apoptosis (Dai et al., 2019a). Fluorescence microscope Nikon Eclipse ci&NIKON DS-U3 was used to detect FITC-labeled fluorescent secondary antibody (EX WL: 465–495 nm, EM WL: 515–555 nm), and the fluorescence intensity was analyzed by CaseViewer software. Fluorescence signal intensity was analyzed using ImageProPlus software and expressed as integrated optical density (IOD). The relative IOD value (IOD_R) is the ratio of the fluorescence intensity of FITC (IOD_F) to the fluorescence intensity of DAPI (IOD_D).

Statistical analyses

Statistical analyses were performed using Graph-Pad Prism 5.0 software (GraphPad Software, San Diego, CA, USA). A one-way or two-way ANOVA analysis was used to compare the differences among more than two groups. P values < 0.05 were considered statistically significant.

The qseB and qseC genes are conserved among 15 standard strains of G. parasuis

The standard strains of G. parasuis 1–15 were used as the templates to amplify the qseB and qseC genes by PCR, and the target bands were obtained (qseB is 672 bp; qseC is 1,398 bp). It was shown that the 15 standard strains of G. parasuis serotype all have qseB and qseC genes (Figs. 1A/1B).

Validation of the qseB/C genes deletion and complementation

The recombinant plasmid pLQ4 was identified by restriction enzyme digestion (Fig. 2A). The recombinant plasmid pLQ4 was digested into a vector fragment (5,600 bp) and an insert fragment (2,840 bp). The in-frame nonpolar ΔqseBC mutant was constructed by natural transformation and confirmed by PCR (Fig. 2B). The identities of SC1401, ΔqseBC and C-ΔqseBC were confirmed in all strains by amplifying the 822-bp fragment encoding 16S rRNA using primers P11/P12. A 2061-bp fragment of the qseBC gene using primers P9/P10 was amplified from the wild type and complement, but not the mutant. The kan fragment (935 bp) was amplified using primers P7/P8 from the mutant and complement, but not from the wild-type strain SC1401, indicating that the gene mutant and its complemented strain were successfully constructed.

The transcription of the qseB and qseC flanking genes were further examined by RT-qPCR to assess the possible polar effects caused by deletion of the qseB and qseC genes. As shown in SC1401 and ΔqseBC, both the upstream gene A4U84_RS03675 and downstream gene groES were not significantly affected (Fig. 2C), indicating no polar effect.

The ΔqseBC mutant decreased adherence and invasion abilities

In order to further study the effect of G. parasuis two-component system QseBC on the interaction between bacteria and host cells, the PAM cells were co-incubated with SC1401, ΔqseBC and ΔqseC to compare their adhesion and invasion abilities. As shown in Figs. 3A/3B, using bacteria in the same growth state (taking OD₆₀₀nm as a reference) and the same batch of plated cells, when MOI =100, the average number of cells adhered to PAM cells by SC1401, ΔqseBC and ΔqseC were 311.67 ×10⁵ CFU/well, 178.67 × 10⁵ CFU/well and 223.33 ×10⁵ CFU/well, respectively. As compared to the parent strain, both mutant strains showed a significant decrease. In addition to the decreased number of adhesions, deletion of the qseB/C genes also significant impaired the ability of the bacteria to invade the PAM cells (p < 0.05). The adhesion and invasion levels were recovered in the complemented strain C-ΔqseBC.

In addition, the decreased adhesion of ΔqseBC and ΔqseC to mouse-derived MLE-12 cells was observed using the IIF assay. The HpGbpA polyclonal antibody (1:500 dil.) prepared and stored in our laboratory was used as the primary antibody, and the commercial Cy3-labeled goat anti-mouse IgG (Servicebio, Gent, Belgium) was used as the secondary antibody to climb the slides of MLE-12 cells. Indirect immunofluorescence assay was performed. The DAPI (blue) column represents the nuclear staining results; the CY3 column (red) represents the bacterial staining results; the third column is the combined graph of the two (observing the results of bacterial adherent cells). Fluorescence signal intensities are displayed as integrated optical density (IOD) values and analyzed by ImageProPlus software. The relative value of IOD_R was calculated by the ratio of Cy3-fluorescence intensity (IOD_C) and IOD of DAPI (IOD_D). As shown in Figs. 3C/3D, the relative fluorescence intensity (RFI) of the deletion strain was significantly lower than that of the parent strain, and the red fluorescence was localized on the cell surface, which was consistent with the location of bacteria adhering to the cell surface. The results of this part further confirmed that the QseBC two-component system was involved in the regulation of bacterial adhesion to cells.

In summary, our results indicated that ΔqseBC resulted in a significant decrease in the adhesion and invasion abilities of the bacteria to the PAM and MLE-12 cells used in the experiment, suggesting that QseBC affects the ability of the bacteria to interact with PAM and MLE-12 cells.

Decreased virulence of the △qseBC mutant in mice

The SC1401 and the △qseC or △qseBC strain were compared in a murine model of infection. After intraperitoneal injection with SC1401, the survival rate was 12.5% (1 mouse), whereas the survival rate was 50% (4 mice) in both the △qseBC and △qseC groups. No anomalies were observed in the PBS control group throughout the entire virulence assays. In short, 87.5% of mice inoculated with SC1401 died as compared to only 50% of mice inoculated with △qseBC and △qseC, respectively (Fig. 4A), demonstrating that QseBC contributes to the virulence of G. parasuis.

Lung and spleen tissues of hosts were harvested for histological pathological observation using H&E staining. As shown in Figs. 4B/4C, pathological lesions were present in all treated groups. The mice in wild-type and complemented groups showed severe congestion of alveolar wall capillaries. However, mice in ΔqseC and ΔqseBC groups demonstrated less severe pathological changes.

MPO immunofluorescence analysis

In this experiment, through the IF analysis of the lung pathological tissue of the infected mice, it was found that 4 days after infection (4 dpi), the IOD of the lung tissue of the mice in each treatment group was significantly higher than that of the blank group (p < 0.05), suggesting that the activity of MPO was increased. The IOD_R of SC1401 (average IOD_R: 0.625) and complemented groups (average IOD_R: 0.582) were significantly higher than that of the △qseBC mutant group (average IOD_R: 0.190) and △qseC mutant group (average IOD_R: 0.371), indicating that △qseBC and △qseC mutants caused significantly lower secondary inflammatory active substances in the lungs of mice than the parent strain, and the inflammatory damage to the target tissues may be more minor (Fig. 5). The evaluation results of the above pathological sections were further confirmed.

Analysis of tissue apoptosis level in challenged mice

In order to directly observe the pathological damage at the level of cell death, the level of bacterial-induced apoptosis in mouse lungs was quantitatively analyzed by TUNEL staining. As shown in Fig. 6, the IOD of the FITC-labeled fluorescent secondary antibody was significantly enhanced in all treatment groups. Taking the IOD _D of DAPI in each group as a reference, the average IOD_R values of mice in the ΔqseBC and ΔqseC challenge groups were only 18.88% and 16.78% of the parental strain challenge groups, respectively. The mean IOD_R of ΔqseBC and ΔqseC were 0.027 and 0.024, respectively, while the IOD_R of the wild-type SC1401 group was 0.143. Replenishing strain group can compensate for a certain degree of cell damage ability (average IOD_R = 0.098). However, the connection between QseBC and some apoptosis-mediated pathways is uncertain and needs to be further studied in the future.