Bordetella bronchiseptica promotes adherence, colonization, and cytotoxicity of Streptococcus suis in a porcine precision-cut lung slice model

ABSTRACT Bordetella (B.) bronchiseptica and Streptococcus (S.) suis are major pathogens in pigs, which are frequently isolated from co-infections in the respiratory tract and contribute to the porcine respiratory disease complex (PRDC). Despite the high impact of co-infections on respiratory diseases of swine (and other hosts), very little is known about pathogen-pathogen-host interactions and the mechanisms of pathogenesis. In the present study, we established a porcine precision-cut lung slice (PCLS) model to analyze the effects of B. bronchiseptica infection on adherence, colonization, and cytotoxic effects of S. suis. We hypothesized that induction of ciliostasis by a clinical isolate of B. bronchiseptica may promote subsequent infection with a virulent S. suis serotype 2 strain. To investigate this theory, we monitored the ciliary activity by light microscopy, measured the release of lactate dehydrogenase, and calculated the number of PCLS-associated bacteria. To study the role of the pore-forming toxin suilysin (SLY) in S. suis-induced cytotoxicity, we included a SLY-negative isogenic mutant and the complemented mutant strain. Furthermore, we analyzed infected PCLS by histopathology, immunofluorescence microscopy, and field emission scanning electron microscopy. Our results showed that pre-infection with B. bronchiseptica promoted adherence, colonization, and, as a consequence of the increased colonization, the cytotoxic effects of S. suis, probably by reduction of the ciliary activity. Moreover, cytotoxicity induced by S. suis is strictly dependent on the presence of SLY. Though the underlying molecular mechanisms remain to be fully clarified, our results clearly support the hypothesis that B. bronchiseptica paves the way for S. suis infection.


Introduction
Respiratory diseases in pigs are often caused by bacterial and viral co-infections. However, very little is known about pathogen-pathogen-host interactions and the mechanisms of pathogenesis. The porcine respiratory disease complex (PRDC) is a paradigm for such infections. It is a major cause of mortality and reduced weight gain in the swine population worldwide and, therefore, a serious economic concern for the pig industry. Presumably, pathogenesis of PRDC is based on synergistic interactions of primary and secondary opportunistic pathogens [1][2][3]. Primary pathogens, including porcine reproductive and respiratory syndrome virus, swine influenza virus (swIAV), Mycoplasma hyopneumoniae, and Bordetella (B.) bronchiseptica, can cause an infection by themselves, which is usually mild. However, they can also pave the way for secondary pathogens, such as Pasteurella multocida, Streptococcus (S.) suis, and Glaesserella parasuis, which are frequent commensals in the porcine respiratory tract. The latter, also known as pathobionts [4], not to state that G. parasuis and S. suis are unable to establish infection without a primary pathogen. While they do complicate infection with other pathogens, they are also capable of causing primary systemic infection [1].
In the present study, we have analyzed interactions of two major bacterial pathogens involved in the PRDC, B. bronchiseptica and S. suis. The former is a Gram-negative bacterium contributing to several secondary bacterial infections [5][6][7][8][9]. In contrast to the closely related human pathogen B. pertussis, B. bronchiseptica has a broad host range, including wild, domestic, and companion animals, as well as humans. Additionally, it plays a role in many respiratory diseases, e.g., atrophic rhinitis in swine and kennel cough in dogs. B. bronchiseptica is ubiquitous in pig populations and is frequently isolated in combination with other pathogens from cases of pneumonia [10]. Virulence factors of this pathogen include, e.g., filamentous hemagglutinin, fimbriae, pertactin, and toxins, like the dermonecrotic toxin (DNT) or the tracheal cytotoxin (TCT). Most of these are regulated by a two-component signal transduction system in response to environmental conditions (reviewed in [11]).
S. suis is a frequent colonizer of the upper respiratory tract of pigs and considered a pathobiont or secondary pathogen, which can become invasive when the respiratory epithelium has been damaged, e.g., by prior infection with a primary viral or bacterial pathogen [4]. Then, S. suis can enter the bloodstream and cause severe systemic diseases, such as meningitis, arthritis, endocarditis, and septicemia [12], resulting in high economic losses for the pig industry. Moreover, as an emerging zoonotic agent, S. suis can also cause meningitis and septicemia in humans [13,14]. Most of the virulent strains are positive for suilysin (SLY), a secreted extracellular acting pore-forming protein that contributes to virulence of S. suis by damaging epithelial cells and promotion of bacterial adherence and invasion [15][16][17] (reviewed in [18]).
In our previous studies, we have shown in coinfections of porcine precision-cut lung slices (PCLS) that swIAV can induce ciliostasis, thereby promoting adherence and invasion of S. suis [19]. This prompted us to investigate possible effects of other (primary) respiratory pathogens on infection of S. suis. Earlier experimental infection studies by Vecht et al. revealed that B. bronchiseptica could predispose piglets to infection with S. suis, resulting in more severe clinical signs and elevated bacterial loads in the tissue [6,7]. However, the mechanisms of how B. bronchiseptica led to increased colonization of the respiratory epithelium by S. suis are still unclear. Possible mechanisms might be the use of adhesins (e.g., filamentous hemagglutinin) secreted by B. bronchiseptica [20] or facilitated attachment to respiratory epithelial cells due to ciliostasis [21][22][23], probably induced by the TCT [24]. Thus, we assumed that B. bronchiseptica might promote S. suis infection by a similar effect as swIAV, i.e. by reduction of ciliary activity. In order to investigate this hypothesis, we first established a PCLS model for coinfection with B. bronchiseptica and S. suis. PCLS have been proven as a suitable model to investigate respiratory infections [25][26][27]. The main advantage of PCLS is the preservation of the original cellular and structural organization of the lung tissue. Moreover, ciliary beating at the bronchiolar surface can be assessed by light microscopy. Using PCLS, we were able to show that prior infection with B. bronchiseptica leads to ciliostasis and enhances subsequent adherence, colonization, and, as a consequence of the higher colonization, the cytotoxicity of S. suis.

Porcine precision-cut lung slices
For the preparation of PCLS, lungs from apparently healthy pigs of different ages were obtained from a local slaughterhouse (Leine-Fleisch GmbH, Laatzen, Germany). As previously described [19,25,26], the cranial, middle, and intermediate lung lobes were filled with 1.5% (w/v) low-melting agarose (GERBU, Heidelberg, Germany) in RPMI 1640 medium (Sigma-Aldrich, Taufkirchen, Germany) and kept on ice until agarose became solidified. Afterward, cylindrical pieces of lung tissue with a bronchiole in the middle were punched out with a tissue-coring tool and cut into approximately 300 µm thick slices using a Krumdieck tissue slicer (model MD 4000-01; TSE Systems, Chesterfield, MO, USA). Slices were collected in RPMI 1640 medium (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with antibiotics and antimycotics (1 µg/ml clotrimazole, 10 µg/ml enrofloxacin, 50 µg/ml kanamycin,100 U/ml penicillin, 100 µg/ml streptomycin, 50 µg/ml gentamicin, 2.5 µg/ ml amphotericin B) and bubbled for approximately 2 hours (h) with a normoxic gas mixture at 37°C to remove agarose from the slices [28]. Then, PCLS were collected in 24-well plates (Greiner Bio-One, Frickenhausen, Germany) and incubated in RPMI 1640 medium supplemented with antibiotics and antimycotics at 37°C and 5% CO 2 for one day. The next day, slices with good ciliary activity (at least 80-90%, determined by light microscopy) were chosen for experiments, washed with phosphate-buffered saline (PBS; Sigma-Aldrich), and incubated in RPMI 1640 medium without any antibiotics and antimycotics for one day.

Bacterial strains
We used a clinical lung isolate of B. bronchiseptica (B. bronchiseptica 1263/2/18), obtained from a swine herd with unspecific clinical symptoms, such as reduced weight gain and sporadic cases of anemic animals. It was grown on Columbia agar supplemented with 7% (v/v) sheep blood (Oxoid™, Thermo Fisher Scientific) for 48 h at 37°C under aerobic conditions.
For infection experiments, cryo-conserved bacterial stocks were prepared. For this, B. bronchiseptica was grown overnight at 37°C under aerobic conditions in Brain Heart Infusion (BHI) broth (Bacto™, Becton Dickinson, Heidelberg, Germany), adjusted to an optical density at 600 nm (OD 600 ) of 0.05 in prewarmed BHI broth and further incubated until early exponential growth phase (OD 600 of 0.6). Bacteria were centrifuged (5000 × g for 10 min) and resuspended in BHI broth with 15% (v/v) glycerol. Aliquots were then frozen in liquid nitrogen and transferred to −80°C for long-term storage.
The virulent SLY-positive S. suis serotype 2 wildtype strain 10 (S. suis 10 wt) was kindly provided by H. Smith (Lelystad, Netherlands) [29]. The isogenic SLY-deficient mutant of S. suis 10 (S. suis 10∆sly) was constructed by insertion of an erythromycin cassette in the sly gene [30]. The complemented SLY-mutant strain (S. suis 10cS148) was generated by chromosomal complementation of S. suis 10∆sly, carrying a silent mutation of the sly gene to delimit this strain from the wt [15]. All S. suis strains were grown on Columbia agar supplemented with 7% (v/v) sheep blood (Oxoid™, Thermo Fisher Scientific) overnight at 37°C under aerobic conditions. Cryo-conserved bacterial stocks were prepared from liquid cultures in Todd-Hewitt Broth (THB; Bacto™, Becton Dickinson) at the late-exponential growth phase (OD 600 of 1.0), as previously described [19].

Infection of PCLS with B. bronchiseptica and S. suis
Prior to infection experiments, PCLS were maintained without antibiotics and antimycotics for one day and washed two times with PBS.
To establish the infection with B. bronchiseptica, PCLS were infected with low doses of B. bronchiseptica (approximately 10 2 -10 4 CFU/well) in 500 µl RPMI 1640 medium in a 24-well plate for 4 h at 37°C and 5% CO 2 . In order to remove non-adherent bacteria, slices were washed two times with PBS and fresh medium was added for further incubation up to 48 hours post-infection (hpi).
For co-infection experiments, PCLS were pre-infected with approximately 10 4 CFU/well of B. bronchiseptica in 500 µl RPMI 1640 medium in a 24-well plate for 4 h at 37° C and 5% CO 2 and then washed two times with PBS to remove non-adherent bacteria. Fresh medium was added, and slices were incubated for another 20 h at 37°C and 5% CO 2 . PCLS were again washed two times with PBS and then co-infected with approximately 10 7 CFU/well of S. suis 10 wt, S. suis 10∆sly, and S. suis 10cS148, respectively, in 500 µl RPMI 1640 medium for 4 h at 37°C and 5% CO 2 . Afterward, slices were washed two times with PBS to remove non-adherent streptococci, fresh medium was added, and slices were further incubated at 37°C and 5% CO 2 up to 48 hpi with S. suis.

Ciliary activity
For infection experiments, we chose slices showing rapidly moving cilia on at least 80% of the whole luminal surface of the bronchiole. Ciliary activity of at least two slices was estimated using light microscopy (Leica DMi1; Leica, Wetzlar, Germany) and the activity determined before infection was set as 100%. Ciliary movement of at least two slices was monitored at different times after infection with B. bronchiseptica and S. suis, respectively, and the experiment was repeated at least three times.

Cytotoxicity assay
To determine the extent of cytotoxicity caused by B. bronchiseptica and/or S. suis, we measured the release of lactate dehydrogenase (LDH) at different times using CytoTox 96® Non-Radioactive Cytotoxicity Assay (Promega, Mannheim, Germany). Supernatants of infected and uninfected PCLS were collected and the cytotoxicity assay was performed as recommended by the manufacturer. LDH release of infected PCLS was normalized to uninfected PCLS and results were expressed as percentage LDH release compared to uninfected PCLS lysed with 10% (v/v) Triton® X 100 (Carl Roth, Karlsruhe, Germany) in RPMI 1640 medium. All experiments were performed in duplicates and repeated at least three times.

Bacterial adherence and colonization
To quantify PCLS-associated bacteria (surface-adherent and/or intracellular bacteria), PCLS were washed two times with PBS and then homogenized in PBS using Lysing Matrix D (MP Biomedicals, Irvine, CA, USA) and the FastPrep-24™ 5 G Instrument (3 × 30 sec, speed 5; MP Biomedicals). Afterward, bacterial number in the lysate was determined by serial dilution and replicate plating on agar plates. To determine the number of B. bronchiseptica in mono-infection experiments, the lysate was plated on Columbia agar supplemented with 7% (v/v) sheep blood (Oxoid™, Thermo Fisher Scientific).
To differentiate S. suis from B. bronchiseptica, we used blood agar plates containing 7% (v/v) bovine blood and Oxoid™ Staph/Strepselective supplement (Thermo Fisher Scientific; referred to as "Staph/Strep-agar plate"). Both, B. bronchiseptica and S. suis were incubated for 48 h at 37°C under aerobic conditions. All experiments were performed in duplicates and repeated at least three times.
Samples were analyzed using the Nikon Eclipse Ti-S (Nikon, Tokyo, Japan), equipped with the objectives Plan Fluor 10x/0.30 DIC and 60x/0.50-1.25 Oil. Brightness and contrast were adjusted using ImageJ/ Fiji 1.52p software (National Institute of Health, USA).

Histological examination
Hematoxylin and eosin (HE) staining of paraffin sections was performed following standard procedures [31]. For examination of the HE stained PCLS, we used the Zeiss AXIO Imager.M2 (Carl Zeiss, Stockholm, Sweden) equipped with the camera AxioCam MRm Rev.3 FireWire (D) and the objective Plan-Apochromat 63x/1.4 Oil.

Field emission scanning electron microscopy
Sample preparation and examination were performed as described previously [32] with some modifications. Infected PCLS were fixed with 5% (v/v) formaldehyde and 2% (v/v) glutaraldehyde in HEPES buffer (0.1 M HEPES, 0.01 M CaCl 2 , 0.01 M MgCl 2 , 0.09 M sucrose; pH 6.9) at 7°C until dehydration process. Samples were washed twice with TE buffer (0.01 M Tris, 1 mM EDTA; pH 6.9) and dehydrated with graded series of ethanol (10%, 30%, 50%, 70%, 90%, and 100%) on ice for 15 min for each dilution. Samples in 100% ethanol were allowed to reach RT and then exposed to critical point drying with liquid CO 2 (Bal-Tec CPD 030, Balzers, Liechtenstein). Subsequently, samples were covered with a gold-palladium film (approximately 10 nm) by sputter coating (Bal-Tec SCD500, Balzers) before examination with the field emission scanning electron microscope (FESEM) Carl Zeiss Merlin™ (Carl Zeiss), equipped with the Everhart Thornley SEdetector and the inlense SE-detector in a 50:50 ratio at an acceleration voltage of 5 kV.

Statistical analyses
All experiments were repeated at least three times and the data in the figures are shown as the means ± standard deviation (means ± SD). All statistical analyses were carried out using GraphPad Prism version 8.0.1 for Windows (GraphPad Software, San Diego, CA, USA). Normal (Gaussian) distribution of the data was tested by Shapiro-Wilk test and statistical significance was determined by t-test or by one-way ANOVA followed by Tukey, Dunnett, or Sidak post-hoc test; p < 0.05 was considered significant.

Establishment of defined conditions for PCLS infection with B. bronchiseptica
Our first objective was to establish defined conditions for the co-infection with B. bronchiseptica and S. suis in PCLS. Since B. bronchiseptica can cause reduction of ciliary activity of respiratory epithelial cells, as well as more severe mucosal damage, we first established conditions at which only ciliary activity was reduced without any further damage. For this, we infected the PCLS with low doses of B. bronchiseptica (10 2 -10 4 CFU/well) for up to 48 h. Even the lowest infection dose of 10 2 CFU/well reduced the ciliary activity at 24 hpi significantly, and 10 4 CFU/well led to complete ciliostasis (Figure 1(a)). At 48 hpi, no beating of the cilia was observed in PCLS infected with any of the infection doses (Figure 1  (a)). By measuring the release of LDH in the supernatant of infected PCLS, we evaluated the extent of cytotoxicity induced by B. bronchiseptica. Notably, we detected only low amounts of LDH in the supernatant and even the highest infection dose of 10 4 CFU/well caused only a maximum of 10% LDH release at 48 hpi (Figure 1(b)). HE staining and immunofluorescence staining of PCLS infected with 10 4 CFU/ well confirmed these results, as almost no damage of the bronchiolar epithelium (loss of ciliated cells) was detectable at 24 hpi (Figure 1(c,d)).
After 48 h, all infection doses led to a small number of detached ciliated cells ( Figure S1, indicated by arrows), but no substantial loss of the bronchiolar epithelium was observed. Furthermore, immunofluorescence analysis showed that B. bronchiseptica adhered preferentially to the cilia (Figure 1(d)). Counting the number of PCLSassociated bacteria revealed a time-and dosedependent increase of adherent and/or invasive bacteria until 24 hpi, but the recovered CFU at 48 hpi was independent of the initial infection dose (Figure 2(a)). FESEM confirmed the high colonization capacity of B. bronchiseptica to the lung tissue, in particular to the cilia (Figure 2(b)). Taken together, we established an infection dose of B. bronchiseptica (10 4 CFU/well for 24 h), which is sufficient for bacterial colonization of the lung tissue and for a reproducible reduction of the ciliary activity without causing substantial damage to the bronchiolar epithelium.

Prior infection with B. bronchiseptica promotes adherence and colonization of S. suis
Next, we investigated effects of prior infection with B. bronchiseptica on subsequent adherence and colonization of S. suis. PCLS were pre-infected with 10 4 CFU/ well of B. bronchiseptica for 24 h. We controlled the ciliary activity by light microscopy and confirmed a reduced motility due to infection with B. bronchiseptica ( Figure S2) before PCLS were coinfected with 10 7 CFU/well of S. suis for up to 48 h. Mono-and co-infected PCLS were homogenized to determine the number of tissue-associated S. suis. Interestingly, at 24 and at 48 hpi significantly more streptococci were counted in the lysates of PCLS preinfected with B. bronchiseptica compared to PCLS mono-infected with S. suis (Figure 3(a)). Similar results were obtained by plating the supernatant of mono-and co-infected PCLS ( Figure S3). FESEM of PCLS coinfected with B. bronchiseptica and S. suis for 48 h allowed us to visualize the co-colonization of the lung tissue by both pathogens, even though this situation was only rarely observed ( Figure  3(b)).
Immunofluorescence staining of mono-and coinfected PCLS showed an enhanced signal for streptococci in PCLS pre-infected with B. bronchiseptica at 24 hpi compared to the mono-infected sample (Figure 4). At 4 hpi no difference in the signal for S. suis was detectable between mono-and co-infected PCLS (Figure 4), confirming the plating results of the tissue lysates. Moreover, immunofluorescence analysis revealed that B. bronchiseptica bound preferentially to the cilia, whereas S. suis was mainly associated with the alveolar epithelium ( Figure 4). Nevertheless, cocolonization of both pathogens was also observed sometimes ( Figure 4). Taken together, we found that pre-infection of PCLS with B. bronchiseptica enhances the colonization capacity of S. suis. Notably, both pathogens seem to adhere to and colonize different areas of the tissue.

Prior infection with B. bronchiseptica promotes cytotoxic effects of S. suis
Next, we wanted to investigate whether the enhanced adherence and colonization of S. suis due to preinfection with B. bronchiseptica resulted in higher cytotoxicity caused by S. suis. Moreover, we wanted to evaluate the role of SLY, the pore-forming toxin secreted by S. suis, which is considered a virulence-associated factor, in PCLS co-infected with B. bronchiseptica and S. suis.
After pre-infection with B. bronchiseptica for 24 h, PCLS were co-infected with 10 7 CFU/well of the virulent S. suis serotype 2 wild-type strain 10 (S. suis 10 wt), its isogenic SLY-deficient mutant (S. suis 10Δsly), or the complemented SLY-mutant strain (S. suis 10cS148) for up to 48 h. We measured the release of LDH in the supernatant of PCLS either mono-infected with S. suis or co-infected with both pathogens to determine the cytotoxic effects caused by S. suis. We observed, that cytotoxicity induced by S. suis alone was strictly dependent on the presence of SLY, as almost no release of LDH was detected when PCLS were infected with S. suis 10Δsly, whereas S. suis 10cS148 restored the phenotype of the S. suis 10 wt strain ( Figure 5). Both, the SLYdeficient mutant strain and the complemented SLYmutant strain showed similar colonization capacities compared to the wild-type strain ( Figure S4). Interestingly, SLY-dependent release of LDH was significantly higher at 24 hpi in PCLS pre-infected with B. bronchiseptica compared to PCLS infected with S. suis alone ( Figure 5). The extent of cytotoxicity in PCLS coinfected with B. bronchiseptica and S. suis 10Δsly was higher compared to PCLS mono-infected with S. suis 10Δsly, but lower than in PCLS co-infected with B. bronchiseptica and the SLY-positive strains ( Figure  5). Immunoblot analysis of SLY expression in the supernatant of mono-and co-infected PCLS revealed a higher amount of SLY in the supernatant of PCLS pre-infected with B. bronchiseptica ( Figure 5, inlay). Thus, we could show that the enhanced colonization capacity of S. suis, due to pre-infection with B. bronchiseptica, goes along with an increased amount of SLY in the supernatant of co-infected PCLS, resulting in higher cytotoxic effects caused by the toxin. In summary, these results show that B. bronchiseptica can promote infection of S. suis by enhancing its adherence and colonization, and, consequently, its cytotoxicity.

Discussion
Respiratory diseases in pigs are often caused by a combination of different pathogens, whereby infection with one pathogen can pave the way for secondary pathogens [1]. B. bronchiseptica and S. suis are two major bacterial pathogens in pigs, contributing to the PRDC. S. suis is a pathobiont, which is considered to become invasive and to cause systemic infections only when other pathogens and/or certain environmental factors promote its translocation through the respiratory epithelium [4,33,34]. To calculate the number of bacteria attached to the tissue, infected PCLS were homogenized at 4, 24, and 48 hpi and the lysate was plated on blood agar plates to determine CFU/ml. All experiments were repeated three times and means ± SD are shown. Significant differences between the infection doses are indicated by * p < 0.05 and ** p < 0.01; one-way ANOVA followed by Tukey post-hoc test. Significant differences between 4 hpi and 24 hpi or 48 hpi are indicated by #### p < 0.0001; one-way ANOVA followed by In a previous study of our group, reduction of the ciliary activity by swIAV subtype H3N2 promoted adherence of S. suis and its invasion into deeper tissues [19]. Like swIAV, B. bronchiseptica is known to contribute to secondary bacterial infections [5][6][7], probably by reduction of ciliary activity due to the TCT [21][22][23][24]. Thus, the objective of our study was to establish a co-infection model with B. bronchiseptica and S. suis in PCLS, to analyze the effects of a primary infection with B. bronchiseptica on subsequent infection with an opportunistic pathogen, such as S. suis. Using a relatively low dose of B. bronchiseptica allowed us to focus on the effects of reduced ciliary activity without affecting the integrity of the respiratory epithelial  barrier. Higher doses might lead to destruction of the respiratory epithelium, as B. bronchiseptica is also known to cause extrusion of ciliated cells [35][36][37]. This would represent a completely different starting condition for the co-infection with an opportunistic pathogen. However, both scenarios are conceivable under natural conditions.
Next to the establishment of an ex vivo co-infection model with B. bronchiseptica and S. suis, we were interested in the role of SLY, a virulence-associated factor of S. suis, in such a co-infection scenario. We found that cytotoxicity induced by S. suis was strictly dependent on the presence of SLY, as almost no release of LDH was detectable when PCLS were mono-infected with the SLY-deficient mutant strain. This is in good agreement with previous studies performed with cell lines and primary cells of varying host and tissue origin [15,38] (reviewed in [18]). In PCLS co-infected with B. bronchiseptica and the SLY-deficient mutant strain, it becomes apparent that the cytotoxic effect of B. bronchiseptica is independent of SLY production. A similar effect was observed in porcine respiratory epithelial cells grown under air-liquid interface conditions when co-infected with swIAV and S. suis [38]. Strikingly, co-infection of PCLS with B. bronchiseptica and SLY-positive streptococci led to significant increase in cytotoxicity at 24 hpi compared to PCLS monoinfected with either of the strains, which seems not to be solely an additive effect. However, it seems plausible, that this effect is attributed to a higher number of streptococci in co-infected PCLS and, consequently, a higher concentration of SLY present in the supernatant of co-infected PCLS. Indeed, we demonstrated that pre-infection with B. bronchiseptica enhanced the colonization of PCLS by S. suis, resulting in a higher amount of SLY present in the supernatant of coinfected PCLS and in higher cytotoxic effects caused by SLY.
Whether pre-infection with B. bronchiseptica only promotes the adherence of S. suis (attachment to cells) or also its invasiveness (capacity to penetrate into deeper tissues para-and/or transcellularly) remains to be elucidated, as we could not distinguish between bacteria attached to the tissue and intracellular bacteria in this study. To address this aspect in future studies, transwell cell culture systems, such as air-liquid interface cultures with primary porcine respiratory epithelial cells [15,38], may be applied.
The precise mechanism of how B. bronchiseptica promotes the colonization of S. suis (and other pathogens) remains to be elucidated. It is plausible to assume that ciliostasis induced by B. bronchiseptica leads to impairment of the mucociliary clearance and thereby supports attachment of secondary bacteria to ciliated cells and/or invasion into deeper tissue [21]. Reduction of the ciliary activity might be attributed to the TCT Figure 5. Cell damage in PCLS co-infected with B. bronchiseptica and S. suis. PCLS were pre-infected with 10 4 CFU/well of B. bronchiseptica for 24 h. Subsequently, PCLS were co-infected with 10 7 CFU/well of S. suis serotype 2 wild-type strain 10 (S. suis 10 wt), its isogenic SLY-deficient mutant strain (S. suis 10Δsly), and the complemented SLY-mutant strain (S. suis 10cS148), respectively, for up to 48 h. Cytotoxicity was determined by measuring the release of LDH in the supernatant of mono-and coinfected PCLS. Results are expressed as percentage LDH release compared to PCLS lysed with 10% Triton® X 100. The experiment was repeated at least three times and means ± SD are shown. Significant differences between mono-and co-infection are indicated by ** p < 0.01 and *** p < 0.001; one-way ANOVA followed by Sidak post-hoc test. Below, immunoblot analysis of SLY-expression in the supernatant of mono-and co-infected PCLS at 24 hpi is shown.
[24], a virulence factor released by B. bronchiseptica. In contrast to B. bronchiseptica, infection with S. suis had no impact on the ciliary activity ( Figure S2).
Damage to the bronchial/bronchiolar epithelium, i.e., the loss of ciliated cells, has the same effect of compromising the mucociliary clearance [39]. Moreover, loss of the uppermost cell layer might facilitate the attachment of secondary bacteria by exposing other receptors (adhesins) present on deeper cell layers of the bronchial/bronchiolar epithelium [39]. However, in our study, immunofluorescence analysis and histological examination did not reveal any substantial loss of ciliated cells in PCLS infected with 10 4 CFU/well of B. bronchiseptica.
Another explanation for the enhanced adherence of S. suis might be the exploitation of adhesins secreted by B. bronchiseptica [40]. For the closely relative human pathogen B. pertussis, it was shown that pretreatment with filamentous hemagglutinin, an adhesin, which is also secreted by B. bronchiseptica, increased the adherence of S. pneumoniae and Haemophilus influenzae to ciliated cells [20]. In co-infection studies with S. suis and swIAV, interaction of the viral hemagglutinin with the bacterial capsular sialic acids was identified to promote adherence of S. suis [19,41,42]. Whether or not interactions between the filamentous hemagglutinin of B. bronchiseptica and the capsular sialic acids or other surface-associated structures of our S. suis strain occur, has to be investigated in future studies. However, a direct binding of S. suis to B. bronchiseptica, as it was reported for the binding of S. suis to swIAVinfected cells [19,38,41], is rather unlikely, as a direct interaction between both pathogens was not observed by FESEM. Moreover, immunofluorescence analysis showed that B. bronchiseptica and S. suis mainly attach to different areas of the lung tissue, which also means that they do not compete for binding sites in the respiratory tract. It might also be possible that B. bronchiseptica promotes streptococcal growth indirectly by providing nutrients, as it has been reported for influenza virus and S. pneumoniae [43].
In conclusion, we established an ex vivo coinfection model with B. bronchiseptica and S. suis in PCLS for analyses of interactions between both pathogens and primary host cells. In particular, the model is suitable to study effects of (primary) pathogeninduced ciliostasis on subsequent infection with a secondary pathogen. Our results showed that preinfection with B. bronchiseptica promotes infection with S. suis by enhancing adherence and colonization of S. suis, which leads to an increased cytotoxicity caused by S. suis. Despite the contribution of B. bronchiseptica to S. suis infection, cytotoxicity of S. suis depends on the presence of the pore-forming toxin SLY. Hence, in co-infection of PCLS, the ciliostatic effect of B. bronchiseptica and the cytotoxic effect of SLY may act synergistically to impair cellular functions. In future studies, it will be interesting to address the mechanisms how B. bronchiseptica contributes to infection with secondary bacterial agents and to investigate the interactions of bacterialbacterial co-infections of the porcine respiratory tract in more detail, as they represent a common economic problem in the pig industry. Finally, understanding of these interactions may well be translated to general pathogenicity concepts of respiratory coinfections in other host species, including humans.