ABSTRACT
Streptococcus suis is an important zoonotic pathogen which can infect humans and pigs worldwide, posing a potential risk to global public health. Suilysin, a pore-forming cholesterol-dependent cytolysin, is considered to play an important role in the pathogenesis of S. suis infections. It is known that infection with influenza A viruses may favor susceptibility to secondary bacterial infection, resulting in more severe disease and increased mortality. However, the molecular mechanisms underlying these coinfections are incompletely understood. Applying highly differentiated primary porcine respiratory epithelial cells grown under air-liquid interface (ALI) conditions, we analyzed the contribution of swine influenza viruses (SIV) to the virulence of S. suis, with a special focus on its cytolytic toxin, suilysin. We found that during secondary bacterial infection, suilysin of S. suis contributed to the damage of well-differentiated respiratory epithelial cells in the early stage of infection, whereas the cytotoxic effects induced by SIV became prominent at later stages of infection. Prior infection by SIV enhanced the adherence to and colonization of porcine airway epithelial cells by a wild-type (wt) S. suis strain and a suilysin-negative S. suis mutant in a sialic acid-dependent manner. A striking difference was observed with respect to bacterial invasion. After bacterial monoinfection, only the wt S. suis strain showed an invasive phenotype, whereas the mutant remained adherent. When the epithelial cells were preinfected with SIV, the suilysin-negative mutant also showed an invasion capacity. Therefore, we propose that coinfection with SIV may compensate for the lack of suilysin in the adherence and invasion process of suilysin-negative S. suis.
INTRODUCTION
Respiratory diseases are frequent and increasing, accounting for nearly one-quarter of all deaths in the world (1). To avoid the detrimental effects of foreign particles, including pathogens, the respiratory epithelium functions as a barrier. A network of intercellular tight junctions prevents the paracellular entry of microorganisms, and a mucociliary clearance system transports foreign substances out of the airways (2). Coinfections of the respiratory tract by viruses and bacteria are a common clinical manifestation (3, 4). Numerous studies have investigated the complications of viral infections and how they increase susceptibility to and morbidity from secondary bacterial infections, for example, by disrupting epithelial barriers (5, 6), modulating immune responses and inflammatory reactions (7, 8), or increasing bacterial attachment or colonization (9, 10). During the 1918 influenza pandemic, thousands of fatal cases were due to secondary bacterial pneumonia, particularly pneumonia caused by Streptococcus pneumoniae (11). During the influenza A (H1N1) virus pandemic in 2009, in more than half of the fatal cases reported from intensive care units (ICU), bacterial coinfections were confirmed (12, 13). In the last decade, the influence of influenza virus infections on secondary bacterial infections in the airway tract has been studied in more detail (14–16). On the one hand, infection by influenza virus may alter the airway epithelial physiology and immune response (17, 18); on the other hand, it may increase bacterial colonization and invasion, resulting in a more severe infection in the respiratory epithelial cells (10). However, the mechanisms underlying the pathogenic effects of coinfections are incompletely understood.
Coinfections with influenza viruses may affect the course of Streptococcus suis infections, as has been shown recently in a precision-cut lung slice model and in immortalized cells (6, 19, 20). S. suis is an important bacterial pathogen causing a number of infectious disease syndromes in pigs worldwide, including arthritis, septicemia, meningitis, and pneumonia (21, 22), and it has been described to be an emerging zoonotic pathogen posing a potential risk for global public health (21, 23). Suilysin (SLY), the yet only known cytotoxic protein secreted by S. suis, has been studied for its contribution to pathogenesis at different stages in the development of systemic disease induced by S. suis (24). It has been suggested that SLY facilitates the colonization and the establishment of the initial stages of infection in pigs by promoting host cell lysis and the invasion of S. suis through the epithelium when bacteria have colonized the upper respiratory tract (25, 26). Furthermore, S. suis isolates lacking the suilysin gene have been shown to be avirulent in a mouse infection model when applied via the intraperitoneal route but virulent in pigs after intravenous injection, which indicates that suilysin is not essential for S. suis to cause systemic infection in pigs once the bacteria have entered the blood vessels (27). The prevalent ratios of suilysin-positive and suilysin-negative strains worldwide vary, and not all virulent strains of S. suis seem to produce suilysin (28, 29). Nevertheless, the possibility that all suilysin-negative strains are noncytotoxic cannot be ruled out, as other yet unknown cytolysins might exist. Therefore, analyzing the biological role of suilysin and potential alternative virulence factors in different stages of bacterial infection may help to provide a better understanding of the pathogenesis of noncytotoxic S. suis strains.
In this study, porcine well-differentiated airway epithelial cells were applied to analyze the effect of swine influenza virus (SIV) infection on a secondary infection by a suilysin-positive wild-type strain and a noncytotoxic (suilysin-negative) S. suis mutant in vitro. We demonstrate that prior infection by swine influenza virus may compensate for the lack of suilysin in the invasion of the respiratory tract cells by S. suis. Our findings suggest a scenario of viral-bacterial interactions in the coinfection of the airway that may also be relevant for other pathogens.
RESULTS
Differences in the time kinetics of S. suis- and SIV-induced cytotoxicity.Both S. suis and influenza virus have a cytotoxic effect on epithelial cells (26, 30). To find out how the two pathogens contribute to the cytotoxic effect in a coinfection scenario, the release of cellular lactate dehydrogenase (LDH) was determined during the course of a coinfection with SIV and S. suis. We used air-liquid interface (ALI) cultures of differentiated porcine bronchial epithelial cells (PBEC) and infected them with SIV at a concentration of 5 × 104 50% tissue culture infective doses (TCID50)/filter. After an adsorption time of 2 h, nonattached viruses were washed away and PBEC were maintained under ALI conditions. At 24 h after virus infection, cells were infected with the S. suis wild type (wt) or the suilysin-deficient mutant 10Δsly at 2.5 × 107 CFU/filter. After incubation for 4 h and a thorough washing step, nonadherent bacteria were removed and incubation was continued until 24 or 48 h postinfection (hpi) under ALI conditions. As shown in Fig. 1, at 4 hpi the release of LDH was significantly increased in PBEC infected with the wt S. suis strain but not after monoinfection with the suilysin-deficient mutant. This finding is consistent with the results of our previous study on S. suis monoinfection of well-differentiated airway epithelial cells (26). Viral-bacterial coinfection, in any of the combinations analyzed, did not result in an increase in the cytotoxic effect compared to that in virus-monoinfected PBEC. Furthermore, SIV monoinfection induced no cytotoxicity in PBEC at 4 h after bacterial infection, which represented 28 h after viral infection in our experiment. A significant virus-induced cytotoxic effect was detected at 24 h and 48 h after bacterial infection, i.e., 48 h and 72 h after virus infection, respectively, when the amount of LDH was increased 5- and 10-fold, respectively. These results indicate that the cytotoxic effects of an S. suis infection occur much earlier than the SIV-induced cytotoxicity. Coinfection with the suilysin-deficient mutant 10Δsly did not affect the SIV-induced cytotoxicity. Interestingly, even though both the S. suis wt and SIV induced a cytotoxic effect in epithelial cells, no synergistic effects were observed after coinfection with SIV and the S. suis wt strain at any time point. Therefore, our data suggest that in the early stage of coinfection with SIV and S. suis, suilysin is the key factor inducing a cytotoxic effect in well-differentiated respiratory epithelial cells, whereas SIV is responsible for such effects in the later stage of infection.
Cytotoxic effects induced by viral and bacterial coinfection on well-differentiated porcine bronchial epithelial cells. The cytotoxic effects of viral and bacterial coinfection on PBEC were quantified by a standard LDH-release assay. The time-dependent cytotoxic effects of H3N2 and S. suis coinfection on PBEC were determined. Results are expressed as the percent cytotoxicity of experimental LDH release compared to the maximum LDH release of the ALI culture. Results are expressed as means ± SEM, and significance was determined using one-way ANOVA and the Tukey multiple-comparison test. *, P < 0.05; **, P < 0.01; ***, P < 0.0001; ns, not significant. Experiments were performed at least three times. Ctr, control.
SIV enhances the adherence of wt and suilysin-deficient S. suis to PBEC.The initial step in the development of pneumonia is the adherence of bacteria to airway epithelial cells (31). We compared the adherence capacity of an S. suis wt strain (strain 10) and its suilysin-deficient mutant, 10Δsly, during coinfection with influenza viruses. For this, epithelial cells were analyzed at 4 h after bacterial infection by immunofluorescence microscopy, and bacterial adherence was quantified by measuring the area of the epithelial cell surfaces positive for green fluorescent bacteria, as shown in Fig. 2A and B. The S. suis wt strain was more efficient in adherence to the airway cells than the 10Δsly mutant, which is consistent with the findings of previous studies (25, 26, 32). In the case of respiratory epithelial cells preinfected with SIV, bacterial adherence was enhanced for both the wt and suilysin-deficient mutant 10Δsly. In both cases, bound bacteria (green fluorescence) were predominantly detected in areas of cells infected with SIV (red fluorescence), as shown in Fig. 2A. Quantification of the fluorescent signals on epithelial cells indicated that the number of adherent bacteria in samples coinfected by SIV and wt S. suis was higher than that in samples coinfected by SIV and suilysin-deficient mutant 10Δsly. However, the latter value was still significantly higher than that for cells that were monoinfected by either the wt strain or the 10Δsly mutant (Fig. 2B). This result is most likely due to the presence of sialic acid on the capsular polysaccharide of wt strain 10 and the 10Δsly mutant. Recent studies have shown that the sialic acid binding activities of influenza viruses recognize the capsular sialic acid of S. suis, resulting in bacterial adherence to influenza virus-infected cells (6, 10, 19, 20). The 10Δsly mutant is encapsulated and, therefore, can benefit from this adherence mechanism, too. In this way, the 10Δsly mutant is able to compensate for the disadvantages in the adherence capacity related to the lack of the cytotoxic protein suilysin. Taken together, SIV preinfection enables S. suis to adhere to respiratory epithelial cells even in the absence of suilysin.
Adherence of S. suis to well-differentiated porcine bronchial epithelial cells preinfected with H3N2. PBEC were preinfected with 5 × 104 TCID50/filter H3N2 for 24 h, subsequently inoculated with S. suis wt strain 10 or the 10Δsly mutant at 2.5 × 107/filter from the apical compartment for 4 h, and then washed thoroughly to remove nonadherent bacteria and fixed for immunostaining. PBEC monoinfected with S. suis wt, 10Δsly, or H3N2 and mock-infected cells served as controls. (A) Immunostaining detection of adherence of streptococci to the bronchiolar epithelium. Streptococci are labeled in green, nucleoproteins of H3N2 are stained in red, and nuclei are shown in blue (DAPI). Bars, 100 μm. (B) To quantify the bacteria attached to PBEC, the areas containing green fluorescent bacteria were determined from the assay whose results are presented in panel A. Results are expressed as the means ± SEM, and significance was determined using one-way-ANOVA and the Tukey multiple-comparison test. *, P < 0.05; **, P < 0.01; ***, P < 0.0001. Three areas were randomly chosen for each sample, and experiments were performed at least three times.
Viral coinfection enhances the colonization of PBEC by S. suis.To get information about the colonization capacity of the wt and mutant S. suis strains under mono- and coinfection conditions, the incubation time after bacterial infection was extended to 24 and 48 h. After washing off nonadherent bacteria, cells were either processed for immunostaining (Fig. 3A) or lysed to determine the number of cell-associated streptococci (Fig. 3B). The results obtained by immunofluorescence microscopy are shown for samples analyzed at 24 h after bacterial infection (Fig. 3A). As is evident from the distribution of green fluorescent signals, which are indicative of the presence of bacteria, after monoinfections, the S. suis wt was more efficient than the 10Δsly mutant in colonizing epithelial cells. A difference was also detectable between monoinfected (Fig. 3A, top) and coinfected (Fig. 3A, bottom) cells. Preinfection with SIV resulted in enhanced colonization of both wt and mutant bacteria. As shown by β-tubulin staining (red), streptococci preferentially adhered to ciliated cells. To determine the total number of cell-associated bacteria, cells from mono- and coinfected filters were collected, lysed with 1% saponin, and used to determine the number of CFU. As shown in Fig. 3B, at both 24 and 48 hpi, the number of wt S. suis bacteria detected on airway epithelial cells was significantly higher than the number of 10Δsly mutant bacteria detected. For both strains, the number of cell-associated bacteria was higher in samples coinfected with SIV than in monoinfected cells. These differences were detectable at both time points postinfection, but at 48 hpi they were more pronounced. It should be noted that the quantitative assay does not differentiate between adherent and internalized bacteria. However, prior to internalization, bacteria need to adhere and, thus, also represent adherent bacteria. Our results indicate that prior infection with SIV enhances the colonization of the bronchial epithelium by both the S. suis wt and the suilysin-deficient mutant.
Colonization of S. suis on PBEC preinfected with SIV. PBEC were preinfected with 5 × 104 TCID50/filter H3N2 for 24 h, subsequently inoculated with S. suis wt strain 10 or the 10Δsly mutant at 2.5 × 107/filter from the apical compartment for 4 h, and then washed thoroughly to remove nonadherent bacteria and further incubated under ALI conditions until 24 hpi. Before fixation, the infected cells were washed with PBS to remove nonadherent bacteria. PBEC monoinfected with S. suis strain 10 wt, 10Δsly, or H3N2 and mock-infected cells served as controls. (A) Immunostaining detection of colonization of streptococci in mono- and coinfected PBEC of the bronchiolar epithelium. Streptococci are labeled in green, the cilia of ciliated cells are stained in red, and nuclei are shown in blue (DAPI). Bars, 50 μm. (B) To quantify the colonized bacteria in PBEC, mono- and coinfected cells were lysed by saponin. Then, cell lysates were plated on Columbia sheep blood agar and the numbers of CFU were determined. Results are expressed as the percentage of bacterial colonization, which was calculated as the number of CFU in the cell lysates compared to the number of CFU used in the inoculum, and are shown as the means ± SEM. Significance was determined using one-way-ANOVA and the Tukey multiple-comparison test. ***, P < 0.0001. Experiments were performed at least three times.
SIV infection renders the airway epithelium susceptible to invasion by noncytotoxic S. suis.Next, we were interested to know whether viral-bacterial coinfection has a direct effect on the invasion process. For this purpose, we chose the time point 48 h after bacterial infection (i.e., 72 h after viral infection). As demonstrated by the quantification results shown in Fig. 3B, at that time point the difference in colonization between mono- and coinfected samples was more pronounced than that at 24 h after bacterial infection. The respective fluorescence microscopy pictures are shown in Fig. 4A and confirm the results of the adherence and colonization analysis; i.e., more wt than mutant streptococci were detected, and more bacteria were observed after preinfection with SIV that in streptococcal monoinfections. To evaluate the invasion capacity of wt and mutant bacteria, vertical sections of these samples were immunostained for viral (red) and streptococcal (green) antigens (Fig. 4). As shown in Fig. 4B, the epithelial cell layer of SIV-infected samples (Fig. 4B, bottom) was somewhat thinner than that of airway cells that were not infected (designated “mock”) or that were monoinfected by bacteria (Fig. 4B, top). This effect of an influenza virus infection on differentiated airway epithelial cells has been reported recently (33) and is more pronounced at later times after SIV infection (7 days postinfection [dpi]). This may be explained by the loss of infected ciliated cells due to apoptosis (33). To compensate for the loss of cells, basal cells start to differentiate into specialized cells. In this way the airway epithelium can maintain its barrier function; however, for some time the epithelial cell layer is thinner than it is in a well-differentiated state. Because of these virus-induced changes, it was interesting to find out whether the invasion capacities of wt and mutant streptococci were affected under coinfection conditions. The monoinfected samples (Fig. 4B, top) showed the difference in invasion efficiency between wt and mutant S. suis that has been reported recently (26). While the mutant bacteria were mainly detectable on top of the epithelial cells, wt streptococci invaded into deeper areas of the epithelial cell layer. In the coinfected samples, wt S. suis showed an invasion phenotype similar to that in monoinfected samples. With respect to the absolute distance that wt streptococci invaded into the epithelial cell layer, there may not be a major difference between mono- and coinfected cells. As the cell layer of coinfected samples was thinner than that after bacterial monoinfection, the relative invasiveness of wt S. suis was distinctively increased in the coinfected samples. Therefore, we assumed that wt streptococci invade the epithelium in coinfected samples earlier than they do in monoinfected samples. A striking difference was detectable in PBEC infected by the suilysin-deficient mutant. In contrast to the location on top of the epithelial cell layer after bacterial monoinfection, under coinfection conditions, the S. suis mutant showed an invasive phenotype similar to that of the wt bacteria (tissue-invasive bacteria are indicated by white arrows in Fig. 4B). We conclude that the suilysin-deficient S. suis mutant may exploit cells preinfected by influenza viruses to adhere to and then invade the airway epithelial cell barrier, thereby overcoming the lack of cytotoxic activity.
Invasion of the bronchiolar epithelium by S. suis after preinfection with SIV. PBEC were preinfected with 5 × 104 TCID50/filter H3N2 for 24 h, subsequently inoculated with S. suis wt strain 10 or the 10Δsly mutant at 2.5 × 107/filter from the apical compartment for 4 h, and then washed thoroughly to remove nonadherent bacteria and further incubated under ALI conditions until 48 hpi. Before fixation, the infected cells were washed with PBS to remove nonadherent bacteria. PBEC monoinfected with S. suis wt strain 10, the 10Δsly mutant, or H3N2 and mock-infected cells served as controls. (A) Immunostaining detection for colocalization of streptococci with H3N2-infected cells in the bronchiolar epithelium. Streptococci are labeled in green, nucleoproteins of H3N2 are stained in red, and nuclei are shown in blue (DAPI). The areas in the squares in the center of the panels in the bottom row are magnified in the insets at the top left of those panels. Bars, 25 μm. (B) Immunostaining detection for invasion of streptococci into the bronchiolar epithelium, with vertical sections of PBEC being shown. Streptococci are labeled in green, nucleoproteins of H3N2 are stained in red, and nuclei are shown in blue (DAPI). White arrows indicate tissue-invasive bacteria, and the dashed lines show the location of the supporting membrane. Bars, 50 μm. Experiments were performed at least three times.
DISCUSSION
The aim of this study was to apply an in vitro ALI infection model to gain insights into the interactions between SIV and S. suis during infection of differentiated respiratory epithelial cells. Although some coinfection studies of respiratory pathogens have been performed previously (7, 34, 35), the porcine ALI culture system has not yet been used to investigate viral-bacterial coinfections so far.
Influenza virus infections may cause increased susceptibility to secondary bacterial pneumonia, resulting in increased morbidity and mortality rates in the human population; the mechanisms underlying this copathogenesis are not well understood (13, 14). For S. suis, it has been shown that the interaction of influenza A virus hemagglutinins with capsular sialic acids results in the binding of streptococci to virus-infected cells (6, 19, 20). This adherence mechanism is more efficient than the binding mediated by bacterial adhesins (6, 10). A streptococcal protein that contributes to the bacterium-host cell association is suilysin, though it is not a typical adhesin, as it is not cell bound but is secreted (36, 37). This cholesterol-dependent cytotoxin can induce apoptosis, resulting in epithelial damage. In addition, it is able to mediate the attachment of S. suis to the surface of the respiratory epithelium (26, 36, 37). Nevertheless, suilysin-negative clinical isolates are also frequently found in swine populations worldwide (28, 29). As influenza viruses not only have a binding activity but also induce apoptosis, it was interesting to analyze whether preinfection by influenza viruses can be exploited by suilysin-negative strains to initiate an infection.
The binding of pathogens to differentiated airway epithelial cells is much more difficult to achieve than that to immortalized cells due to the mucociliary clearance function of the airway epithelium. This protection mechanism comprises the collaboration of mucus-producing cells and ciliated cells. If any of these cell types are injured, this will impair the mucociliary clearance function, which may facilitate infection by pathogens. The apoptotic effect of influenza A virus infections results in the loss of ciliated cells (33, 38). As a consequence, in the areas of infection, the mucociliary clearance function is impaired until the loss is compensated for by basal cells that differentiate into specialized cells (39). During this regeneration period, it is likely that S. suis has more time for adherence to the airway epithelium and, thus, may benefit from coinfection with influenza A viruses. Furthermore, prior viral infection may expose additional cellular receptors that are hidden by the mucus layer in intact cells. The major advantage of the coinfection, however, is the enhanced adherence efficiency. The sialic acid-dependent interaction between encapsulated streptococci and influenza virus hemagglutinins expressed on the surface of virus-infected cells is a very effective way for the attachment of bacteria to differentiated epithelial cells (19). Though suilysin has a positive effect on the adherence of S. suis to uninfected airway epithelial cells, this binding is less efficient than that to virus-infected cells, which is evident from a comparison of the binding of suilysin-deficient bacteria to virus-infected cells and the binding of wt streptococci to uninfected cells.
Successful bacterial colonization requires the prior adherence of the bacteria to cells. As both suilysin-positive and suilysin-negative strains of S. suis show an increased adherence to influenza virus-infected cells, it is not surprising that they also show an enhanced colonization efficiency. The interaction between capsular sialic acids of S. suis and influenza virus hemagglutinins expressed on the surface of virus-infected cells appears to be not only a short-term advantage but also a long-term benefit. However, factors other than the enhanced adherence may also contribute to the enhanced colonization efficiency. Nutrients are also crucial for bacterial growth on airway epithelial cells (40). From studies with S. pneumoniae, it has been reported that the sialic acids released by the bacterial neuraminidase from mucins or other cell surface sialoglycoconjugates can be used by the pneumococci as nutrients to support colonization (14). S. suis lacks a neuraminidase; however, influenza A viruses contain a neuraminidase. This viral surface glycoprotein is expressed—together with the hemagglutinin—on viral particles and infected cells and releases sialic acids from cell surface glycoconjugates. Whether or not these released sialic acids are used as nutrients by adherent S. suis remains to be determined. As there is no culture medium in the apical compartment of the ALI culture system used in our study, the proliferation of streptococci relies on the metabolism of nutrients provided by the epithelial cells (26). As influenza viruses are able to induce apoptosis (33), it is conceivable that potential nutrients are released from dying cells (41) and might be utilized by S. suis. However, the virus-dependent cytotoxicity was observed only late in infection; therefore, this possibility appears to be less likely, as far as coinfection by influenza viruses and suilysin-negative S. suis bacteria is concerned.
The enhanced adherence and colonization efficiencies observed after coinfections by influenza viruses and S. suis also affect the subsequent step in a successful infection by streptococci, invasion. The more bacteria that adhere to the epithelial surface, the more likely that invasion events will occur. For the bacterial pathogen, a coinfection with influenza A viruses provides a further advantage, because virus infection induces apoptosis of virus-infected cells, which are mainly ciliated cells. It has been proposed previously that the loss of these specialized cells enables bacterial pathogens to get access to subepithelial cells and, thus, to spread to other parts of the host organism (38, 42). This view on the mechanism of pathogenesis does not take into account the fact that influenza virus infection does not impair the barrier function of the airway epithelium (33). The basal cells try to compensate for the loss of cells by initiating a regeneration process in which they differentiate into specialized cells (33, 39). Differentiation of the cells lasts more than a week. In this time, the epithelial cell layer is able to maintain the barrier function, but, as discussed above, the mucociliary clearance function is impaired (43). Another consequence of the regeneration process is that the cell layer has a reduced thickness. Therefore, for wt S. suis it is much easier to get across the epithelial barrier of an influenza virus-infected epithelium than across a thick cell layer of uninfected cells. The result from Fig. 4 was obtained at 72 h after virus infection. At 7 days after infection by influenza A viruses, the epithelial cell layer is thinner than it is at 3 dpi (33); therefore, at this later time point it will be even easier for wt S. suis to invade the airway epithelium at the foci of virus infection.
So far, the mechanisms that enable a bacterium to invade mucosal surfaces are not well-known. A surprising result of our study was that an S. suis mutant lacking suilysin acquires an invasion capacity when it adheres to areas of the airway epithelial cells that have been preinfected by influenza viruses. However, it has to be noted that in our study it was not possible to clearly differentiate between inter- and intracellular invasion. We have the impression that the former is more likely, but this has to be further clarified in future studies. Suilysin is considered a virulence-associated factor that facilitates S. suis invasion, as has been shown for both immortalized cells and differentiated airway epithelial cells (26, 36, 37). The cytolytic activity of suilysin contributes significantly to this effect, as shown in our previous study, in which a mutant expressing a point-mutated suilysin lacking cytolytic activity was significantly less invasive than the wt streptococci (26). Nevertheless, there was a significant difference between the invasion rates of two mutants lacking only the cytotoxic activity and mutants lacking the whole protein (26). This finding suggests that there may be invasion-promoting effects of suilysin which do not depend on the cytolytic activity.
Taken together, our results provide new insights into the role of suilysin in the mechanism of tissue invasion. We propose that suilysin-negative S. suis strains can become invasive in a coinfection scenario with influenza A viruses, which might explain why suilysin-negative strains can cause clinical infections. Finally, these results may contribute to our understanding of viral-bacterial interactions in airway coinfections.
MATERIALS AND METHODS
If not stated otherwise, all materials were purchased from Sigma-Aldrich.
Cells, influenza virus, and bacteria. (i) Cells.Madin-Darby canine kidney (MDCK) cells (44) were maintained in Eagle’s minimal essential medium (EMEM; Life Technologies) supplemented with 10% fetal calf serum and 5 mM glutamine (Life Technologies). The cells were incubated in a humidified atmosphere containing 5% CO2 at 37°C and passaged every 2 to 3 days.
(ii) Virus.Swine influenza virus (SIV) of the H3N2 subtype (A/sw/Herford/IDT5932/2007; H3N2) was obtained from Ralf Dürrwald (45). Virus stocks were propagated by infection of MDCK cells at a low multiplicity of infection (MOI) and incubation in EMEM containing 1 μg/ml acetylated trypsin. Supernatants were clarified by low-speed centrifugation (200 × g, 10 min, room temperature) and stored at −80°C.
(iii) Bacteria.Virulent Streptococcus suis serotype 2 wild-type strain 10 (designated the wt) was kindly provided by H. Smith, Lelystad, The Netherlands (46). The corresponding suilysin-deficient mutant of wt strain 10 (designated the 10Δsly mutant) was constructed as described in a previous study by insertion of an erythromycin resistance cassette into the suilysin gene (37). Streptococci were grown in Todd-Hewitt broth medium (THB; Becton, Dickinson Diagnostics) or on Columbia agar supplemented with 7% sheep blood (Oxoid) overnight under aerobic conditions at 37°C. In all infection experiments, cryoconserved bacterial stocks were used and prepared as previously reported (26). The numbers of viable bacteria were determined by serial plating on Columbia agar plates supplemented with 7% sheep blood and counting the number of CFU the next day.
Porcine airway epithelial cell culture.Porcine lungs were obtained from pigs at a local slaughterhouse. Primary porcine bronchial epithelial cells (PBEC) were harvested from the bronchi as previously described (47). Briefly, after treatment of tissue segments with antibiotics and proteases, PBEC were harvested by scraping the cells from the luminal surface of the bronchus with a scalpel and were cultured with bronchial epithelial cell growth medium (BEGM). The BEGM was modified as previously described (48, 49). When cell monolayers had reached a confluence of about 80%, primary PBEC were transferred to type IV collagen-coated transwell polycarbonate membranes (0.4-μm pore size; Corning Costar) at a density of 2.5 × 105 cells per filter and maintained with ALI medium modified as described previously (50). After the primary PBEC had reached confluence, the cells were maintained under ALI conditions for at least 4 weeks at 37°C in a humidified 5% CO2 atmosphere, and the apical surface was washed once per week with Hanks’ balanced salt solution (HBSS; Life Technologies). PBEC were screened for porcine-specific respiratory tract pathogens, including porcine circovirus 2, porcine reproductive and respiratory syndrome virus, porcine cytomegalovirus, porcine influenza A virus, porcine respiratory coronavirus, Mycoplasma hyorhinis, and Mycoplasma hyopneumoniae, by PCR. The PBEC used in this study were free from the above-mentioned pathogens.
Infection of well-differentiated epithelial cells by swine influenza virus and S. suis.For each treatment, three transwell filters were used, and all experiments were repeated at least three times. The coinfection of well-differentiated porcine airway epithelial cells was analyzed after primary infection with influenza virus, which was performed 24 h prior to secondary infection with S. suis. Briefly, PBEC (differentiated for at least 4 weeks) were maintained without antibiotics and antimycotics 1 day before viral infection, and the influenza virus H3N2 strain (5 × 104 TCID50/filter) was applied to the apical surface of PBEC that had been washed three times with phosphate-buffered saline (PBS) supplemented with calcium (PBS+) before infection. Cells inoculated with ALI medium only served as mock-infected cells. At 24 h after primary viral infection, infected PBEC were washed with PBS+ and inoculated with or without bacteria from the apical side of the filter with approximately 2.5 × 107 CFU of the S. suis wt or 10Δsly strain in 70 μl ALI medium (multiplicity of infection [MOI], approximately 50 bacteria per one epithelial cell). Cells inoculated with ALI medium only served as mock-infected controls. Bacterial infection was performed for 4 h in a humidified atmosphere containing 5% CO2 at 37°C. Afterwards, infected and mock-infected cells were washed three times with PBS+ to remove nonadhered bacteria. All cells were maintained under ALI conditions for up to 72 h in a humidified atmosphere containing 5% CO2 at 37°C. If not stated otherwise, all the time points stated in the study were the times (in hours) after bacterial infection.
Cytotoxicity assay.Cytotoxicity was detected by an LDH-release assay as described previously (36). To quantify the relative cellular damage of well-differentiated epithelial cells, results were normalized by the maximum LDH release from the ALI culture and expressed as the percent cytotoxicity of LDH release compared to the maximum LDH release of the ALI culture. Briefly, supernatants were collected from the apical compartments of coinfected, monoinfected, and mock-infected PBEC prior to the washing step at 4 h postinfection (hpi). For the determination of cytotoxicity at 24 and 48 hpi, 70 μl PBS+ was added to the apical filter compartment, and cells were incubated on a horizontal shaker for 5 min to collect the supernatant. The ALI cultures were permeabilized with 1% Triton X-100 for 4 h in a humidified atmosphere containing 5% CO2 at 37°C, followed by incubation on a horizontal shaker for 5 min to collect the maximum LDH released into the supernatant. LDH release was determined using a CytoTox 96 nonradioactive cytotoxicity assay (Promega). The experiments were performed in triplicate and repeated at least three times.
Bacterial colonization assays.To investigate streptococcal colonization of PBEC, colonizing streptococci were collected at 24 and 48 hpi. After the cells had been washed three times with PBS+ to remove nonadherent bacteria from the apical compartment, 50 μl/filter 0.05% trypsin-EDTA (final concentration) was added to the apical side of the filter at 37°C for 5 min, followed by the addition of 1% fresh saponin solution (200 μl/filter) at 37°C for 10 to 20 min. Then, the cell lysates were gently pipetted, serially diluted in PBS, and plated on Columbia agar supplemented with 7% sheep blood to determine the number of CFU, indicating the number of colonizing bacteria, as described previously (10). The results were indicated as the percentage of bacterial colonization, which was calculated as the number of CFU in the cell lysates compared to the number of CFU used in the inoculum.
Immunofluorescence analysis.ALI cultures were fixed with 3% formaldehyde for 20 min, followed by 5 min of incubation with 0.1 M glycine and three washing steps with PBS. Then, the samples were permeabilized with 0.2% Triton X-100 for 20 min at room temperature, followed by three washing steps with PBS. All antibodies were diluted in 1% bovine serum albumin in PBS and incubated with the samples for 1 h at room temperature. All the substances were incubated from both sides of the ALI cultures. Ciliated cells were visualized using a Cy3-labeled monoclonal antibody against β-tubulin (1:400; Sigma-Aldrich). For detection of virus particles and streptococci, monoclonal antibodies against the influenza A virus nucleoprotein (NP; 1:750; AbDSeroTec) and a rabbit anti-S. suis antiserum (1:200) (51) were used, respectively, followed by incubation with fluorescent secondary antibody (1:1,000; Alexa Fluor 568 anti-mouse IgG [H+L] antibody and Alexa Fluor 488 anti-rabbit IgG [H+L] antibody [Life Technologies]). The nuclei of the PBEC were stained by 1 μg/ml DAPI (4′,6-diamidino-2-phenylindole), and the membrane of the transwell filters was cut down, embedded in Mowiol mucoadhesive, and stored at 4°C for further analysis.
Samples were analyzed by using an inverse immunofluorescent microscope (Nikon Eclipse Ti-S) equipped with 10× (numerical aperture [NA], 0.30) and 40× (NA, 0.60) Plan Fluor objectives (Nikon). The area of the epithelial cell surfaces positive for green fluorescent bacteria was analyzed by applying analySIS (version 3.2) software (Soft Imaging System) to quantify bacterial adherence. Three areas were randomly chosen for each sample, all experiments were repeated at least three times, and results are presented as the percentage of bacterial fluorescence. Confocal immunofluorescence microscopy of samples was performed using a TCS SP5 confocal laser scanning microscope equipped with a 63× (NA, 1.30) glycerin HC PL Apo objective and a 63× (NA, 1.40) oil HCX PL Apo objective (Leica). Image stacks with a z or y distance of 0.5 μm per plane were acquired using a 1-Airy-unit pinhole diameter in sequential imaging mode to avoid light scattering. Maximum-intensity projections were calculated for display purposes, and brightness and contrast were adjusted.
Statistical analyses.If not stated otherwise, experiments were performed at least three times and the results are expressed as the means with standard deviations. Data were analyzed by one-way analysis of variance (ANOVA) and the Tukey multiple-comparison test, using GraphPad Prism (version 5) software. A P value of <0.05 was considered significant.
ACKNOWLEDGMENTS
This work was performed by F.M. in partial fulfillment of the requirements for a Dr. Med. Vet. degree from the University of Veterinary Medicine Hannover. Part of this work was contributed by D.V. in partial fulfillment of the requirements for a Ph.D. degree from the University of Veterinary Medicine Hannover.
This work was supported by grants from the Deutsche Forschungsgemeinschaft (DFG) to G.H. (He 1168/15-1 and He1168/19-1) and to P.V.-W. (Va2391/7-1). This project has received funding to P.V.-W. from the European Union’s Horizon 2020 research and innovation program under grant agreement no. 727966. This study was supported by a grant to X.C. from the National Key Technology Support Program of China (2015BAD11B02).
We thank Hilde E. Smith (Wageningen Bioveterinary Research, Lelystad, The Netherlands) for providing S. suis strain 10 and Ralf Duerrwald (Robert Koch Institute, Berlin, Germany) for the SIV.
FOOTNOTES
- Received 6 May 2019.
- Accepted 6 May 2019.
- Accepted manuscript posted online 28 May 2019.
- Copyright © 2019 American Society for Microbiology.
