Rabbit polyclonal antibody against LC3B was obtained from Sigma-Aldrich (Sigma, St. Louis, MO, USA), goat polyclonal antibodies against Beclin-1, Atg5, and Atg7, and GAPDH antibody were obtained from Santa Cruz Biotechnology (California, USA). Goat anti-mouse and anti-rabbit antibodies labeled with horseradish peroxidase (HRP) were acquired from Boster (Wuhan, China). The enhanced chemiluminescent (ECL) chromogenic kit was purchased from Biouniquer (Nanjing, China). Recombinant adenovirus tandom expressing GFP-RFP-LC3 was purchased from HANBIO (Shanghai, China), which was used to increase the expression of LC3 and facilitate the observation of fluorescence.

Three standard serotypes of H. parasuis (serovars 4, 5, and 11; designated Hps4, Hps5, and Hps11) were used in this study. Bacteria were cultured at 37°C on Trypticase Soy Agar and in Trypticase Soy Broth (TSA and TSB, respectively; OXOID, Hampshire, England) supplemented with 0.01% nicotinamide adenine dinucleotide (NAD, Nanjing, JS, China) and 5% (v/v) inactivated bovine serum.

Bacteria were grown overnight and then inoculated in fresh TSB medium (1:100). Optical density at 600 nm was measured at 1 h intervals to establish growth curves. Bacteria from exponential phase cultures were serially diluted and plated onto TSA to determine the relationship between OD and viable bacterial counts (colony forming units).

Porcine kidney (PK-15) cells were cultured at 37°C in a humidified 5% CO₂ incubator and maintained in Dulbecco's Modified Eagle Media (DMEM, Gibco), supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco) and antibiotics (100 U/ml of penicillin G and 100 mg/ml of streptomycin, Sigma, USA). Cells were sub-cultured when 90% cell confluence was attained. In 24-well tissue culture plates, cells were diluted in culture medium to a concentration of 10⁵ cells per well and incubated to 90% confluence.

Bacteria were grown overnight in TSB broth at 37°C with shaking as previously described (Cue et al., 1998), diluted 1:100 in fresh medium and incubated until reaching an optical density of 0.7 at 600 nm. Bacteria were pelleted by centrifugation at 8000 g for 1min, washed 3 times with PBS, and resuspended in DMEM at 0.25 OD (0.1OD = 1 × 10⁸ cells/ml) for the infection experiments (Yuan et al., 2012). PK-15 cells cultivated overnight in serum-containing medium were washed with sterile PBS and transferred to serum-free and antibiotic-free medium. The cells were infected with H. parasuis suspensions at different multiplicities of infection (MOI). At various time points post-infection, cells were treated with different methods as follows.

Adherence and invasion assays were performed as previously described (Cue et al., 1998; Vanier et al., 2006; Zhang et al., 2013) with modifications. Confluent monolayers of PK-15 cells were grown in 24-well plates and infected with H. parasuis suspensions at an MOI of 100. The plates were centrifuged at 800 × g for 10 min to promote contact of H. parasuis with the cell monolayer surface, incubated for 1 h in 5% CO₂ at 37°C to allow bacterial adherence, and then vigorously washed five times with PBS to remove non-specifically attached bacteria.

For the adherence assay, after the removal of non-specifically attached bacteria, the plates were incubated for 10 min at 37°C with 100 μl of 0.025% trypsin. Following incubation, cells removed by scraping were transferred to sterile centrifuge tubes and centrifuged for 1 min at 12,000 rpm. The supernatant was removed and the pellet was resuspended with an equivalent volume of TSB. Serial dilutions of the cell lysate (100 μl) were plated onto Trypticase Soy Agar and incubated for 48 h at 37°C.

For the invasion assay, after the removal of non-specifically attached bacteria, culture medium containing two antibiotics (100 mg of penicillin G/ml and 5 mg of gentamicin/ml, Sigma) was added to each well, and the plates were incubated for 1 h at 37°C in 5% CO₂ to kill any remaining extracellular H. parasuis. The monolayers were washed three times and cells were harvested as described above.

Bacterial adhesion/invasion was evaluated as the average number of bacteria per well (CFU, colony forming unit). All assays were performed in triplicate and repeated three times.

Cell samples were lysed in RIPA buffer (Cell signaling Technologies) on ice for 30 min, then supernatant was collected through centrifugation at 12,000 rpm for 5 min. The lysate mixed with loading buffer was incubated for 8 min in a boiling water bath. Sample was loaded onto a 10% SDS-polyacrylamide mini-gel and subjected to electrophoresis. The proteins were transferred onto NC membranes (BioTrace^TM NT, Pall Corporation, United States) and blocked for 2 h at room temperature using 10% non-fat milk blocking buffer. Membranes were incubated overnight at 4°C with primary antibodies diluted to appropriate concentration in accordance with the instruction with PBST. After three rinses with PBST, the membranes were incubated for 45 min at room temperature with horseradish peroxidase-conjugated secondary antibody diluted 1:10000 (Wu et al., 2002). Bound antibody was visualized using an enhanced chemiluminescence detection kit (Super Signal West Pico; Pierce). For quantification, protein band density was analyzed using the image processing program ImageJ (Schneider et al., 2012).

PK-15 cells were grown on coverslips in a 24-well plate and infected with adenovirus at an MOI of 1. After 24 h, a subset of the cells were pretreated with the autophagy activator rapamycin for 12 h, or the autophagy inhibitor 3-MA for 3 h, as positive and negative controls, respectively. Adenovirus infected/non-treated cells and adenovirus infected/3-MA treated were infected with H. parasuis at 0.1 MOI. Cells were washed five times with sterile PBS to remove non-specifically attached bacteria. Coverslips were mounted on microscope slides and samples were examined using a Zeiss LSM 700 confocal microscope.

Cells in 6-well plates were infected with adenovirus at an MOI of 1 for 24 h and treated with rapamycin, 3-MA+ H. parasuis, H. parasuis or left untreated, respectively. Cells were harvested, centrifuged for 1 min at 300 × g, resuspended, and fixed in 2.5% glutaraldehyde at 4°C for 12 h, post-fixed in 1% osmium tetroxide, and finally dehydrated in a graded series of ethanol concentrations. Samples were treated in propylene oxide and embedded in epoxy resin for thin sectioning. Sections were double stained with uranyl acetate and lead citrate before being observed by TEM (Faulkner and Garduno, 2002; Zhang et al., 2015).

siRNA fragments specifically targeting Atg5, Atg7, Beclin-1, and one nonspecific random fragment (NC) were synthesized by Invitrogen. All siRNA sequences are listed in Table 1. For optimizing the conditions for gene silencing, PK-15 cells were seeded in 24-well plates at 1 × 10⁵ cells per well. At approximately 70% confluence, cells were incubated with several concentrations of siRNA fragments for different lengths of time, following the manufacturer's instructions. 24 h after transfection, cells were harvested and lysed for Western blot analysis. Then, the cells transfected with optimal siRNA fragments for Atg5/Atg7 (0.5 μg/well) and Beclin-1 (0.17 μg/well), were infected with H. parasuis suspensions at 0.1 MOI for 6 h, and then harvested and lysed for Western blot analysis.

Three biological replicates were performed for every experiment and results were recorded as mean ± SD and GraphPad Prism 5 was used to analyze difference between groups (one-way ANOVA; Tukey's post-hoc test, ^***p < 0.001, ^**p < 0.01, ^*p < 0.05). ImageJ software was used to confirm the band strength of WB and fold difference compared with control was labeled with number.

To determine whether infection by H. parasuis can induce autophagy, PK-15 cells were infected with adenovirus. After confirming successful expression of GFP-RFP-LC3, the PK-15 cells were infected with a highly virulent strain Hps5, in a dose- and time dependent manner (Figures S1, S2). We observed that Hps5 infection with a bacterial:cell multiplicity of infection (MOI) of 1:10 induced the most significant LC3- II level at 6 h (Figure S2, p < 0.001). Surprisingly, compared with control group, the level of LC3-II was decreased at 1–2 h and increased at 4–6 h during the infection (Figure S2).

Autophagy induced in PK-15 cells was further confirmed with western blotting, confocal laser scanning and transmission electron microscopy. PK-15 cells were first infected with adenovirus, then cells were either (i) infected with Hps5, or (ii) treated with rapamycin for 12 h, or (iii) treated with 3-MA for 3 h before infected with Hps5, or (iv) left untreated and uninfected. As shown in Figures 1A,B, cells infected with Hps5 or treated with rapamycin exhibited increased LC3 punctuation. Consistent with these results, western blot analysis showed that LC3-II levels was increased significantly in cells infected with H. parasuis (Figure 2).

Transmission electron microscopy results showed that the autophagosome, a double membrane structure with heterogeneous contents, is apparent in cells infected by Hps5 (Figure 3B) or treated with rapamycin (Figure 3C). Uninfected/untreated cells (−) (Figure 3A) and cells treated with 3-MA before infection with H. parasuis (Figure 3D) have no obvious autophagosomes. In summary, these results demonstrate that Hps5 infection induces autophagy in PK-15 cells.

Adherence and invasion assays were performed to determine if the autophagy induced by H. parasuis results from cell adherence or cell invasion. Briefly, PK-15 cells were infected at MOI 100 with strains Hps5 (highly virulent), Hps4 (less virulent), and Hps11 (avirulent) for 1 h. As shown in Figure 4A, Hps5 adhered to PK-15 cells was 6-fold more than Hps4 and 17-fold more than Hps11; In Figure 4B, invasion Hps5 in PK-15 cells was 5-fold more than Hps4 and 8-fold more than Hps11.

To explore the relationship between autophagy and H. parasuis pathogenesis, PK-15 cells preinfected with adenovirus at 0.1 MOI were infected with H. parasuis strains having different virulence: Hps5 (highly virulent), Hps4 (less virulent), and Hps11 (avirulent). Six hours after infection, cells were collected for western blot analysis. Using the mock-infected control as a baseline, the level ratio of LC3 II/GAPDH were calculated. The ratios were 1.50, 1.28, and 1.15 for the infection with Hps 5, Hps 4, and Hps 11, respectively (Figure 5A).

In another experiment, PK-15 cells pre-reated with adenovirus were infected with viable and UV-inactivated Hps5 at MOI 100, incubated, and then harvested for Western blot analysis to detect LC3-II level. The results showed that viable H. parasuis induces more autophagy than either the mock-infected control or inactivated Hps5 (Figure 5B). Given these results, we concluded that autophagy is predominately induced in porcine kidney epithelial cells by the presence of alive bacteria in cells.

Although autophagy ordinarily functions as a defense against bacterial invasion, some pathogens subvert the autophagic pathway to obtain nutrition for survival and replication, and ultimately to establish a persistent infection. In order to explore the role of autophagy in Hps5 infection, PK-15 cells were infected with adenovirus. After confirming successful expression of GFP-RFP-LC3, the cells were incubated with the autophagy activator rapamycin (1 μM, 12 h). Having confirmed that autophagy was induced in pretreated cells in comparison with untreated cells (Figure 6A), the treated and untreated cells were infected with Hps5 at MOI 100 and then the invasion assay was performed as described above. As shown in Figure 6B, the number of intracellular bacteria decreased significantly in rapamycin treated cells (p < 0.05). This result suggests that autophagy inhibited the invasion of H. parasuis.

The classical autophagy pathway is mediated by a large number of autophagy related genes (Atg), among which are Atg5, Atg7, and Beclin-1 (Atg6) (Ohsumi and Mizushima, 2004; Yuan et al., 2012). To determine whether this pathway is involved when H. parasuis infection induces autophagy, siRNA fragments specifically targeting Atg5 (siRNA-21/siRNA-42), Atg7 (siRNA-138/siRNA-846), Beclin-1 (siRNA-101/siRNA-125) (Table 1) and one nonspecific random fragment (NC) were synthesized and used in gene expression silencing experiments. The concentration and timing of interference effects were firstly determined by western blot analysis, the best interference result was shown in Figure 7A. After transfection using optimal conditions, PK-15 cells were infected by Hps5 (highly virulent), and untransfected cells were infected as controls. The level of LC3-II was significantly reduced by all three specific interfering fragments (Figure 7B). We conclude that Atg5, Atg7, and Beclin-1 play an important roles in the signaling pathway of the autophagy induced in PK-15 cells by H. parasuis infection.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.