Brucella Melitensis 16M Regulates the Effect of AIR Domain on Inflammatory Factors, Autophagy, and Apoptosis in Mouse Macrophage through the ROS Signaling Pathway

Brucellosis is a highly contagious zoonosis caused by Brucella. Brucella can invade and persist inside host cells, which results in chronic infection. We constructed AIR interference and overexpression lentiviruses to acquire AIR interference, overexpression, and rescue stable expression cell lines. We also established a Brucella melitensis 16M-infected macrophage model, which was treated with either the vehicle control or NAC (ROS scavenger N-acetylcysteine (NAC) for 0, 3, 6, 12, and 24 h. Confocal laser microscopy, transmission electron microscopy, fluorescence quantitative PCR, flow cytometry, ELISA, and Western blot were used to detect inflammation, cell autophagy and apoptosis-related protein expression levels, ROS levels, and the distribution of mitochondria. It was found that after interference and overexpression of AIR, ROS release was significantly changed, and mitochondria became abnormally aggregated. B. melitensis 16M activated the NLRP3/AIM2 inflammatory complex, and induced RAW264.7 cells to secrete IL-1β and IL-18 through the ROS pathway. B. melitensis 16M also altered autophagy-related gene expression, increased autophagy activity, and induced cell apoptosis through the ROS pathway. The results showed that after B. melitensis 16M infection, ROS induced apoptosis, inflammation, and autophagy while AIR inhibited autophagosome maturation and autophagy initiation. Autophagy negatively regulated the activation of inflammasomes and prevented inflammation from occurring. In addition, mitophagy could promote cell apoptosis.


Introduction
Brucella is a facultative intracellular parasite that invades and persists inside host macrophages [1]. Macrophages, as important immune cells in the body, play an important role in removing

Bacterial Strains and Growth Conditions
Brucella melitensis 16M (Chinese Center for Disease Control and Prevention; Beijing, China) was cultured in tryptic soy agar (TSA) or tryptic soy broth (TSB) medium (Oxoid, UK) in either a static 37˚C incubator or a 37˚C shaker. E. coli strain DH5α (Chinese Center for Disease Control and Prevention; Beijing, China) was cultured with Luria-Bertani (LB) culture medium containing antibiotics (50 mg/ml of ampicillin) ( Table 1).

Cells and Growth Conditions
HEK-293FT cells and mouse macrophages RAW264.7 (Shanghai Institute of Cell Biology, Chinese Academy of Sciences, Shanghai, China) ( Table 1), were cultured in DMEM medium containing 10% FBS (Gibco, USA) and set in a 37˚C, 5% CO 2 incubator.

Plasmids
Lentiviral overexpression vector pLEX-AIR was constructed by our lab. pLEX-MCS, pLEX--EGFP, and enveloping and packaging vectors (pSPAX2 and pMD2.G) were purchased from Xinjiang Academy of Animal Science (Xinjiang, China). Lentiviral RNAi vector pLL3.7-AIR was constructed by our lab. pLL3.7-EGFP and helper plasmids PRSV-REV and PCMV-VSVG were purchased from Xinjiang Academy of Animal Science (Xinjiang, China). pMD19-T cloning vector was purchased from Takara (Dalian, China).

AIR Interference Vector and Overexpression Vector and Identification of Stable Expression Cell Lines
According to the GenBank sequence of the mouse TECPR1 (Accession No. NM_027410), we designed PCR amplification primers (Table 2) for the AIR gene using Premier 5.0 software (Premier, Canada) while qRT-PCR primers for the mouse AIR gene (Accession No: NM_027410.1) and β-actin (Accession No: EF095208. 1), were designed according to the Gen-Bank sequences ( Table 2). Total RNA was extracted (RNA extraction kit, Cwbiotech, China) from cultured murine macrophages RAW264.7, and cDNA was synthesized from RNA through reverse transcription (HiFi-Script cDNA first strand synthesis kit, Cwbiotech, China). The AIR gene was amplified using AIR-F and AIR-R primers, and was ligated to the pMD19-T vector, which was digested by BamH I and Xho I and ligated to pLEX-MCS to generate pLEX-AIR. We also constructed pLL3.7-AIR using the similar subcloning strategy. HEK-293FT and RAW264.7 cells were transfected with the lentiviral packaging plasmids pLEX-AIR and pLL3.7-AIR, with the ratio of 5 ml of plasmids for 2 × 10 5 cells. Plasmids and cells were thoroughly mixed and placed in a 37˚C, 5% CO 2 incubator. AIR stable cell interference (I-A

qRT-PCR Detection
According to the B. melitensis 16M infection models, at 0, 3, 6, 12, and 24 h after infection, the total RNA was extracted and reverse transcribed to cDNA, with β-actin as a reference gene.

Transmission Electron Microscopy
At 6 and 12 h post-infection, 4% glutaraldehyde was added in the cells. Cells were fixed overnight at 4˚C. Glutaraldehyde was removed and replaced with 1% osmium tetroxide at room temperature for 1 h. Cells were dehydrated with a gradient concentration of ethanol and infiltrated with dehydrating agent and epoxy resin (ratio of 3:1, 1:1, and 1:3, 1 h for each step). Samples were embedded and sectioned with a microtome followed by double-staining with uranyl acetate and lead citrate. Finally, samples were observed under transmission electron microscopy (JEOL, Japan).

Flow Cytometry
At 3, 6, 12, and 24 h post-infection, cells were digested with EDTA-free trypsin and washed with PBS containing 2% BSA 2 times, centrifuged at 800 rpm for 5 min and resuspended with 500 μl of binding buffer mixed with 5 μl of Annexin V-FITC and 5 μl of Propidium Iodide. After incubation in the dark at room temperature for 15 min, an additional 2 ml of PBS was added into the cell suspension and detected by flow cytometry (Life Technology, USA).

Western Blotting
Protein lysate samples were separated by 12% SDS-PAGE gel and transferred to the NC membrane for 40 min (100 mM of Tris-HCl, 150 Mm of NaCl, 0.05% Tween 20, and pH 7.2). The NC membrane was blotted with 5% non-fat milk in TBST for 1 h, followed by incubation with a primary antibody diluted with TBST for 2 h at 37˚C. The membrane was washed with TBST 3 times and incubated with a secondary antibody at 37˚C for 2 h. After washing with PBS 3 times, the NC membrane was stained using a HRP-DAB kit (Zhongshan Golden Bridge, Beijing, China). The primary antibodies that we used are: mouse anti-p62 monoclonal antibody (Abcam, UK), rabbit anti-LC3A/B monoclonal antibody, rabbit anti-NLRP3 monoclonal antibody, rabbit anti-Caspase-3 monoclonal antibody (Cell Signaling Technology, USA), rabbit anti-caspase-1 p10 monoclonal antibody, and rabbit anti-β-actin monoclonal antibody (Jackson ImmunoResearch, USA). Secondary antibodies were Cy3-conjugated donkey anti-rabbit monoclonal antibody (Jackson ImmunoResearch, USA), HRP-conjugated goat anti-rabbit IgG, and HRP-conjugated goat anti-mouse IgG (Bioworld, USA).

Statistical Analysis
SPSS Statistics 17.0 software (IBM, USA) was used for analysis. All experimental data were denoted as mean ± standard deviation (" x ± SD). Independent sample t test was used for comparison within the group. One-way ANOVA was used for comparison between different groups. All tests were repeated 3 times.

AIR Gene Expression Levels in Differently Treated Cells
Relative gene expression of AIR was calculated using the 2 -ΔΔCt method ( Table 3 and Fig 1). Compared with the control group (NC), in the interference group I-A, AIR expression in cell I-A-1 and I-A-2 was significantly decreased (p < 0.01). The highest interfering efficiency reached 65%. In the overexpression O-A group and rescue OA-IA group, AIR expression in O-A-LC3 cells and OA-IA-LC3 cells was significantly increased after a lentiviral pLEX-AIR infection (p < 0.01). The results showed that we successfully constructed AIR interference, overexpression, and rescue cell lines.

ROS Levels in Differently Treated Cells Infected with B. melitensis 16M
Brucella-induced ROS production in RAW264.7 cells in a time-dependent manner. AIR inhibited the production of ROS, while overexpression or rescue of AIR promoted ROS production (Fig 2a). After NAC pretreatment, ROS production was significantly decreased (P < 0.01) (Fig  2b). After B. melitensis 16M infection for 6 and 12 h, ROS production in the cells was detected with confocal laser microscopy. ROS production in noninfected control cells was significantly lower than in the experimental group, and ROS production in O-A and I-A was significantly higher than the control cells while ROS production in the OA-IA group had no significant change. NAC-pretreated cells exhibited lower ROS production than the experimental group (S2 Fig).

Distribution of Mitochondria in Different Cell Groups Infected with B. melitensis 16M
At 6 and 12 h after B. melitensis 16M infection, a confocal laser microscope was used to detect the distribution of mitochondria (labeled with Mito-ID 1 RED) in the cells. I-A and O-A cells showed an abnormally significant higher accumulation of mitochondria compared with the control cells, negative control cells, and OA-IA cells (P < 0.01). The results indicated that ROS could cause cell damage by activating the inflammasomes. Therefore, mitochondria-derived ROS could activate the NLRP3 inflammasomes. After NAC pretreatment, abnormal aggregation of mitochondria in I-A and O-A cells was significantly higher than in the control cells, negative control cells, and OA-IA cells, but lower than in the untreated cells (P < 0.01) (S3 Fig  and Fig 3). TEM images showed that after B. melitensis 16M infection for 6 hours and 12 hours, I-A, O-A, and OA-IA cells showed significant abnormal aggregation of mitochondria. Mitochondria were enlarged and became round shaped and mitochondria showed multifocal vacuolar degeneration, with a lighter matrix and less cristae. The cristae structure became less clear and more disorganized, and some showed multiple cystic crests. The center of the mitochondria showed small particles or large vacuoles, indicative of damaged mitochondria. After NAC pretreatment, mitochondria showed similar aggregation and damage, but to a lesser degree compared with the untreated cells (S4 Fig).
NLRP3 protein exhibited an apparent change at 6 and 24 h after B. melitensis 16M infection, while the AIM2 protein underwent a marked change at 6 h. At 6 h after B. melitensis 16M infection, NLRP3 and AIM2 in the O-A and OA-IA cells were significantly higher than in the control cells (P < 0.01), indicating that at 6 h B. melitensis 16M infection activated NLRP3 and the AIM2 inflammasome. At 24 h after B. melitensis 16M infection, NLRP3 and AIM2 in the I-A, O-A, and OA-IA cells were lower than in the control group and NLRP3 and AIM2 in the I-A cells were significantly different from those in the OA-IA cells (P < 0.01), suggesting B.
Mark: -The expression is inhibited; + The expression is increased.
Note: Compared with the control group,* indicates a significant difference (P < 0.05) and ** indicates an extremely significant difference (P < 0.01).  5). B. melitensis 16M infection Lysates were left at room temperature for 5 min so that the protein was completely isolated from the nucleic acid complex, and then the cells were collected. Total RNA was extracted and the concentration and purity was measured. A total of 2 μg of RNA was quantified and reverse transcribed into cDNA, which was used as the template for qRT-PCR testing.
doi:10.1371/journal.pone.0167486.g001   induced the maturation and secretion of downstream proinflammatory cytokines. Secretion of IL-lβ depends on the activation of inflammatory complexes. Different pathogens activate different inflammatory complexes and induce the secretion of IL-lβ so no single molecule was involved in Brucella-induced IL-lβ and IL-18 secretion, but rather a variety of inflammatory complexes were involved in the regulation of IL-lβ and IL-18 during different infectious periods. Results from the NAC-pretreated group showed that NAC could cause Caspase-1 activation, resulting in the increased secretion of inflammatory cytokines, and thereby activating the NLRP3 and AIM2 inflammasomes (Fig 5). Caspase-1 p10 protein expression in the experimental groups was all lower than in the control group. Conversely, NAC pretreatment upregulated the Caspase-1 p10 protein level, suggesting that NAC activated Caspase-1 and thereby increased the secretion of inflammatory cytokines.

Co-localization of LC3 and Mitochondria in B. melitensis 16M-Infected Cells
After B. melitensis 16M infection, GFP-LC3 was dispersed in the control cells while experimental groups showed punctate aggregates (Fig 6), which indicated cell autophagy upon the infection of B. melitensis 16M. After AIR interference, GFP-LC3 punctate aggregates became more significant (Fig 6), suggesting that interference of AIR promoted autophagy. NAC pretreatment decreased intracellular GFP-LC3 punctate aggregates (Fig 6), indicating that autophagy was inhibited. I-A and O-A cells showed less abnormal mitochondrial aggregates than the control cells (Fig 6), suggesting that NAC could reduce the mitochondrial damage caused by B. melitensis 16M.

Expression Level of Autophagy-Associated Genes in B. melitensis 16M-Infected Cells
Relative expression levels of LC3A, LC3B, and p62 were calculated using the 2 -ΔΔCt method.
Cell autophagy-related gene expression levels in B. melitensis 16M-infected but non-NACtreated cells are shown in Fig 7. Compared with the control group, LC3A in the interference group reached the highest at 6 h; LC3B was first reduced and then gradually increased and  peaked at 6 h. Compared with the control group, P62 expression in the interference and overexpression groups was suppressed at 6 h, but was increased at other time points, indicating that autophagy was promoted at 6 h but was inhibited otherwise. After NAC pretreatment (Fig  7), compared to the negative control group, LC3A levels in the interference and overexpression groups were decreased, but the difference was not significant. LC3B in the interference group was gradually increased and peaked at 6 h, and the difference was significant (p < 0.01); however, the LC3B protein level dropped afterwards. P62 in the interference group reached the highest point at 6 h, and the difference was significant (p < 0.05), suggesting that NAC inhibited autophagy.

P62 Protein Levels in B. melitensis 16M-Treated Cells
Western blot showed the expression level of p62 at different time periods after B. melitensis 16M infection (S7 Fig). At 6 h, p62 protein expression in the I-A group was significantly lower than in the control group. It showed that after the AIR gene was silenced, compared with the control group, the relative expression level of the p62 protein showed a significant change, indicating that after AIR was silenced, autophagy activity was increased. After NAC pretreatment, at 6 h, p62 expression in the I-A group was increased, suggesting that NAC pretreatment of I-A cells inhibited autophagy activity. Compared with the control group, the overexpression and rescue groups showed no significant change in p62 protein levels.

Expression Levels of Apoptosis-Associated Genes in B. melitensis 16M-Infected Cells
Relative expression of Bcl-2 and Bax in the transcription levels were calculated using the 2 -ΔΔCt method.
Expression levels of the apoptosis-associated genes in the B. melitensis 16M-infected but NAC-untreated cells are shown in Fig 8. At 6 h, the ratio of Bax/Bcl-2 in the I-A, O-A, and OA-IA groups was significantly higher than in the control group (P < 0.01).
Expression levels of apoptosis-associated genes in the B. melitensis 16M-infected and NACtreated cells are shown in Fig 8. At 6 and 24 h, the ratio of Bax/Bcl-2 in the experimental groups was significantly lower than in the control group (P < 0.05) or (P < 0.01). At 12 h, the ratio of Bax/Bcl-2 in the B. melitensis 16M-infected experimental groups was significantly higher than that in the control group (P < 0.01).

Apoptosis in B. melitensis 16M-Infected Cells
At 6 h after B. melitensis 16M infection, confocal laser microscopy images of apoptotic cells are shown in Fig 9. Compared with the control group, after B. melitensis 16M infection, the greenand red-stained cells increased. After NAC pretreatment, compared with the control group, B. melitensis 16M infection resulted in fewer green-and red-stained cells. Phase contrast images showed the occurrence of cell apoptosis and marked the changes in cell morphology. Apoptotic cells were separated from surrounding cells and exhibited a smaller cell size, deformation, and cell body shrinkage. Cells were elongated and fragmented while nuclei were condensed or fragmented.
Transmission electron microscopy images of apoptotic cells are shown in S8 Fig. Compared with the control group, cell apoptosis in the experimental group was more significant. Control cells became smaller in size and the cytoplasm appeared dense. Chromatin underwent condensation into compact patches in the nucleus and the cell membrane began blebbing. However, apoptosis in the experimental group was more severe. Chromatins in the apoptotic cells were highly condensed and marginalized, with characteristics of late apoptosis. Nuclei broke apart into apoptotic bodies. After NAC pretreatment, cell apoptosis and apoptotic cell numbers were reduced with only a small fraction of cells showing nuclei degradation. Flow cytometry results showed apoptotic rates in different treatment groups (S9 Fig and Fig 10). With the elongation of the B. melitensis 16M infection time, the apoptosis rate was also increased. NAC pretreatment significantly reduced the apoptotic rate, indicating that B. melitensis 16M infection in macrophages induced apoptosis via the ROS pathway.

Caspase-3 Protein Levels in B. melitensis 16M-Infected Cells
Western blot was performed to detect the expression levels of Caspase-3 protein. ImageJ software (NIH, USA) was used to analyze Caspase-3 abundance relative to the internal control. The results showed that Caspase-3 protein expression levels in the experimental group were significantly lower than that in the control group (P < 0.01). After NAC pretreatment, Caspase-3 protein expression levels in the experimental groups were generally decreased, but were still higher than the in control group (S10 Fig), which suggested that B. melitensis 16M induced apoptosis of RAW264.7 cells via the ROS pathway.

Discussion
In the innate immune cell cytoplasma there is a class of multi-protein complexes called inflammation complexes, which can not only quickly identify various intracellular pathogenic microorganisms and their products, but also serves as a sensor for the host metabolic stress stimuli [27]. These inflammation complexes activate Caspase-1, affect the maturation and secretion of certain cytokines, regulate the immune response and inflammation, and play a critical role in innate immunity defense against pathogens [28,29]. These inflammation complexes can be activated by multiple stimulants [30]. In this experiment, a macrophage RAW264.7 model was used to observe the effects of B. melitensis 16M on the activation of inflammatory complexes and the role of ROS therein. Brucella induced ROS production in RAW264.7 cells in a timedependent manner. Overexpression or rescue of AIR also promoted ROS production. Caspase-1 could activate NLRP3 and the AIM2 inflammasome, thereby causing inflammation. Caspase-1 was consistent with the changes of ASC, which also confirmed that Brucella infection, similar with other intracellular parasites, activated the NLRP3 inflammasome and shared the same molecular mechanisms with a variety of other pathogenic microorganisms, especially facultative intracellular parasites. After the AIR gene was silenced, Brucella-induced atypical autophagy was further increased, demonstrating that the AIR genes were not only involved in inflammation, but also in autophagy.
Excess ROS can alter the activity of a particular enzyme by redox reaction, and participate in the regulation of autophagy and programmed cell death, and thus, exert adverse effects on the body [31]. We found that after B. melitensis 16M infection, GFP-LC3 was dispersed in the control cells while experimental groups showed punctate aggregates, which indicated cell autophagy upon the infection of B. melitensis 16M. Mitochondria are important intracellular energy production organelles, and the major source of ROS. The structure, function, and dynamics of mitochondria and ROS from mitochondria are closely related with the autophagic process [32]. We confirmed that after AIR interference, GFP-LC3 punctate aggregates became more significant, indicating that the interference of AIR promoted autophagy. Furthermore, we found that NAC pretreatment decreased intracellular GFP-LC3 punctate aggregates, suggesting that autophagy was inhibited. I-A and O-A cells showed less abnormal mitochondrial aggregates than the control cells, suggesting that NAC could reduce mitochondrial damage caused by B. melitensis 16M. If the body is dysfunctional, it may cause mitochondrial dysfunction and excessive ROS can damage the mitochondria itself and other cellular components [33].
Studies have shown that after cells undergo apoptosis, ROS production is increased, and the increase of ROS could promote cell apoptosis. Compared with the control group, B. melitensis 16M infection resulted in higher cell apoptosis, suggesting that AIR domain were involved in cell apoptosis. The dynamics of ROS determines the different reactions of cells in response to the same signal and inherent ROS levels can determine the sensitivity of certain cells to a particular signal [34]. In our current findings, the ratio of Bax/Bcl-2 in the I-A, O-A, and OA-IA groups was significantly higher than in the control group, indicating that Bcl-2 can inhibit ROS induction, which is consistent with previous research results [35]. After receiving the signals, different levels of ROS determine cell apoptosis, necrosis, or the conversion from apoptosis to necrosis [36]. Cell apoptosis was significantly reduced in B. melitensis 16M-infected and NAC-pretreated cells, indicating that there was a connection between AIR and ROS. Thus, it was substantiated that B. melitensis 16M infection-induced cell apoptosis via the ROS pathway.

Conclusions
In conclusion, B. melitensis 16M is capable of regulating the effects of the AIR domain on inflammatory factors, autophagy, and apoptosis in mouse macrophage via the ROS signaling pathway. The AIR domain participates in the inflammatory response and activation of NLRP3 promoted by B. melitensis 16M through the ROS signaling pathway. After the interference of the AIR domain, B. melitensis 16M-induced atypical autophagy was promoted, indicating that the AIR domain is not only involved in the cellular inflammatory response, but also in the autophagy. In addition, B. melitensis 16M promotes apoptosis, which was positively correlated with the time of infection. The apoptotic rate of the NAC-pretreatment group was significantly lower than that of the untreated group, which showed that AIR could affect B. melitensis 16M apoptosis induced by B. melitensis 16M via ROS pathway. The study provides a reliable theoretical basis for the systematic exposition of the molecular mechanism of B. melitensis 16Mpersistent infection.
Supporting Information S1 Fig. RAW264.7 cells were transfected with recombinant lentivirus. RAW264.7 cells were seeded into 6-well plates and infected with 3-5 mL lentivirus. Polybrene (1 mg/mL) was added to the cell medium at a final concentration of 1 μg/mL, and was mixed well, and then placed in a 37˚C, 5% CO 2 incubator. After 12 h, the cell morphology was observed and infected with a second round of lentivirus (or control medium) and polybrene. Cells were placed back in a 37˚C, 5% CO 2 incubator. Cell fluorescence was observed 48 h later. After trypsin was discarded, cells were fixed with 4% glutaraldehyde. Cells were fixed again with 1% osmium tetroxide, followed by ethanol dehydration and penetration of the epoxy resin. Samples were sectioned with microtome and stained with uranyl acetate and lead citrate. Both the untreated and NAC-pretreated groups were infected with B. Melitensis 16M. At 6 and 12 h after infection, the cells were digested with 0.25% trypsin. After trypsin was discarded, cells were fixed with 4% glutaraldehyde. Cells were fixed again with 1% osmium tetroxide, followed by ethanol dehydration, and penetration of the epoxy resin. Samples were sectioned with microtome, and stained with uranyl acetate and lead citrate. Autophagosome were observed under a transmission electron microscope. (A, a)