Peripheral blood mononuclear cells (PBMCs) were prepared from healthy adult donors by centrifugation over Ficoll-Histopaque (Sigma, St. Louis, MO, USA) as described (21). All the donors had signed an informed consent before their blood was used in this study. M/MΦs cells were enriched from PBMCs using Ficoll-Percoll gradients (GE Healthcare) and further purified by anti-CD14 magnetic beads with column purification according to the manufacturer’s instructions (Miltenyi Biotec; purity of cells >95%). The enriched cells were cultured in RPMI 1640 (Invitrogen) supplemented with 10% fetal bovine serum (FBS, Hyclone).

The human monocytic cell line THP-1 and HEK-293 cells were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). THP-1 were maintained in RPMI 1640 medium and HEK-293 cells were maintained in Dulbecco’s minimum essential medium, both supplemented with 10% FBS at 37°C with 5% CO₂.

Brucella suis 1330 (ATCC 23444) and its derived mutants were cultured in tryptic soy broth at 37°C. Omp25-deficient mutant (Δomp25 B. suis) were obtained using suicide plasmid PCVD442 carrying omp25 gene interrupted by kanamycin resistance gene as previously described (20). In this study, all live Brucella experiments were performed in biosafety level 3 facilities according to standard procedures. Either deleted omp25 gene or expressed Omp25 protein was omp25, not omp25b, omp25c, omp23d, or omp22 (22). The native omp25 gene (GenBank No. U39397.1) was amplified by PCR from B. suis 1330 using omp25 specific Primers with Flag tag encoding sequence (Table S1 in Supplementary Material), then omp25 gene was inserted into the vector pShuttle-CMV (Clontech) to construct the adenoviral shuttle vector pShuttle-CMV-Omp25. The linearized pShuttle-CMV-Omp25 and pShuttle-CMV were transformed into BJ5183-AD-1 by electroporation to recombine with pAd-Easy-1, followed by amplification and extraction of the recombinant vector named pAd-Omp25 and pAd-Blank. The recombinant adenovirus named recombinant adenoviruses expressing Omp25 protein (rAd-Omp25) and rAd-Blank were obtained by transferring the linearized pAd-Omp25 and pAd-Blank to HEK-293 cells. The titer of rAd-Omp25 and rAd-Blank were measured by the method of TCID₅₀ as described in previous studies (23).

Human THP-1 cells were cultured with 100 nM 1,25-dihydroxyvitamin D3 (VD3, Sigma) for 72 h before Brucella infection as previously described (24). THP-1 cells and human M/MΦs were infected with B. suis as previously described (24, 25). Briefly, 1 × 10⁶ cells/ml or 2 × 10⁵ cells/ml of VD3-treated THP-1 cells or human M/MΦs were cultured in RPMI 1640 medium with 10% heat-inactivated FBS and infected with B. suis 1330 or mutated B. suis strains at a multiplicity of infection of 50 for 1 h at 37°C in a 5% CO₂ atmosphere. At the end of the incubation time (time 0 p.i.), cells were washed three times with sterile PBS to remove uninternalized bacteria, and then cells were cultured in standard medium with 50 µg/ml of gentamicin (Sigma) and 50 µg/ml of streptomycin (Sigma) to kill remaining extracellular bacteria.

To monitor Brucella survival in cells, infected cells (5 × 10⁵ cells/well) were washed with PBS and lysed in 0.2% Triton X-100 (Sigma). Serial dilutions of lysates were rapidly plated on tryptic soy agar plates to count and calculate the number of intracellular viable bacteria in CFU per well.

Sequences of predicted mature human miRNAs were obtained from miRBase.¹ Il12A and il12B sequences were from the National Center for Biotechnology Information Nucleotide database.² MiRNAs predicted to bind to 3′ untranslated regions (UTRs) of il12A and il12B genes were selected by target analysis through the prediction algorithms Targetscan human and RNAhybrid.

MicroRNAs and IL-12 p40 or p35 subunit mRNA was quantified by quantitative polymerase chain reaction (Q-PCR). Total cellular RNA was isolated by TRIZOL according to the manufactory’s protocol (26). RNA concentration and purity were measured using a NanoDrop spectrophotometer (Thermo), and equal amounts of RNA (10 ng for miRNA or mRNA) were used for Q-PCR analysis. MiRNA and mRNA levels were detected by SYBR Green kit using Bio-Rad IQ5 Real-time system and calculated using RNU6B and β-actin as endogenous control, respectively, following the 2^−ΔΔCt method. Primers used in this assay were presented in Table S1 in Supplementary Material.

Cells were lysed in RIPA plus protease/phosphorylase inhibitors (Sigma) on ice. The cytosol and nuclear fraction were isolated using NE-PER Nuclear and Cytoplasmic Extraction Reagents according to the manufactory’s instruction (Thermo). Protein from each extract were separated on 4–12% SDS-PAGE and transferred to polyvinyl difluoride membranes (Millipore). After blocking with TBS plus 0.05% Tween 20 containing 5% non-fat milk for 1 h, the membranes were incubated with specific primary antibodies at 4°C overnight. Primary antibodies include anti-IL-12 p40/p35 (Santa Cruz), anti-TAB2, anti-phospho-IκB, anti-IκB, anti-p65, anti-TRAF6, anti-IRAK1, anti-IRAK2, anti-Histone H3 (Cell Signaling Technology), anti-FLAG M2 (Sigma), anti-Omp25, and anti-β-actin (Tianjin Sungene Biotech) antibodies. HRP-conjugated anti-mouse IgG or anti-rabbit IgG (Santa Cruz) were used as secondary antibodies. Pierce^® Fast Western Blot Kit, ECL Substrate (Thermo) was used for chemiluminescent detection.

Human il12A or il12B promoter sequence was cloned into pGL3 basic vector (Promega) to construct the promoter report vector pGL-il12A, pGL-il12B using specific primers (Table S1 in Supplementary Material). The pGL3 basic vector served as a control. THP-1 cells were transfected with a mixture of pGL-il12A, pGL-il12B, or NF-κB activity reporter plasmid and pRL-TK-renilla-luciferase plasmid (Promega, for normalization) using Amaxa nucleofection technology (Human monocyte Nucleofector Kit, Lonza). Cells were then infected with rAd-Blank or rAd-Omp25. At 24 h post infection, luciferase activities were determined via Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer’s instructions.

The full-length 3′ UTR of the human il12A gene (GenBank accession no. NM_000882) or il12B gene (GenBank accession no. NM_002187) was amplified from human THP-1 cDNA using specific primers (Table S1 in Supplementary Material) and cloned into pMIR-REPORT Luciferase vector (Ambion). The segment (base pairs 984–1450) of the il12A 3′UTR containing the mutated miR-21-5p target sequence (ATAAGCT to TATTCGA) and the segment (base pairs 1030–2347) of the il12B 3′UTR containing the mutated miR-23b target sequence (AATGTGA to TTACACT) were also cloned into the pMIR-REPORT Luciferase vector (Ambion). The primers for mutants were also showed in Table S1 in Supplementary Material. HEK-293 cells were transfected with the indicated luciferase reporter plasmid (500 ng), pRL-TK-Renilla-luciferase plasmid (100 ng), and the indicated miRNA mimics (final concentration, 10 or 50 nM) by Lipofectamine 2000 (Invitrogen). At 48 h post transfection, cellular extracts were prepared and measured luciferase activities by the Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer’s instructions.

Cell surface molecules and IL-12 expression in monocyte/macrophages were determined by flow cytometric analysis as described previously (27, 28). Briefly, 1 × 10⁶ cells were infected with WT B. suis, Δomp25 B. suis, rAd-Omp25 or rAd-Blank for 24 or 48 h to harvest the cells for immune staining. Cells were first stained for cell surface molecules using APC-anti-human CD14 Ab and PE-anti-PD-1 or PE-anti-PD-L1 Ab (Biolegend) in staining buffer and were then treated with fixation and permeabilization buffer (Biolegend) for intracellular staining according to the manufacturer’s instruction. Stained cells were analyzed by flow cytometry (BD Accuri C6) and further analyzed on flowJo V7.6 software as previous (29). Commercial IL-12 p70 ELISA kits (Biolegend) were used to measure the levels of IL-12 p70 in cell-free culture supernatants according to the manufacturer’s instructions.

All experiments were performed at least three times, and results are representative of three independent experiments. Prism software (GraphPad Software) was used for statistical analysis. Data were presented as means ± SEM. Statistical comparisons of the results were made using ANOVA followed by Bonferroni post hoc test or performed by unpaired Student’s t-test. Pearson’s correlation coefficients were calculated to examine the correlations between PD-1⁺ or PD-L1⁺ cells with IL-12⁺ cells. P < 0.05 or less was considered significant.

Omp25 is an outer membrane protein on the surface of Brucella, and Omp25-deficient mutants have been shown to exhibit enhanced ability to secret TNF-α compared with WT Brucella (5, 20). Since IL-12 is also an important cytokine for clearance of intracellular pathogens, we investigated the effects of Omp25 on the expression of IL-12. We first confirmed the presence of Omp25 in WT B. suis-infected cells and the absence of Omp25 in the corresponding Omp25-deficient mutant (Δomp25 B. suis)-infected cells (Figure S1 in Supplementary Material). Next, we observed the intracellular survival of WT B. suis and Δomp25 B. suis, and compared the production of IL-12 in WT B. suis- and Δomp25 B. suis-infected human THP-1 cells. As would be expected, the numbers of intracellular viable bacteria were similar for WT and Δomp25 B. suis infection (Figure S2 in Supplementary Material), but Δomp25 B. suis-infected cells produced more IL-12, ranging from 210 to 256 pg/ml, than that of WT (Figure 1A, P < 0.01). Similarly, the level of IL-12 p40 mRNA was higher in the Δomp25 B. suis-infected cells than that in WT B. suis-infected cells (Figure 1B), suggesting that the absence of Omp25 might improve B. suis’s ability to induce IL-12. To further confirm whether Omp25 play a key role in inhibition of IL-12 production, THP-1 cells were infected with recombinant adenoviruses rAd-Omp25 to express Omp25, and then infection of Δomp25 B. suis and measured the production of IL-12. Results showed that the IL-12 p70 production of rAd-Omp25-infected cells was significantly lower than that of control blank adenoviruses (rAd-Blank)-infected cells upon Δomp25 B. suis infection (Figure 1C), in which cells were infected with rAd-Omp25 to complement Omp25 (Figure S1 in Supplementary Material). These results indicated that the presence of Omp25 inhibited B. suis’s ability to induce IL-12.

To investigate the direct effects of Omp25 protein on IL-12 p70 production in macrophage, we examined the LPS/R848-induced IL-12 p70 production in THP-1 cells infected with rAd-Omp25 or control blank adenoviruses (rAd-Blank). We found that Omp25 was stably expressed in the THP-1 cells infected with rAd-Omp25 (Figure 2A), which inhibited IL-12 p70 production induced either by LPS/R848 or recombinant adenoviruses themselves (Figure 2B). We also observed that the expression of IL-12 p35 and p40 subunit induced by LPS/R848 was lower in rAd-Omp25-infected cells compared to that in rAd-Blank-infected cells (Figures 2C–E). Similarly, upon LPS/R848 stimulation, the il12B (p40) mRNA level of rAd-Omp25-infected cells was lower than that of rAd-Blank-infected cells, but il12A (p35) mRNA level was not significantly difference between these two cells (Figure 2F). These results indicated that Omp25 can directly inhibit the production of LPS/R848-induced IL-12 p70 through suppressing p40 subunit expression at both protein and mRNA levels, yet suppressing p35 subunit only at protein level in THP-1 cells. Our data also suggested that the expression of both p35 and p40 subunits are regulated at the posttranscriptional level in Omp25-expressed cells.

To investigate the effects of Omp25 on the transcription of p35 and p40 subunit, we examined the promoter activities of il12A and il12B and NF-κB transcriptional activities in rAd-Omp25-infected cells and rAd-Blank-infected cells. Omp25 expression did not appear to affect IL-12 p35 transcription (Figure 3A), but did inhibit IL-12 p40 transcription (Figure 3B). Upon LPS/R848 treatment, the transcriptional activities of NF-κB were lower in rAd-Omp25-infected cells than that in rAd-Blank-infected cells (Figure 3C). In line with the changes of NF-κB activities, the phosphorylation levels of IκB and the levels of nuclear NF-κB p65 unit were lower in rAd-Omp25-infected cells than that in rAd-Blank-infected cells at 0, 3, and 6 h after LPS/R848 treatment (Figure 3D). Given the key roles of NF-κB p65 unit in the transcription of il12B, these results further demonstrated that Omp25 inhibits the expression of p40 subunit but not p35 subunit at the transcriptional level.

To investigate the potential role of microRNAs in regulating IL-12 p40 and p35 in Omp25-expressed cells, we used multiple prediction algorithms, including Targetscan human, miRanda, and RNA hybrid, to search for conserved microRNAs whose target sites exist in the il12A and il12B 3′ UTRs of human and other mammals. In total, seven miRNA–mRNA target duplexes were predicted by these algorithms (Figure S3 in Supplementary Material), including miR-590-5p, miR-21-5p, miR-340-5p, miR-23a, miR-23b, miR-23c, and miR-494-3p. To determine which miRNAs might be regulated by Omp25 in THP-1 cells, we examined the expression profile of these miRNAs, as well as some other miRNAs that have been reported to regulate NF-κB signaling, such as miR-146a, miR-146b, and miR-155. Q-PCR profiling of 10 miRNAs revealed that miR-21-5p, miR-23b and miR-155 were higher in rAd-Omp25-infected cells than that in control rAd-Blank-infected cells (Figure 4A). This result suggested that these three miRNAs might be involved in the regulation of IL-12 expression in Omp25-expressed-THP-1 cells.

To identify the regulatory roles of miR-21-5p, miR-23b, and miR-155 in the inhibitory effects of Omp25 in THP-1 cells, we first examined the characteristics of miR-21-5p, miR-23b and miR-155 in THP-1 cells after expression of Omp25. A detailed time-course experiment showed that miR-23b, miR-155 and miR-21-5p levels were significantly increased at 16, 24 and 36 h as shown in the Figures 4B–D left panels after infecting rAd-Omp25, respectively. To further understand the kinetics of these miRNAs in the process of Omp25 expression, the fold changes of miRNAs expression were calculated by comparing the value of Omp25-expressing cells to that of control cells without Omp25 in parallel. MiR-23b showed an average increase of 2.4-fold after 16 h, rising to 5.9-fold by 36 h (Figure 4B right panel). MiR-21-5p expression showed increases of 2.3-fold by 16 h of infection and further increased at 24 and 36 h after infection (Figure 4C right panel). MiR-155 increased 2.7-fold by 16 h of infection and further increased by 4.72-fold and 5.77-fold at 24 and 36 h, respectively (Figure 4D right panel). However, no significant change was observed for the expression of miR-146a, which served as a negative control (Figure 4E). These results confirmed that the expressions of miR-21-5p, miR-23b and miR-155 are induced by Omp25, consistent with the original screen data shown in Figure 4A.

Given above results showing that Omp25 inhibits LPS/R848-induced IL-12 expression and upregulates miR-21-5p, miR-23b and miR-155 in THP-1 cells, we hypothesized that miR-21-5p, miR-23b and miR-155 likely played important roles in the Omp25 inhibition of LPS/R848-induced IL-12 p70 production. To test this hypothesis, we transfected cells with miR-21-5p mimics, miR-23b mimics or miR-155 mimics and stimulated the transfected cells with LPS/R848 for 24 h. IL-12 p40 was significantly decreased in the cells transfected with miR-23b mimics or miR-155 mimics, but not in the miR-21-5p mimics-transfected cells (Figure 5A). IL-12 p35 was significantly decreased in the cells transfected with miR-21-5p mimics, but not in the miR-23b or miR-155 mimics-transfected cells (Figure 5B). After treatment with miR-21-5p, miR-23b or miR-155 inhibitors, the expression of the respective miRNAs was downregulated over 2.5-fold, compared with control-transfected cells (Figure S4 in Supplementary Material). Inhibition of miR-23b or miR-155 induction suppressed the Omp25-induced reduction of IL-12 p40, while having no discernible effect on the Omp25-induced reduction of IL-12 p35 (Figures 5C,D). Inhibition of miR-21-5p induction suppressed the Omp25-induced reduction of IL-12 p35, while having no discernible effect on the Omp25-induced reduction of IL-12 p40 (Figures 5C,D). Consequently, inhibition of miR-21-5p, miR-23b or miR-155 induction did not significantly suppress the Omp25-induced reduction of IL-12 p70, but inhibition of both miR-21-5p and miR-23b induction, or inhibition of both miR-21-5p and miR-155 induction, significantly improved LPS/R848-induced IL-12 p70 production in the rAd-Omp25-infected THP-1 cells (Figure 5E). Notably, in cells transfected with a mix of miR-21-5p, miR-23b and miR-155 inhibitors, the inhibitory effects of Omp25 on LPS/R848-induced IL-12 p70 production were further attenuated, resulting in a level of IL-12 p70 that was higher than other inhibitor-treated cell after rAd-Omp25 infection (Figure 5E). These results demonstrated that downregulation of miR-21-5p, miR-23b, and miR-155 can block the inhibitory effects of Omp25 on LPS/R848-induced IL-12 p70.

Inhibition of miR-21-5p and -23b attenuates the inhibition effects of Omp25 on LPS/R848-induced IL-12 p35 and p40 expression, respectively. To confirm the possibility that IL-12 p35 and p40 were regulated posttranscriptionally by these Omp25-inducible miRNAs, we constructed reporter plasmids encoding the complete WT 3′UTR of the human il12A and il12B mRNA downstream of the firefly luciferase gene (il12A or il12B WT-3′UTR); parallel plasmids containing mismatches in the predicted miR-21-5p or -23b binding sites (miR-21-5p or 23b MT-3′UTR) of the 3′UTR region were also made (Figure S5A in Supplementary Material). In HEK-293 cells transfected with the reporter plasmids together with the indicated miRNA mimics and an internal control, pRL-TK-Renilla-luciferase, we observed that miR-21-5p and miR-23b reduced the expression of luciferase from il12A and il12B WT-3′UTR compared with either control miRNA mimics transfections (Figure S5B in Supplementary Material). However, miR-21-5p and miR-23b did not reduce luciferase expression from their respective MT-3′UTR plasmids containing specific mutations in their cognate binding sites (Figure S5B in Supplementary Material). These experiments validated the regulatory potential of miR-21-5p and miR-23b via binding with the il12A and il12B WT-3′UTR, respectively, and further confirmed that Omp25-induced miR-21-5p and miR-23b inhibit the expression of IL-12 p35 and p40 induced by LPS/R848 at the posttranscriptional level, respectively.

MiR-155 is an NF-κB transactivational target and involved in a negative feedback loop through downregulation of TAB2 or other genes. Our data suggested that miR-155 might be an additional regulator at the transcriptional level in rAd-Omp25-infected cells. We thus hypothesized that miR-155 likely plays a role in Omp25’s inhibition of IL-12 p40 transcription by targeting some key molecules involved in the activation of NF-κB pathway. Indeed, we found that miR-155 inhibitor markedly decreased the suppression of Omp25 to the transcription of il12B compared with the inhibitor control, but miR-146a and miR-23b inhibitors did not affect the transcription of il12B (Figure 6A). Consequently, miR-155 inhibitor attenuated the suppression of Omp25 on IL-12 p40 production, and together with miR-23b inhibitor, it further blocked the inhibitory effects of Omp25 on the production of IL-12 p40 in THP-1 cells (Figure 6B). We also found that TAB2 was decreased in rAd-Omp25-infected cells compared with rAd-Blank-infected cells, but no change was observed for TRAF6, IRAK1, IRAK2, or IκB (Figure 6C). MiR-155 mimics decreased the level of TAB2 protein expression, suggesting that TAB2 is the target of miR-155 in THP-1 cells (Figure 6D). To further identify the regulatory roles of miR-155 in TAB2 in Omp25-expressed cells, cells were transfected with miR-155 inhibitors or an inhibitor control and infected with rAd-Omp25 or rAd-Blank, followed by stimulation with LPS/R848 for 0 and 3 h. In the presence of inhibitor control, the levels of phospho-IκB and TAB2 were decreased in rAd-Omp25-infected cells compared with rAd-Blank-infected cells, but the reduction of phospho-IκB and TAB2 were partly reversed in the presence of miR-155 inhibitor (Figure 6E). These results suggested that Omp25-induced miR-155 inhibits the transcription of il12B via targeting TAB2, and downregulation of miR-155 and -23b can block the suppression of Omp25 on IL-12 p40 expression at transcriptional and posttranscriptional levels, likely by targeting TAB2 and il12B 3′UTR, respectively.

Now that Omp25 can directly induce miR-21-5p, miR-23b and miR-155 in monocytic THP-1 cells, we suspected if WT B. suis could induce a higher level of these miRNAs than the Omp25-deficient mutant. Indeed, we found that the levels of miR-21-5p, miR-23b and miR-155 were higher in the human M/MΦs infected with WT B. suis than that in the M/MΦs infected with Omp25-deficient mutant (Δomp25 B. suis) (Figure 7A). Upon stimulation with LPS/R848, Δomp25 B. suis-infected human M/MΦs produced more IL-12 than WT B. suis-infected human M/MΦs (Figure 7B). These results further demonstrated that Omp25 plays a key role in the dysfunction of Brucella-infected monocyte/macrophages.

Several studies have showed that PD-1 is upregulated in monocyte/macrophages and associated with dysfunction during different infections (30, 31). To determine whether PD-1 is also upregulated and involved in the regulation of IL-12 production in M/MΦs during Brucella infection, we examined PD-1 expression on the M/MΦs infected with WT or Δomp25 B. suis, or on the M/MΦs infected with rAd-Omp25 or rAd-Blank. Results showed more PD-1 positive CD14⁺ M/MΦs in WT B. suis-infected cells compared to that of Δomp25 B. suis-infected cells (Figure 8A). Similarly, the percentage of PD-1⁺ cells in rAd-Omp25-infected M/MΦs was higher than that in rAd-Blank-infected cells (Figure 8B). In contrast to the suppression of IL-12 production in PD-1⁺ cells during LPS/R848 stimulation, the IL-12 p70 production of PD-1⁻ cells was significantly greater (Figure 8C). Notably, Omp25 significantly increased PD-1 expression in M/MΦs, as well as decreased of IL-12 production.

To determine whether the upregulated PD-1 expression on the surface of M/MΦs correlated with intracellular IL-12 production during Brucella infection, M/MΦs derived from different donors were infected with different doses of Brucella and stimulated with LPS/R848, then the expression of PD-1, PD-L1, and IL-12 were determined and compared by Pearson correlation analysis. As shown in Figure 8D, the upregulation of PD-1 in M/MΦs was inversely associated with IL-12 production (r = −0.5737, P = 0.0102) (representative dot plot experiment in inset). Interestingly, although the expression of PD-L1 was also upregulated in infected M/MΦs, it did not significantly correlate with IL-12 production (Figure 8E).

Because PD-1⁺ cells displayed impaired IL-12 production, we next sought to define the relationship between PD-1 and miRNAs (miR-21-5p, miR-23b and miR-155) upregulation as a potential mechanism underlying Omp25-mediated IL-12 inhibition. As an initial approach, we investigated whether blockade of the PD-1 pathway affected Omp25-induced miRNAs expression in M/MΦs. THP-1 cells and M/MΦs isolated from healthy donors were infected with WT B. suis and preincubated with anti-PD-1, anti-PD-L1, or control IgG Abs overnight followed by stimulation with LPS/R848. Incubation of THP-1 cells or healthy M/MΦs with anti-PD-1 or anti-PD-L1 Ab resulted in significant suppression of miR-21-5p, miR-23b and miR-155 compared with the control IgG (Figures 9A,B). Omp25-induced IL-12 inhibition was also reduced by either anti-PD-1 or anti-PD-L1 treatment (Figure 9C). Together, these data suggested that modulation of PD-1 signaling can significantly improve the dysfunction of Brucella-infected M/MΦs and promote IL-12 production.