Diagnoses of acute brucellosis were made based on history, clinical presentation, and laboratory tests (1). Patients who had contact history with cattle, sheep, or deer and clinical manifestation of intermittent fever with or without arthralgia during the last 3 months were subjected to laboratory testing. Rose bengal test positive, serum agglutination titer >1:160, and PCR test [Diagnostic Kit for Brucella DNA Kit (Jiangsu Bioperfectus Technologies, Taizhou, China)] positive were confirmation of brucellosis. All the patients were also examined by chest tomography and tuberculosis T-SPOT test to exclude tuberculosis. Twenty-five patients diagnosed with acute brucellosis and 15 healthy individuals were included in this study (Table 1). All patients were treated with rifampicin and doxycycline for 6 weeks, and five of them were recalled for posttreatment follow-up and collection of blood samples. The demographic data and clinical characteristics before and after treatment are shown in Table 2. The venous blood from patients and healthy controls (HC) in terms of diagnosis were collected before and after treatment. The study was approved by the Human Research Ethics Committee of the First Hospital of Jilin University, Jilin Province, China. All adult subjects provided written informed consent, and no children were included in the study.

The smooth strain B. abortus 2308 was grown overnight in tryptic soy broth, harvested by centrifugation, and washed twice in phosphate-buffered saline (PBS). Bacterial numbers in cultures were estimated by comparing the OD at 600 nm with a standard curve, but the actual concentration of inocula was checked by plating on tryptic soy agar (TSA) plates. All live Brucella manipulations were performed in biosafety level 3 facilities (China CDC, Beijing). To prepare HK-Br, bacteria were washed in sterile PBS, heat killed at 70°C for 30 min, aliquoted, and stored at −70°C until used. The absence of B. abortus viability after heat killing was verified by the absence of bacterial growth on TSA.

PBMCs were isolated by density centrifugation using Lymphoprep (Fresenius Kabi Norge AS, Halden, Norway) as previously described (24). Monocytes were purified from PBMCs using a Human CD14 MicroBead Kit (Miltenyi Biotec, Bergisch Gladbach, Germany) as previously described (25). The purity of the CD14⁺ cells was >95% as determined by flow cytometry.

Purified monocytes (2 × 10⁶/ml) were stimulated with 1.0 µg/ml LPS from E. coli O111:B4 (Sigma-Aldrich Inc., St. Louis, MO, USA) as previously described (26, 27). Monocytes were differentiated to M1/M2 macrophages as previously described with modification (25). Briefly, purified monocytes (2 × 10⁶/ml) were cultured with GM-CSF (400IU/ml) or M-CSF (50 ng/ml) for 5 days in RPMI 1640 medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% heat-inactivated fetal calf serum, 100 IU/ml penicillin and 100 µg/ml streptomycin. For M1 polarization, GM-CSF-induced macrophages were cultured with LPS (100 ng/ml) and IFN-γ (20 ng/ml) for an additional 24 h. For M2 polarization, M-CSF-induced macrophages were further cultured with IL-4 (25 ng/ml) and IL-13 (25 ng/ml) for an additional 24 h. All the recombination human cytokines were purchased from R&D (Minneapolis, MN, USA).

To investigate the involvement of autophagy in monocytes and macrophage polarization, purified monocytes were pretreated with an autophagy inhibitor, 3 µM 3-MA (Sigma-Aldrich, St. Louis, MO, USA) for 3 h, and then cultured for 24 h with HK-Br (MOI = 100:1), 100 nM rapamycin (Sigma-Aldrich, St. Louis, MO, USA) was used as a positive control for autophagy as previous described (12).

Cell staining and flow cytometry analysis were performed as described (25). Briefly, differentiated M1 and M2 macrophages were characterized by staining with the following antibodies: mouse anti-human CD14-FITC, mouse anti-human CD16-PE-Cy7, mouse anti-human CD80-PE, mouse anti-Human CD86-FITC, mouse anti-human CD163-PE, and mouse anti-Human CD206-FITC. Markers of monocytes were mouse anti-human CD282-FITC, CD283-PE, and mouse anti-human CD284-APC. Mouse anti-human LC3B-PE was used as a marker for the autophagy-related protein (28). Intracellular staining was performed using the following mouse anti-human mAb: TNF-α, IL-6, IL-1β, and IL-10. All the antibodies were purchased from BD Biosciences (San Jose, CA, USA), except for anti-LC3B antibody (Cell Signaling Technology, Danvers, MA, USA), and used at a dilution of 1:50. Flow cytometry was performed using a BD Biosciences FACS CANTOII flow cytometer (BD Biosciences). The data acquired were analyzed with FlowJo (Treestar software, Ashland, OR, USA).

Western blotting was performed according to the published method (9, 12). Samples (50 µg total protein/lane) were loaded onto SDS-PAGE gels in this experiment. The primary (rabbit anti-human LC3, rabbit anti-human Beclin-1 and rabbit anti-human β-actin antibody) and secondary (anti-rabbit horseradish peroxidase-labeled antibody) antibodies were purchased from Cell Signaling Technology and used at a dilution of 1:1,000 and 1:2,000, respectively. The images were obtained using a CanoScan LiDE 100 scanner (Canon). The results were quantified using Image-J software.

Enzyme-linked immunosorbent assay was performed with human IL-1β, IL-6, IL-10, and TNF-α ELISA Ready-SET-Go Kits (eBioscience, San Diego, CA, USA) according to the manufacturer’s instructions.

All data were analyzed using the D’Agostino and Pearson omnibus normality test. P values of <0.05 were considered statistically significant. Mean values were compared using either a paired t-test (two groups) or ANOVA (more than two groups), followed by a Bonferroni correction for multiple comparisons test. All statistical tests were performed using GraphPad Prism software (San Diego, CA, USA).

Brucella establishes infection in host monocytes and macrophages. To determine whether Brucella infection affects the frequency and phenotype of monocyte subsets, the monocytes were divided into three subsets based on CD14 and CD16 expression (29) as gating strategy shown in Figure 1A. We observed that the proportion of CD14⁺⁺CD16⁻ monocytes in patients with Brucella infection was significantly higher than in HC, whereas the proportion of CD14⁺CD16⁺ monocytes was significantly lower. There was no significant difference in the CD14⁺⁺CD16⁺ monocyte subset between the two groups (Figure 1B).

We further compared the frequency and phenotype of monocyte subsets in five patients with brucellosis before and after treatment with rifampicin and doxycycline. Results showed that the proportion of CD14⁺⁺CD16⁻ monocytes significantly decreased after 6 weeks of antibiotic treatment, whereas both CD14⁺CD16⁺ and CD14⁺⁺CD16⁺ monocytes increased significantly (Figure 1C).

To further characterize the phenotypic difference of monocyte subsets between patients with Brucella infection and healthy control individuals, cell surface expression of antigen presenting markers: CD1a and CD1b, costimulatory molecules: CD40, CD80, and CD86, M2 macrophage markers: CD163 and CD206, and TLR: CD282 (TLR-2), CD283 (TLR-3), and CD284 (TLR-4) were examined by flow cytometry. Results in Figure 1D showed that the expression levels of CD40, CD282, and CD284 on all three monocyte subsets are statistically higher in patients with Brucella infection than in healthy control individuals, whereas no significant differences were found in CD1a, CD1b, CD80, CD86, CD163, CD206, and CD283 expression levels for the various monocyte subsets between patients and controls.

Recent studies have shown that many human diseases result from overzealous innate immune responses including autophagy (30). Therefore, we investigated whether Brucella infection induces monocyte autophagy by quantify the expression of an autophagy-related protein LC3B using flow cytometry in patients with Brucella infection and in healthy control individuals. Results shown in Figure 2A demonstrated a higher LC3B expression level on the entire monocyte population from patients with Brucella infection than that from healthy control individuals, as well as monocyte subsets that are either CD14⁺⁺CD16⁺ or CD14⁺⁺CD16⁻ (Figure 2B). It was reported that rifampicin inhibits rapamycin-induced autophagy (31). We further examined whether anti-Brucella treatment could modulate LC3B expression on monocytes. Figure 2C showed a significant downregulation of LC3B expression levels on the overall monocyte population after 6 weeks of treatment of brucellosis with rifampicin and doxycycline (P < 0.001), and the same is true for each of the three monocyte subsets (Figure 2D).

To determine whether Brucella infection induces LC3-dependent autophagy in monocytes, purified monocytes from healthy donors were cultured for 24 h with HK-Br (a possible autophagy inducer) in the presence or absence of 3-MA, and rapamycin was used as a positive control for autophagy (12). LC3B expression levels of monocytes were determined by flow cytometry. The results showed that both HK-Br and rapamycin significantly induced LC3B expression of monocytes compared with untreated monocytes, whereas 3-MA significantly inhibited HK-Br-induced LC3B expression (Figure 3A). Beclin-1 was regarded as an important marker of autophagy as LC3 (12). Most autophagic responses were Beclin-1 dependent. The expression levels of LC3-I/II and Beclin-1 were also examined by western blotting, and β-actin was used as internal reference protein. The ratio of LC3-II/LC3-1 and Beclin-1/β-actin were calculated. Consistently, both HK-Br and rapamycin increased the ratios of LC3-II/LC3-I and Beclin-1/β-actin, and 3-MA decreased these ratios (Figure 3B). Collectively, these results indicate that Brucella infection induces autophagy-related protein LC3B expression in monocytes.

In response to pathogens, monocytes produce pro-inflammatory cytokines including IL-1β, IL-6, IL-8, and TNF-α, and anti-inflammatory cytokine such as IL-10 and TGF-β. To determine whether Brucella infection regulates monocyte function, purified monocytes from healthy donors and brucellosis patients were stimulated by LPS, with or without pretreatment with 3-MA. The expression and secretion of pro-inflammatory (IL-1β, IL-6, TNF-α) and anti-inflammatory (IL-10) cytokines were then examined by flow cytometry and by ELISA. Results in Figure 4 showed that upon LPS stimulation, monocytes from brucellosis patients expressed lower levels of TNF-α (Figure 4A), IL-6 (Figure 4B), IL-1β (Figure 4C), and IL-10 (Figure 4D) than those from HC. Conversely, 3-MA treatment that inhibited autophagy led to elevated expression of TNF-α (Figure 4E), IL-6 (Figure 4F), IL-1β (Figure 4G), and IL-10 (Figure 4H) than those from untreated monocytes from brucellosis patients. These results indicated that cytokine expression and production from monocytes is impaired after Brucella infection, but partially restored by the autophagy inhibitor.

To further test whether Brucella infection alters monocyte polarization to M1/M2 macrophages, isolated monocytes were polarized to M1 and M2 macrophages in the presence or absence of 3-MA, and analyzed by flow cytometry. Results in Figure 5 showed that M1 and M2 macrophages polarization was phenotypically different between monocytes obtained from patients with Brucella and those from healthy individuals. As expected, the polarization to M1/M2 macrophages in the presence of 3-MA was similar irrespective of the source of monocytes. Morphologically, M1 macrophages polarized from healthy individuals had a round-shaped appearance, while M2 macrophages exhibited a long spindle shape. In contrast, M1 macrophages polarized from patients with Brucella infection had a smaller round-shaped appearance, and M2 macrophages lacked the long spindle shape (Figure 5A). CD80 and CD86 were highly expressed on M1 macrophages, while CD163 and CD206 were highly expressed on M2 macrophages in healthy individuals. However, there were no obvious differences in CD80 and CD86 expression levels between M1 and M2 macrophages in patients with Brucella infection. Additionally, CD80 and CD86 expression levels of M1 macrophages in patients with Brucella infection were significantly lower than those of healthy individuals (Figure 5B). In addition, the expression levels CD163 and CD206 on M2 macrophages were higher than on M1 macrophages in patients with Brucella infection (Figure 5B). CD163 expression level on M2 macrophages in patients with Brucella infection was significantly lower than those of healthy control individuals, but CD206 expression level on M2 macrophages exhibited no such a difference. Moreover, 3-MA restored CD80 and CD86 expression levels on M1 macrophages, and CD163 and CD206 expression levels on M2 macrophages in brucellosis patients (Figure 5C).

The production of TNF-α and IL-10 are reflective of the pro-inflammatory and anti-inflammatory functions of M1 and M2 macrophages, respectively. Their production by M1 and M2 macrophages was quantified by ELISA and summarized in Figures 5D,E. As expected, M1 macrophages produced TNF-α and M2 macrophages secreted IL-10 in healthy control individuals. However, there were no obvious differences in TNF-α and IL-10 secretion by M1 and M2 macrophages in patients with Brucella infection, and their respective levels were both lower than those in healthy individuals (Figure 5D). However, 3-MA treatment restored TNF-α production by M1 macrophages and IL-10 secretion by M2 macrophages in brucellosis patients (Figure 5E).

Taken together, these results indicate that the potential of monocytes to polarize into M1 or M2 macrophages is impaired in brucellosis patients, and it was recovered by 3-MA treatment.

To determine whether Brucella-induced autophagy of monocytes plays a role in macrophage polarization, purified monocytes from HC were pretreated for 24 h with HK-Br. Rapamycin and 3-MA were also used as a positive control for autophagy, and an autophagy inhibitor, respectively. The pretreated monocytes were then polarized to M1/M2 macrophages as described above. M1 macrophages polarized from rapamycin-pretreated monocytes had a round-shaped appearance similar to those that received no pretreatment, while M2 macrophages from rapamycin-pretreated monocytes exhibited a smaller spindle-like shape. Interestingly, both M1 and M2 macrophages polarized from HK-Br-pretreated monocytes exhibited a smaller round shape. However, 3-MA treatment restored the morphology of M2 macrophage, but not M1 macrophage (Figure 6A). The expression levels of CD80 and CD86 on M1 macrophages (Figures 6B,C) and TNF-α production (Figure 6F) from M1 macrophages polarized from rapamycin-pretreated monocytes were similar to the controls; however, CD163 expression and IL-10 production of M2 macrophages polarized from rapamycin-pretreated monocytes were much lower than those with no pretreatment. CD86 expression on M1 macrophages and CD163 expression on M2 macrophages polarized from HK-Br-pretreated monocytes were significantly lower than those that received no pretreatment. Additionally, 3-MA treatment partially relieves the HK-Br-mediated downregulation of CD163 expression (Figure 6D) on M2 macrophages, but not CD206 expression (Figure 6E). TNF-α production by M1 macrophages and IL-10 production by M2 macrophages polarized from HK-Br-pretreated monocytes were much lower than those without pretreatment. The HK-Br-mediated inhibition of IL-10 production by M2 macrophages was recovered with 3-MA treatment (Figure 6G). Collectively, these results suggest that Brucella-induced autophagy affected the polarization of M2 macrophages, but not M1 macrophages.