Exosomes released by Brucella‐infected macrophages inhibit the intracellular survival of Brucella by promoting the polarization of M1 macrophages

Abstract Exosomes, membrane vesicles released extracellularly from cells, contain nucleic acids, proteins, lipids and other components, allowing the transfer of material information between cells. Recent studies reported the role of exosomes in pathogenic microbial infection and host immune mechanisms. Brucella‐invasive bodies can survive in host cells for a long time and cause chronic infection, which causes tissue damage. Whether exosomes are involved in host anti‐Brucella congenital immune responses has not been reported. Here, we extracted and identified exosomes secreted by Brucella melitensis M5 (Exo‐M5)‐infected macrophages, and performed in vivo and in vitro studies to examine the effects of exosomes carrying antigen on the polarization of macrophages and immune activation. Exo‐M5 promoted the polarization of M1 macrophages, which induced the significant secretion of M1 cytokines (tumour necrosis factor‐α and interferon‐γ) through NF‐κB signalling pathways and inhibited the secretion of M2 cytokines (IL‐10), thereby inhibiting the intracellular survival of Brucella. Exo‐M5 activated innate immunity and promoted the release of IgG2a antibodies that protected mice from Brucella infection and reduced the parasitaemia of Brucella in the spleen. Furthermore, Exo‐M5 contained Brucella antigen components, including Omp31 and OmpA. These results demonstrated that exosomes have an important role in immune responses against Brucella, which might help elucidate the mechanisms of host immunity against Brucella infection and aid the search for Brucella biomarkers and the development of new vaccine candidates.


INTRODUCTION
Brucellosis is a serious zoonosis caused by Brucella that results in abortion and infertility in several mammal species including humans (Yi et al., 2021). Brucella is a facultative intracellular bacteria with high selectivity and pathogenicity to its natural hosts. Brucella uses its surface virulence factors (lipopolysaccharides, outer membrane proteins and type IV secretion system) to parasitize macrophages for a long duration, resulting in persistent infection. Studies reported that exosomes that carry antigens and immunogenic factors during | 1525 EXOSOMES PROMOT THE M1 MACROPHAGES POLARIZATION infection by pathogenic microorganisms might induce immune responses in hosts. Exosomes are tiny vesicles released extracellularly from cells. Intracellular pathogenic bacteria can survive in the host cell for a long time after invading the host and cause chronic infections, which can be harmful to the host (Zhao et al., 2020). At present, few studies have examined the relationship between exosomes and intracellular pathogenic bacteria, though it was shown that intracellular bacteria such as Mycobacterium tuberculosis, Listeria monocytogenes and Salmonella induced cells to secrete microvesicles (Garcia-del Portillo et al., 1997). Other studies confirmed that microvesicles carrying M. tuberculosis antigens established protective immunity in mice (Giri & Schorey, 2008). Whether Brucella an intracellular bacteria -can induce macrophages to release exosomes requires further study.
Innate immunity is an important host defence mechanism that resists infection by pathogenic microorganisms. Macrophage polarization is a form of innate immune defence induced by external stimuli. Macrophages can be polarized into two subtypes: stimulation by interferon gamma (IFNγ) or lipopolysaccharides (LPS) leads to M1 macrophages phenotype, characterized by the production of nitric oxide synthase (iNOS) and inflammatory cytokines such as tumour necrosis factor alpha (TNFα) and interleukin-6 (IL-6). It mainly exhibits pro-inflammatory activity and promotes the killing of pathogenic microorganisms by the host.
In contrast, the Th2 cytokines interleukin-4 (IL-4) and IL-13 activate signal transducer and activator of transcription 6 (STAT6) to promote development of M2 macrophages phenotype. It secretes anti-inflammatory cytokines such as IL-4, IL-10 and transforming growth factorβ (TGFβ) and high expression of arginase 1 (Arg1) and mannose receptor (CD206), which play important roles in tissue repair, allergic inflammation and helminth infection (Higuchi et al., 2016). Furthermore, M1 and M2 macrophages can be polarized to each other and their balance is important for cell homeostasis.
Exosomes carry material information between cells and participate in the activation of innate immunity (Wang et al., 2019). Studies reported that exosomes affect the biological function of macrophages by regulating their polarization, thereby affecting the activation of innate immunity. Other studies showed that exosomes mediated inflammatory responses of host cells through the NF-κB pathway (Gao et al., 2016), which can induce macrophages to polarize to the M1 type, thereby indirectly promoting the release of Th1 type cytokines (Dai et al., 2020). In addition, innate immunity, as well as subsequent anti-pathogenic effects, induced by exosomes mainly relies on the transmission of immunologically active substances between immune cells via exosomes. It has been reported that LPS-pre-exosome can improve the regulation of macrophage polarization and secretion of inflammatory cytokines through the Toll-like receptor (TLR)4/TLR2/NF-κB signalling pathway, thereby promoting NLRP3 inflammation to activate and enhance cell phagocytic ability (Ti et al., 2015). However, Brucella induces host innate immunity mostly in the early stages of infection, after which 90% of the bacteria are eliminated. Whether Brucella can induce the release of exosomes from macrophages that then induce macrophage polarization to affect the survival of bacterial cells requires further study.
This study examined how exosomes carrying Brucella antigens initiated innate immune responses in the early stages of Brucella infection, and investigated the effects of macrophage polarization induced by exosomes on the survival of Brucella melitensis M5 (B. melitensis M5) in cells. This will lay a theoretical foundation for the prevention and treatment of brucellosis and the development of new Brucella vaccine candidates.

Identification of exosomes by different methods
Using nanoparticle tracking analysis (NTA) to detect nanoparticle sizes and transmission electron microscope (TEM) observation, we found that the exosomes collected from B. melitensis M5-infected macrophages (EXO-M5; Figure 1A) or the exosomes collected from macrophages (EXO; Figure 1B) were saucer-like threedimensional double-layer membrane structures with diameters of 30-120 nm ( Figure 1C). The number of the EXO-M5 was slightly more than that of the EXO. The background was clean and conformed to the definition of exosomes.
Western blotting detected the surface markers CD63 and TSG101 on exosomes, and exosomes released from macrophages infected with B. melitensis M5 positively expressed specific markers ( Figure 1D).

Exosomes promote M1 polarization in macrophages by carrying Brucella
In order to test whether exosomes bearing B. melitensis are associated with signalling and promoting polarization in macrophages, macrophages derived from bone marrow were incubated with Exo or Exo-M5 for 12 and 48 h and then the expression levels of macrophage marker molecule M1 CD86 and M2 CD206 were measured using flow cytometry analysis. Furthermore, macrophages incubated with IFNγ and LPS or IL-4 were used as positive controls to stimulate macrophage M1/M2 polarization. We found that the Exo-M5 group significantly induced CD86 expression (Figure 2A,C) and promoted macrophage polarization of the M1 type, but Exo did not ( Figure 2B,D).

Exosomes promote the expression of M1-type macrophage-related factors by carrying Brucella
To explore the effects of exosomes on the transcriptional expression of M1/M2-related factors in macrophages, we used qRT-PCR to detect the M1-related factors TNFα and NOS2 and the M2-related factors IL-10 and ARG1 induced by exosomes. Compared with the phosphate-buffered saline (PBS), Exo-M5 group significantly induced the transcription of TNFα and NOS2 at 4, 12 and 24 h time points examined (p < 0.05 or p < 0.001, Figure 3A,B), which reached a maximum at 12 h. In addition, Exo-M5 did not significantly induce the expression of IL-10 and ARG1 (p > 0.05, Figure 3C,D). Exo group has no effect on the release of M1-and M2-related factors. This confirms that exosomes can promote the expression of M1type macrophage marker-related factors by carrying Brucella.

Exo-M5 regulates the activation of M1/M2 molecular signalling pathway
To detect whether Exo-M5 regulates macrophage polarization through the macrophage polarization-related signalling pathway, western blotting was used to detect activation of the M1/M2 gene signalling pathways. Exo-M5 activated the M1 signalling pathway via NF-κB p65 and phosphorylated p-p65, which was time dependent and reached a peak at 12 h (p < 0.001, Figure 4A). However, Exo-M5 did not activate the M2 type signal pathway via STAT6 and p-STAT6 (p > 0.05, Figure 4B). These results show that infectious exosomes can induce the activation of NF-κB pathway, which may be closely related with M1-type macrophages.

Detection of Brucella intracellular survival
To explore the effect of exosome-induced macrophage polarization on the intracellular survival of Brucella, we used B. melitensis M5 to infect macrophages pretreated with exosomes and counted intracellular colonyforming units (CFUs) at different time points. The results showed that at 4, 12 and 24 h of B. melitensis M5 infection, Exo-M5 pre-treated macrophages significantly inhibited the intracellular CFUs of B. melitensis M5 (p < 0.05 or p < 0.01, Figure 5) compared with the PBS control group and that this was time dependent with a gradual downward trend overall. The results confirmed that exosome-induced M1-type macrophages inhibited the intracellular survival of Brucella and that the secretion of Th1 type cytokines may be a key factor in eliminating intracellular bacteria.

Detection of the relevant immune index in mice
To determine whether Brucella-infected exosomes affected the release of M1 (TNFα) and M2 (IL-10) cytokines in mice, we immunized mice with exosomes for 5, 15, 25 and 35 days and then measured TNFα, IL-10 and IgG2a protein levels ( Figure 6A). Results showed that Brucella-infected exosomes significantly induced the release of TNFα compared with the PBS control group, and reached a maximum at 12 days after immunization (p < 0.05 or p < 0.01, Figure 6B), but had no effect on the secretion of IL-10 (p > 0.05, Figure 6C). However, exosomes released by uninfected macrophages did not induce macrophage polarization in vivo.
The Exo-M5 group significantly induced the release of IFNγ and IgG2a (p < 0.05 or p < 0.01, Figure 6D,E) on days 5, 15 and 25 after immunization compared with the PBS control group. However, 35 days after immunization, the ability of Exo-M5 to induce innate immunity gradually decreased. This suggests that Exo-M5 induces high levels of Th1-type immune responses in mice.

F I G U R E 2
Flow cytometric analysis of the expression of M1/M2 surface marker molecules (CD86/CD206) in Exo and Exo-M5 incubated macrophages. The cells incubated with Exo, Exo-M5, IFNγ and LPS or IL-4 (control group) for 12 and 48 h and polarized macrophages were counted via flow cytometry. CD86 (A and C) and CD206 (B and D) were used as makers to identify M1-type or M2-type macrophages. All treatments were repeated three times with n = 3/time point. Statistical significance was determined by one-way ANOVA. Data were presented as means ± SD. ***p < 0.001, ns p > 0.05.

Effect of exosomes on immune protection in mice
Next, we investigated whether Brucella-infected exosome-mediated macrophage polarization affected the survival rate and spleen CFUs of mice after an injection of B. melitensis M5. Mice in different immunization groups were infected with a maximum dose of 1 × 10 9 CFUs 35 days after immunization, and the survival of the mice was recorded. Nine days after Brucella infection, the survival rate of mice in the PBS and Exo groups was 0%, whereas that of mice in the Exo-M5 group was 42.8% ( Figure 7A). Therefore, macrophage exosomes infected by B. melitensis M5 improved the resistance of mice against Brucella infection, thereby improving their survival rate.
At 35 days after Exo and Exo-M5 immunization, the spleens of mice in each group were isolated at 1, 3, 5 and 7 weeks after B. melitensis M5 (1 × 10 6 ) infection and the spleen CFU count of each group was calculated. The results showed that compared with the PBS group, the amount of bacteria in the spleens of mice in the Exo-M5 group was significantly reduced at each time point (p < 0.05 or p < 0.01, Figure 7B). There was no significant difference in the Exo groups at each time point. In conclusion, macrophage exosomes infected by B. melitensis M5 reduced damage to the spleen, to varying degrees, caused by Brucella infection. Exosomes also reduced the survival of Brucella in mice, and the combined adjuvant was beneficial for improving the immune protective effect of the exosomes.

Identification of Brucella components in exosomes
After 56-carboxyfluorescein diacetatesuccinimidyl ester (CFSE)-labelled B. melitensis M5 and PKH26labelled macrophages were co-cultured, exosomes were extracted and observed with a confocal microscope. Red-stained microparticles contained greenstained B. melitensis M5 components (Figure 8), indicating that RAW264.7 macrophages lysed Brucella after phagocytosis and that some of the lysed bacterial components were released extracellularly through the exosome pathway.

F I G U R E 3 Exosomes derived from
Brucella melitensis M5-infected macrophages promote M1-related factors mRNA transcription in macrophages. The transcription level of TNFα (A), NOS2 (B), IL-10 (C) and ARG1 (D) were assessed using the qRT-PCR. PBS-treated samples were used as a control group. All treatments were repeated three times with n = 3/time point. Data were presented as means ± SD. Statistical significance was determined by one-way ANOVA. *p < 0.05. ***p < 0.001. ns p > 0.05.

Analysis of protein aggregation in exosomes
Next, we investigated the components in the RAW264.7 macrophage exosomes. Label-free quantitation technologies were used to analyse the protein profiles of exosomes secreted from macrophages (with or without Brucella infection). Overall, 1236 proteins were identified, including 1156 proteins derived from host cells and 80 proteins derived from B. melitensis M5. Particularly, we found that 80 proteins contained in infectious exosomes were specific to B. melitensis M5, including outer membrane proteins, LPS-related proteins, transporters and other proteins (Table 1). Thus, Brucella antigens in these exosomes may be important stimulators that mediate macrophage polarization.

DISCUSSION
Macrophages, an important component of immune responses, remove pathogens and maintain an immune steady state. Macrophages have strong plasticity and after a host is stimulated by pathogenic microorganisms, they can differentiate into different subtypes (M1/ M2), which have different biological functions (Wang et al., 2010). M1-type macrophages mediate Th1-type immune response and promote the expression of M1type-related factor NO and Th1 cytokines such as TNFα, IL-6 and IL-12, whereas M2 macrophages mediate Th2-type immune responses and release Th2 cytokines such as IL-4 and IL-10. M1-and M2-type macrophages can also be transformed in a special microenvironment (Verreck et al., 2004). However, macrophages F I G U R E 4 Exosomes released from Brucella melitensis M5-infected cells (Exo-M5) induce the activation of NF-κB p65 signalling pathway. Exosome were incubated with macrophage for 12 h and 48 h. Western blot analysis of p65, p-p65 (A, a) and STAT6, p-STAT6 protein expression in exosome (B, b). All treatments were repeated three times with n = 3/time point. Data were presented as means ± SD. Statistical significance was determined by one-way ANOVA. ***p < 0.001.

F I G U R E 5
The intracellular bacterial survival was tested by measuring CFU. Recipient cells pre-treated with Exo-M5 or Exo, PBS as a control group and then the CFU level in macrophages were measured at different time points post-infection. All treatments were repeated three times with n = 3/time point. Data were presented as means ± SD. Statistical significance was determined by one-way ANOVA. *p < 0.05. **p < 0.01. are important Brucella host cells and Brucella was reported to polarize macrophages to an M1 type in the early stages of infection, leading to increased levels of the Th1 immune factor IFNγ and a rapid decrease in the bacterial load of the mouse spleen (Konrad et al., 2005).
Exosomes are an important medium for transmitting material information, and they have functions in tumours, autoimmune diseases and pathogenic microorganism infections. Although research on the role of exosomes in pathogenic microbial infections is in the preliminary stage, exosomes might have great potential for the prevention and treatment of infectious diseases. Previous studies found that M. tuberculosis induced immune cells to produce a large number of exosomes. In this study, we discovered that B. melitensis, an intracellular bacteria, induced macrophages to release exosomes by a time-dependent mechanism. Exosomes are important substance signal carriers released by cells, and increasing numbers of studies have reported that exosomes affected the biological functions of macrophages by regulating the polarization of macrophages. A study by Moradi-Chaleshtori et al. (2021) showed that exosome-mediated miR-33 transfer induced M1 polarization in mouse macrophages and exerted antitumor effects in a 4T1 breast cancer cell line. Cheng et al. (2017) found exosomes from M1-polarized macrophages potentiate the cancer vaccine by creating a pro-inflammatory microenvironment in the lymph node.
In this study, we found that exosomes secreted from B. melitensis M5-infected macrophages induced bone marrow-derived mononuclear macrophages to polarize to an M1 type and promoting the release of F I G U R E 6 Immune effects of exosomes in mice. Schematic diagram of the immunization schedule and overall experimental design (A). The mouse serum were harvested and TNF-ɑ (B), IL-10 (C), IFNγ (D) and IgG2a (E) production (pg/ml) were assessed by ELISA assay. All treatments were repeated three times with n = 3/time point. Data were presented as means ± SD. Statistical significance was determined by one-way ANOVA. *p < 0.05. **p < 0.01. Th1-type immune factors (NOS2, TNFα). NF-κB is an important immunogenic regulatory pathway that regulates the secretion of the pro-inflammatory cytokines IL-1β, IL-6 and IL-12 among others (Shah et al., 2011). Some studies have found that many pathogenic microbial infections are able to promote M1 polarization of macrophages by activating the Akt, mTOR and NF-κB pathway (Gary & Elizabeth, 2009;Lin et al., 2018); the activation of NF-κB pathway by Exo-M5 was also detected in our experiments, suggesting that Exo-M5 may mediate M1 polarization of macrophages/monocytes via NF-κB activation. In addition, LPS and IFNγ are important activators of macrophages and their combination can strongly induce the activation of NF-κB pathway; therefore, many scholars believe that the activation of NF-κB signalling pathway can promote the secretion of Th1-type cytokines (Chow et al., 2012). Studies suggested that Th1-type factors can induce immune cells to produce reactive oxygen species (ROS), which can directly kill pathogenic microorganisms and help mobilize natural killer cells to perform important antibacterial functions (Jiang et al., 1993). An increase in the ROS concentration is disadvantageous to the intracellular proliferation of Brucella (Li et al., 2016). Our study found that exosome-induced M1 macrophages inhibited the intracellular survival of Brucella, indicating Th1 cytokine-mediated ROS release may be an important factor in the inhibition of Brucella reproduction. In addition, animal experiments demonstrated that exosomes induced the release of M1 cytokines in vivo, thereby reducing the number of Brucella in the spleen, by a mechanism that might involve M1-type cytokines F I G U R E 7 Mice immunized with Exo-M5 were protected against Brucella infection. Mice immunized with different groups of exosomes were infected with B. melitensis M5 (1 × 10 9 ) for 15 days to observe the survival rate of mice in different groups. Mice immunized with different groups of exosomes were infected with B. melitensis M5 (1 × 10 6 ) for 1, 3, 5 and 7 weeks infection. Then, each group of mice was sacrificed and mice spleens CFU (B) were measured. All treatments were repeated three times with n = 3/time point. Data were presented as means ± SD. Statistical significance was determined by one-way ANOVA. *p < 0.05. **p < 0.01. (IFNγ, IL-12) released by M1 macrophages that directly kill Brucella in the host, as well as enhance the phagocytosis and bactericidal ability of macrophages, which can rapidly clear Brucella from the host (Xu et al., 2021). Therefore, exosome-mediated macrophage polarization is a new means of removing Brucella from a host. Exosomes are important carriers of proteins, nucleic acids and lipids, and allow the transfer of information between cells (Mathivanan et al., 2010).

F I G U R E 8 Detection of
Exosomes carrying bacterial antigens have an important role in the host immune response. In this study, we confirmed that the exosomes were loaded with a critical antigenic protein from Brucella that affected the polarization of macrophages. We showed that Brucella antigens, indicated by green fluorescence, were present in exosomes released from macrophages using confocal laser scanning microscopy. Furthermore, we also found 80 important Brucella antigens using label-free protein histology, including Omp31, Omp19, Omp25b, L7/L12, Cu/Zn-SOD and others in exosomes released by Brucella-infected macrophages. Therefore, these proteins in the exosomes may be an important factor in inducing macrophage polarization. Omp31 and Omp25B are immunogenic and were reported to induce humoral and cellular immune responses, as well as have protective effects against Brucella (Clausse et al., 2013). A study reported that mice immunized with peptides of OMP31 containing T epitope, B epitope or TB epitopes are of high immunogenicity, which can protect mice from B. melitensis infection in the lung (Zhang et al., 2019) and that a stable memory immune response was produced (Ranjbar et al., 2016). Therefore, exosomes containing antigens can inhibit the reproduction of Brucella in mice and induce the M1 polarization of macrophages in vivo to promote Th1-type immunity. The secretion of cytokines, such as TNFα and IFNγ, inhibit the replication of bacteria and exosomes carrying antigens can induce the host to produce IgG, resulting in protection against infection. However, how the antigen molecules in the exosomes interact with the host and how they activate the immune response mechanisms require further exploration.
In summary, our study confirmed that exosomes released by B. melitensis M5-infected macrophages induced bone marrow-derived mononuclear macrophages to polarize to M1 and promoted the secretion of Th1 cytokines, thereby indirectly inhibiting the survival of B. melitensis M5 in vivo and in vitro. In addition, exosomes stimulated strong immune responses in mice that were protective against infection. These immunostimulatory effects may be related to Brucella antigens present in the exosomes. Taken together, these findings provide a basis to explore the role of exosomes in host immune responses and aid the development of new Brucella vaccines.

Bacterial strains, cells and mice
Brucella melitensis M5 was provided by the Chinese Center for Disease Control and Prevention. RAW264.7 macrophages were obtained from the Type Culture Collection of the Chinese Academy of Sciences (China, Beijing). BALB/c mice aged 6 weeks were purchased from the Experimental Animal Center of Xinjiang Medical University. All animal care in this study was carried out in accordance with institutional animal care guidelines and related laws.

Preparation of mouse bone marrowderived mononuclear macrophages
Bone marrow cells were isolated from mouse femurs and tibias and cultured in RPMI 1640 medium. On the second day, mouse M-CSF (RPMI 1640 + 10% T A B L E 1 List of major Brucella melitensis M5 antigen proteins in exosomes. FBS + M-CSF) was added at a concentration of 10 ng/ ml to induce bone marrow monocytes to differentiate into monocytes/macrophages. On the seventh day, the differentiated bone marrow-derived mononuclear macrophages were collected.

Protein ID
Brucella melitensis M5 infection of RAW264.7 macrophages RAW264.7 macrophages were infected with B. melitensis M5 for 1 h at a multiplicity of infection (MOI; bacteria: cell) of 100:1. Cells were incubated for 1 h and then washed three times with PBS. Next, gentamicin (25 μg/ml) was added into the cell culture medium and incubated for 45 min to kill Brucella outside the cells.
The culture medium was discarded and replaced. At 24 h post-Brucella infection, the supernatants were collected from each plate.

Extraction and identification of exosomes
The extraction and identification methods for exosomes were reported previously (Chen et al., 2020). RAW264.7 macrophages were infected with B. melitensis M5 and cultured at an MOI of 100:1 (bacteria/cell) for 1 h. After adding gentamicin sulphate for 30 min, the cell culture supernatant was collected at 24 h after infection. Then, supernatants were centrifuged at 1000 × g for 5 min, 5000 × g for 5 min and 14,000 × g for 30 min to eliminate the cell debris. The supernatant was passed through a 0.22μm filter to eliminate Brucella and then centrifuged further for 2 h at 12,000 × g to pellet the exosomes. Brucella uninfected cells were also used as controls to extract exosomes using the same procedure. The purified exosomes were identified by NTA, TEM and western blotting as described by Chen (Chen et al., 2020). The total protein concentration of exosomes was measured using a BCA assay kit (Pierce, Rochford, IL). Before the experiment, all isolated exosomes were stored at −80°C.

Flow cytometric analysis
The bone marrow-derived mononuclear macrophages were incubated with exosomes, and then macrophages were collected at 12 and 48 h. In addition, IFNγ and LPS or IL-4-stimulated cells were added to the cells as a positive control. The expressions of the M1-type macrophage marker molecule CD86 and M2-type macrophage marker molecule CD206 were detected by flow cytometry with IgG2a as an isotype control. After the macrophages of different treatment groups were fixed and ruptured, the corresponding antibodies were added, incubated, protected from light at 4°C for 30 min, washed with 1× binding buffer and tested on the flow cytometer.

Real-time PCR
Exosomes were co-incubated with 5 mmol L −1 ATP and peritoneal macrophages or bone marrow-derived mononuclear macrophages. At 8 and 24 h, Trizol lysate was added, RNA was extracted according to the RNA extraction kit instructions and the total RNA concentration was measured. Then, a cDNA synthesis kit was used for the reverse transcription of the total RNA.
A fluorescence real-time quantitative PCR instrument was used to detect the expression levels of mRNAs for M1/M2-related genes. The RT-PCR system included 1 μl cDNA, 5 μl SYBR green mix, 0.2 μl of upstream and downstream primers and 3.6 μl sterile water. The reaction conditions were pre-denaturation at 95°C for 30 s, denaturation at 95°C for 30 s, extension at 60°C for 35 s, with 45 cycles. GAPDH was used as an internal control.

Western blotting
Bone marrow-derived mononuclear macrophages and peritoneal macrophages were incubated with exosomes, and protein samples were collected at 12 and 48 h. In addition, IFNγ and LPS or IL-4 stimulated cells were added to the cells as a positive control. Western blotting was used to detect the protein expressions of NF-κB p65, p-p65 and STAT6, p-STAT6. All collected protein samples were tested for related protein expression levels according to the above-mentioned western blotting method.

CFU detection
Exosomes released from different pre-treated cells were incubated with bone marrow-derived mononuclear macrophages for 8 h and then infected with B. melitensis M5 for 4, 8, 12, 24 or 48 h. Then, these cells were lysed with 0.5 ml of 0.2% Tween 20 at each of post-infected time point, incubated for 15-30 min on ice, followed by rinsing each well with 0.5 ml of PBS. Viable bacteria were quantified by serial dilution in sterile PBS and plating on TSA and then they were incubated at 37°C for 72 h followed by colony counting.

Immunization of mice with exosomes
Three immunization groups were formed: exosomes released from Brucella-infected RAW264.7 macrophages combined with adjuvant (Exo-M5); exosomes released from normal RAW264.7 macrophages combined with adjuvant (Exo) and PBS alone. The immunization dose for each group was 100 μl (0.3 μg/μl). Immunization was performed two times at 15 days intervals and 100 μl was injected each time.

Brucella challenge
Mice were inoculated with exosomes 35 days later. The maximum lethal dose of B. melitensis M5 (1 × 10 9 CFUs) was suspended in 0.1 ml PBS and then injected subcutaneously into the abdominal cavity of mice. The survival of mice and the mental state of the surviving mice were noted. Of note, 0.1 ml of PBS was injected as a control. Mice were inoculated with exosomes 35 days later and the optimal dose of B. melitensis M5 was injected into the abdominal cavity of the mice.

Detection of the immune index and measurement of the spleen CFUs in mice
On days 5, 15, 25 and 35 after immunization, the mice in each group were tail-clamped to collect blood and serum. Mouse IgG2a and IFNγ ELISAs were performed in accordance with the manufacturer's instructions. Three mice in each group were sacrificed by cervical dislocation, spleens were removed and spleen cells were isolated and counted. The number of spleen lymphocytes was adjusted to 1 × 10 6 cells/ml and they were added to a 96-well plate. Then, 10 μl of inactivated B. melitensis M5 (10 7 CFUs/well) was added for 68 h. Next, the lymphocyte culture supernatant was collected and used for mouse IFNγ ELISA.
At 5, 15, 25 and 35 days after B. melitensis M5 infection, three Brucella-infected mice per group were humanely sacrificed, and spleens were removed and weighed. The spleen was placed into a centrifuge tube containing PBS buffer, 1 ml of 0.2% Triton X-100, several small steel balls and homogenized in a homogenizer. Then, 100 μl of the diluted homogenate was diluted in sterile saline and plated onto TSA. The plates were incubated at 37°C and Brucella CFUs were enumerated. The results were presented as log 10 CFUs per spleen.

Chemical staining
CFSE (Sigma, USA)-stained B. melitensis M5 was added to PKH26 (Sigma, USA)-stained RAW264.7 macrophages and mixed. Brucella particles were added at 50 times the number of cells. After 12 h of co-cultivation, the supernatant was collected by centrifugation and the staining results were observed under a confocal microscope.

Exosomal proteomics analysis
In this experiment, the identified exosomes were sent to Applied Protein Technology Co., Ltd. (Shanghai, China) for label-free quantitative proteomics analysis. The exosome samples were divided into two groups: exosomes released from normal RAW264.7 macrophages (uninfected group) and exosomes released from Brucella infected RAW264.7 macrophages (infected group), with three replicates per group.

Data analyses
All original data were analysed using Microsoft Excel software, and GraphPad Prism statistical software was used for graphing and one-way analysis of variance. The results are displayed using the mean ± standard deviation (M ± SD), and differences were considered statistically significant when p values were <0.05 and <0.01.

A C K N O W L E D G E M E N T S
This work was supported by the Open Project Program of Jiangsu Key Laboratory of Zoonosis (No. R2104), Youth Innovation and Talent Training Program of Shihezi University (CXPY202109), supported by the National Natural Science Foundation of China (No. U1803236). We thank Liwen Bianji (Edanz) (www. liwen bianji.cn/) for editing the language of a draft of this manuscript.