Vaccine adjuvant activity of a TLR4-activating synthetic glycolipid by promoting autophagy

Toll-like receptors (TLRs) play crucial roles in host immune defenses. Recently, TLR-mediated autophagy is reported to promote immune responses via increasing antigen processing and presentation in antigen presenting cells. The present study examined whether the synthetic TLR4 activator (CCL-34) could induce autophagy to promote innate and adaptive immunity. In addition, the potential of CCL-34 as an immune adjuvant in vivo was also investigated. Our data using RAW264.7 cells and bone marrow-derived macrophages showed that CCL-34 induced autophagy through a TLR4-NF-κB pathway. The autophagy-related molecules (Nrf2, p62 and Beclin 1) were activated in RAW264.7 cells and bone marrow-derived macrophages under CCL-34 treatment. CCL-34-stimulated macrophages exhibited significant antigen-processing activity and induced the proliferation of antigen-specific CD4+T cells as well as the production of activated T cell-related cytokines, IL-2 and IFN-γ. Furthermore, CCL-34 immunization in mice induced infiltration of monocytes in the peritoneal cavity and elevation of antigen-specific IgG in the serum. CCL-34 treatment in vivo did not cause toxicity based on serum biochemical profiles. Notably, the antigen-specific responses induced by CCL-34 were attenuated by the autophagy inhibitor, 3-methyladenine. In summary, we demonstrated CCL-34 can induce autophagy to promote antigen-specific immune responses and act as an efficient adjuvant.

We next investigated whether CCL-34-induced autophagy is also TLR4-dependent. BMDMs were generated from either wild-type (C3H/HeN) or TLR4-defective (C3H/HeJ) mice and treated with CCL-34. As shown in Fig. 1G, the LC3-II protein was increased in the C3H/HeN BMDMs under CCL-34 treatment, whereas the LC3-II protein in CCL-34-treated C3H/HeJ BMDMs was not altered. In addition, the LC3-II protein level induced by CCL-34 in RAW264.7 cells was reduced upon co-treatment of TAK-242, a small-molecule inhibitor of TLR4 (Fig. S1A). Taken together, the results demonstrate that CCL-34 induces autophagy in a TLR4-dependent manner.
Involvement of the Nrf2-p62 axis, Beclin 1 induction and NF-κB activity in CCL34-induced autophagy. It is known that TLR4-mediated autophagy can be triggered by LPS 13 , and previous studies have shown that activation of the Nrf2-p62 axis is involved in TLR4-induced autophagy 33,34 . Nrf2 activation leads to transcriptional upregulation of p62, and the P62 protein can interact with LC3 to form an aggresome-like induced structure (ALIS), which promotes autophagosome formation. To investigate whether the Nrf2-p62 axis plays a role in CCL-34-induced autophagy, the expression of Nrf2 and p62 in CCL-34-treated RAW264.7 cells was measured using LPS as the positive control. The protein level of Nrf2 was elevated in CCL-34-treated RAW264.7 cells ( Fig. 2A), and both the mRNA and protein levels of p62 were significantly increased under CCL-34 treatment (Fig. 2B,C). Furthermore, a key inducer of autophagy 35,36 , Beclin 1, was also increased under CCL-34 treatment in BMDMs (Fig. 2D).
CCL-34 induced the APC ability of macrophages and enhanced the Ag-specific T cell response. To address the immune-stimulating effect of CCL-34, we further investigated whether CCL-34 could promote Ag processing and presentation in APCs. The expression of MHC class II and the costimulatory signals (CD86 and CD80) in CCL-34-treated BMDMs were measured. As shown in Fig. 3A,B, the MHC-II + CD86 + and MHC-II + CD80 + cell populations were elevated in BMDMs under CCL-34 treatment. Furthermore, a critical cytokine produced by APCs during antigen presentation, IL-12, was elevated both in RNA and protein level in CCL-34-treated BMDMs (Fig. 3C). The Ag processing ability of CCL-34-treated macrophages was further determined using DQ ovalbumin (DQ-OVA) which is a self-quenched fluorescently labeled ovalbumin (OVA) and used as a model Ag. DQ-OVA emits green fluorescence (DQ-OVAgreen) upon proteolytic digestion, whereas red fluorescence (DQ-OVAred) is emitted when the digested fragments of DQ-OVA accumulate in organelles at high concentrations 38 . As shown in Fig. 3D, CCL-34-treated BMDMs exhibited a higher percentage of DQ-OVAgreen and DQ-OVAred fluorescence than the vehicle control, indicating that CCL-34-activated macrophages had an elevated Ag processing ability. Furthermore, CCL-34 treatment induced a higher percentage of DQ-OVAgreen + RAW264.7 cells than vehicle group while the increased DQ-OVAgreen population was suppressed by the TLR4 inhibitor, TAK-242. This observation indicates that CCL-34 promotes Ag processing ability in a TLR4-dependent manner (Fig. S1B).
We further investigated whether the CCL-34-mediated increase in CD86 and CD80 levels and induction of Ag processing could lead to activation of Ag-specific T cells. OVA-specific CFSE-labeled CD4 + T cells were cocultured with BMDMs pretreated with OVA 323-339 peptide alone or in combination with CCL-34, and the decrease in CFSE fluorescence was an indication of T cell proliferation. Only CCL-34 treatment of BMDMs resulted in the proliferation of OVA-specific CD4 + T cells and the production of activated T cell-related cytokines, IL-2 and IFN-γ ( Fig. 3E-H). Taken together, these results demonstrate that CCL-34 can stimulate macrophages to activate Ag-specific T cell responses.
Autophagy was involved in the CCL-34-mediated Ag presentation to T cells. We next determined whether the CCL-34-induced enhancement of Ag processing and presentation by macrophages is regulated by autophagy. BMDMs were treated with CCL-34 in combination with or without the classical autophagy inhibitor, 3-MA, and analyzed for their Ag-presenting activity. As shown in Fig. 4A, DQ-OVAgreen fluorescence was decreased in CCL-34-stimulated BMDMs upon cotreatment with 3-MA, indicating that autophagy is involved BMDMs generated from C57BL/6 were incubated with candidate drugs for 24 hours and the Beclin 1 protein was measured by immunoblotting (n = 5). (E) NF-κB activation is involved in CCL-34-mediated autophagy. RAW264.7 cells were treated with Bay11-7082 (10 μM) for 1 hour and then incubated with the candidate drugs for 24 hours. LC3-II protein was detected by immunoblotting (n = 3). The results were plotted using GraphPad Prism version 8.1.0 (www. graphpad.com). All the data are shown as the mean ± SD, and *p < 0.05 indicates a significant difference versus the medium control or vehicle control analyzed using Student's t-test. Total protein (50 μg for RAW 264.  www.nature.com/scientificreports www.nature.com/scientificreports/ in Ag processing in CCL-34-stimulated macrophages. Moreover, the increased proliferation of OVA-specific CD4 + T cells as well as the production of IL-2 and IFN-γ mediated by CCL-34-activated macrophages were also suppressed under treatment with 3-MA ( Fig. 4B-D). These data indicate that autophagy plays a role in CCL-34-stimulated Ag processing and presentation to activate Ag-specific T cell responses.

CCL-34 functions as an immune adjuvant in vivo.
We next investigated whether CCL-34 can serve as an adjuvant in vivo. C57BL/6 mice were immunized with CCL-34 and DQ-OVA via the intraperitoneal injections. We found that cells with the ability to process Ag (DQ-OVAgreen + and DQ-OVAred + cells) were significantly increased in peritoneal cavity cells (PECs) but not in spleen, mesenteric lymph nodes (MLN) or other lymph nodes at 24 hours after injection (Fig. 5A). In addition, the total cell number of PECs in the CCL-34/ OVA-injected mice was higher than in the OVA alone group (Fig. 5B). Upon further phenotyping of PECs by flow cytometry, the percentage of monocytes (SSC low/med CD11b + Ly6C + ) was significantly increased in CCL-34/ OVA-injected mice (Fig. 5C). Notably, the percentages of CD80 + and CD40 + monocytes were also higher in CCL-34/OVA-injected mice (Fig. 5D). The percentage of large peritoneal macrophages (LPMs) was reduced following by the increased percentage of small peritoneal macrophages (SPMs) (Fig. 5E). Overall, CCL-34 enhances recruitment and activation of innate immune cells, and promotes antigen processing in vivo.
To further demonstrate that the activation of innate immunity induced by CCL-34 leads to activation of adaptive immune responses, the antibody responses in both C3H/HeN and C57BL/6 mice after immunization with CCL-34 and/or OVA were examined. The CCL-34/OVA-immunized mice showed higher levels of OVA-specific IgG in serum than the OVA-immunized group in both mouse strains, indicating that such immune responses were not strain-dependent (Fig. 5F,G). The adjuvant activity of alum, the positive control, is notably higher than CCL-34 (Fig. 5F,G). In addition, the level of IgG1 and IgG2a, the important markers for T-helper 2 and T-helper 1 immune responses respectively, also showed the increasing tendency in CCL-34/OVA-immunized mice (Fig. S2A,B). The CD3 + CD4 + T cells were also increased in the spleen of CCL-34/OVA-immunized mice (Fig. 5H). Furthermore, to investigate whether autophagy participates in CCL-34-mediated immune responses, the mice were immunized with CCL-34 in combination with or without the autophagy inhibitor, 3-MA. As shown in Fig. 5G, the induction of OVA-specific IgG by CCL-34 was suppressed by 3-MA treatment. These data demonstrate that CCL-34 can function as an adjuvant in vivo by promoting autophagy. Additionally, CCL-34 did not show any toxicity based on the serum biochemistry data (Fig. 6).

Discussion
Our previous studies found that the synthetic glycolipid, CCL-34, can activate macrophages and induce maturations of DCs in a TLR4-dependent manner [28][29][30] . CCL-34 also exhibits anticancer activity via macrophage-released NO-mediated thoc1 downregulation in cancer cells 39 . In this study, we revealed that CCL-34 can induce TLR4-mediated autophagy and enhance Ag processing in macrophages. Furthermore, we also demonstrated CCL-34 could induce Ag-specific immune responses both ex vivo and in vivo.
Previous studies have shown that TLR4 is an environmental sensor of autophagy and that LPS-induced autophagy can overcome the mycobacterial phagosome block 13 . In addition, TLR4, TLR2 and TLR7 are known to promote autophagy in immune cells as a mechanism of pathogen elimination 24 . Several molecules and pathways have been reported to be involved in the regulation of TLR-mediated autophagy. TLR signaling increases the interaction of MyD88 and Trif with Beclin 1, leading to the induction of autophagy 40 . The autophagy-promoting molecule p62 (SQSTM1), which enhances the formation of ALIS and autophagosomes, is required for mycobactericidal activities 33,41 . The MyD88-NF-κB-DRAM1 axis is also critical for autophagic defenses against intracellular pathogens 37 . Our data in this study demonstrated the induction of Nrf2-p62 and Beclin 1 in CCL-34-activated macrophages ( Fig. 2A-D). Furthermore, we also demonstrated that CCL-34-mediated autophagy was NF-κB-dependent (Fig. 2E). These data indicate that CCL-34-induced autophagy may play pivotal roles in TLR4-mediated mycobactericidal activities.
Vaccines play an important role in the prevention of infectious diseases by inducing pathogen-specific responses. One of the strategies for vaccine development is to combine recombinant antigens with adjuvants that modulate the immunogenicity of antigens as well as the microenvironments for activation of innate immunity 42 .
To design safer and more effective vaccines, enhancing the immune effects of adjuvants via well-characterized mechanisms is necessary. In both preclinical studies and clinical trials, TLR4 agonists have well-defined mechanisms and have been proven to be potent immune stimulants to facilitate Ag delivery and generate appropriate microenvironments for the activation of APCs 43 . Therefore, TLR4 agonists are considered as potentially safe, universal and effective vaccine adjuvants for clinical applications [44][45][46] . Increasing autophagy-mediated Ag presentation has been demonstrated as a simple and powerful strategy to improve vaccine efficiency, as in the case of the BCG vaccine 22 . Autophagy-inducing small molecules or peptides have been demonstrated as potential adjuvants through their enhancement of Ag delivery, processing and presentation [47][48][49] . Combination with TLR2-stimulating peptide or LPS in BCG vaccines can induce autophagy in APCs and promote immunogenicity days. The CFSE dilution, as an indicator of OT-II CD4 + T cell proliferation, was detected by flow cytometry. (E) A representative histogram; (F) The quantitative data (n = 4). (G,H) Co-culture of CCL-34-activated macrophages with CD4 + T cells increases the production of IL-2 and IFN-γ. The production of IL-2 (G) and IFN-γ (H) was measured using ELISA (n = 4). The results were plotted using GraphPad Prism version 8.1.0 (www.graphpad.com). All the data are shown as the mean ± SD, and *p < 0.05 indicates significant difference versus medium control or vehicle control analyzed using Student's t-test. The gating strategies are presented in Supplementary Fig. 5A www.nature.com/scientificreports www.nature.com/scientificreports/ in mice 22,23 . Therefore, small molecules selectively triggering TLRs-mediated autophagy presumably increase immunogenicity and can potentially be developed as mycobacterial vaccines and adjuvants. In this study, we demonstrated that the TLR4 activator CCL-34 can induce autophagy, promote Ag-specific immune responses and act as an efficient adjuvant.
Our previous data have demonstrated that CCL-34 activates the TLR4 signaling pathway and induces cytokines related to the immune response 29,30 . In this study, we further showed that the Ag-specific T cells were  (C,D) The production of IL-2 and IFN-γ induced by co-culture of CCL-34 with CD4 + T cells is suppressed by 3-MA. The production of IL-2 (n = 6) (C) and IFN-γ (n = 7) (D) was measured using ELISA. The results were plotted using GraphPad Prism version 8.1.0 (www.graphpad. com). All the data are shown as the mean ± SD, and *p < 0.05 indicates significant difference versus the medium control or vehicle control analyzed using Student's t-test. The gating strategies are presented in Supplementary  Fig. 6A www.nature.com/scientificreports www.nature.com/scientificreports/ On the other hand, the adjuvant activity of CCL-34 in current study was investigated using the drug concentration, drug delivery method and immunization route based on our previous report 28 . Considering further improvement of adjuvant activity of CCL-34 in vivo, more effective immunization, more dosages and alternative immunization routes, such as the subcutaneous or intradermal injection 51 , may be tested in the future study. Moreover, the vaccine formulations can also play a critical role in vaccine efficiency via controlling the vaccine biodistribution and presentation to immune cells. The water-in-oil emulsion and liposomes are commonly used in TLR4-based adjuvant systems, such as AS01, AS02 and AS15 42,51 . Therefore, the water-in-oil emulsion and liposomes is suggested to develop potential formulation of CCL-34.
A previous study using i.p. injection as the immunization route showed that macrophages and neutrophils were the main cell population recruited into peritoneal cavity and then migrated into mesenteric lymph nodes (MLN) 52 . Our data also showed that macrophages and monocytes are the major innate immune cells recruited in peritoneal cavity of animal model (Fig. 5C). Therefore, we focused on the functions of macrophages in this study. LPMs (F4/80 high MHC-II low ) are the major macrophages that regulate the homeostasis of the peritoneal cavity and they migrate to the omentum during immune stimulation. SPMs (F4/80 low MHC-II high ) are the minor subset in the unstimulated peritoneal cavity but they become the major population for the secretion of cytokines and NO under immune stimulation or infection [53][54][55][56][57][58] . Previous studies showed that upon LPS stimulation, LPMs disappeared while the number of SPMs and monocytes increased in mice stimulated with LPS, TLR-4 agonists or Alum using i.p. injection 54,57,59,60 . LPS-stimulated LPMs may migrate into lymph nodes, serve as APCs, and trigger adaptive immune responses 54,57 . Furthermore, recent study found SPMs have the ability to present antigens to CD4 + T cells in MLNs 60 . In our study, we found SPMs were increased accompanied by the decrease of LPMs under CCL-34 immunization (Fig. 5E), indicating that both LPMs and/or SPMs may function as the APCs in CCL-34-immunized mice. On the other hand, CCL-34 was shown to induce DC maturation ex vivo in our previous study 30 . The role of DCs in CCL-34-mediated adjuvant activity in animal model remains further investigation.
CD4 + T cells were also significantly increased in the spleens of CCL-34-immunized mice whereas the percentage of CD4 + T cells was not affected by 3-MA treatment (Fig. 5H). Notably, the percentage of CD4 + T cells was increased under treatment with CCL-34 alone without antigens (Fig. 5H). Therefore, we hypothesized that CCL-34 may induce an antigen-or autophagy-independent increase of CD4 + T cells. In addition, although the percentage of CD4 + T cells was increased under treatment of CCL-34 and CCL-34 plus 3-MA, the functions of T cells in vivo were not examined in our study. Considering the critical role of autophagy in regulating T cell responses 6 , whether the functions of CCL-34-induced T cells are suppressed by 3-MA in vivo can be further explored in the future. Taken together, our data indicate that CCL-34 regulates both LPM and SPM functions and induces the percentage of CD4 + T cell which is similar to the immune stimulating response of LPS.
Although LPS is a natural adjuvant that activates TLR4 signaling pathways and has a profound effect on CD4 + T cell responses, its powerful adjuvant activity is associated with toxicity 61 . A high-dose LPS treatment can induce acute inflammation and lead to sepsis in vivo 62 . However, in the case of MPLA, an LPS-derived TLR4 agonist, toxicity and immunogenicity are not always linked 61,63 . CCL-34 is a synthetic TLR4 agonist that has II high ), T cells (CD45 + CD3 + ) and B cells (CD45 + B220 + ) are shown. www.nature.com/scientificreports www.nature.com/scientificreports/ a defined structure, and its structure is less complicated compared with that of other TLR4 agonists. Notably, our data showed that the dose used for an in vivo experiment in CCL-34-treated mice did not cause toxicity based on serum biochemical data, including GOT (glutamate oxaloacetate transaminase), GPT (glutamic pyruvate transaminase), BUN (blood urea nitrogen) and CRE (creatinine) measurements (Fig. 6). In addition, activation of TLR4 by CCL-34 does not lead to high secretion of proinflammatory cytokines as LPS treatment dose 29 . Therefore, the immune response induced by CCL-34 may be safer than that induced by LPS.
Together, our results demonstrated that CCL-34 induced autophagy in macrophages and functioned as an immune adjuvant in the induction of Ag-specific immune responses. Although some of the underlying mechanisms of CCL-34-mediated T cell activation and SPM/LPM regulation were not fully investigated in vivo, our results revealed that autophagy was involved in CCL-34-mediated Ag processing and humoral immune responses, providing the potential application of CCL-34 as a vaccine adjuvant. These findings also highlight the potential clinical application of TLR agonist and/or autophagy-inducing agents as vaccines or adjuvants.
Mice. The OT-II mice were purchased from the Jackson Laboratory, and the C56BL/6, C3H/HeN and C3H/ Immunoblotting. Cells were lysed with modified RIPA buffer (90 mM Tris, 150 mM NaCl, 1% NP40, 0.25% sodium deoxycholate, 5 mM EDTA, and 1 mM EGTA) containing protease inhibitor cocktails (Sigma-Aldrich). The protein quantity was determined by Bradford assay (Sigma-Aldrich). Total protein (50 μg for RAW 264.7 cells or 25 μg for BMDM) was analyzed on SDS-PAGE gel and transferred onto PVDF membranes (Bio-Rad). The membranes were blocked with 5% non-fat milk or 5% BSA (Sigma-Aldrich), incubated with the indicated primary antibodies at 4 °C overnight, and then incubated with appropriate peroxidase-conjugated secondary antibodies. The membranes were reacted with an enhanced chemiluminescence reagent (Merck Millipore, MA, USA) and the signals were detected using a Luminescence/Fluorescence Imaging System (Fujifilm, Tokyo, Japan).
Fluorescence microscopy. Cells were grown on glass coverslips and stimulated with the candidate drugs. Cells were fixed with 4% formaldehyde, stained with the blue nuclear chromatin stain 4′, 6-diamidino-2-phenylindole, dihydrochloride (DAPI) (Sigma-Aldrich) and mounted using ProLong Diamond Antifade Mountant (Thermo Fisher Scientific). The samples were pictured using an Olympus BX61 Microscope (Olympus Corporation, Tokyo, Japan). The percentage of cells with LC3 punctate dots was calculated to quantitate autophagy. A minimum of 150 cells per sample were counted under different experimental conditions. MTT cell proliferation/viability assay. RAW264.7 cells (10 4 ) were cultured in 96-well plates and treated with the candidate drugs for 24 hours. Then the cells were incubated with 0.5 mg/mL MTT reagent (Sigma-Aldrich) for 4 hours, followed by incubation with solubilization buffer (12% SDS in 45% DMF, pH=4.7). The absorbance at 550 nm (reference 650 nm) was measured by ELISA reader (Bio-Rad, Hercules, USA).
Differentiation of murine bone marrow-derived macrophages. The bone marrow cells were isolated from C56BL/6, C3H/HeN and C3H/HeJ mice, and differentiated as described previously 64 . In brief, bone marrow cells (10 7 cells) were cultured in complete medium. On day 2 and day 4, new medium supplemented with M-CSF (20 ng/mL, R&D Systems) was added. On day 6, non-adherent cells and loosely adherent cells were gently washed with PBS. The adherent cells were detached using 2.5 mM EDTA and reseeded for the following experiments.
RNA extraction and the quantitative reverse transcription polymerase chain reaction (qRT-PCR). Total RNA was extracted using TRIzol reagent according to the manufacturer's instructions (Thermo Fisher Scientific), and the total RNA was reverse transcribed using the ThermoScript RT-PCR system (Thermo Fisher Scientific). The cDNA product was analyzed by qRT-PCR using Fast SYBR Green Master Mix (Thermo Fisher Scientific). The primer sequences were as follows. Mouse p62: ACAGCCAGAGGAACAGAT (sense) and ACAAGAATGCCAAGACACT (antisense). Mouse Il12a: TATCTCTATGGTCAGCGTTCC (sense) and TGGTCTTCAGCAGGTTTCG (antisense). Mouse Il12b: TCATCAGGGACATCATCAAACC (sense) and TGAGGGAGAAGTAGGAATGGG (antisense). Mouse GAPDH: 5′-TGTGATGGGTGTGAACCACGAG (sense) and TGCTGTTGAAGTAGCAGGAGAC (antisense). All assays were performed in triplicate using the Applied Biosystems Model 7000 instrument (Thermo Fisher Scientific). The data are quantitated using 2 −ΔCt (ΔCt = CtTarget gene-CtGAPDH; Ct: cycle number when the fluorescent value of the sample is equal to the threshold value).

Measurement of Ag processing in vitro.
BMDMs were stimulated with the candidate drugs in combination with 5 μg DQ-OVA (Thermo Fisher Scientific) for 24 hours. The cells were stained with anti-F4/80-PECy7 (BioLegend 123113), and analyzed by FACSCanto (BD Biosciences).
Ag-specific T cell activation assay. BMDMs isolated from C56BL/6 mice were stimulated with the candidate drugs in the presence of 5 μM OVA-peptide 323-339 (AnaSpec, CA, USA) for 24 hours. The CD4 + T cells were isolated from the spleen and mesenteric lymph nodes (MLN) of OT-II mice using CD4 (L3T4) MicroBeads (Miltenyi Biotec 130-049-201), and the CD4 + T cells were stained with 2 μM CellTrace CFSE Cell Proliferation Kit (Thermo Fisher Scientific) for 5 minutes at 37 °C. The stimulated BMDMs (2.5×10 4 ) and CD4 + CFSE labeled OT-II T cells (5×10 4 ) were resuspended and cocultured in 96-well U bottom plates for 3 days and 5 days, respectively. The supernatant was collected on day 3 for IL-2 detection. On day 5, the supernatants were collected for IFN-γ detection, and the proliferating T cells were stained with anti-F4/80. Dead cells were excluded by staining Mouse immunizations. Eight-to 10-week-old C57BL/6 or C3H/HeN mice were immunized i.p. with PBS alone, ovalbumin (100 μg/mouse, Sigma-Aldrich) in PBS, OVA (100 μg/mouse) plus Al(OH) 3 (80 mg/kg, Thermo Fisher Scientific), OVA (100 μg/mouse) plus CCL-34 (4 mg/kg), or OVA (100 μg/mice) plus vehicle (10% DMSO) in a volume of 100 μl. The method and dose of OVA immunization were chosen according to a previous publication [66][67][68] . For the short-term immunization, the mice were sacrificed 24 hours after injection. For the long-term immunization, the mice were injected with the above-indicated drugs on day 0, day 7 and day 14, and sacrificed on day 21. For the combined treatment with autophagy inhibitor, 3-MA (20 mg/kg) was administered intraperitoneally 30 minutes before immunization as previously described 69 . The i.p. injections were carried out in awake mice without anaesthesia procedure. Measurement of serum OVA-specific IgG. The levels of OVA-specific IgG in serum were measured using ELISA as described previously 70 . In brief, OVA-specific IgG was bound to an ELISA plate (Nunc, Roskilde, Denmark) coated with 2 mg/mL OVA and detected by enzyme reaction after incubating with polyclonal HRP-conjugated anti-mouse IgG (Sigma-Aldrich A9044) and tetramethylbenzidine. After the enzyme reaction, absorbance was measured at 450 nm (reference 540 nm).

Isolation of mouse peritoneal cavity cells (PECs
Data and statistical analyses. All the data from at least three independent experiments were analyzed with Student's t-test by GraphPad Prism version 8.1.0 for Windows (GraphPad Software Inc., La Jolla, USA, www. graphpad.com). All the results are shown as the mean ± SD, and statistical significance is indicated by * (p < 0.05) for comparison between untreated and treated groups.