The co-administration of adjuvants is required to induce T-cell immunity to protein antigens. Unfortunately, the lack of understanding the molecular and cellular mechanisms of adjuvant-induced immune responses hinders rational vaccine design and deployment, especially for emerging infectious diseases, or vaccines where antigen sparing (e.g., influenza) is an objective. Adjuvants in vaccines licensed for humans include aluminum salts, oil-in-water emulsions (MF59, AS03), liposomes (AS01), CpG oligonucleotides, and virosomes. While many of these induce strong antibody responses, most of them only weakly provoke Th1 and especially Th17 immunity, which is crucial for defense against intracellular pathogens, e.g., Mycobacterium tuberculosis (1). The use of molecularly defined adjuvants is a key component required for successful development of novel recombinant subunit vaccines. However, despite recent progress (2), to date, such adjuvants for safe and efficient use in humans are still lacking (3).

DDA/TDB [also known as CAF01 (4)] is a next generation synthetic adjuvant. DDA/TDB has been successfully evaluated in phase I clinical studies for vaccination with, e.g., the recombinant M. tuberculosis fusion protein Ag85B-ESAT-6 (H1) (5) and with the recombinant chlamydia protein CTH522 (6). DDA/TDB is a mixture of cationic dimethyldioctadecylammonium (DDA) surfactant lipid-based liposomes containing trehalose-6,6-dibehenate (TDB), the synthetic analog of the mycobacterial cord factor trehalose-6,6-dimycolate (TDM). In contrast to, e.g., aluminum salts and the TLR9-triggering adjuvant CpG, it was shown that DDA/TDB induces strong and long-lasting Th17 memory responses in mice (7–11). TDB and TDM bind to the C-type lectin receptor Mincle (Clec4e) and activate antigen-presenting cells (APC) via the FcRγ-Syk-Card9 pathway (10, 12, 13). Consequently, deletion of the Mincle receptor or the FcRγ-Card9 pathway in mice strongly reduces the Th17 adjuvant activity of TDB (13, 14) and TDM (15). In addition, MyD88-dependent signaling via IL-1R is required for induction of DDA/TDB-mediated generation of vaccine-induced immunity (14). Expression of Mincle is strongly regulated by cytokines, with IL-4 suppressing Mincle messenger RNA (mRNA) and protein levels in mouse and human macrophages (16, 17), whereas tumor necrosis factor (TNF) is essential for the upregulation of Mincle and Th17 adjuvanticity by DDA/TDB (18). In addition to the cord factor analog TDB, other trehalose esters have been explored as Mincle agonists in the search for novel Th17-inducing adjuvants (19–25).

Beyond the definition of innate immune receptors and pathways essential for adjuvant activity, the question of which APC are required for the generation of T-cell responses after immunization is pivotal for rational vaccine design. Current thinking suggests that vaccine immunogenicity depends on direct dendritic cell (DC) targeting and activating capacity of the administered adjuvant. However, following injection of fluorescently labeled DDA/TDB, only a minute fraction of CD11c⁺ lymph node DC contained antigen/adjuvant and became activated (26). Therefore, whether this exquisite targeting of DC in the lymph node is the major prerequisite for a successful adjuvant remains unclear. Many adjuvant formulations, including the classical human adjuvant aluminum hydroxide cause strong recruitment of innate immune cells to the SOI (27, 28). Surprisingly, aluminum hydroxide has no stimulatory effect on DC in vitro (29). Instead, in vivo experiments revealed that, in the presence of aluminum hydroxide, antigen was taken up by recruited inflammatory monocytes and transported to the dLN. These cells differentiate into inflammatory DC and contribute to adaptive immune responses (29). Subcutaneous immunization with DDA/TDB liposomes in mice causes robust and sustained local swelling and cellular infiltration (10, 13, 14). However, the detailed cellular composition and contribution of the local inflammatory response to successful adjuvanticity are unknown. Here, we investigated the recruitment of APC to the SOI and its dependence on the Mincle receptor and MyD88-signaling and identified CCR2⁺ monocytes, but not neutrophils, as critical for the induction of Th1/Th17 responses by the DDA/TDB adjuvant in vivo.

The new generation synthetic glycolipid adjuvant DDA/TDB induces robust and persistent Th1/Th17 responses in animal models (9, 10, 14, 37) and promotes long-lived T-cell responses to the MTB fusion protein H1 in humans (5). In this manuscript, we have investigated the cellular recruitment to the injection site, the signaling mechanisms required, and the functional role of the most abundantly recruited cell types. Neutrophils and monocytes formed the major component of the cellular infiltrate and were recruited in a partially Mincle- and MyD88-dependent manner. The major finding of this study is the identification of a pivotal function of CCR2⁺ monocytes in the promotion of Th1/Th17-biased antigen-specific immunity by DDA/TDB. This requirement for monocytes was selective for DDA/TDB, as the Th1-inducing effect of the TLR9 ligand CpG ODN was not affected in Ccr2^−/− mice. Neutrophils, although being the most abundantly recruited cell type after DDA/TDB injection, were dispensable for adjuvanticity, as shown by antibody-mediated depletion and confirmed in neutrophil-deficient Mcl-1 conditional knockout mice. Thus, our results show a dichotomy in the functional roles of monocytes and neutrophils, with implications for future adjuvant optimization and investigations of cellular and molecular mechanisms.

Neutrophils can contribute to the development of adaptive immune responses by release of chemokines and cytokines attracting monocytes and DC and by the transport of antigen or whole microorganisms to lymphoid tissues (38, 39). However, strong neutrophil infiltration can also be associated with more severe side effects after vaccination, ranging from pain at the injection site to severe tissue damage (40). Therefore, our finding that a lack of neutrophils did not hinder the adjuvant effect of DDA/TDB implies the possibility of reducing unwanted inflammation due to granulocyte infiltration by developing and selecting adjuvants with a selective targeting of monocytes. On the other hand, the tendency for stronger Th1/Th17 induction observed here in mice depleted of neutrophils with 1A8 antibody ( Figure 3 ) may indicate a negative regulatory effect of neutrophils, consistent with triggering of IL-10 release by the cord factor analog TDB (41) or the previously described chemokine binding and degrading function of neutrophil extracellular traps (42).

CCR2-dependent monocyte recruitment plays an essential role in the control of several bacterial and fungal infections by inducing protective immune responses (43–49). In addition, Ccr2^−/− mice show diminished IFNγ responses upon immunization with Complete Freund’s Adjuvant (43), whose activity is partially due to the TDM of the killed mycobacteria and, similar to TDB, mediated by Mincle-Card9- and IL1R-MyD88-dependent signaling (15). In contrast, soluble CpG ODN-induced generation of Th1 cells was not affected by CCR2 deficiency ( Figure 4 ). This differential requirement is consistent with a crucial role for monocytes in promoting the response to particulate vaccines in the dLN, where they may present antigen to T cells directly, transfer the antigen to DC, or create the appropriate cytokine milieu through release of soluble mediators.

Whether monocytes can directly prime naive T-cell responses is debated. Following immunization with aluminum hydroxide as adjuvant, monocytes recruited to the SOI take up antigen and differentiate to monocyte-derived DC (moDC) during their migration to the dLN (29). During infection with Leishmania major, moDC controls the induction of protective Th1 responses (50). Recruited monocytes differentiate into moDC and migrate to the dLN, and only these moDCs capture and present L. major-derived antigens. On the other hand, evidence against direct priming capabilities of moDC has been reported using a yeast vaccine strain expressing a model antigen (51). Despite association of yeast and moDC, peptide:MHCII complexes could not be detected on these cells. Instead, the material had to be transferred to LN-resident DC or migratory skin-derived DC, which were essential for priming of naive T cells (51). During infection with M. tuberculosis, inflammatory monocytes were necessary for CD4⁺ T-cell priming because they transported bacteria to the dLN. However, by transferring MHCII-deficient monocytes into monocyte-depleted hosts, it was shown that only conventional DC (cDC) were able to induce antigen-specific CD4⁺ T-cell proliferation (52). Recently, Bosteels et al. described that cDC2 in inflammatory conditions acquire a phenotype sharing characteristics with moDC (e.g., expression of CD64) and termed these cells inf-cDC2 (53). Mixed bone marrow chimeras showed that CCR2 was required for the presence of inf-cDC2 in the lungs after respiratory virus infection (53) and in the draining lymph nodes after immunization with the adjuvant AS01 (54).

In our experimental system, flow cytometry revealed no differences in cellular composition of the dLN of Ccr2^−/− mice (except for the monocytes), and the numbers of adjuvant-positive cells were comparable. Thus, DDA/TDB adjuvant is targeted equally well to APC in the dLN in the absence of monocytes. Therefore, the essential function of the monocytes recruited to the SOI and the dLN may be to provide cytokines and mediators acting as signal 3 to direct the T-cell response in the dLN towards Th1/Th17. Candidates to be tested in future experiments include IL-1, shown by us and others to direct Th1/Th17 induction by DDA/TDB (14, 15), and IL-6, which is produced by human monocytes and important for Th17 differentiation (55).

The rapid and striking increase in cell numbers in the dLN following DDA/TDB administration is likely due to trapping of cells rather than to proliferation. A shutdown phase, leading to a dramatic reduction in efferent cell output, has been described (56) and shown to be mediated by TNFα and IL-6 (57). We detected IL-6 release already 2 h p.i. in peritoneal lavage fluids. In addition, IFNα/β-mediated upregulation of CD69 on T cells contributes to their retention within lymphoid organs (58); however, we have not addressed induction of type I interferon upon immunization here. Trapping of cells in the dLN maximizes the number of naive antigen-specific T cells available for encounter with antigen-loaded APC. Cell trapping was a MyD88-dependent effect predominantly induced by DDA during the first 24 h, whereas a further Mincle- and MyD88-dependent increase in cell numbers at d3 p.i. was only observed when TDB is incorporated into the liposomes. Remarkably, CCR2 deficiency had no effect on lymph node cellularity following a single immunization with DDA/TDB, indicating that the factors mediating lymph node shutdown are produced independently of monocytes and are insufficient to direct Th1/Th17 responses.

In conclusion, our study provides evidence that Th17 immunity induced by a Mincle-targeting adjuvant depends on the recruitment of CCR2⁺ inflammatory monocytes. In contrast, the abundantly and rapidly recruited neutrophils were completely dispensable for the adjuvant effect. How exactly the monocytes promote Th1/Th17 generation remains to be elucidated. Since the Mincle- and MyD88-dependent increase in lymph node cellularity and the appearance of vaccine-containing APC in the draining lymph node were not compromised in CCR2-deficient mice, the contribution of monocytes may consist in the activation of APC or provision of signal 3 to T cells.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

The animal study was reviewed and approved by Animal Protection Committee Regierung von Mittelfranken.

CD performed most experiments. CD and RL designed experiments and analyzed data. PM and LH performed experiments. PM, CL, LH, DC, PA, DD, and MM provided reagents and contributed to writing of the manuscript. CD and RL wrote the manuscript. All authors contributed to the article and approved the submitted version.

PA and DC are co-inventors of patents relating to cationic liposomes. All rights have been signed to Statens Serum Institute.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.