The Novel Toll-Like Receptor 2 Agonist SUP3 Enhances Antigen Presentation and T Cell Activation by Dendritic Cells

Dendritic cells (DCs) are highly specialized antigen-presenting cells that play crucial roles in innate and adaptive immunity. Previous studies suggested that Toll-like receptor (TLR) agonists could be used as potential adjuvants, as activation of TLRs can boost DC-induced immune responses. TLR2 agonists have been shown to enhance DC-mediated immune responses. However, classical TLR2 agonists such as Pam3CSK4 are not stable enough in vivo, which limits their clinical applications. In this study, a novel structurally stable TLR2 agonist named SUP3 was designed. Functional analysis showed that SUP3 induced much stronger antitumor response than Pam3CSK4 by promoting cytotoxic T lymphocytes activation in vivo. This effect was achieved through the following mechanisms: SUP3 strongly enhanced the ability of antigen cross-presentation by DCs and subsequent T cell activation. SUP3 upregulated the expression of costimulatory molecules on DCs and increased antigen deposition in draining lymph nodes. More interestingly, SUP3 induced less amount of pro-inflammatory cytokine production in vivo compared to other TLR agonists such as lipopolysaccharide. Taken together, SUP3 could serve as a novel promising immune adjuvant in vaccine development and immune modulations.

countries) in oncological conditions (3)(4)(5)(6)(7). TLR2 is a conserved molecule expressed by both murine and human conventional DCs (cDCs), and TLR2 agonists have been shown to have great potential to be utilized as adjuvants to enhance cDC-mediated immune responses (8). However, until now TLR2 agonists have not been approved as clinical adjuvants, possibly due to their poor stabilities and thus weak potency in vivo (1,2,9). It is therefore desirable to design new TLR2 agonist, if it would be used as an adjuvant in immunotherapy. Innate immunity represents the first line of defense against invading pathogens. Innate immune cells like DCs sense danger signals and initiate responses against them. As the most efficient antigen-presenting cells (APCs), cDCs are in charge of recognizing antigens and promoting adaptive immunity through activation of naïve T cells (10,11). Stimulation of DC upon recognition of antigens via multiple pattern recognition receptors (PRRs) upregulates the expression of costimulatory molecules and induces the production of various cytokines and chemokines, contributing to T cell activation and inflammatory responses. Therefore, DCs bridge innate and adaptive immunity and play pivotal roles in orchestrating immune responses (12). Murine splenic DCs are heterogeneous populations consisting of CD8 + cDCs, CD11b + cDCs, and plasmacytoid DCs (pDCs) (13,14). CD8 + cDCs are specialized in antigen cross-presentation via major histocompatibility complex (MHC)-I to activate CD8 + T cells and ultimately differentiate them into cytotoxic T lymphocytes (CTLs), suggesting CD8 + cDCs as essential initiators of cellular immunity. Moreover, CD8 + cDCs can polarize CD4 + T cells into Th1 phenotype by direct antigen presentation through MHC-II pathway and by producing IL-12p70. The CD11b + cDCs mainly uptake exogenous antigens and directly present them to CD4 + T cells via MHC-II to induce differentiation of multiple subsets of helper T cells, including Th2 and Th17, which are crucial in defense against extracellular pathogens like parasites and bacteria (15). In contrast, pDCs are characterized by secreting large amount of type I interferons (IFNs) upon viral infection, but have limited antigen presentation ability. The heterogeneity of DCs highlights the concept of division of labor through functional specialization of DC subsets (15).
Pathogen-associated molecular patterns induce functional maturation of immune cells through binding on specific PRRs and enhancing the cellular functions. For instance, ligation of TLR on DCs effectively program DCs into an active status and dramatically enhance DC functions. Therefore, TLR agonists are widely used as stimulators of DCs and enhancers of DC-mediated immune responses (2,16,17). TLR2 is a special member among TLR families that needs to form heterodimers with TLR1 or TLR6 to recognize triacyl (Pam3CSK4) or diacyl (Pam2CSK4) lipopeptides, respectively (18,19). Cell surface receptor TLR2 recognizes lipopeptides mainly derived from cell wall components of bacteria to elicit innate signaling events. TLR2 represents one of important sensors to pathogens and TLR2 deficiency impairs host defense against bacterial infection (20,21). As TLR2 is conserved in mouse cDCs and their human counterparts, especially between murine CD8 + cDCs and the human equivalent CD141 + cDCs (22), agonists targeting TLR2 could be promising adjuvants in designing vaccines through activating DCs for cancer immunotherapy. Previous studies showed that conjugation of antigens and TLR agonists targeting APCs achieved dramatically more potent responses than combination of antigens and TLR agonists (23)(24)(25)(26). Unlike other TLR ligands, the intrinsic peptide component of TLR2 ligands provides the possibility to directly conjugate antigenic peptides within TLR2 agonists. Notably, the expression of TLR2 by DCs, providing the potential for using TLR2 agonists as adjuvants in DC-mediated immune responses, and so far increasing number of TLR2 agonists have been developed (24, [27][28][29][30]. Recently, it has been reported that TLR2 agonists could be used directly (31)(32)(33)(34) or as structurally modified forms (26) in cancer treatment, suggesting that TLR2 agonists can be potential effective enhancers for cancer immunotherapies. Administration of TLR2 agonists could enhance effector and memory T cell responses, leading to elevated efficacy of vaccination and tumor rejection (24, 30,31,33,35). Moreover, TLR2 agonists could sensitize B cell lymphoma to chemotherapeutic agents via upregulating costimulatory molecules to increase their sensitivity to NK cell and CTL cytotoxicity (32), or inducing caspase 8-dependent apoptosis (34). These facts have made TLR2 agonists to be attractive adjuvants in the therapy of cancers (28,30,35). However, current available TLR2 agonists have limited applications due to their rapid degradation in plasma where abundant esterase exists, which breaking ester bonds connecting glycerol backbone and palmitic acid chains, resulting in diminished stimulation of TLR2. Hence, it is necessary to develop new TLR2 agonists with higher stability in vivo and can be further utilized as immune adjuvants.
In this study, we designed a new TLR2 agonist termed SUP3 based on the structure of Pam3CSK4 (Pam3 hereafter), which is a chemically more stable ligand for TLR2/TLR1 heterodimers. Functional analysis revealed that SUP3 had stronger capabilities than Pam3 in DC-mediated activation of CTL and T-dependent antibody production. Further studies of its activity and effects on immune responses showed that SUP3 enhanced antigen presentation ability of cDCs and subsequent T cell activation without leading to excessive inflammatory responses. Our studies demonstrated that SUP3 could function as a potential immune adjuvant in vaccination and immunotherapy.

MaTerials anD MeThODs
Mice C57BL/6 mice were maintained in a specific pathogen-free facility at Tsinghua University. OT-I and OT-II transgenic mice were purchased from Jackson Laboratory. Mice were sacrificed at the age of 6~8 weeks for experiments. The experiments using mice were approved by the Institutional Animal Care and Use Committees at Tsinghua University.
DCs were incubated with OVA-AF488 (5 µg/mL) at 37°C or on ice (negative control) for 30 min. Cells were washed twice with cold medium and analyzed by flow cytometry.

Tumor Model
Tumor model was applied as previously described with some modifications (42

OT-ii activation In Vivo
Carboxyfluorescein diacetate succinimidyl ester-stained OT-II CD4 + T cells (2 × 10 6 cells per mouse) were transferred into mice 1 day prior immunization. Then the recipients were subcutaneously (s.c.) injected with OVA alone (100 µg per mouse), or OVA mixed with SUP3 (2 nmol per mouse) or Pam3 (2 nmol per mouse) at groin. Three days after immunization, OT-II CD4 + T cells from inguinal lymph nodes were analyzed by flow cytometry.

antigen Deposition
Wild-type mice were s.c. injected with OVA-AF647 alone (25 µg per mouse; Molecular Probes O-34784), or mixtures of OVA-AF647 with SUP3 (2 nmol per mouse) or Pam3 (2 nmol per mouse) at groin. Inguinal lymph nodes were isolated 24 h post-injection, and then immune cells were analyzed by flow cytometry.

statistical analysis
Data were analyzed by ANOVA followed by Bonferroni multiple comparison test or by unpaired, two-tailed Student's t test with GraphPad Prism 5. Data were presented as mean ± SD or ± SEM, and p < 0.05 (*) was considered statistically significant.

Design of Tlr2 agonist sUP3
Pam3 is a lipopeptide and serves as a TLR2 agonist through binding to TLR2/TLR1 heterodimer. Although Pam3 is widely used as a TLR2 agonist, some intrinsic problems remain in detailed mechanism studies and drug development. The two ester groups of Pam3 can be easily hydrolyzed in plasma, and the hydroxyl group in serine could be readily oxidized in physiological conditions ( Figure S1A in Supplementary Material). All of these facts made Pam3 less stable in vivo. In order to overcome the disadvantages of the current existing TLR2 ligands, we designed a new TLR2 agonist that should be chemically and metabolically more stable. Based on the structural features of Pam3 and TLR1/2 receptors co-crystal, we replaced the two ester-linked palmitoyl groups with chemically more stable bioisostere, carbamate-linked C14 groups (Figures S1A,B in Supplementary Material). In addition, we displaced serine residue in Pam3 with glycine residue to eliminate this chiral center (Figures S1A,B in Supplementary Material). Since the third lysine residue in Pam3 is crucial for H bonding, we kept three lysine residues in the new molecule and removed the forth lysine residue to further reduce the molecular weight of the new TLR2 agonist, (2S,5S,8S,14R,18R)-2,5,8-tris(4-aminobutyl)-4,7,10,13,21pentaoxo-14-palmitamido-18-((tetradecylcarbamoyl)oxy)-20-oxa-16-thia-3,6,9,12,22-pentaazahexatriacontan-1-oic acid. It is termed SUP3 (Figure 1A; Figures S1A,B in Supplementary Material). Chemical characterization of SUP3 by 1 H nuclear magnetic resonance and electrospray ionization mass spectrum analysis further confirm the structure ( Figures S2A,B in Supplementary Material). The affinity of SUP3 is comparable with Pam3 ( Figure S3A in Supplementary Material), whereas the plasma stability of SUP3 is much higher than Pam3 ( Figure S3B in Supplementary Material).
To validate the specificity of SUP3 for TLR2, we generated Tlr2 −/− RAW 264.7 cells with CRISPR-Cas9 system ( Figure 1B). Both Pam3 and SUP3 stimulation induced large amount of TNFα production by WT RAW cells, indicating that SUP3 could function similarly as Pam3 in triggering TLR2 response ( Figure 1C). TLR2 deficiency abolished the effect of Pam3 and SUP3 on RAW cells, evidenced by no secretion of TNFα by Tlr2 −/− RAW cells upon challenge by Pam3 and SUP3 ( Figure 1C). It demonstrated that TLR2 was the receptor for SUP3. To further determine the bioactivity of SUP3 as a TLR2 agonist, we determined the cytokine production by GMDCs upon TLR2 stimulation. GMDCs were generated from bone marrow cells cultured in the presence of GM-CSF and interleukin-4 (44,45). It has been demonstrated that GMDCs are potent pro-inflammatory cytokine producers upon TLR stimulation and GMDCs express TLR2 and respond to TLR2 ligands. We found that SUP3 exhibit equal ability as Pam3 in activating GMDCs, as the production of IL-6 and TNFα were comparable by GMDCs upon stimulation by SUP3 and Pam3 ( Figure 1D). Taken together, these data suggested that SUP3, the newly designed TLR agonist, engaged TLR2 specifically as well as Pam3.

sUP3 induced More Vigorous cTl responses Than Pam3 without excessive Production of inflammatory cytokines In Vivo
The consequence of immunization or vaccination is the acquirement of preventive or therapeutic effects on infections or tumors, via generating antigen-specific adaptive immune responses and memory. Elimination of tumors or intracellular infection requires antigen-specific CD8 + CTLs, the major executors in cellular immunity (46,47). In mice, the differentiation and activation of CD8 + CTL is launched through antigen cross-presentation  by CD8 + cDC (46,47). Some TLR agonists have been shown to enhance antigen cross-presentation by cDCs and the subsequent CD8 + T cell activation (30,35,(48)(49)(50)(51). To investigate whether SUP3 has an effect on CTL responses, we determined the activity of SUP3 in antigen-specific cellular immunity. Mice were immunized by intravenous injection of OVA coated splenocytes (OCS), as these cell-associated OVA antigens can be cross-presented efficiently by CD8 + cDC to CD8 + T cells (52). Seven days postimmunization, the percentages of OVA antigen-specific CD8 + T cells in spleens were determined by pentamer (SIINFEKL/ H-2Kb) staining. The effects of various doses of SUP3 and Pam3 on the induction of CTL were determined and 2 nmol was selected as the optimal dose in the following assays ( Figures S4A,B in Supplementary Material). Administration of SUP3 at a high dose of 10 nmol resulted in decreased CTL induction and enlarged spleens, whereas treatment with Pam3 at the same dose did not have the same effect (data now shown), indicating that SUP3 was much more potent than Pam3 and the mice could not tolerate such a high dose of agonists. We found that OCS formulated with SUP3 induced significantly more antigen-specific CD8 + T cells than that with Pam3 in vivo (Figures 2A,B). To further determine whether the enhanced CTL activation by SUP3 can lead to more vigorous antitumor activity, we immunized mice as described above and then challenged by intravenous inoculation of OVAexpressing B16 melanoma cells (B16-OVA) (53). We found that mice immunized with OCS and SUP3 showed markedly reduced number of tumors in the lungs compared to those with OCS alone or OCS with Pam3 16 days after the challenge (Figures 2C,D). It indicated that immunization with SUP3 conferred more efficient CTL responses, thus more effective protection against tumor formation. Taken together, these results demonstrated that SUP3 was far more effective than Pam3 in inducing CTL responses in vivo. Sustainable stimulation by TLR agonists such as LPS, often lead to elevated inflammatory cytokines in vivo, which may result in severe adverse reactions. To test whether this is also the case for   SUP3 and Pam3, we determined the level of serum inflammatory cytokines upon SUP3 or Pam3 administration. The inflammatory cytokines including IL-6, TNFα, and IL-12p40 and chemokine MIP1α were all maintained at low levels even after SUP3 or Pam3 stimulation (Figure 2E), suggesting that both SUP3 and Pam3 acted as mild TLR agonists for inflammatory responses. Overall, SUP3 promoted stronger CTL response without inducing excessive inflammatory cytokine production, suggesting that SUP3 could function as a better adjuvant than Pam3 with potential for clinical applications.
sUP3 enhanced cross-Presentation of OVa antigen by cD8 + cDcs In Vitro Cytotoxic T lymphocyte is one of the major effector cell types in cellular immunity, which can be induced by CD8 + cDCs via antigen cross-presentation to naïve CD8 + T cells (11,13,14,47). It is well known that TLR activation can enhance the function of DCs (54). It is likely that SUP3 promoted stronger CTL responses through enhancing the function of CD8 + cDCs. To investigate whether SUP3 played any roles in antigen cross-presentation by splenic CD8 + cDCs and the subsequent activation of naïve CD8 + T cells, we examined the antigen cross-presentation capacity of CD8 + cDCs with or without SUP3 by using OVA protein antigen and the OVA-specific T cell receptor (TCR) transgenic OT-I CD8 + T cells. The CD8 + cDCs pulsed with OVA peptide SIINFEKL specifically recognized by OT-I T cells were used as positive control. SIINFEKL peptide pulsed CD8 + cDCs presented antigen peptide directly without antigen processing and induced strong OT-I T cell proliferation ( Figure 3A). Upon stimulation of OVA protein pulsed CD8 + cDCs by TLR2 agonists, the proliferation of cocultured OT-I cells increased significantly, evidenced   by increased numbers of divided OT-I T cells (Figures 3A,B) and enhanced production of IFN-γ ( Figure 3C) compared to that without stimulation by TLR2 agonists. These results indicated that SUP3 and Pam3 were comparable in enhancing cross-presentation by CD8 + cDCs in vitro, thus leading to enhanced antigen-specific CD8 + T cell activation. It is interesting to note that unlike the results described above, in which SUP3 induced much stronger antigen-specific CD8 + T cell proliferation and antitumor CTL responses in vivo, SUP3 and Pam3 induced similar levels of OVA-specific CD8 + T cell activation in vitro.
This difference is likely due to the improved stability of SUP3 compared to Pam3 in vivo. Exogenous antigens are internalized, degraded into short peptides, uploaded onto MHC-I and finally transported to cell surface for recognition by cognate T cells (55)(56)(57). The first step of antigen presentation is internalization of extracellular materials. To clarify whether antigen uptake process is affected by TLR2 engagement, we determined the endocytosis of fluorescenceconjugated soluble OVA (OVA-Alexa Fluor 488, OVA-AF488) by CD8 + cDCs. The results showed that the uptake of OVA by CD8 + Function of Novel TLR2 Agonist Frontiers in Immunology | www.frontiersin.org February 2017 | Volume 8 | Article 158 cDCs was not influenced by TLR2 activation (Figures 3D,E), indicating that enhanced antigen presentation by CD8 + cDCs was not due to facilitated antigen uptake process.
sUP3 Upregulated expression of costimulatory Molecules on cD8 + cDcs and induced low level Production of inflammatory cytokines Costimulatory molecules, which serve as the second signal for T cell activation, can be efficiently upregulated upon TLR activation. To investigate whether SUP3 affected the expression of costimulatory molecules on CD8 + cDCs, we stimulated purified DCs by SUP3 or Pam3 in vitro and stained with antibodies to costimulatory molecules as well as MHC molecules. We found that both SUP3 and Pam3 induced substantial upregulation of CD40, CD80, CD86, and MHC-II on CD8 + cDCs, but not MHC-I (Figure 4A), demonstrating that SUP3 induced effective costimulation as well as Pam3 in vitro. We then tested whether SUP3 had similar effects on splenic CD8 + cDCs in vivo.
Upon in vivo administration of SUP3 or Pam3 via intravenous (i.v.) route, the expressions of CD40 and CD86 increased on splenic CD8 + cDCs from both SUP3-or Pam3-treated mice, indicating the ability of SUP3 to promote DC activation in vivo ( Figure 4B). Toll-like receptor stimulation is required for functional maturation of DCs, which is essential for T cell activation and initiation of adaptive immunity. Antigen recognition without danger signals like TLR ligands leads to tolerance but not antigen processing for subsequent presentation (17,58). It is well established that TLR-induced activation of nuclear factorkappa B (NF-κB) and mitogen-activated protein kinase (MAPK) signaling events contribute to activation of responding cells. To determine whether SUP3 could activate classical Toll-dependent signaling pathways in CD8 + cDCs, we treated purified CD8 + DCs by TLR2 agonists at various time points (0-120 min) and detected the expression of the relevant signaling molecules and their phosphorylated forms by immunoblotting. We found SUP3 stimulation induced phosphorylation of p65 and degradation of IκBα and enhanced phosphorylation of ERK, JNK, and p38 in CD8 + cDCs (Figure 4C), indicating that similar to Pam3, SUP3 was able to induce activation of NF-κB and MAPK pathways. Taken together, SUP3 could induce activation of canonical TLR signaling events in CD8 + cDCs.
In addition to costimulation, secretion of cytokines by DCs is another major feature of DC activation upon TLR engagement. Cytokines like bioactive IL-12p70 predominately produced by CD8 + cDCs drive functional differentiation of activated T cells and provide the third signal for DC-induced T cell activation (59). As SUP3 could facilitate CD8 + T cell activation (Figures 3A-C), we therefore examined whether SUP3 could induce inflammatory cytokines production by CD8 + cDCs. We activated purified CD8 + cDCs by SUP3, Pam3, or CpG in vitro and quantified the amounts of cytokines in the supernatant by ELISA. Interestingly, in contrast to CpG stimulation which induced large amount of IL-12p70, IL-6, and TNFα production by CD8 + cDCs, SUP3 and Pam3 induced no IL-12p70 and only low levels of IL-6 and TNFα ( Figure 4D). These results demonstrated that SUP3 acted as a mild agonist in inflammatory responses. Collectively, our results demonstrated that SUP3 acted as an efficient enhancer of antigen cross-presentation by CD8 + cDCs through upregulating surface costimulatory molecules and promoting activation of NF-κB and MAPK signaling pathways, without inducing large amount of inflammatory cytokines. Thus, SUP3 fulfilled the properties of a good adjuvant with strong immunostimulatory capacity for T cell activation and minimal induction of inflammatory response.

sUP3 enhanced cD4 + T cell activation and Production of antibodies against TD antigens
In addition to cellular immunity, the other arm of adaptive immunity is humoral immunity mediated by antibodies. Production of TD antigen-specific antibodies requires the assistance of Th2 subtypes, which is mainly induced by CD11b + cDCs (15,60,61). Exogenous antigens are routinely presented via MHC-II pathway by CD11b + cDCs and ultimately recognized by CD4 + T cells for helper T cell differentiation (11). CD11b + cDCs expressed comparable levels of TLR2 to that of CD8 + cDCs ( Figure S5A in Supplementary Material). We predicted that SUP3 might enhance the functions of CD11b + cDCs thus Th2mediated production of TD antigen-specific antibodies. To test this hypothesis, we immunized mice via intraperitoneal (i.p.) injection of TD antigen NP-KLH with or without TLR agonists. Formulation with alum was regarded as positive control in this setting, as alum is always used as an efficient adjuvant in antibody induction. Serum antibody titers were determined every week post-immunization. As expected, SUP3 induced higher titers of TD antigen-specific antibodies including IgM, total IgG and high affinity IgG than Pam3 did ( Figure 5A). This result suggested that SUP3 remained functioning for a longer time than Pam3 in vivo, thus resulting in higher levels of antibody production.
As mentioned above, CD4 + Th2 cells polarized by CD11b + cDCs are critical in TD antigen-specific antibodies production. Thus, SUP3 might enhance TD antibody production via facilitating CD4 + T cell activation in vivo. To test this hypothesis, purified CD4 + OT-II T cells from spleens of OVA-specific TCR transgenic mice were labeled by CFSE and transferred into WT mice for 1 day, followed by subcutaneous (s.c.) immunization with soluble OVA proteins with TLR agonists for another 3 days. Treatment with SUP3 induced more proliferation of OT-II CD4 + T cells than that with Pam3 in draining lymph nodes (Figures 5B,C), demonstrating the ability of SUP3 to induce stronger CD4 + T cell activation in vivo, thus leading to enhanced TD antibodies production.

sUP3 Facilitated antigen Presentation by cD11b + cDcs In Vitro
As CD4 + T cell activation is enhanced upon SUP3 treatment, we speculated that SUP3 functioned through improving functions   of CD11b + cDCs. To test this speculation, we evaluated OVA antigen presentation by SUP3 pretreated CD11b + cDCs to OT-II CD4 + T cells. An enhanced OT-II CD4 + T cell proliferation was observed upon stimulation of CD11b + cDCs by SUP3 in vitro (Figures 5D,E), indicating that SUP3 strengthened antigen presentation ability of CD11b + cDCs. The maturation of CD11b + cDC is marked by upregulation of costimulatory molecules. Intravenous administration of SUP3 and Pam3, both induced maturation of splenic CD11b + cDC indicated by CD86 upregulation ( Figure 5F). Furthermore, similar to CD8 + cDCs, CD11b + cDCs produced lower level of IL-6, TNFα, and MIP1α upon stimulation by SUP3 and Pam3, implying a weak ability of SUP3 in inducing inflammatory responses ( Figure 5G).
Collectively, these results demonstrated that SUP3 facilitated antigen presentation by CD11b + cDCs and subsequent activation of OT-II CD4 + T cells, which led to higher levels of TD antibody production in vivo. These observations again demonstrated the physiological advantages of SUP3 over Pam3 as an immune adjuvant.

sUP3 enhanced antigen Deposition in aPcs in Draining lymph nodes
As immune adjuvants enhance immunogenicity of simultaneously administrated antigens via increasing antigen deposition in lymph nodes, and prolonging antigen presentation (62,63), we proposed that SUP3 might prolong the duration of antigen deposition in draining lymph nodes. To test this, we s.c. immunized mice with OVA-AF647 at groin, or co-administrated with either SUP3 or Pam3 for 24 h. Antigen deposition in various immune cell populations from inguinal lymph nodes was then assessed. After immunization of OVA with SUP3, increased numbers of antigen-containing CD8 + cDCs, CD11b + cDCs, macrophages, and monocytes were observed compared to those with OVA alone or OVA with Pam3 (Figures 6A,B). Moreover, the total number of cells containing antigens increased more upon administration of SUP3 over Pam3 ( Figure 6B). This finding indicated that SUP3 facilitated antigen deposition in cDCs, macrophages and monocytes in draining lymph nodes, thus contributed to enhanced T cell activation.
sUP3 Directly Promoted stronger B cell activation In Vitro B lymphocytes can function as APCs that directly respond to invading pathogens through recognition via innate sensors. In addition to myeloid cells including DCs, macrophages, and monocytes, B cells are only lymphocytes positive for TLR2 ( Figure S5A in Supplementary Material). TLR ligation on B cells drives fast IgM production to eliminate pathogens before the formation of effective antibody-mediated adaptive immunity, which requires 5-7 days to develop (64). The direct TLR triggered IgM production by B cells, as a thymus independent (TI) pathway, compensates the delayed formation of humoral immunity to control pathogens. To evaluate B cell response to SUP3, we analyzed the expression of activation markers and production of IgM upon treatment by SUP3 in vitro. Engagement of TLR2 by SUP3 and Pam3 induced marked upregulation of CD25, CD40, and CD86 on B lymphocytes (Figure 7A), implying a direct effect of SUP3 on B cell activation. In addition, more antigen-independent production of IgM was induced by SUP3 than that induced by Pam3 (Figure 7B), indicating that SUP3 could induce a stronger and longer lasting B cell response than Pam3. This finding is in line with the better stability of SUP3 compared to Pam3. Our results demonstrated that TLR2 agonists could enhance B cell function not only in T cell-dependent but also in T-independent manners.

DiscUssiOn
Our newly designed TLR2 agonist SUP3 is structurally similar to classical TLR2 ligand Pam3. The chemical modification by replacing oxygen with nitrogen in Pam3 stabilizes the covalent bonds between glycerol backbone and palmitic acid chains. In addition, glycine instead of lysine reduces the chiral center and further simplifies the structure. These improvements overcome current disadvantages of Pam3, such as weak stability and structural complexity. Meanwhile, the specificity and affinity to TLR2 and bioactivity to promote TLR2-dependent cytokines are retained. Moreover, the peptide components of SUP3 make it easy and convenient to be conjugated with antigenic peptides. This type of conjugation could easily confer adjuvant activity to the conjugated antigens and induce markedly enhanced immune responses toward antigens without additional adjuvant formulations (6,25,65,66). Intriguingly, this special feature is not shared with any other TLR agonists except those of TLR2. Adjuvants of vaccines are defined as components capable of enhancing and shaping antigen-specific immune responses (1). Ideal adjuvants should possess properties including safety and effectiveness with defined and well-controlled functions (1). An effective adjuvant should help to reduce required dose of antigens or the number of immunizations and have significant improvement in antibody-as well as cell-mediated immune responses (1). Although both SUP3 and Pam3 possess these features of an ideal adjuvant, SUP3 is chemically more stable and structurally simpler than Pam3. SUP3 also exhibited clear advantages compared to Pam3 in biological activities, suggesting that SUP3 could be a good substitution of Pam3 in vaccinations and cancer treatment.
Recognition of microbial or viral components by PRRs on DCs elicits immediate host response and ultimately long-lasting adaptive immunity (17). DCs are one of the first encounters for non-self components, such as pathogens and tumor antigens. The initial responses of DCs to these pathogens determine the shape of adaptive immunity. Thus DCs occupy an indispensable position in orchestrating immune responses to eliminate pathogens and tumors. Consequently, immune modulations targeting DCs could be an effective strategy to harness immune system to endow protective immunity. Cancer immunotherapy with DCs is favorable due to fewer side effects, although with limited benefit observed from recent studies (67).
Enhancement of adaptive immunity is a key parameter of adjuvants in the context of vaccines. Shaping adaptive immunity  always requires several days of time. In this study, we observed that more antigen-specific CTLs were generated in the presence of SUP3 than that with Pam3 during the induction of effector T cell differentiation by OCS in vivo. Consistent with this observation, the presence of SUP3 induced much stronger protective response against tumor challenge, as revealed by significantly reduced tumor formation in the lungs. Those effects of SUP3 could be achieved through enhancing antigen cross-presentation by CD8 + cDCs, which are responsible for CTL induction and subsequent clearance of tumor cells. SUP3 significantly enhanced the surface expression of costimulatory molecules on cDCs, providing one of the essential signals for T cell activation. Intriguingly, SUP3 enhanced antigen cross-presentation by CD8 + cDCs was accompanied by a low to moderate levels of inflammatory cytokines, which is different from effects of many other TLR agonists used as adjuvant such as LPS. This functional feature of SUP3 can be considered as an advantage for a good adjuvant, as over production of inflammatory cytokines could be detrimental. Altogether, SUP3 can serve as a potential adjuvant in therapeutic or prophylactic vaccines.
Antigen-specific antibody-mediated humoral immunity is the other arm of adaptive immunity. Indeed, SUP3 not only enhances CTL-mediated cellular immunity but also facilitates the production of various types of specific antibodies to TD antigens. As mentioned previously, TD antibody production requires effector CD4 + T cells activated through MHC-II antigen presentation pathway mainly promoted by CD11b + cDCs. Addition of SUP3 enhanced antigen presentation ability of CD11b + cDCs in in vitro assays and administration of SUP3 enhanced CD4 + T cell activation in vivo. Thus, SUP3 induced more antibody production than Pam3 in T-dependent manner through enhancing antigen presentation by CD11b + cDCs. Moreover, an enhanced TI antibody production by B cells was also observed after SUP3 administration. B cells are capable of recognizing and responding to pathogens directly via innate immune receptors like TLR. TLR activation of B cells could ultimately result in IgM secretion, which rapidly neutralizes pathogens, namely the TI antibody response (68). In this study, SUP3 directly promoted IgM production by B cells, suggesting its effects on both innate and adaptive immunity. Overall, SUP3 enhanced cellular immunity as well as TD and TI antibody production. This functional property of SUP3 made it a promising ideal adjuvant in immunotherapy and vaccination.
The more vigorous immune stimulatory activities of SUP3 compared to Pam3 in CTL induction, rejection of B16-OVA tumors, CD4 + T cell activation, and the production of TD antibodies could all attribute to its better stability and sustainable activity over Pam3 in vivo. Thus, SUP3 could act as an effective adjuvant in immunization.
An ideal adjuvant should have minimum or even none adverse effects. Generally, TLR agonists are potent immune stimuli capable of inducing large amount of inflammatory cytokines, which may lead to cytokine storm to damage the hosts before inducing protective immune responses. Our study and previous studies showed that engagement of TLR2 by triacyl lipopeptides did not promote severe inflammatory responses, which would be beneficial for its clinical applications (65,69,70). Ligation of TLR2 also activated PI3K-Akt-β-catenin axis to induce IL-10 expression, which is a regulatory cytokine to suppress inflammation (70). TLR2 agonists cloud also induce rapid degradation of IRAK1, which is an essential component of inflammatory response (71). These effects might contribute to the weaker inflammatory properties of TLR2 agonists. In addition, TLR2 ligation signal via adaptor MyD88 to promote downstream events, whereas LPS recognition by TLR4 can activate both MyD88 and TRIF adaptors mediated pathways (72,73). Thus, LPS could induce stronger inflammatory responses via two pathways than that induced by TLR2 agonists. Meanwhile, this immune regulatory effect did not attenuate T cell activation by DCs (74). Again, SUP3 could be considered as a novel effective immune adjuvant.
Adjuvants are not only applied in vaccination but also in cancer immunotherapy. CTLs induced by DCs are major knights to combat tumors. Enhanced DC activation will lead to vigorous antitumor cytotoxicity by CTLs. Our studies proved that SUP3 conferred more protective immunity from tumor metastasis to the lungs. Tumors escape from immune attack by diminishing CTL responses. PD-1/PD-L1 is important check-point molecules during immune responses, which suppress antitumor immune responses. Blocking and eliminating their inhibitory effects by neutralizing antibodies significantly enhanced antitumor immune responses and had been used in clinical trials for cancer therapies (75)(76)(77). However, this treatment is not always effective for some patients due to various reasons. A combination of an effective immune adjuvant together with anti-PD-1/PD-L1 antibodies may improve the efficacy of the treatment. Our synthetic TLR2 agonist SUP3 could be an outstanding candidate for this purpose. Further tests are needed to determine the effectiveness of this strategy.
Overall, in this study, we developed a novel TLR2 agonist SUP3, which is more stable than Pam3 and could enhance DC functions without inducing excessive inflammation, the most favorable property of an ideal adjuvant. As DCs are essential initiators of antigen-specific immune responses, the enhanced activation and function of DCs by TLR2 agonists should improve immune responses to pathogens or tumors. Our study suggested that the new TLR2 agonist SUP3 could be a promising adjuvant in vaccination and immune modulations.

aUThOr cOnTriBUTiOns
The study was designed by XG, NW, YS, JH, XBL, and LW. XG, XL (4th author), TW, YZ, XL (7th author) and JH preformed experiments and acquired the data. XG, NW, YS, XL (7th author), JH, XBL, and LW analyzed and interpreted the data. Manuscript was written by XG, NW, YS, XBL, and LW. This study was supervised by LW and XBL.