The synthetic TLR2 agonist BPPcysMPEG leads to efficient cross-priming against co-administered and linked antigens

Authors


Abstract

The property of DC to generate effective CTL responses is influenced by TLR signaling. TLR ligands contain molecular signatures associated with pathogens, have an impact on the antigen processing and presentation by DC, and are being exploited as potential adjuvants. We hypothesized that the TLR2/6 heterodimer agonist S-[2,3-bispalmitoyiloxy-(2R)-propyl]-R-cysteinyl-amido-monomethoxyl polyethylene glycol (BPP), a synthetic derivative of the Mycoplasma macrophage activating lipopeptide-2, is a potent adjuvant for cross-priming against cellular antigens. Systemic administration of BPP-induced maturation of CD8α+ DC and CD8α DC in the spleen and resulted in enhanced cross-presentation of intravenously co-administered antigen in mice. In addition, administration of BPP and cell-associated OVA generated an effective CTL response against OVA in vivo in a CD4+ T helper cell-dependent manner, but independent of IFN-α. Delivering antigenic peptides directly linked to BPP led to superior CTL immunity as compared to giving antigens and adjuvants admixed. In contrast to other TLR ligands, such as CpG, systemic activation of DC with BPP did not result in shut-down of antigen presentation by splenic DC subsets, although cross-priming against subsequently encountered antigens was reduced. Together, our data provide evidence that BPP is a potent stimulus to generate CTL via cross-priming.

Introduction

DC are considered professional antigen presenting cells, which bridge the gap between innate and adaptive immunity 1, 2. DC capture antigen from the sites of infection, migrate to draining secondary lymphoid organs and present antigens to naïve T cells, thereby priming adaptive immune responses 3, 4. Exogenous antigens are presented in the context of MHC class II molecules; endogenous antigens follow the proteasomal degradation pathway and are presented on MHC class I molecules to CD8+ T cells 5. Some DC have the ability to divert exogenously acquired antigens to the MHC class I processing pathway, a phenomenon termed as cross-presentation. Differential expression of the cell surface markers CD4 and CD8 on spleen DC allows to classify them as CD8+, CD4+ and CD8CD4 DC 6, 7. Among the different splenic DC subsets, only the CD8α+ DC constitutively cross-present 8–11. Different DC subsets may therefore influence the intensity and nature of the immune response. In tissues, DC continuously samples their microenvironment and encounter various microbial products, which are then recognized via pattern recognition receptors. The most well characterized among these are the TLR.

TLR ligands act as potential adjuvants that control DC maturation and influence the magnitude of T-cell responses 12–14. Thus far, 13 mammalian TLR have been identified. Mostly, TLR ligands are non-peptidic components of pathogens ranging from components of bacterial cell walls to viral and bacterial DNA or RNA. It has been demonstrated that certain lipopeptides of bacteria are recognized by TLR2 and can potentially induce immune activation 15. In this regard, a 2-kDa lipopeptide obtained from the cell membrane of Mycoplasma fermentas was shown to stimulate macrophages and induce pro-inflammatory chemokines and cytokines 16, 17. The macrophage activating lipopeptide-2 (MALP-2) triggers global immune activation on mucosal associated lymphoid tissues and acts as potential mucosal adjuvant 18. Enhanced B-cell and T-cell responses are induced against the antigens co-administered with MALP-2 19. MALP-2 is recognized by TLR2/6 heterodimers and signals via a MyD88-dependent signaling pathway 20. Almost all known DC subpopulations express TLR2 on their surface 21. In vitro treatment of DC with MALP-2 promotes their activation and maturation, as well as a shift of their protein pattern from proteasomes to immunoproteasomes, which correlates with an increased proteolytic activity and improved antigen presentation capacities 22. In addition, studies using synthetic derivatives of MALP-2 have demonstrated their potential adjuvanticity in triggering humoral responses and cellular immune responses on mucosal surfaces. S-[2,3-bispalmitoyiloxy-(2R)-propyl]-R-cysteinyl-amido-monomethoxyl polyethylene glycol (BPPcysMPEG) is a pegylated synthetic derivative of macrophage-activating lipopeptide-2, which exhibits improved pharmacological properties (e.g. solubility) and retains its agonistic capacities for the TLR 2/6 heterodimer. Given its chemical structure, BPPcysMPEG, to which we here refer as BPP, can be linked to antigens, thereby becoming a potential candidate for simultaneous targeting and activation of DC subpopulations. We aimed to understand the adjuvant effects of BPP on the cross-presentation capacity of DC. We hypothesized BPP to be a potent adjuvant for cross-priming against cellular antigens. In the present study, we show that systemic administration of BPP in mice generates a CD4+ T helper cell-dependent effective cytotoxic immune response against exogenous antigens. We found that in contrast to CpG, BPP exhibits unique features in terms of regulating cross-presentation and cross-priming by DC.

Results

BPP-induced maturation of BM-derived and splenic DC subsets

To test whether BM-derived DC (BMDC) respond to stimulation with BPP, we generated CD11c+CD24+ and CD11c+CD172a+ BMDC 23, 24 from C57BL/6 (B6) and TLR2−/− mice in cultures supplemented with Flt3-L and measured upregulation of CD86 surface expression. Both CD11c+CD24+ as well CD11c+CD172a+ BMDC increased their CD86 expression after stimulation with BPP, LPS or CpG, whereas the response to BPP was explicitly absent in BMDC from TLR2−/− mice (Fig. 1A).

Figure 1.

Maturation of BMDC and splenic DC subsets after TLR stimulation. (A) BMDC of B6 (WT) or TLR2−/− mice from Flt3-L cultures were treated with BPP (100 ng/mL), LPS (1 μg/mL) or CpG (10 nmol/L) and the surface expression of CD86 assessed by flow cytometry. The panels on the left show the gating used. (B) The surface expression of MHC II, CD80 and CD86 was compared on splenic CD8α+ and CD8α DC subpopulations isolated from mice (n=3 per group, pooled analysis), which were left untreated (grey histogram) or treated with 1 μg BPP (black line) or 1 μg LPS (dashed line) 12 h previously. (C) Surface expression profile of CD80 and CD86 on CD11+ DC of B6 (WT), MyD88−/− and TLR2−/−mice (n=3 per group, pooled analysis) untreated (grey histogram) or treated with 1 μg BPP (black line) or 25 nmol CpG (dotted line, only WT). (D) GM-CSF-cultured BMDC of B6 (WT) or TLR2−/− mice were treated with 100 ng/mL BPP, 1 μg/mL LPS or 10 nmol/L CpG and cytokine secretion determined using a cytokine bead array. n.d., not detectable. The results are representative data obtained in one (A+C), two (B) or three (D) independent experiments.

Macrophages use CD36, which is also expressed on CD8α+ DC, for efficient MALP-2 binding and signaling 25, 26. To test whether BPP might have a more profound impact on the spleen-derived CD8α+ DC, we injected mice with BPP or LPS i.v., isolated CD8α+ DC and CD8α DC from the spleen 12 h later and measured surface expression of MHC class II (MHC II), CD80 and CD86. This revealed extensive upregulation of all these molecules on both the CD8α+ DC as well as CD8α DC (Fig. 1B) with similar kinetics, suggesting that all the splenic DC subpopulations responded to BPP in vivo. This response was TLR2 and MyD88 dependent, respectively, as the splenic CD11c+ DC of the TLR2−/− and MyD88−/− mice did not show CD86 upregulation (Fig. 1C). To confirm the DC activation, B6 and TLR2−/− BMDC from GM-CSF cultures were treated with BPP, LPS or CpG and the secretion of inflammatory cytokines such as IL-6 and TNF-α was assessed. BPP, LPS and CpG-induced secretion of both cytokines in B6 BMDC and this response was again specifically abrogated in TLR2−/− BMDC after BPP treatment (Fig. 1D). Taken together, BPP leads to increased expression of co-stimulatory molecules and secretion of inflammatory cytokine by BMDC and spleen DC populations in a TLR2-dependent manner.

Enhanced antigen presentation upon co-administration of antigen and BPP

TLR ligands serve as potential adjuvants providing additional co-stimulatory signals for T-cell priming. First, we assessed the potential of BPP to enhance T-cell stimulation mediated by spleen-derived CD11c+ DC. We used soluble OVA, which allowed us to assess whether antigen presentation via MHC I and II resulted in increased proliferation of OT-I and OT-II T cells. Enriched (75% pure) CD11c+ DC pulsed with soluble low-endotoxin OVA (EndoGrade™ Ovalbumin) + BPP induced more OT-I and OT-II cell proliferation as compared to DC, which where pulsed with OVA only (Fig. 2A). The absence of this increase in DC lacking TLR2 again confirmed the specificity of these BPP-mediated effects. In order to determine the effects of BPP on antigen presentation in vivo, we co-administered BPP (1 μg/mouse) and 3 mg OVA (Sigma, grade VI) into B6 mice i.v. and purified splenic CD8α+ DC and CD8α DC 16 h later by cell sorting. These DC were then used to stimulate CFSE-labeled OT-I cells. As expected, presentation of soluble OVA to OT-I cells was restricted to CD8α+ DC subset. Although small, there was a statistically significant enhancement in the cross-presentation of OVA by CD8α+ DC in mice that had received BPP (Fig. 2B). In contrast, only spleen-derived CD8α DC significantly increased T-cell stimulation via MHC II following BPP and antigen co-administration in vivo (Fig. 2B). To confirm the specificity of BPP, we injected mice with 3 mg low-endotoxin OVA with or without BPP, sorted the CD8α+ DC and CD8α DC and used them to stimulate OT-I and OT-II cells, respectively. This again resulted in increased OT-I and OT-II cell proliferation and this effect was not detectable in TLR2−/− mice (Fig. 2C). To further confirm enhanced cross-presentation in vivo, we adoptively transferred CFSE-labeled OT-I along with soluble OVA with or without BPP i.v. and determined OT-I proliferation in lymph nodes and spleen after 55 h. As expected, co-administration of BPP and OVA resulted in increased expansion of adoptively transferred OT-I cells in secondary lymphoid organs (Fig. 3). We conclude that BPP leads to a TLR2-dependent antigen-specific increase in CD4+ and CD8+ T-cell stimulation against co-administered soluble antigens.

Figure 2.

Effects of BPP on presentation of soluble antigens by spleen DC subsets. (A) 1×104 enriched CD11c+ DC (75% pure) per well from B6 or TLR2−/− mice were pulsed with 100 μg/mL special low-endotoxin OVA with or without 5 nmol/L CpG or 100 ng/mL BPP and used to stimulate 5×104 OT-I (left panel) or OT-II cells (right panel) in vitro. The number of proliferating OT-I and OT-II cells (CFSElow) were determined by flow cytometry after 65 h of co-culture. (B) Sorted splenic CD8α+ and CD8α DC subsets (1×104) from B6 mice systemically challenged with 3 mg of soluble OVA in the presence or absence of BPP (1 μg/mouse) were cultured with 5×104 CFSE-labeled OT-I (left panel) or OT-II cells (right panel) and T-cell proliferation assessed as in (A). The lower panels show representative CFSE-proliferation profiles. (C) Sorted splenic DC subsets (1×104) from B6 or TLR2−/− mice systemically challenged with 3 mg of special low-endotoxin soluble OVA in the presence or absence of BPP (1 μg/mouse) were cultured with 5×104 CFSE-labeled OT-I (left panel) or OT-II cells (right panel) and T-cell proliferation measured as in (A). Data are the mean ± SEM of triplicate measurements for each condition (*p<0.05, **p<0.01). Data shown were obtained in two (A+C) or representative for four (B) independent experiments.

Figure 3.

Clonal expansion of adoptively transferred OT-I cells in secondary lymphoid organs. 1–3×106 CFSE-labeled OT-I cells were adoptively transferred into groups of mice (n=3 mice per group) together with 3 mg soluble OVA in the presence or absence of 1 μg/mouse BPP. The control group (PBS) received only OT-I cells. The numbers of proliferated OT-I cells (CFSElow) were determined by flow cytometry (representative plots are shown) in the inguinal lymph node and spleen of these mice after 50 h. Data are represented as the mean ± SEM obtained in three independent experiments (**p<0.01).

BPP-mediated activation of DC induces potent CTL responses in vivo

We next assessed the efficiency of BPP for effective cross-priming. We adoptively transferred purified OT-I cells and immunized recipient mice with OVA-coated splenocytes (OCS) generated from the spleen of bm1 mice together with 1 μg/mouse BPP the next day. The cell-associated antigen was used for two reasons: it reduces the amount of antigen required to activate CD8+ T cells, considering that cell-associated OVA is presented 50 000-fold more efficiently to CD8+ T cells than soluble OVA 27, which would minimize potential effects from endotoxin contamination. Second, use of OCS would enable us to study the impact of BPP on CTL priming, considering that OCS + adjuvant are able to induce efficient CTL responses via cross-priming. Seventy-two hours after the immunization, splenocytes were re-stimulated with OVA257–264ex vivo and the frequency of IFN-γ-producing CD8+ T cells was estimated by intracellular cytokine staining. Increased numbers of IFN-γ+ OT-I cells were observed in mice primed with OCS + BPP as compared to mice that had received OCS only (Fig. 4A). The latter presented a significant background response, which we consider to result from a lower stimulatory threshold of the adoptively transferred TCR-transgenic OT-I cells for gaining effector cell function as compared to endogenous T cells. Again, the BPP response was absent in TLR2−/− mice primed with OCS.

Figure 4.

Cross-priming in the presence of BPP leads to generation of helper cell-dependent effector CTL. (A) 1×106 OT-I cells were adoptively transferred into B6 or TLR2−/− mice (n=3 per mice group) and 1 day later, the mice were immunized i.v. with OCS with or without BPP. Three days after immunization, the percentage of IFN-γ-producing OT-I cells in the spleen of the recipient was assessed after in vitro re-stimulation of splenocytes with OVA257–264. The dot plots represent the percentage of IFN-γ+ events among those gated on CD8+Vα2+ OT-I cells. Results are representative of two independent experiments. (B) B6 or TLR2−/− mice (n=3 or 4 mice per group) were immunized with 2×107 OCS with or without BPP and cytotoxicity assessed in vivo. (C) To study the requirement of CD4+ T-cell help, B6 and MHC class II−/− mice (n=3–4 mice per group) were immunized with OCS only, 2×107 OCS + 1 μg BPP or 2×107 OCS + 25 nmol CpG and CTL-mediated killing in the spleen assessed in vivo on day 5. (D) The role of IFN-α for CTL-priming and cytotoxic killing was studied by immunizing B6 and IFN-αR−/− mice (n=3 or 4 mice per group) as described in (C). The bars represent the mean ± SEM across two (B) or three (C+D) independent experiments; (A) is representative for two independent experiments (*p<0.05).

To further proof the generation of CTL by BPP from an endogenous T-cell repertoire, OCS were injected with or without BPP i.v. into naïve B6 and TLR2−/− mice. Five days later, generation of CTL was quantified in an in vivo killing assay. The results confirmed that efficient cross-priming of antigen-specific CTL from the endogenous T-cell repertoire in the presence of BPP and that this response was mediated via TLR2 (Fig. 4B). Taken together, endogenous CD8+ T cells primed by BPP-activated DC become cytotoxic T cells and are efficient in eliminating target cells.

BPP-induced cross-priming of CD8+ T cells by DC is T helper cell dependent

Cross-priming of CD8+ T cells by DC is mostly dependent on the provision of help by CD4 T cells 28–30, although some studies using TLR agonists showed that this requirement may not be absolute 31, 32. To test if the impact of BPP on generating CTL response was dependent on CD4 T-cell help, we performed in vivo cytotoxicity assays in MHC II−/− mice, which had been previously immunized with OCS with or without BPP or CpG. CTL immunity was markedly reduced in MHC II−/− mice, even in the presence of BPP, suggesting that BPP mediated cross-priming is to a large extent dependent on CD4+ T-cell help (Fig. 4C). On the other hand, CpG, which was previously shown to induce helper-independent priming of CTL 32, induced readily detectable, though slightly reduced, CTL responses in MHC II−/− mice, suggesting that MHC II−/− mice have no major intrinsic impairment in developing effector CTL-responses.

IFN-α is not crucial for BPP-mediated cross-priming

Type I interferons are major inflammatory mediators important for driving CTL responses against viral pathogens and antigens delivered together with TLR agonist 33. Durand and colleagues have shown impaired CTL responses in IFN-α/β-deficient mice, when LPS, poly (I:C) or CpG was used as adjuvant 34, 35. To test the role of IFN-α for BPP-mediated CTL priming, IFN-αR−/− mice and B6 mice were injected i.v. with OCS with or without BPP or CpG and priming was assessed by an in vivo cytotoxicity assay. Interestingly, CTL responses observed in the B6 and IFN-αR−/− mice were comparable and there was no difference whether BPP or CpG was used as an adjuvant (Fig. 4D). These data suggest that IFN-α or IFN-α signaling is not a crucial factor for BPP-mediated cross-priming under experimental conditions testes here.

Pre-activation of DC with BPP does not impair cross-presentation of subsequently encountered antigens

Maturation of DC with TLR agonists prior to antigen challenge can lead to inhibition of cross-presentation and classical MHC II presentation 36, 37. We assessed the effects of BPP pre-treatment on antigen presentation and T-cell stimulation against subsequently administered antigens. For this, CD8α+ DC and CD8α DC isolated from BPP pre-treated B6 mice were enriched by FACS sorting and co-incubated with irradiated OCS or pulsed for 45 min with soluble OVA (to asses MHC II presentation) and then used to stimulate CFSE-labeled OT-I or OT-II cells, respectively. As shown in Fig. 5A and B, and in contrast to pre-treatment of mice with CpG, DC obtained from BPP-treated mice maintained T-cell stimulation via cross-presentation and MHC II presentation. To better evaluate antigen capture and cross-presentation of CD8α+ DC after BPP pre-treatment in vivo, B6 mice were injected with BPP or CpG and received OCS i.v. the next day. One day later, CD8α+ DC were enriched by FACS-sorting and used in graded numbers to stimulate CFSE-labeled OT-I cells. CpG pre-treatment resulted in a marked reduction of OT-I proliferation, whereas BPP pre-treatment even caused a slight increase in T-cell proliferation, as compared to CD8α+ DC exposed to OCS only (Fig. 5C). Taken together, these findings suggest that BPP-matured spleen-derived DC, in contrast to CpG-treated DC, do not significantly downregulate cross-presentation.

Figure 5.

Impact of DC activation with BPP and BPP-OVA on antigen presentation and priming. Spleen-derived CD8α+ DC and CD8α DC subsets isolated from PBS-treated, BPP-treated (1 μg/mouse) and CpG-treated (25 nmol/mouse) mice (n=3 mice per group) were pulsed with 2×107 OCS or 100 μg/mL of soluble OVA. 1–2×104 antigen-loaded DC were then co-cultured with 5×104 CFSE-labeled (A) OT-I or (B) OT-II cells and proliferation was analyzed after 60 h. (C) B6 mice (n=3 mice per group) received PBS, 1 μg BPP or 20 nmol CpG i.v. The next day, mice were injected with 2×107 OCS and 16 h later, splenic CD8α+ DC were used to stimulate CFSE-labeled OT-I cells and T-cell proliferation was measured after 60 h. (D) B6 mice (n=3 mice per group) received PBS or 1 μg BPP/mouse i.v. The next day, the animals were treated with 2×107 OCS with or without 1 μg BPP. Five days later, CTL activity was determined in an in vivo cytotoxicity assay. (E) Groups of mice (n=3 or 4 mice per group) were immunized at day 0 and day 7 via footpad injection with 3 mg OVA, 3 mg OVA + 3 μg BPP, OVA257–264 + OVA323–339 (10 μg each) with or without 10 μg BPP, 10 μg BPP-OVA323–339 or 10 μg BPP-OVA. On day 12, the CTL response in the draining lymph nodes of these mice was measured in an in vivo cytotoxicity assay. The data show results obtained in one experiment (mean ± SEM) and are representative of at least two independent experiments (**p<0.01, ***p<0.001).

Pre-activation with BPP impairs cross-priming

Considering the above results, we wished to determine whether BPP-matured DC were still efficient in cross-priming. B6 mice were systemically challenged with BPP and 16 h later immunized with OCS alone or in combination with BPP. Cross-priming of CD8+ T cells was assessed 5 days later in an in vivo cytotoxicity assay, as described above. OVA-specific CTL response was observed in PBS pre-treated mice, which were then immunized with OCS in the presence of BPP. In contrast, when antigens were delivered in combination with BPP, the CTL priming in mice pre-treated with BPP was reduce to control mice which were not primed (Fig. 5D). The above findings suggest that BPP-matured DC, although able to cross-present, are impaired to cross-prime CTL responses.

Antigenic epitopes chemically linked to BPPcysMPEG promote efficient CTL priming

Finally, we sought to study the effectiveness of delivering antigenic peptides, which were directly linked to BPP, for inducing CTL responses. To this end, a synthetic peptide encompassing OVA257–264 and OVA323–339 was chemically linked to BPP and this compound was termed BPP-OVA. Single administration of BPP-OVA failed to mount efficient CTL responses. However, in mice, which were immunized twice with BPPcysOVA via footpad injections at an interval of 1 wk, effective CTL-mediated cytotoxicity was observed (Fig. 5E). On the other hand, only a weak CTL activity was detectable when mice were immunized with comparable amounts of OVA257–264 and OVA323–339 peptides mixed and co-administered with BPP (Fig. 5E). Administration of OVA257–264 plus OVA323–339 peptide or injection of only OVA protein (Sigma) resulted in negligible CTL-responses, similar to priming with a control construct containing only the MHC-class II peptide (BPP-OVA323–339). These results support the notion that antigens linked to TLR agonists induce better immune responses compared to those obtained when giving antigens and adjuvants admixed.

Discussion

In this study, we examined the effects of the synthetic TLR2/6 ligand BPP to efficiently prime a CTL response against exogenous antigens and the potential of linking peptide antigens to BPP in order to generate CTL immunity. In addition, our studies reveal some distinct characteristics of BPP for DC maturation and presentation on subsequently administered antigen via MHC I molecules as compared to other TLR ligands.

TLR agonists are preferential candidates for the usage as adjuvants for T-cell priming 14, 38. Activating DC by TLR ligation alters their cytokine secretion profiles, increases the surface expression of co-stimulatory molecules on DC and creates a favourable cytokine environment for T-cell priming. This is in contrast to T-cell stimulation with immature DC, which often leads to tolerance 39, 40. Several reasons lead us to study BPP. First, a study reported a prominent function of CD36 for efficient MALP-2 signaling and that a mutation in CD36 receptor makes macrophages insensitive to MALP-2 25. Given that CD36 is exclusively expressed on CD8α+ DC 26, a DC subset specifically equipped for constitutive cross-presentation 10, we speculated that BPP would preferentially stimulate this DC subpopulation. Second, BPP is chemically synthesized, which allows for generation of ample amounts of the adjuvants and linking it with various antigenic peptides without contamination of other bacterial products. Finally, MALP-2 has been characterized as an effective mucosal adjuvant and synthesis of BPP further improved solubility and pharmacokinetic features of the adjuvant. Systemic administration of BPP resulted in the phenotypic maturation of CD8α+ DC as well as CD8α DC, which could be a result of a secondary mechanism, such as cytokines released by macrophages and other bystander cells. Alternatively, the systemic concentrations obtained following intravenous application of BPP may overcome the requirement for CD36 signaling on DC and macrophages.

Although MALP-2 was shown to induce maturation of BMDC 22, the effects of MALP-2 on cross-presentation have not been systematically studied in vivo. Here, we provide evidence that BPP, like other TLR agonists, act on DC to enhance their MHC II-mediated antigen presentation and cross-presentation properties, as determined by increased T-cell proliferation both in vitro and in vivo. More importantly, when given together with cellular antigen, BPP induced an effective antigen-specific CTL response against a co-administered antigen. Our data suggest that the ability of BPP to promote cross-priming is dependent on CD4+ T-cell help. Other TLR ligands, such as CpG, are able to overcome this requirement or have been shown to mediate their adjuvant effect by inducing type-I interferons 32, 33, 35. We speculate that the enhanced CD4+ helper cell response after systemic BPP administration contributes at least partially to the adjuvant effect on CTL priming. Our experiments using TLR2-deficient cells strongly indicate the TLR2-dependent action of BPP, as any detectable effect was reduced down to control levels in these animals (Fig. 1, 2 and 4). In addition, if the BPP effects were to be caused by LPS-contamination, we would have expected to observe shut-down of cross-presentation after BPP pre-treatment 36, but this was not the case. In addition, there was no cytokine secretion in TLR2−/− BMDC after BPP exposure, thought LPS and CpG resulted in cytokine release, suggesting that no unspecific effects were caused by BPP. Finally, we used specifically low-endotoxin OVA in several fundamental experiments and primed mice with OCS in order to further reduce the amount of antigen required for priming and to avoid potential bias through endotoxin contamination.

The term “DC maturation” is often used to describe the transition of naïve DC into functionally immunogenic DC 41. Another hallmark of DC maturation is the progressive reduction in the phagocytosis of particulate antigens. The existing view is that immature DC capture antigens from the periphery, migrate to the secondary lymphoid organs during a process, in which they lose their ability to phagocytose, and finally give rise to mature DC that are exceedingly potent in priming T cells. Recent reports have demonstrated that once DC are matured with TLR ligands, they become inefficient in cross-presentation and MHC II presentation 36, 37. These studies proposed reduced phagocytosis and downregulation of MHC II synthesis after DC maturation as underlying mechanisms. It was surprising to observe in our experiments that BPP-activated DC were still capable of cross-presentation and MHC II presentation. In agreement with this finding, experiments using fluorescent latex beads indicated that BPP signaling has only limited impact on phagocytosis (data not shown). However, priming of CD8+ T cells against subsequently administered cellular antigens was impaired in vivo, even when DC stimulation with BPP plus antigen was repeated. One logical explanation for this observation could be that DC initially upregulate molecules needed for T-cell activation and at later time points possibly accumulate inhibitory signals such as PD-1 that inhibit T-cell function 42. Therefore, despite continuous antigen presentation by BPP-matured DC, CD8+ T cells did not become effectors or they acquired a regulatory phenotype. Alternatively, we speculate that TLR stimulation leads to exhaustion of DC with limited cytokine production or that TLR tolerance causes unresponsiveness of DC to repeated TLR signaling. Taken together, our data provide further evidence that stimulation of T-cell responses is not only controlled by down-modulating antigen uptake and presentation upon TLR stimulation, but also at the level of DC and T-cell interaction. More detailed studies are required in order to dissect the mechanism responsible for the differential regulation of antigen presentation by various TLR ligands.

Targeting antigens to DC and DC subsets is an attractive strategy to induce T-cell response and has been used efficiently for generating MHC class I and II responses 39, 43, 44. The co-stimulation signal for effective priming (via, e.g. CD40L), however, had to be given non-specifically in many cases. Alternatively, TLR ligands, such as CpG, have been successfully linked to model antigens in order to improve T-cell immunity. Given that most of the TLR ligands are non-peptidic components of infectious agents, which make them difficult to link to antigenic epitopes, lipopeptides have been identified as interesting candidates for vaccine strategies 15, 18. Elegant studies have highlighted the need for TLR ligands to be present along with antigen for efficient antigen presentation 45 and the inferior effects of inflammatory stimuli for CD4+ T-cell priming 46. These observations, together with the helper cell requirement for BPP, led us to study the outcomes of linking appropriate antigenic epitopes to BPP in order to attain better immune responses. Our data suggest that MHC class I and MHC class II epitopes of OVA can be chemically linked to BPP and that this construct improves immunization of mice as compared to the responses observed in mice immunized with OVA protein or MHC I and II peptides mixed along with BPP. Although two consecutive immunizations were necessary for this outcome, it should be emphasized that roughly twice as much peptide antigen was given to the control mice as compared to the BPP-OVA immunized mice.

In conclusion, our results show that BPP stimulation of DC via TLR2/6 leads to generation of an effective CTL response and that linking antigenic peptides to BPP is a promising strategy for vaccination (e.g. viral infections, cancer). Our studies suggest that mechanisms beyond downregulation of macropinocytosis and phagocytosis contribute to shut-down of cross-priming after TLR-mediated DC maturation. These observations contribute to a better understanding of the unique events induced by TLR ligands and for the design of antigen targeting strategies in vivo.

Materials and methods

Mice

C57BL/6 (B6) mice were purchased from Charles River Laboratories, Sulzfeld, Germany, and all the mice were used between 6 and 12 wk of age. OT-I, OT-II, MHC class II-deficient (MHC II−/−) 47, and B6.C-H-2Kbm1 (bm1) mice have been described previously 36 and were kindly provided by Dr. Bill Heath, The Walter Eliza Hall Institute of Medical Research, Australia. The OT-I and OT-II mice express MHC class I- and II-restricted TCR (Vα2/Vβ5.1+) specific for the H-2Kb-restricted OVA peptides OVA257–264 and OVA323–339, respectively. B6.C-H-2Kbm−1 mice express a mutated H-2Kb molecule that prevents presentation of OVA257−264 peptides to the respective TCR-transgenic T cells. All mice were bred and maintained at the animal facility of Hannover Medical School. Type I IFN receptor-deficient mice (IFN-αR−/−) and TLR2−/− mice were derived from the Helmholtz Center for Infection Research, Braunschweig, Germany. All experiments were reviewed and approved by an institutional review committee.

Antibodies and antigens

CD4 (GK1.5), CD8a (53–6.7), MHC II (2G9), CD11c (HL3), CD80 (16–10A1), IFN-γ (XMG1.2), CD86 (GL1), CD24 (ML5), CD172a (P84), Vα2 TCR (B20.1) specific antibodies were from BD PharMingen, Germany. Chicken egg ovalbumin (OVA, grade VI) was purchased from Sigma, Deisenhofen, Germany or EndoGrade Ovalbumin, endotoxin concentration <1 EU/mg, obtained from Hyglos, Regensburg, Germany.

TLR ligands

LPS from Escherichia coli serotype 026:B6 was ordered from Sigma. Phosophorothioate-stabilized CpG (5′ TTC ATG ACG TTC CTG ATG CT) oligonucleotides were ordered from TIB Molbiol, Berlin, Germany. BPPcysMPEG (BPP), International Patent Classification (IPC): A61K 47/48 (2006.01), Pub.No.: WO/2007/059931), a pegylated derivative of MALP-2 and BPPcysOVA (BPP-OVA), a compound resulting from BPPcysMPEG coupling to a synthetic peptide encompassing the MHC class I (OVA257–264) and II (OVA323–339) restricted epitopes from OVA, were synthesized at the Helmholtz Centre for Infection Research, Germany. Peptide synthesis resinTentaGel S RAM (Rapp Polymere, Tübingen, Germany) (100 μmol) was Fmoc-deprotected with 20% piperidin in DMF, washed with DMF and treated with Fmoc-NH-(PEG)27-COOH (Novabiochem/Merck, Nottingham, UK) (463.4 mg, 300 μmol), TBTU (96.3 mg, 300 μmol) and diisopropylethyl amine (105 μL, 600 μmol) in DMF for 18 h 1. The OVA-specific MHC I and II peptide sequences ISQAVHAAHAEINEAGRAA-SIINFEKL were assembled on the solid support with a Pioneer automatic peptide synthesizer (Applied Biosystems) employing Fmoc chemistry with TBTU/diisopropylethyl amine activation. Side chain protections were as follows: Glu and Ser: t-Bu, Asn, Gln and His: Trt, Arg: Pbf, Lys: Boc. Double couplings of 1 h each were employed throughout the synthesis. After final Fmoc-deprotection and washing with DMF and DCM, a test cleavage with 10 mg of the product was carried out by a 3-h treatment with TFA containing 3% triisopropylsilane and 2% water (1 mL). An HPLC and MALDI analysis of the crude product revealed a satisfying quality with the major product showing the expected molecular mass. BPPcys (Fmoc-protected S-(2(R),3 bis(palmitoyloxy)propyl]-L-cystein) (125.7 mg, 150 μmol), TBTU (48 mg, 150 μmol) and diisopropylethyl amine (53 μL, 300 μmol) in DMF were added to the solid support. After an incubation of 18 h the resin was washed with DMF followed by DCM. Cleavage from the support and side chain deprotection was carried out by a 5-h treatment with TFA containing 3% triisopropylsilane and 2% water (10 mL). After removing the resin by filtration and evaporation of most of the TFA, the crude product was precipitated with t-butylmethyl ether. The peptide was purified by preparative HPLC (RP-18) with water/acetonitrile gradients containing 0.1% TFA and characterized by MALDI-MS. The BPP-OVA molecule exhibits a molecular weight of 4818 Da.

Primary in vitro studies were performed with spleen and lymph node cells isolated from OT-I and OT-II animals for the estimation of the biological activity of BPP-OVA in comparison to the parenteral compound BPP. Therefore, cells were labeled with CFSE (5 min in 1 μM CFSE) and then plated in a 96-well plate in a concentration of 0.5×106per well. After restimulation with three different concentrations (0.1, 1 and 10 μg/mL) of BPP-OVA for 5 days, cells were stained with for CD4 or CD8 and analyzed by flow cytometry. The reduction in CFSE-labeling was determined as a measure of cycling and proliferation of these cells, thereby confirming the biological activity of BPP-OVA.

DC isolation, sorting and culture

DC enrichment was done as described 4, 10, 36. After digesting the splenic fragments with collagenase/DNase, the light density cells were enriched by density gradient centrifugation on a 1.078 g/cm3 Nycodenz layer for 10 min at 1700 g. DC were further enriched by using an optimized mAb cocktail (anti-CD3, anti-B220, anti-GR1, anti-CD19, anti-Ery) and anti-rat IgG coupled magnetic beads for depleting unwanted cells. The obtained cells were stained with anti-mouse CD11c (HL3) and anti-mouse CD8 (53–6.7), antibodies, and CD11c+ cells were sorted into CD8+, CD8 DC subsets on a FACSAria (BD Biosciences, Heidelberg, Germany) or Mo-Flo (Beckman Coulter GmbH, Germany). Purity of DC subsets was usually 95–98%.

Generation of BMDC

BMDC were generated by flushing out the bone-marrow from tibia and fibula of B6 or TLR2−/− mice. Red blood cells were lysed and cells were filtered through a 70-micron nylon mesh, counted and plated at a concentration of 1×106 cells/mL of complete RPMI supplemented with 20 ng/mL rGM-CSF. Medium was changed on day 3 and day 5 and usually 7-day-old cultures were used for experiments. Alternatively, cultures containing 1.5×106 cells/mL were supplemented with Flt3-L (200 ng/mL) for 9 days without any change of medium 23. For measurement of IL-6 and TNF-α, a cytokine bead array (FlowCytomix, Bender Medsystems, Germany) was performed according to the manufacturer's instructions. Samples were measure in triplicates.

T-cell proliferation in vitro

OVA-specific TCR transgenic CD8+ and CD4+ T cells were purified from the lymph nodes of OT-I and OT-II mice respectively, according to the instructions provided by the manufacturer (CD8+ and CD4+ T-cell enrichment set, BD™ IMag, BD Biosciences). The cells were purified using BD™ Imagnet and the efficiency of enrichment was routinely 85−95%, as determined by flow cytometry. The obtained cells were labeled with 5 μM carboxyfluorescein diacetate succinimidyl ester (CFSE). The antigen-loaded DC (1×104), usually pulsed with 100 μg/mL OVA, were washed three times and co-cultured with 5×104 T cells at 37°C for 60–65 h. The cells from the co-culture were washed and stained with anti-mouse CD8 (53–6.7) or anti-mouse CD4 (GK1.5) antibodies and the number of proliferating CFSElow OT-I or OT-II cells was analyzed by flow cytometry in duplicates or triplicates 30, 36.

In vivo proliferation of OVA-specific T cells

1×106 CFSE-labeled OT-I cells were injected intravenously into B6 mice previous to a challenge with OVA (3 mg/mouse) with or without BPP. Proliferation of adoptively transferred OT-I (CD8+Vα2+) cells in the inguinal lymph nodes and spleen was quantified 50–58 h after the antigen challenge by determining the number of CFSElow OT-I cells.

Preparation of OCS from bm1 mice

Single cell suspension of spleen cells obtained from bm1 mice was γ-irradiated (1000 rads) and splenocytes were re-suspended and incubated in medium containing 10 mg/mL of special low-endotoxin OVA for 10 min at 37°C. After thorough washes, these OCS were used as a source of cellular antigen (2×107 cells/ mouse) 36.

In vivo cytotoxicity assay

Splenocytes from B6 mice were split into two equal parts. One half was pulsed for 1 h at 37°C with 1 μg/mL of OVA257–264 and then labeled with a high concentration (2.5 μM) of CFSE (CFSEhi population). The remaining splenocytes were labeled with a low concentration (0.25 μM) of CFSE (CFSElow population). Equal numbers of cells from each population were mixed and 2×107 cells in total were adoptively transferred by i.v. injection into mice 5 days after priming with OCS together with 1 μg of BPP or CpG. Four hours later, spleen cell suspensions were analyzed by flow cytometry and each population was distinguished by fluorescence intensity. Percent OVA-specific lysis was determined by loss of the peptide-pulsed CFSEhi population compared with the control CFSElow population and was calculated using the following formula: % specific lysis=[1(runprimed/rprimed)]×100, where r=(%CFSElow/% CFSEhi). The corresponding ratio in naïve mice defined the 0% lysis levels 30.

Intracellular IFN-γ cytokine staining

Purified OT-I cells (2×106) were injected (i.v.) into B6 or TLR2−/− mice, which 24 h later were immunized (i.v.) with OCS (2×107) with or without BPPcysMPEG. After 3 days, mice were sacrificed and single cell suspension of splenocytes was made in complete RPMI. 2×106 splenocytes in 200 μL of medium containing 5 μg/mL of Brefeldin A and 50 U/mL of recombinant IL-2 (BD PharMingen, Heidelberg, Germany) were seeded into U-bottom 96-well plate and pulsed with 1 μg/mL of OVA257–264 for 5 h at 37°C. The cells were washed and stained for intracellular cytokines according to the manufacturer's instructions of the fixation permeabilization kit (BD Cytofix/Cytoperm™). After staining for surface markers with fluorochrome conjugated anti-mouse CD8 and anti-mouseVα2 antibodies, the cells were permeabilized and the intracellular IFN-γ was stained using fluorochrome conjugated anti-mouse IFN-γ (XMG1.2) antibody. The cells were analyzed on a FACSCalibur (BD Biosciences).

Statistical analysis

All data are presented as mean ± SEM. Student's t-test was used for statistical evaluation and a p-value below 0.05 was considered significant.

Acknowledgements

The authors acknowledge the excellent technical assistance of the members of the MHH Core Facility for Cell Sorting. The authors thank Marion Hitzigrath for excellent technical help. Furthermore, the authors thank Werner Tegge of the Helmholtz Centre for Infection Research for the production of the BPP-OVA construct. C. K. P. was supported by the Hannover Biomedical Research School (HBRS) and J. K. K by the GK1441. G. M. N. B., T. E. and C. A. G. were supported by a grant from the Excellence Cluster “From Regenerative Biology to reconstructive Therapy” by the German Research Foundation (EXC 62/1) and H. C. and G. M. N. B were supported by the Collaborative Grant SFB 587 (German Research Foundation).

Conflict of interest: T. E. and C. A. G. submitted a patent “BPPcysOvamPEG as mucosal immune-modulator” (EP 09016050.8, 28.12.2009). The other authors declare that they have no conflict of interest.

Ancillary