Immunology and Cell Biology (2014) 92, 377–383 & 2014 Australasian Society for Immunology Inc. All rights reserved 0818-9641/14 www.nature.com/icb

ORIGINAL ARTICLE

The use of a TLR2 agonist-based adjuvant for enhancing effector and memory CD8 T-cell responses Brendon Y Chua1, Matthew R Olson1, Sammy Bedoui, Toshiki Sekiya, Chinn Y Wong, Stephen J Turner and David C Jackson We have previously shown that the immunogenicity of protein antigens can be significantly enhanced if electrostatically associated with the Toll-like receptor-2 agonist-based lipopeptide R4Pam2Cys. The precise mechanisms and effectiveness of the cytotoxic T-lymphocyte (CTL)-mediated response facilitated by this agonist, however, have not been studied. Here we show that priming by dendritic cells (DCs) in the draining lymph nodes of animals vaccinated with antigen delivered using R4Pam2Cys results in significantly improved T-cell proliferation and induces their differentiation into polyfunctional effector CTLs characterised by granzyme B expression and the ability to secrete interferon-c, interleukin-2 and tumor necrosis factor-a 7 days after vaccination. After 30 days, frequencies of antigen-specific CD62lowCD127high (effector memory), CD62highCD127high (central memory) and CD43lowCD27high CD8 þ T cells, a phenotype associated with strong recall responses against respiratory infections, are also increased compared with responses obtained with antigens formulated in the adjuvants Alum (alhydrogel) and CFA (complete Freund’s adjuvant). The phenotypic changes observed in these mice vaccinated using R4Pam2Cys further correlated with their ability to recall specific T cells into the lung to mediate the reduction of pulmonary viral titres following challenge with a chimeric influenza virus containing the KbOVA257–264 epitope compared with animals vaccinated using Alum or CFA. The findings from this study not only demonstrate that better T-cell responses can be elicited using R4Pam2Cys compared with classically utilised adjuvants but also highlight the potential effectiveness of this lipopeptide-based adjuvant particularly against viral infections that require resolution through cell-mediated immunity. Immunology and Cell Biology (2014) 92, 377–383; doi:10.1038/icb.2013.102; published online 7 January 2014 Keywords: adjuvants; vaccination; CD8 T cells; TLR2; anti-viral immunity

The discovery that agonists of Toll-like receptors (TLRs) can act as adjuvants is one approach that is being increasingly used to enhance the responses obtained by vaccination and forms the basis for a new generation of rationally designed vaccines that can induce and maintain cellular, particularly cytotoxic CD8 þ T lymphocyte (CTL), responses. Chemical conjugation or ligation of TLR agonists to antigens can facilitate their delivery to dendritic cells (DCs), enabling improved antigen uptake, upregulation of surface MHC, increased expression of co-stimulatory molecules and the secretion of proinflammatory cytokines.1,2 TLR-dependent activation of DCs can promote CD8 þ T-cell responses resulting in antigen-driven proliferation3 and the acquisition of effector functions including the ability to secrete multiple cytokines and kill infected cells.4–7 Studies also show that improved and more consistent CTL responses can be evoked through the use of antigens that are bound directly, for example, covalently, to a TLR agonist. This enhanced efficacy is possibly due to the simultaneous delivery of both agonist and antigen to DCs, which results in more efficient cross-presentation.8,9 TLR agonists that have been used in vaccine design include those that

target TLR2 (di- and tri-acylated palmitic acid-based derivatives),4,9–11 TLR5 (flagellin),12 TLR9 (CpG oligodeoxynucleotides)13 and TLR7/8 (imidazoquinolines).7,14 The methods that are available for covalent coupling of antigen to agonist involve chemical modifications requiring reagents and processes that can be complex and could potentially prove costly for vaccine manufacture. Our solution to simplify this process is to use a cationic lipopeptide (R4Pam2Cys) containing the TLR2 agonist dipalmitoyl-S-glyceryl-cysteine that can bind electrostatically to soluble protein antigens.15,16 This approach provides a means of forming closely associated complexes of antigen and TLR2 agonist without resorting to the complexities of covalent chemistry and allows robust antibody and CD8 þ T-cell responses to be elicited by soluble protein antigens.4,16 Our initial proof-of-concept studies have demonstrated that inoculation of mice with the model antigen ovalbumin (OVA) formulated with R4Pam2Cys induces greater numbers of endogenous OVA257-specific interferon-g (IFN-g)-secreting CD8 þ T cells than those that are induced by inoculation with protein alone. We have

Department of Microbiology and Immunology, University of Melbourne, Parkville, Victoria, Australia 1These authors contributed equally to this work. Correspondence: Dr BY Chua, Department of Microbiology and Immunology, University of Melbourne, Royal Parade, Parkville, Victoria 3010, Australia. E-mail: [email protected] Received 3 November 2013; revised 1 December 2013; accepted 2 December 2013; published online 7 January 2014

Enhancement of T-cell-mediated immunity BY Chua et al 378

now further characterised the primary and recall CD8 þ T-cell responses that are induced with this formulation and compared these with responses induced using OVA formulated with other more traditional adjuvants. By adoptive transfer of TCR transgenic CD8 þ T cells, we have determined how inoculation of OVA with R4Pam2Cys impacts DC antigen presentation, T-cell proliferation and acquisition of effector function and development of immunological memory. Using a chimeric influenza virus containing the KbOVA257–264 epitope, we have also determined whether qualitative changes in T-cell responses are associated with improved recall responses to facilitate viral clearance, allowing us to delineate immune correlates with a state of effective anti-viral immunity. RESULTS Endogenous CD8 þ T-cell responses induced by R4Pam2Cys-OVA are TLR2 dependent We first established a requirement for TLR2 in mediating CD8 þ T-cell responses by comparing the endogenous responses elicited in TLR2-deficient (TLR2 / ) mice with those elicited in C57BL/6 mice. Inoculation of C57BL/6 mice with OVA formulated with R4Pam2Cys (R4Pam2Cys-OVA) resulted in a greater than eightfold increase in the number of IFN-g-secreting CD8 þ T cells in the spleen compared with the number induced by OVA alone (Figure 1a). This enhanced response was however abrogated in the absence of TLR2 because similarly low numbers of IFN-g-producing CD8 þ T cells were detected in TLR2 /  mice inoculated with OVA alone or with R4Pam2Cys-OVA (Figure 1b). The ability of both strains of mice to mount OVA257 epitope-specific responses following infection with the influenza virus PR8-OVA, which carries this epitope,17 indicates that T-cell priming per se was not affected in the absence of TLR2 (Figures 1a and b). In vivo T-cell proliferation is mediated by enhanced presentation of antigen to DCs Given the importance of DCs in naive T-cell activation and their ability to respond to TLR signals, we examined in vivo antigen presentation in draining lymph nodes following inoculation with

Figure 1 Endogenous CD8 þ T-cell responses in mice vaccinated with R4Pam2Cys protein formulations. (a) C57BL/6 mice or (b) TLR2-deficient (TLR2 / ) mice (n ¼ 3–5 per group) were inoculated subcutaneously at the base of the tail with 25 mg of OVA alone, OVA formulated with R4Pam2Cys or infected intranasally with chimeric PR8-OVA influenza virus. Spleens of vaccinated mice were removed 10 days later and numbers of IFN-gsecreting OVA257-specific CD8 þ T cells were determined in an intracellular cytokine-staining assay. Bar graphs show the mean of the total number (±s.d.) of antigen-specific cytokine-secreting cells per organ. Immunology and Cell Biology

R4Pam2Cys-OVA. Carboxyfluorescein succinimidyl ester (CFSE)labelled CD8 þ T cells from OT-1 transgenic mice (OT-1 T cells) expressing the Va2 TCR specific for OVA257 presented in association with H2kb were adoptively transferred into naive C57BL/6 mice 1 day before inoculation with 3, 6 or 25 mg of OVA alone or with the same amount of OVA conjugated to R4Pam2Cys. Analysis of CFSE-labelled OT-1 T cells in the draining lymph nodes of inoculated animals was carried out 3 days later. Although OVA alone induced the division of CFSE-labelled OT-1 T cells in a concentration-dependent manner (Figure 2a), equivalent amounts of antigen formulated with R4Pam2Cys resulted in higher levels of proliferation. Improved responses associated with R4Pam2Cys were also reflected in significantly higher frequencies of T cells when cell numbers in the nodes from each animal were taken into account; B2- to 10-fold increases in cell frequencies were observed using R4Pam2Cys–OVA compared with OVA alone (Figure 2b). In fact, similar proliferative levels achieved using 25 mg of OVA alone were achieved using 3 mg of R4Pam2Cys-OVA, demonstrating a strong dose-sparing effect. We hypothesised that R4Pam2Cys facilitates the induction of immune responses by activating DCs in vivo, a view that is consistent with observations demonstrating that Pam2Cys promotes DC uptake of antigen, induces DC maturation and promotes the secretion of proinflammatory cytokines in vitro.4,11,16 We therefore set out to determine whether or not DCs were the principal APCs facilitating CTL priming in an ex vivo antigen presentation assay. One day following inoculation of antigen, DCs were isolated from draining lymph nodes and cocultured with CFSE-labelled OT-1 T cells. CFSE dilution profiles of T cells present in cultures after 3 days demonstrated that DCs from OVA-vaccinated mice induced the proliferation of B68% of T cells (Figure 2c). DCs obtained from animals inoculated with R4Pam2Cys-OVA, however, triggered the proliferation of B90% of T cells. Taken together with the fact that these responses could be titrated by varying the number of DC per well (Figure 2d) and the in vivo data shown in Figures 2a and b, these results are consistent with the idea that R4Pam2Cys directly enhances the ability of DC to initiate priming of antigen-specific CD8 þ T-cell responses. Antigen-specific expansion and differentiation into effector T cells To investigate the effects of Pam2Cys on the development of effector responses, we adoptively transferred a lesser number of non-CFSElabelled OT-1 T cells into mice before inoculation with R4Pam2CysOVA. Using a congenic marker to distinguish these cells in vivo, OT-1 T cells were found to represent B0.03–0.27% of the total CD8 þ T-cell population 7 days following inoculation with phosphatebuffered saline, OVA or R4Pam2Cys alone (Figure 3a). However, inoculation with R4Pam2Cys-OVA resulted in B10-fold increase of this population, which was also reflected in absolute cell numbers (Figure 3b). We also observed that B50% of OT-1 T cells in mice inoculated with R4Pam2Cys-OVA responded to antigen by secreting proinflammatory cytokines compared with B20% in mice inoculated with OVA alone (Figure 3c). Marked differences in the cytokine profiles of these cells were also observed where a significantly larger proportion (B25%) of OT-1 T cells secreting multiple cytokines, IFN-g, interleukin-2 (IL-2) and tumor necrosis factor-a, ensued following inoculation with R4Pam2Cys-OVA compared with B5% using OVA alone. The responses that were elicited by the use of this lipopeptide were also accompanied by increased expression of lytic proteins, particularly granzyme B, as reflected by significantly higher percentages and

Enhancement of T-cell-mediated immunity BY Chua et al 379

Figure 2 In vivo antigen presentation in vaccinated mice. (a) A total of 1  106 CFSE-labelled OT-1 CD8 þ T cells were transferred intravenously into naive C57BL/6 mice (n ¼ 3 per group) and inoculated subcutaneously at the base of the tail the next day with various amounts of OVA alone or formulated with R4Pam2Cys. Inguinal lymph nodes were harvested 3 days later and CD45.1 þ CD8 þ T cells were analysed for CFSE levels by flow cytometry. Figures within each histogram indicate the percentage (±s.d.) of proliferating cells. (b) Results from (a) are presented as bar graphs that depict the mean of the total number (±s.d.) of OT-1 T cells. (c) DCs isolated from the popliteal lymph nodes of mice 24 h after subcutaneous inoculation via the footpad were incubated at various concentrations with 5  104 CFSE-labelled OT-1 T cells for 72 h. Proliferating OT-1 T cells were then determined by measuring CFSE levels. Representative histograms of OT-1 T-cell CFSE levels showing the percentage (±s.d.) of proliferating cells from cocultures containing 6  103 DCs. (d) Also presented is the frequency (±s.d.) of proliferating CFSE-labelled OT-1 T cells cocultured with titrated numbers of DCs.

Figure 3 Inoculation leads to enhanced in vivo antigen-specific CD8 þ T-cell proliferation and function. A total of 1  104 OT-1 CD8 þ T cells were transferred intravenously into naive mice (n ¼ 3–4 per group) 1 day before subcutaneous inoculation via the base of the tail with 25 mg of OVA alone or formulated with R4Pam2Cys. Some mice received phosphate-buffered saline alone or an equivalent amount of R4Pam2Cys without antigen. Spleens were harvested 7 days later and the (a) average percentage and (b) total number (±s.d.) of OT-1 T cells determined. (c) Cytokine secretion from OT-1 T cells in the presence of OVA257 peptide antigen was measured in an intracellular cytokine-staining assay for IFN-g, IL-2 and tumor necrosis factor-a. (d) OT-1 T cells were also stained for their coexpression of granzymes A and B. (e) Results of granzyme þ OT-1 T cells are also represented as bar graphs that depict the mean (±s.d.) of the total number of cells present in the spleen. A full colour version of this figure is available at the Immunology and Cell Biology journal online. Immunology and Cell Biology

Enhancement of T-cell-mediated immunity BY Chua et al 380

frequencies of granzyme B-expressing OT-1 cells in mice vaccinated with R4Pam2Cys-OVA (Figures 3d and e). Although a similar trend in the expression of granzyme A was also observed in R4Pam2Cys-OVAvaccinated animals, expression levels of this protein varied among OT-1 cells analysed. Comparison of primary and memory T-cell responses induced by antigens formulated with other adjuvants Given the enhancement of CTL responses induced by the use of R4Pam2Cys, it was important to compare its relative effectiveness to other well-studied and traditionally used adjuvants. We therefore compared alhydrogel (Alum) and complete Freund’s adjuvant (CFA) with R4Pam2Cys in their ability to induce OT-1 T-cell proliferation when inoculated in association with OVA. The highest frequencies of antigen-specific OT-1 cells present after 7 days, distinguished by the congenic marker CD45.1, were induced using R4Pam2Cys (Figure 4a). These responses were almost twofold greater than those induced by CFA and B13-fold greater than responses induced using Alum. Activation of an efficient T-cell response is characterised by an expansion phase dominated by highly functional effector populations. This is followed by a contraction phase and differentiation into effector memory (TEM) and eventually central memory T cells (TCM) that have the capacity to respond strongly upon re-encounter with antigen. In analysing T-cell frequencies 30 days after vaccination, decreases in OT-1 numbers from all vaccinated groups were observed (Figure 4b) reminiscent of this contraction phase and possibly also reflecting the transition of effector cells into stable memory populations. In fact, the trend towards higher numbers of T cells in animals inoculated with R4Pam2Cys suggests that there may be differences in the size of the memory pool induced by R4Pam2Cys compared with the other two adjuvants. We therefore examined the OT-1 T-cell populations in mice 30 days after inoculation with the different antigen–adjuvant formulations for the activation marker CD62L18 and the IL-7 receptor CD127, which is linked with the development of long-term memory cells.19 Although it would seem that higher percentages of effector T cells defined by CD62LlowCD127low expression were induced by OVA formulated in CFA or Alum compared with R4Pam2Cys (Figure 5a), absolute cell numbers indicated that effector frequencies were similar in all mice

Figure 4 Comparison of proliferative responses induced by inoculation of antigen formulated with different adjuvants. A total of 1  104 OT-1 CD8 þ T cells were transferred intravenously into naive C57BL/6 mice (n ¼ 3–4 per group) 1 day before inoculation via the subcutaneous route at the base of the tail with 25 mg of OVA alone, formulated with either Alum, CFA or with R4Pam2Cys. Spleens were harvested 7 (a) or 30 (b) days later and the total number (±s.d.) of CD45.1 þ CD8 þ OT-1 T cells was enumerated. Immunology and Cell Biology

regardless of the adjuvant used (Figure 5b). Higher percentages and frequencies of TEM (CD62LlowCD127high) and TCM cells (CD62highCD127high) were, however, present in animals inoculated with R4Pam2Cys-OVA (Figures 5a and b). In fact, negligible numbers of these populations were present in mice inoculated with antigen in Alum or CFA. These results demonstrate that delivery of antigen by R4Pam2Cys not only drives better proliferative T-cell responses than Alum or CFA but it is also more effective at driving memory T-cell development. We also analysed OT-1 T-cell populations for CD43lowCD27high expression. CD8 þ T cells with this phenotype are associated with strong recall responses elicited by viral respiratory infections.20 Our analysis (Figure 5c) showed that the highest percentages of CD43lowCD27high cells were detected in animals that received R4Pam2Cys-OVA. CD43lowCD27high cell frequencies in these mice were B20-fold higher when compared with mice that received Alum and B5-fold higher than mice that received CFA (Figure 5d). Animals vaccinated with R4Pam2Cys-OVA demonstrate enhanced viral clearance To determine whether the T-cell phenotypes elicited correlated with an ability to mediate anti-viral recall responses, OT-1 T cells were

Figure 5 Phenotype of CD8 T cells induced by inoculation with antigen formulated with various adjuvants. C57BL/6 mice (n ¼ 3–5 per group) received 1  104 OT-1 CD8 þ T cells 1 day before inoculation subcutaneously at the base of the tail with 25 mg of OVA in saline, formulated with either Alum, CFA or R4Pam2Cys. Spleens were harvested 30 days later and CD45.1 þ CD8 þ OT-1 T cells were analysed for coexpression of (a) CD62L and CD127 or (c) CD27 and CD43. Representative dot plots (a, c) are derived from concatenated samples from each group. Bar graphs represent (b) the total frequency (±s.d.) of OT-1 cells distinguished by their memory phenotype profile based on CD62L and CD127 expression and (d) the total number of CD43lowCD27high OT-1 T cells. Where indicated, asterisks (*) in (b) indicate Po0.05 between effector memory and central memory populations of R4Pam2Cys-OVAvaccinated animals compared with the use of other adjuvants. A full colour version of this figure is available at the Immunology and Cell Biology journal online.

Enhancement of T-cell-mediated immunity BY Chua et al 381

Figure 6 Lung viral titres and infiltration of CD8 T cells following virus challenge. OT-1 T cells (1  104) were transferred intravenously into naive C57BL/6 mice 1 day before inoculation via the subcutaneous route at the base of the tail with OVA (25 mg) in saline, formulated with Alum, CFA or mixed with a threefold molar excess of R4Pam2Cys. Animals were challenged intranasally after 28 days with chimeric X31-OVA influenza virus. (a) Titres of influenza virus in lung homogenates 5 days later were determined by plaque formation. (b) Percentages of OT-1 T cells within the CD8 T-cell population (±s.d.) and total numbers (±s.d.) of OT-1 CD8 þ T cells present in lung bronchoalveolar lavage (BAL) washes were also enumerated.

transferred into mice that were then inoculated with OVA in the presence of Alum, CFA or R4Pam2Cys, and subsequently challenged with X31-OVA influenza virus4,17 28 days later. Lungs were removed 5 days following viral challenge and the pulmonary viral titres determined. The highest viral titres were detected in the lungs of mice inoculated with OVA in saline (Figure 6a) and there were no significant differences in titres in animals vaccinated using Alum or CFA. Animals vaccinated with R4Pam2Cys-OVA, however, presented with reduced viral titres when compared with antigen delivered in other formulations. The ability of vaccinated mice to clear infection also correlated with the presence of antigen-specific T cells in the lung where up to a threefold difference in the number of OT-1 T cells was found in the bronchoalveolar washings of mice inoculated with R4Pam2Cys-OVA compared with those that received OVA in saline, Alum or CFA (Figure 6b). Therefore, the fact that these mice were unable to reduce significantly viral loads to the same effect as those that received R4Pam2Cys-OVA is likely to be attributed to the qualitative and functional differences in T-cell responses induced by these adjuvants. DISCUSSION In this study, we show that primary endogenous CD8 þ T-cell responses induced by inoculation of antigen that is electrostatically associated with R4Pam2Cys are dependent on TLR2 signalling. Using CD8 þ T cells obtained from transgenic mice expressing the same antigen-specific TCR, we have demonstrated that inoculation in the presence of R4Pam2Cys enhances the ability of DCs to present antigen to naive T cells in vivo and generates high frequencies of effector and memory T-cell populations. The ability of this lipopeptide to induce generation of TEM and TCM T cells, as defined by the expression of CD62L and CD127, was more effective than Alum or CFA, and was also better at facilitating efficient recall viral-clearing responses. Our results support the view that limiting excessive inflammation, by using controlled amounts of TLR agonists, for example, is conducive for establishing T-cell memory.21 TLR2-mediated priming using antigen mixed with Pam3Cys has been shown to promote CD4 þ T-cell memory development in mice3 and detectable TCM-type CD8 þ T-cell responses have been described in humans who have been vaccinated in a Phase IIa study with nanoparticles containing the Melan-A/MART peptide epitope associated with the TLR7 agonist imiquinod.22 By delineating the phenotype and functional qualities of T cells induced by the targeting of antigen to

TLR2 in this study has further allowed us to correlate these changes with an improved state of anti-viral immunity. In addition, the responses facilitated by the use of R4Pam2Cys were also associated with high numbers of CD8 þ T cells with a CD43lowCD27high phenotype, which were present in low numbers in mice that received Alum or CFA. This phenotype represents a distinct cell population that can mediate strong recall responses against Sendai virus infection independently of their TCM or TEM status.20 In demonstrating a correlation between recall efficiency and the frequency of these cells, we have shown that the expression of these markers together with more conventional effector memory phenotypes could also be a useful predictor of vaccine efficacy and may constitute an alternative method for classifying T-cell recall responses induced by limiting inflammation.21 Although it remains to be seen which DC subsets are involved in facilitating the CD8 þ T-cell responses observed in our study, the ability of R4Pam2Cys to modulate early phases of immune induction is likely to explain why it is a more effective adjuvant than Alum or CFA at inducing CD8 þ T-cell responses. We can speculate that dermal and CD8a þ DCs have an important role in the cross-priming of CD8 þ T cells and activation of CD4 þ T cells based on recent studies using antigens conjugated to TLR7 and TLR2 agonists.8,9,14 In addition, Langerhans cells have also been implicated in crosspresentation of antigens that are delivered with imidazoquinolinebased TLR7 agonists.8 However, because TLR-mediated signalling of CD8 T cells can enhance their effector function by lowering the activation threshold for costimulatory signals delivered by APCs23 and promote their differentiation into memory cells,24 we cannot rule out the possibility that R4Pam2Cys directly engages TLR2 on T cells,25 resulting in augmentation of DC-mediated antigen presentation. Alum has been one of the most widely used adjuvants in humans, and while it may be the adjuvant of choice for formulation with vaccines to induce protective neutralising antibodies, its inability to augment T-cell responses is a limitation for its use particularly in cases where cell-mediated immunity is required.26 The inability of CFA to induce memory CD8 þ T-cell responses is somewhat unexpected when we consider that it contains mycobacterial-derived components and is believed to require TLR signalling for its adjuvant effect.27 However, a study by Gavin et al.28 demonstrated that antigens emulsified in CFA can still induce robust antibody responses in mice that are deficient in TLR signalling, indicating that other immunological and physicochemical mechanisms could account Immunology and Cell Biology

Enhancement of T-cell-mediated immunity BY Chua et al 382

for the adjuvanticity of CFA. Another likely possibility is that inflammation induced by CFA, although able to generate potent short-lived effector cells, may be excessive for memory precursor effector cell formation.29 By characterising the mechanisms and features associated with the induction and development of T-cell responses mediated by R4Pam2Cys, our findings have highlighted the potential of utilising this TLR2-agonist-based adjuvant for eliciting T-cell immunity. It is also however important to bear in mind that anti-inflammatory responses can also be induced by some TLR agonists, which may inhibit the generation of effective cell-mediated responses. TLR9 stimulation by CpG-containing DNA sequences can induce IL-1030 and prostaglandin E31 productions, while stimulation via TLR2 can lead to increased regulatory T-cell function in an IL-10-dependent manner.32 In certain immunosuppressive microenvironments such as the milieu within tumours, these effects could have a major role in dampening any potential antitumour T-cell responses and may in fact help explain the poor efficacies of TLR agonists in clinical trials evaluating their use as immunotherapies or adjuvants for tumour antigens in the treatment of cancer.33 It is therefore possible that when considering the use of R4Pam2Cys or any other TLR-agonist-based adjuvant in humans, additional approaches to minimise any immunosuppressive or immunomodulatory cytokine- and/or cell-mediated effects may be required to optimise the induction of protective T helper type 1-type responses. Moreover, evidence of hyporesponsiveness to the Lyme disease vaccine based on the outer-surface lipoprotein (OspA) of Borrelia burgdorferi, which is recognised by TLR2,34 indicates that TLR polymorphisms and defects in signalling35,36 exist within the population and may have a role in determining the effectiveness of these adjuvants. Nonetheless, the use of simple and easily manufactured adjuvants that allow soluble proteins to induce robust T-cell immunity with clear mechanisms of action address an unmet medical need that may lead to the development of novel clinical applications, particularly for the treatment of afflictions that require resolution through cell-mediated immunity. METHODS Synthesis of the cationic lipopeptide R4Pam2Cys The synthesis, purification and analysis of the branched cationic lipopeptide R4Pam2Cys has been described in detail previously.4

Diego, CA, USA) antibodies. Analysis was performed on a FACS Canto II and data were analysed using FlowJo software (Treestar, Ashland, OR, USA).

Ex vivo antigen presentation by DCs Popliteal lymph nodes were obtained from mice inoculated subcutaneously in the footpad 1 day after inoculation. Pooled lymph nodes were enriched for DCs using antibodies to Gr-1, CD3e, CD19, Thy-1, B220 and erythrocytes (Ter119), followed by magnetic depletion as described.39 Naive OT-1 T cells were obtained from lymph nodes of CD45.1 þ OT-1 mice and enriched with antibodies against Mac-1, F4/80, Gr-1, I-A/E, CD4 and erythrocytes before depletion. Purified OT-1 cells were labeled with 1 mM CFSE (Sigma-Aldrich, Castle Hill, NSW, Australia) and cultured (5  104) for 72 h with various concentrations of purified DCs. Proliferation of propidium iodide-excluded CD8 þ Va2 þ T cells by CFSE dilution was determined by flow cytometry.

In vivo T-cell proliferation CFSE-labelled CD45.1 þ OT-1 cells (1  106) were transferred intravenously into naive mice before inoculations the next day. Inguinal lymph nodes of vaccinated mice were obtained 3 days later and CFSE levels of CD8 þ CD45.1 þ T cells were analysed. To determine the proliferation of T cells 7 days after inoculation, non-CFSE-labelled OT-1 cells (1  104) were transferred. For coadministration of antigen with adjuvant, mice received 25 mg of OVA formulated either in an equal volume of aluminium hydroxide for 3 h (13 mg ml 1; Sigma-Aldrich) or as an emulsion in an equal volume of CFA (Sigma Aldrich) subcutaneously at the base of the tail. Spleens were obtained 30 days later and CD8 þ CD45.1 þ T cells were analysed using CD62L-PECy7 (BioLegend), CD127-APC (eBioscience, San Diego, CA, USA), CD43-FITC and CD27-PerCP Cy5.5 (BioLegend) antibodies.

Influenza virus challenge and lung virus assay Mice were challenged intranasally with 104 PFU of recombinant A/Aichi/2/68 (H3N2) influenza virus containing the KbOVA257–264 epitope (X31-OVA).17 Lymphocytes obtained from lung bronchoalveolar lavage washes were obtained 5 days later and enumerated by flow cytometry. Lungs of mice were homogenised and titres of virus in supernatants were determined using a Madin–Darby canine kidney plaque assay as described previously.40

Statistical analyses Analysis of variance and P-values in this study were obtained using one-way analysis of variance nonparametric analyses and Tukey’s post hoc range tests or two-way analysis of variance analyses and Bonferroni post-tests, performed using the Prism 5 software package (GraphPad Software, La Jolla, CA, USA).

ACKNOWLEDGEMENTS Mice C57BL/6, CD45.1 þ OT-1 TCR transgenic37 and TLR2-deficient38 mice were bred and maintained under specific pathogen-free conditions and used when 6–10 weeks of age. All experiments were approved by The University of Melbourne’s animal ethics committee.

This work was supported by a National Health and Medical Research Council (NHMRC) Program grant (to DCJ and SJT), an Australia Research Council Future Research Fellowship (to SJT), a University of Melbourne Early Career Researcher Grant (to BYC) and an NHMRC Career Development Fellowship (to SB).

Inoculation and measurement of CD8 þ T-cell responses Groups of mice were inoculated subcutaneously at the base of the tail with 25 mg OVA (0.55 nmol) in phosphate-buffered saline or premixed with a threefold molar excess of R4Pam2Cys (1.65 nmol; 3.2 mg) in100 ml (50 ml on each side). A separate group of mice were inoculated intraperitoneally with 1.5  107 plaque-forming units (PFU) of recombinant A/Puerto Rico/8/34 (H1N1) influenza virus strain containing the KbOVA257–264 epitope (PR8OVA).17 After 10 days, an intracellular cytokine-staining assay was performed by incubating splenocytes (1  106) with OVA257–264 peptide (SIINFEKL; 1 mg ml 1) in the presence of BD GolgiPlug (BD Biosciences, San Diego, CA, USA) and recombinant IL-2 (10 U ml 1; Roche, Burgess Hill, West Sussex, UK). Lymphocytes were stained with CD8-PerCP Cy5.5 (BD Biosciences) 6 h later, fixed and permeabilised and further stained with IFN-g-FITC (BD Biosciences), IL-2-PE and tumor necrosis factor-a-APC (BioLegend, San Immunology and Cell Biology

1 Byun EH, Kim WS, Kim JS, Jung ID, Park YM, Kim HJ et al. Mycobacterium tuberculosis Rv0577, a novel TLR2 agonist, induces maturation of dendritic cells and drives Th1 immune response. FASEB J 2012; 26: 2695–2711. 2 Tacken PJ, Zeelenberg IS, Cruz LJ, van Hout-Kuijer MA, van de Glind G, Fokkink RG et al. Targeted delivery of TLR ligands to human and mouse dendritic cells strongly enhances adjuvanticity. Blood 2011; 118: 6836–6844. 3 Chandran SS, Verhoeven D, Teijaro JR, Fenton MJ, Farber DL. TLR2 engagement on dendritic cells promotes high frequency effector and memory CD4 T cell responses. J Immunol 2009; 183: 7832–7841. 4 Chua BY, Pejoski D, Turner SJ, Zeng W, Jackson DC. Soluble proteins induce strong CD8 þ T cell and antibody responses through electrostatic association with simple cationic or anionic lipopeptides that target TLR2. J Immunol 2011; 187: 1692–1701. 5 Day EB, Zeng W, Doherty PC, Jackson DC, Kedzierska K, Turner SJ. The context of epitope presentation can influence functional quality of recalled influenza A virusspecific memory CD8 þ T cells. J Immunol 2007; 179: 2187–2194.

Enhancement of T-cell-mediated immunity BY Chua et al 383 6 Deliyannis G, Kedzierska K, Lau YF, Zeng W, Turner SJ, Jackson DC et al. Intranasal lipopeptide primes lung-resident memory CD8 þ T cells for long-term pulmonary protection against influenza. Eur J Immunol 2006; 36: 770–778. 7 Wille-Reece U, Wu C-Y, Flynn BJ, Kedl RM, Seder RA. Immunization with HIV-1 Gag protein conjugated to a TLR7/8 agonist results in the generation of HIV-1 Gag-specific Th1 and CD8 þ T cell responses. J Immunol 2005; 174: 7676–7683. 8 Oh JZ, Kedl RM. The capacity to induce cross-presentation dictates the success of a TLR7 agonist-conjugate vaccine for eliciting cellular immunity. J Immunol 2010; 185: 4602–4608. 9 Prajeeth CK, Jirmo AC, Krishnaswamy JK, Ebensen T, Guzman CA, Weiss S et al. The synthetic TLR2 agonist BPPcysMPEG leads to efficient cross-priming against coadministered and linked antigens. Eur J Immunol 2010; 40: 1272–1283. 10 Jackson DC, Lau YF, Le T, Suhrbier A, Deliyannis G, Cheers C et al. A totally synthetic vaccine of generic structure that targets Toll-like receptor 2 on dendritic cells and promotes antibody or cytotoxic T cell responses. Proc Natl Acad Sci USA 2004; 101: 15440–15445. 11 Zeng W, Ghosh S, Lau YF, Brown LE, Jackson DC. Highly immunogenic and totally synthetic lipopeptides as self-adjuvanting immunocontraceptive vaccines. J Immunol 2002; 169: 4905–4912. 12 Nguyen CT, Kim SY, Kim MS, Lee SE, Rhee JH. Intranasal immunization with recombinant PspA fused with a flagellin enhances cross-protective immunity against Streptococcus pneumoniae infection in mice. Vaccine 2011; 29: 5731–5739. 13 Maurer T, Heit A, Hochrein H, Ampenberger F, O’Keeffe M, Bauer S et al. CpG-DNA aided cross-presentation of soluble antigens by dendritic cells. Eur J Immunol 2002; 32: 2356–2364. 14 Kastenmu¨ller K, Wille-Reece U, Lindsay RWB, Trager LR, Darrah PA, Flynn BJ et al. Protective T cell immunity in mice following protein-TLR7/8 agonist-conjugate immunization requires aggregation, type I IFN, and multiple DC subsets. J Clin Invest 2011; 121: 1782–1796. 15 Chua BY, Eriksson EM, Brown LE, Zeng W, Gowans EJ, Torresi J et al. A selfadjuvanting lipopeptide-based vaccine candidate for the treatment of hepatitis C virus infection. Vaccine 2008; 26: 4866–4875. 16 Chua BY, Johnson D, Tan A, Earnest-Silveira L, Sekiya T, Chin R et al. Hepatitis C VLPs delivered to dendritic cells by a TLR2 targeting lipopeptide results in enhanced antibody and cell-mediated responses. PLoS One 2012; 7: e47492. 17 Jenkins MR, Webby R, Doherty PC, Turner SJ. Addition of a prominent epitope affects influenza A virus-specific CD8 þ T cell immunodominance hierarchies when antigen is limiting. J Immunol 2006; 177: 2917–2925. 18 Obar JJ, Khanna KM, Lefrancois L. Endogenous naive CD8 þ T cell precursor frequency regulates primary and memory responses to infection. Immunity 2008; 28: 859–869. 19 Kaech SM, Tan JT, Wherry EJ, Konieczny BT, Surh CD, Ahmed R. Selective expression of the interleukin 7 receptor identifies effector CD8 T cells that give rise to long-lived memory cells. Nat Immunol 2003; 4: 1191–1198. 20 Hikono H, Kohlmeier JE, Takamura S, Wittmer ST, Roberts AD, Woodland DL. Activation phenotype, rather than central- or effector-memory phenotype, predicts the recall efficacy of memory CD8 þ T cells. J Exp Med 2007; 204: 1625–1636. 21 Badovinac VP, Messingham KA, Jabbari A, Haring JS, Harty JT. Accelerated CD8 þ T-cell memory and prime-boost response after dendritic-cell vaccination. Nat Med 2005; 11: 748–756. 22 Goldinger SM, Dummer R, Baumgaertner P, Mihic-Probst D, Schwarz K, HammannHaenni A et al. Nano-particle vaccination combined with TLR-7 and -9 ligands triggers

23

24

25

26 27 28

29

30

31

32

33

34

35

36 37

38

39

40

memory and effector CD8 þ T-cell responses in melanoma patients. Eur J Immunol 2012; 42: 3049–3061. Cottalorda A, Verschelde C, Marc¸ais A, Tomkowiak M, Musette P, Uematsu S et al. TLR2 engagement on CD8 T cells lowers the thresholdfor optimal antigen-induced T cell activation. Eur J Immunol 2006; 36: 1684–1693. Quigley M, Martinez J, Huang X, Yang Y. A critical role for direct TLR2-MyD88 signaling in CD8 T-cell clonal expansion and memory formation following vaccinia viral infection. Blood 2009; 113: 2256–2264. Asprodites N, Zheng L, Geng D, Velasco-Gonzalez C, Sanchez-Perez L, Davila E. Engagement of Toll-like receptor-2 on cytotoxic T-lymphocytes occurs in vivo and augments antitumor activity. FASEB J 2008; 22: 3628–3637. Kool M, Fierens K, Lambrecht BN. Alum adjuvant: some of the tricks of the oldest adjuvant. J Med Microbiol 2012; 61: 927–934. Stills HF. Adjuvants and antibody production: dispelling the myths associated with Freund's complete and other adjuvants. ILAR J 2005; 46: 280–293. Gavin AL, Hoebe K, Duong B, Ota T, Martin C, Beutler B et al. Adjuvant-enhanced antibody responses in the absence of toll-like receptor signaling. Science 2006; 314: 1936–1938. Joshi NS, Cui W, Chandele A, Lee HK, Urso DR, Hagman J et al. Inflammation directs memory precursor and short-lived effector CD8( þ ) T cell fates via the graded expression of T-bet transcription factor. Immunity 2007; 27: 281–295. Duramad O, Fearon KL, Chan JH, Kanzler H, Marshall JD, Coffman RL et al. IL-10 regulates plasmacytoid dendritic cell response to CpG-containing immunostimulatory sequences. Blood 2003; 102: 4487–4492. Chen Y, Zhang J, Moore SA, Ballas ZK, Portanova JP, Krieg AM et al. CpG DNA induces cyclooxygenase-2 expression and prostaglandin production. Int Immunol 2001; 13: 1013–1020. Yamazaki S, Okada K, Maruyama A, Matsumoto M, Yagita H, Seya T. TLR2-dependent induction of IL-10 and Foxp3 þ CD25 þ CD4 þ regulatory T cells prevents effective anti-tumor immunity induced by Pam2 lipopeptides in vivo. PLoS One 2011; 6: e18833. Ridnour LA, Cheng RY, Switzer CH, Heinecke JL, Ambs S, Glynn S et al. Molecular pathways: toll-like receptors in the tumor microenvironment—poor prognosis or new therapeutic opportunity. Clin Cancer Res 2013; 19: 1340–1346. Alexopoulou L, Thomas V, Schnare M, Lobet Y, Anguita J, Schoen RT et al. Hyporesponsiveness to vaccination with Borrelia burgdorferi OspA in humans and in TLR1- and TLR2-deficient mice. Nat Med 2002; 8: 878–884. Lorenz E, Mira JP, Cornish KL, Arbour NC, Schwartz DA. A novel polymorphism in the toll-like receptor 2 gene and its potential association with staphylococcal infection. Infect Immun 2000; 68: 6398–6401. Schroder NW, Schumann RR. Single nucleotide polymorphisms of Toll-like receptors and susceptibility to infectious disease. Lancet Infect Dis 2005; 5: 156–164. Kurts C, Heath WR, Carbone FR, Allison J, Miller JF, Kosaka H. Constitutive class I-restricted exogenous presentation of self antigens in vivo. J Exp Med 1996; 184: 923–930. Wooten RM, Ma Y, Yoder RA, Brown JP, Weis JH, Zachary JF et al. Toll-like receptor 2 is required for innate, but not acquired, host defense to Borrelia burgdorferi. J Immunol 2002; 168: 348–355. Belz GT, Bedoui S, Kupresanin F, Carbone FR, Heath WR. Minimal activation of memory CD8 þ T cell by tissue-derived dendritic cells favors the stimulation of naive CD8 þ T cells. Nat Immunol 2007; 8: 1060–1066. Tannock GA, Paul JA, Barry RD. Relative immunogenicity of the cold-adapted influenza virus A/Ann Arbor/6/60 (A/AA/6/60-ca), recombinants of A/AA/6/60-ca, and parental strains with similar surface antigens. Infect Immun 1984; 43: 457–462.

Immunology and Cell Biology