Efficient Immunization and Cross-Priming by Vaccine Adjuvants Containing TLR3 or TLR9 Agonists Complexed to Cationic Liposomes

of February 1, 2016.This information is current as

LiposomesTLR9 Agonists Complexed to Cationicby Vaccine Adjuvants Containing TLR3 or Efficient Immunization and Cross-Priming

Ross Kedl, Angelo Izzo, Catharine Bosio and Steven DowKaren Zaks, Michael Jordan, Amanda Guth, Karen Sellins,

http://www.jimmunol.org/content/176/12/7335doi: 10.4049/jimmunol.176.12.7335

2006; 176:7335-7345; ;J Immunol 

Referenceshttp://www.jimmunol.org/content/176/12/7335.full#ref-list-1

, 16 of which you can access for free at: cites 71 articlesThis article

Subscriptionshttp://jimmunol.org/subscriptions

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/ji/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/cgi/alerts/etocReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2006 by The American Association of9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

by guest on February 1, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

Efficient Immunization and Cross-Priming by VaccineAdjuvants Containing TLR3 or TLR9 Agonists Complexed toCationic Liposomes1

Karen Zaks,3* Michael Jordan,3† Amanda Guth,* Karen Sellins,* Ross Kedl,‡ Angelo Izzo,*Catharine Bosio,* and Steven Dow2*§

Complexing TLR9 agonists such as plasmid DNA to cationic liposomes markedly potentiates their ability to activate innateimmunity. We therefore reasoned that liposomes complexed with DNA or other TLR agonists could be used as effective vaccineadjuvants. To test this hypothesis, the vaccine adjuvant effects of liposomes complexed to TLR agonists were assessed in mice. Wefound that liposomes complexed to nucleic acids (liposome-Ag-nucleic acid complexes; LANAC) were particularly effective ad-juvants for eliciting CD4� and CD8� T cell responses against peptide and protein Ags. Notably, LANAC containing TLR3 orTLR9 agonists effectively cross-primed CD8� T cell responses against even low doses of protein Ags, and this effect was inde-pendent of CD4� T cell help. Ag-specific CD8� T cells elicited by LANAC adjuvants were functionally active and persisted forlong periods of time in tissues. In a therapeutic tumor vaccine model, immunization with the melanoma peptide trp2 and LANACadjuvant controlled the growth of established B16 melanoma tumors. In a prophylactic vaccine model, immunization with theMycobacterium tuberculosis protein ESAT-6 with LANAC adjuvant elicited significant protective immunity against aerosol chal-lenge with virulent M. tuberculosis. These results suggest that certain TLR agonists can be combined with cationic liposomes toproduce uniquely effective vaccine adjuvants capable of eliciting strong T cell responses against protein and peptide Ags. TheJournal of Immunology, 2006, 176: 7335–7345.

S timulation of innate immunity is necessary for generationof effective adaptive immune responses against a varietyof Ags. Pattern recognition receptors, including Toll-like

receptors, play a critical role in the initial activation of key APCsuch as dendritic cells (DC)4 and macrophages (1–4). Althoughmembers of the TLR family all share certain structural and func-tional properties, the signals delivered by various TLRs may elicitqualitatively and quantitatively different immune responses (5).How APCs are activated can therefore have a large impact on thequality and the magnitude of adaptive immune responses. Thus, abetter understanding of the relationship between activation of in-nate and adaptive immunity is of particular importance in the de-sign of new vaccine adjuvants.

The only vaccine adjuvants currently licensed for use in humans(aluminum hydroxide, salts, MF59, and virosomes) augment im-mune responses largely through enhancing Ag delivery (6, 7).Both aluminum hydroxide and MF59 stimulate primarily humoralimmune responses (8). In contrast, other common vaccine adju-vants, such as Freund’s adjuvant (FA), monophosphoryl lipid A(MPL) adjuvant, and CpG oligonucleotides, function primarily asactivators of innate immunity (9). For example, FA and MPL ad-juvant both activate innate immunity via activation of TLR2 orTLR4, whereas the adjuvant effects of CpG oligonucleotides aredependent on TLR9 activation (8–16).

Cationic liposomes have been shown previously to markedlypotentiate activation of innate immunity by TLR9 agonists (bac-terial DNA and CpG oligonucleotides) (17–21). The ability of cat-ionic liposomes to potentiate activation of innate immunity byDNA is due to protection of the DNA from extracellular degrada-tion and to enhanced entry of DNA into the endosomal compart-ment, where TLR9 is selectively expressed (19, 22, 23). The abil-ity of cationic liposomes to promote entry into the cell via theendosomal compartment may also be important to activation byTLR3 agonists such as poly(I:C), inasmuch as TLR3 is also ex-pressed primarily in the endosomal compartment (2–5, 24). Lipo-somes have also been used to introduce protein Ags into the cy-tosol and the MHC class I pathway for generating CD8� T cellresponses (10, 25, 26). However, cationic liposomes alone are rel-atively inert in terms of activating innate immune responses (16,27, 28).

Currently available vaccine adjuvants are generally effective ateliciting Ab responses, but few nonreplicating vaccine adjuvantsare able to generate strong CD8� T cell responses against proteinAgs (8). Immune stimulating complexes are probably the mosteffective adjuvants developed to date for generating cellular im-mune responses, including CD8� T cell responses, against protein

*Department of Microbiology, Immunology, and Pathology and †Department of Pe-diatrics, University of Cincinnati College of Medicine and Cincinnati Children’s Hos-pital, Cincinnati, OH 45229; ‡Integrated Department of Immunology, National JewishMedical and Research Center and the University of Colorado Health Sciences Center,Denver, CO 80206; and §Department of Clinical Sciences, Colorado State University,Ft. Collins, CO 80523

Received for publication September 12, 2005. Accepted for publication March22, 2006.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 This work was supported by Grants CA86224-01 and AI056487-01 from the Na-tional Institutes of Health.2 Address correspondence and reprint requests to Dr. Steven Dow, Department ofMicrobiology, Immunology, and Pathology, College of Veterinary Medicine and Bio-medical Sciences, Colorado State University, Ft. Collins, CO 80523. E-mail address:[email protected] K.Z. and M.J. both contributed equally to this work.4 Abbreviations used in this paper: DC, dendritic cell; FA, Freund’s adjuvant; MPL,monophosphoryl lipid A; LCMV, lymphocytic choriomeningitis virus; LANAC, li-posome-Ag-nucleic acid complex.

The Journal of Immunology

Copyright © 2006 by The American Association of Immunologists, Inc. 0022-1767/06/$02.00

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

Ags (29, 30). Immune stimulating complex adjuvants are thoughtto elicit cell-mediated immunity primarily by enhancing Ag deliv-ery to APC. Use of CpG oligonucleotides as vaccine adjuvants canalso elicit cross-priming when high doses of Ag are administered(14, 31, 32). However, the most efficient and robust CD8� T cellresponses are currently generated by immunization with replicat-ing vaccines, including both viral and bacterial vectored vaccines(7, 8, 11, 33–39).

Therefore, there remains a need for new nonreplicating vaccineadjuvants capable of eliciting strong cellular immune responsesagainst purified Ags. Based on prior observations that cationic li-posomes complexed to CpG oligonucleotides or plasmid DNAcould greatly augment activation of innate immunity, we wonderedwhether liposome-nucleic acid complexes could also be used aseffective vaccine adjuvants. Previous reports have indicated thatcomplexes of liposomes with encapsulated CpG oligonucleotidesor with protamine-DNA complexes could be combined with pro-tein or peptide Ags to elicit effective antitumor immunity in vivo(40–42). In addition, it was reported previously that plasmid DNAand liposome complexes could be used as vaccine adjuvants (43).Therefore, we conducted studies to further investigate the adjuvantproperties of liposome-DNA complexes and to extend the earlierobservations. Our studies have now also included an examinationof the adjuvant properties of other TLR agonists when complexedto liposomes. We also investigated the effectiveness of these ad-juvants in therapeutic and prophylactic vaccination models.

We found that adjuvants consisting of cationic liposomes com-plexed to TLR9 or TLR3 agonists were particularly effective ingenerating strong CD8 and CD4 T cell responses. In contrast, li-posomes complexed to other TLR agonists were much less activeas vaccine adjuvants. Vaccination with liposome-TLR9 agonist ad-juvants elicited functional and long-lived T cells in tissues andgenerated therapeutic antitumor immunity and significant protec-tive immunity against aerosol challenge with Mycobacterium tu-berculosis. Thus, the liposome-TLR agonist adjuvant system de-scribed here represents a novel and potentially clinically effectivenonreplicating vaccine adjuvant system for generating strong cel-lular immune responses against subunit Ags.

Materials and MethodsTLR agonists and Ags

TLR agonists, including LPS, zymosan, R848, poly(I:C), and CpG oligo-nucleotides were purchased from InvivoGen. All TLR agonists exceptR848 were prepared by dissolving at a final concentration of 3 mg/ml insterile PBS, whereas R848 was prepared by dissolving at a final concen-tration of 1 mg/ml in ethanol. Low endotoxin content plasmid DNA wasprepared by Althea Technologies. The plasmid used in these studies(pMB75.6) did not contain a coding gene (44). OVA was purchased fromSigma-Aldrich and was prepared as a 1 mg/ml solution in PBS. The so-lution was filtered through a 10-kDa filter (Centricon; Millipore) before useto remove any small m.w. peptides. The Kb binding peptides OVA8 (OVA)and trp-2 (B16 melanoma) and the I-Ab binding peptide gp 61 from lym-phocytic choriomeningitis virus (LCMV) were synthesized and purified byHPLC by the Macromolecular Resources Department at National JewishMedical and Research Center and were prepared as 1 mg/ml solutions inPBS and sterile filtered before use.

Preparation of cationic liposomes and vaccines

Liposomes were prepared by dissolving the cationic lipid octade-cenolyoxy[ethyl-2-heptadecenyl-3 hydroxyethyl] imidazolinium chloride(Sigma-Aldrich Chemical) and cholesterol (Avanti Polar Lipids) in chlo-roform and adding equimolar concentrations to round-bottom, 15-ml glasstubes to a final concentration of 2 mM. The solution was then dried over-night in a vacuum desiccator to a thin film. The lipids were rehydrated in5% dextrose in water at 50°C for 50 min, followed by incubation for 2 hat room temperature. The liposomes were then extruded through a series of1-, 0.45-, and 0.20-�m filters to form the final liposomes, as descried pre-viously (45). To formulate the vaccines, complexes of liposomes and TLR

agonists were prepared by first diluting the liposomes in 5% dextrose inwater at a concentration of 100 �l of liposomes per 1 ml of dextrosesolution. Next, TLR agonists were added with gentle pipetting to the lipo-some solution at a final concentration of 100 �g of agonist per milliliter ofliposome solution. In the case of complexes prepared with R848, the R848in methanol was first dried down along with lipids, then rehydrated to-gether with the liposomes. Next, peptide or protein Ags were added to thepreformed liposome-TLR agonist complexes and mixed by gentle pipet-ting. Vaccines were prepared at room temperature and administered within30 min of preparation.

Mice and immunizations

Female C57BL/6 mice ages 8–12 wk were used for most experiments. Allmice were purchased from Harlan Sprague Dawley or The Jackson Lab-oratory. CD4�/� mice on the C57BL/6 background were provided by E.Gelfand and P. Marrack (National Jewish Medical and Research Center,Denver, CO). Female MyD88�/� mice on a C57BL/6 � 129 were pro-vided by R. Titus (Colorado State University (CSU), Fort Collins, CO).Control female wild-type control mice on the C57BL/6 � 129 backgroundwere purchased from The Jackson Laboratory. Protocols for the animalexperiments described in this study were approved by Institutional AnimalCare and Use Committees at National Jewish Medical and Research Centerand at Colorado State University. For most experiments, mice were im-munized with 5 �g of peptide or protein Ag, based on prior dose-titrationexperiments. Mice were immunized with 100 �l of liposome-TLR agonistvaccine for i.v. immunization and 200 �l for s.c., i.p., and i.m. immuni-zation. For most experiments, mice were immunized once and then boosted7–10 days later, and cells were collected for analysis on days 4–5 after theboost. In some experiments, mice were primed and boosted then left un-treated for several weeks to months before T cells were analyzed.

Immune responses to immunization with conventional vaccines werealso assessed. One group of mice was immunized s.c. with 5 �g of OVA8peptide in CFA (Sigma-Aldrich) and boosted 1 wk later with peptide inIFA. Other mice were immunized with peptide-pulsed or protein-pulsedDC. Briefly, bone-marrow cells were propagated in high levels of GM-CSFfor 6 days, and the nonadherent cells were collected (�85% CD11c andMHC class IIhigh) and pulsed overnight with 10 �M OVA8 peptide (or 10�g/ml OVA protein), then activated with 1 �g/ml LPS for the last 4 h (46).Mice were each immunized with 1 � 106 Ag-pulsed DC by either the s.c.or i.p. routes, then boosted 7 days later, and spleen cells were analyzed byflow cytometry 5–7 days after that. Another group of mice was immunizedi.v. with 109 PFU of vv-OVA (vaccinia virus encoding full-length OVAprotein), then spleen cells were analyzed 7 days later by flow cytometry.Mice were also immunized with plasmid DNA encoding the OVA protein.For plasmid DNA vaccination, mice each received 100 �g of OVA plasmidDNA injected at several sites i.m. in the cranial tibialis muscles bilaterallyand were then boosted 2 wk later, and T cells were analyzed 1 wk after theboost.

Cell preparation

Single-cell suspensions of spleen cells were prepared by mechanical dis-ruption and screening through a 100-�m nylon mesh screen (BD Bio-sciences), followed by NH4Cl lysis. Lymph node cells were prepared bymechanical screening through a 100-�m nylon mesh screen. Lung and liverlymphocytes were isolated by first mincing the tissues, then digesting in asolution of 2 mg/ml collagenase (type IA; Sigma-Aldrich) plus soybeantrypsin inhibitor (100 �g/ml) and DNase (500 IU/ml) for 1 h at 37°C,followed by mechanical disruption through an 18-gauge needle, as de-scribed previously (47). The cell suspensions were further purified byNH4Cl lysis of RBC and then screening through 70- and 40-�m screens(BD Biosciences). Cells were resuspended in complete medium (MEMEwith essential and nonessential amino acids, penicillin, and streptomycin,and 10% FBS (Invitrogen Life Technologies) before analysis.

Abs and flow cytometric analysis

Directly conjugated Abs used for flow cytometric analysis were purchasedfrom either BD Pharmingen or eBioscience. The following Abs were used:anti-CD8a (APC; clone 53.6.7), anti-CD4 (APC; clone RM4-5), anti-CD44(FITC; clone IM7), anti-CD62L (PE/Cy5; clone Mel-14), anti-CD69 (PE/cy7; clone H1.2F3), anti-I-A/I-E (MHC class II, biotin, or PE; clone M5/114.15.2), followed by either SA-pe/cy5 or SA-Alexa-350 (MolecularProbes), anti-CD11b (PE-Cy5 or APC-Cy7; clone M1/70), anti-CD11c (PEor APC; clone N418), anti-Gr-1 (PeCy7; clone RB6-8C5), B220 (APC-Cy7; clone RA3-6B2), anti-NK1.1-biotin (clone PK 136) or anti-CD3(APC-Cy7; clone 145-2C11). Nonspecific binding of Abs was blocked bypreincubation of cells in normal mouse serum with 40% supernatant fromrat anti-FcRIII hybridoma 24.G2, plus 0.2 �g/ml human IgG. Abs were

7336 TLR AGONIST LIPOSOME ADJUVANTS

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

diluted in FACS buffer (PBS with 2% FBS and 0.1% sodium azide) forstaining. Staining (except for tetramers) was done at 4°C for 20 min, fol-lowed by washing in FACS buffer. In most cases, cells were fixed in 1%paraformaldehyde for 30 min and stored in FACS buffer at 4°C beforeanalysis. Flow cytometry was performed using either a BD FACSCaliburcytometer (BD Biosciences) for four-color analysis or a Cyan MLE cy-tometer (DakoCytomation) for six- to seven-color analysis.

For analysis of CD8� and CD4� T cell responses, gates were drawn toinclude live lymphocytes, based on forward- and side-scatter characteris-tics of spleen cells. For analysis of tetramer-positive cells, the total CD8�

or CD4� T cell population was first gated, and then the percentage oftetramer-positive cells as a percentage of total CD8� or CD4� T cells wascalculated. A minimum of 200,000 total events were collected for tetramerexperiments. Data analysis was done using either CellQuest Software (BDBiosciences) or Summit Software (DakoCytomation).

MHC-peptide tetramers

A soluble MHC Kb tetramer was prepared as described previously (48).Tetramers were loaded with the OVA8 peptide SIINFEKL or the trp-2,peptide SVYDFFVWL. Single-cell suspensions from spleen, lymph nodes,lung, or liver (5 � 105 to 1 � 106 cells in 100 �l of complete medium)were incubated with tetramer at 37°C for 1–1.5 h. Splenocytes from OT-1mice (OVA8-specific TCR transgenic mice; provided by T. Potter, Na-tional Jewish Medical and Research Center, Denver, CO) were used aspositive controls for tetramer staining. Negative controls included spleensfrom naive mice and use of tetramers prepared with irrelevant peptides. Foranalysis of Ag-specific CD4 T cells, MHC class II tetramers were producedwith a covalent linkage to the gp61–80 peptide of LCMV (I-Ab-gp61) asdescribed previously (49). After incubation with tetramers, cells werewashed and stained with Abs to surface determinants, then washed andstained with appropriate streptavidin conjugates for 15 min, then washedand resuspended in FACS buffer or fixed in 1% paraformaldehyde beforeanalysis.

Tracking uptake of fluorescent liposome-DNA complexes in vivo

To track cellular uptake of liposome-DNA complexes by APCs in vivo,liposomes were labeled with the fluorescent dye BODIPY and used toprepare complexes. The labeled liposome-DNA complexes were injectedby the i.v., s.c., and i.p. routes, and, at various time points postinjection,spleen cells, peritoneal cells, and relevant draining lymph node cells werecollected, and single-cell suspensions were prepared. In the case of analysisof spleen and lymph node APC, the cell suspensions were also digested in1 mg/ml collagenase for 20 min at room temperature to release DC. Cellswere then immunostained with Abs to cell surface Ags and analyzed byflow cytometry. Analysis gates were set on all live cells, including thosewith high side-scatter and forward-scatter characteristics to enumerate allcell types that might have bound the labeled liposomes.

CD8� T cell assays

Cytotoxic activity of CD8� T cells elicited by vaccination was assessed byin vivo assay, as described previously (48). Briefly, equal numbers of au-tologous spleen cells from naive donor mice were pulsed for 2 h at 37°Cin vitro with 10 �M OVA8 peptide, or were incubated but unpulsed. Afterpeptide pulsing and washing, the peptide-pulsed and unpulsed spleen cellpopulations were loaded with either 50 �g/ml CFSE (Molecular Probes) or10 �g/ml CSFE for 15 min, respectively. Immediately before tail veininjection, equal numbers of the two spleen cell populations were mixed.Recipient vaccinate and control mice were each injected with equal num-bers (1.5 � 107 spleen cells/mouse) of the two populations of CFSE-la-beled spleen cells. The injected mice were sacrificed 18 h later, and single-cell suspensions were prepared from the spleen. The cells were evaluatedby flow cytometry, and the number of CFSEhigh and CSFElow populationswas determined. The ratio of CFSElow:CFSEhigh cells was calculated todetermine the amount of in vivo lysis of peptide-pulsed target cells.

Cytokine assays

Cytokine production by CD8� spleen cells and lung cells from immunizedmice was assessed by determination of Ag-specific cytokine release intosupernatants and by intracellular cytokine staining. Single-cell suspensionsof spleen cells were cultured overnight at a concentration of 2 � 106 incomplete medium in the presence or absence of 1 �M OVA8 peptide.Supernatants were collected 18 h later and assayed for release of IFN-�,using a commercial ELISA (R&D Systems). Controls included spleen cellsfrom nonimmunized mice and wells incubated without peptides. For in-tracellular cytokine assay, spleen cells were placed in complete mediumwith 10 �g/ml Brefeldin A, then stimulated with 10 �M OVA8 peptide for

5 h at 37°C. Cells were then stained for cell surface determinants, thenfixed and permeabilized for detection of intracellular IFN-�, according tomanufacturer’s directions (BD Pharmingen). Controls included cells incu-bated without peptide and cells from nonimmune mice incubated with andwithout OVA8 peptide.

B16 tumor model

Mice (5/group) were injected s.c. with 1 � 105 B16.F10 cells (provided byI. Fidler, MD Anderson Cancer Center, Houston, TX) and vaccination wasinitiated 7 days later, at which time most mice had palpable tumors. Micewere treated by injection of 5 �g of trp-2 peptide in liposome-plasmidDNA complexes, given by the s.c. or i.p. routes. Treatments were contin-ued weekly for 4 wk or until the mice were euthanized. Tumor growth wasmonitored by calipers every 2–3 days, and the tumor surface area wasdetermined by multiplying two perpendicular tumor dimensions.

M. tuberculosis aerosol protection model

Mice (5 animals/treatment group) were immunized with 10 �g of recom-binant ESAT-6 protein in liposome-DNA complexes by the i.p. or s.c.routes. Another group of mice was immunized ESAT-6 protein in MPLadjuvant (Corixa). Recombinant ESAT-6 Ag was prepared as part of theNational Institutes of Health tuberculosis vaccine contract at CSU. Micewere primed, then boosted at 2 and at 4 wk. Negative control groups in-cluded unvaccinated mice and mice given liposome-DNA complexes with-out Ag, also by the i.p. route. Mice were challenged by aerosolization of100 CFU of M. tuberculosis, Erdman strain, 4 wk after the last vaccination.Thirty days after aerosol challenge, mice were sacrificed, and the numberof viable organisms in lung tissues was determined by homogenizing lungtissues and plating serial dilutions of the homogenates on Middlebrookagar plates. The log10 protective titer was calculated by subtracting theobserved titer from the titer obtained in mice treated with saline only.

Statistical analyses

In experiments with multiple groups of mice, statistical differences be-tween treatment groups were compared using ANOVA and Tukey’s mul-tiple means comparisons test. For comparisons between two treatmentgroups, Student’s t test was used. Statistical analyses were done usingGraphPad software. A p value �0.05 was considered statistically signifi-cant for these analyses.

ResultsImmunization with liposome-Ag-nucleic acid complexes(LANAC) elicits strong CD8� and CD4� T cell responses

The ability of liposome-DNA complexes to function as vaccineadjuvants and elicit CD8� or CD4� T cell responses was firstassessed by immunizing mice with OVA8 or gp61 peptides, re-spectively. The vaccines were formulated using cationic liposomescomplexed to noncoding plasmid DNA, which we demonstratedpreviously could elicit marked activation of innate immunity (20,21). The complexes used in these studies also had a net positivecharge, which likely facilitated binding of protein or peptide Agsto the complexes. Mice were immunized with LANAC formulatedwith noncoding plasmid DNA and with low doses (typically 5 �g)of Ag. Dose ranging studies revealed that strong T cell responsescould be elicited when mice were immunized with LANAC con-taining as little as 1 �g of peptide per mouse, whereas doses above10 �g per mouse did not increase the efficiency of vaccine re-sponses (data not shown). Mice were immunized by the s.c., i.m.,i.p., and i.v. routes to assess the efficiency of vaccination by dif-ferent routes. Mice were immunized twice, 7–10 days apart, andcells were analyzed 5 days after the last immunization using Kb-OVA8 or I-Ab-gp61 tetramers.

Immunization i.p. with 5 �g of OVA8 peptide in LANAC elic-ited a significant ( p � 0.01) increase in the percentage of OVA8-specific CD8� T cells in spleens of immunized mice when com-pared with control mice (Fig. 1A). For example, immunizationwith OVA8 peptide in LANAC generated an average of 5.4% ofOVA8-specific CD8� T cells in the spleen of mice. Immunizationby the i.p. route was the most efficient route of immunization,although immunization by the s.c. or i.m. routes also elicited large

7337The Journal of Immunology

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

numbers of OVA8-specific CD8� T cells, whereas smaller num-bers were elicited by vaccination by the i.v. route (data not shown).Large increases in the numbers of OVA8-specific CD8� T cellswere also observed in peripheral lymph nodes of immunized mice(data not shown). Immunization with liposome-DNA complexesdid not elicit OVA8-specific CD8� T cell responses. Immuniza-tion with peptide in LANAC also appeared to be more efficientthan immunization with peptide in FA or immunization with pep-tide-pulsed DC (Fig. 1B), although direct comparisons could notbe made until the other immunization strategies were carefullyoptimized. Thus, the LANAC adjuvant was very effective for elic-iting CD8� T cell responses against a model peptide Ag.

The ability of LANAC adjuvant to elicit CD4� T cell responseswas assessed next. Mice were immunized twice i.p. with 5 �g ofthe MHC class II-binding LCMV peptide gp61, and CD4� T cellresponses were assessed using I-Ab-gp61 tetramers. Mice immu-nized with the gp61 peptide in LANAC developed a significant( p � 0.001) pronounced Ag-specific CD4 T cell response com-pared with control mice (Fig. 1B). In fact, the magnitude of theCD4� T cell response elicited by gp61-LANAC vaccination ex-ceeded that elicited by infection with live LCMV at the height ofthe immune response (Fig. 1D). Thus, the LANAC adjuvant wasalso very effective for generating CD4� T cell responses againstpeptide Ags.

LANAC vaccines elicit effective cross-priming

One of the major drawbacks to immunization with nonreplicatingvaccines and recombinant Ags is the relative inability of thesevaccines to elicit CD8� T cell responses. Therefore, the ability ofthe LANAC vaccine to generate CD8� T cell responses againstprotein Ags was assessed. In these experiments, C57BL/6 micewere immunized with 5 �g of intact OVA protein, using the sameimmunization protocol described above, and CD8� T cell re-sponses were quantitated using Kb-OVA8 tetramers. Remarkably,the LANAC adjuvant was extremely effective in eliciting OVA8-specific CD8� T cell responses following immunization withwhole OVA (Fig. 2). In fact, CD8� T cell responses to immuni-zation with OVA in LANAC consistently exceeded those elicitedby immunization with equivalent doses (by weight) of OVA8 pep-tide. For example, mice immunized with OVA generated an aver-age of 10.2% OVA8-specific CD8� T cells per total CD8� spleencells, compared with 5.4% OVA8-specific CD8� T cells per totalCD8� in mice immunized with OVA8 peptide in LANAC. Im-munization with LANAC also appeared to be more efficient incross-priming CD8� T cell responses than immunization withOVA-pulsed DC, with a viral vectored vaccine (vv encoding full-length OVA), or with plasmid DNA encoding full-length OVA(Fig. 2).

FIGURE 1. CD8� and CD4� T cell responses to immunization with LANAC. Experiments were conducted to assess the ability of vaccines preparedwith LANAC adjuvants to elicit T cell responses. A, Mice (C57BL/6; 4/group) were each immunized i.p. twice with LANAC containing noncoding plasmidDNA and 5 �g of OVA8 peptide Ag. Five days after the second immunization, CD8� T cell responses in spleen were quantitated using Kb-OVA8 tetramers,as described in Materials and Methods. Briefly, total CD8� T cells were gated for analysis (after excluding MHC class II� cells), and the percentage ofKb-OVA8� cells was plotted vs CD44� expression. A, Representative FACS plot of OVA8-specific T cells elicited by vaccination with peptide in LANACadjuvant is shown. This result is representative of �6 independent experiments. B, CD8� T cell responses to peptide vaccination with LANAC wascompared with those elicited by conventional peptide vaccines, including immunization with peptide in FA and immunization with peptide-pulsed bonemarrow-DC, as described in Materials and Methods. The mean (�SD) OVA8-specific CD8� T cells responses were calculated following a prime and boostimmunization (4 animals/group) and plotted. As a control, another group of mice was immunized with liposome-nucleic acid complexes without peptideAg (LNAC). Immunization with LANAC elicited significantly greater CD8� T cell responses (p � 0.01) than control mice. These results were pooled fromtwo independent experiments. C, CD4� T cell responses were quantitated in C57BL/6 mice (4/group) immunized twice with 5 �g of LCMV gp61 peptidewith LANAC adjuvant, using IAb-gp61 tetramer. A representative FACS plot is presented for a control mouse and a mouse immunized with gp61. D, Themean gp61-specific CD4� T cell response (�SD) in the spleen of mice (4/group) immunized with gp61 peptide in LANAC was compared with that elicitedby infection with live LCMV. In LCMV-infected mice, the T cells were analyzed at the predicted peak of the T cell response to live virus infection (7 dayspostinoculation). Immunization with gp61 LANAC elicited a significantly greater (p � 0.05) CD4� T cell response than LCMV infection, and both gp61vaccination and live virus infection elicited significantly greater CD4 responses than observed in control mice. Similar results were obtained in oneadditional experiment. �, Denotes significant differences (p � 0.05) when LANAC-vaccinated mice were compared with control mice, as determined byANOVA, followed by Tukey multiple means comparison, whereas �� denotes significant differences (p � 0.05) between gp61 LANAC-vaccinated miceand LCMV-infected mice.

7338 TLR AGONIST LIPOSOME ADJUVANTS

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

Generation of CD8� T cell responses by LANAC vaccines isoptimal when the three vaccine components are physicallyassociated

Studies were conducted next to elucidate the contribution of theindividual components of the LANAC adjuvant to overall vaccineefficacy. Mice were therefore immunized with OVA plus lipo-somes only or OVA plus plasmid DNA only, and the CD8� T cellresponses were compared with those elicited by immunizationwith all three components of the complex (OVA plus liposomesplus noncoding plasmid DNA). Mice immunized with OVA plusliposomes alone or OVA plus plasmid DNA alone failed to gen-erate significant CD8� T cell responses against the OVA8 peptide,whereas mice immunized with the three-part complex of lipo-somes, DNA, and OVA (LANAC) generated strong CD8� T cellresponses (Fig. 3A).

To further evaluate the need for physical association of the com-ponents of the LANAC vaccine for optimal immunization, we per-formed experiments where the Ag was administered separatelyfrom the liposome-DNA complexes. When mice were injected i.p.with 5 �g of OVA, followed 5 min later by i.p. injection of lipo-some-DNA complexes, we found that OVA8-specific CD8� T cellresponses were elicited, but with only �50% the efficiency withwhich LANAC elicited CD8� T cell responses (data not shown).

When the order of injection was reversed and liposome-DNA com-plexes were injected first, followed by injection of OVA, the mag-nitude of the CD8� T cell response was only �10% of that fol-lowing immunization with intact LANAC (data not shown).Immunization with OVA administered s.c., followed by adminis-tration of liposome-DNA complexes i.p. did not elicit detectableCD8� T cell responses (data not shown).

Therefore, it appeared that physical association of all three com-ponents of the vaccine was required for optimal efficiency of cross-priming and generation of CD8� T cell responses. Similarly, thethree-part liposome-DNA-Ag complex was also required for opti-mal generation of CD8� T cell responses against peptide Ags (datanot shown). Using labeled OVA and labeled liposomes and fluo-rescence resonance energy transfer analysis, we also observed thatthe OVA protein was in fact physically associated with the lipo-some-DNA complexes (T. Anchordoquy, unpublished data). Theseresults suggest that it is likely that the three components of theLANAC vaccine may physically enter the same APC to elicit Tcell responses.

Adjuvant activity of LANAC is dependent on signaling via aMyD88-dependent pathway

Inclusion of TLR3 or TLR9 agonists with charged liposomes wasable to elicit optimal induction of CD8� T cell responses in ourexperiments. Therefore, to determine whether activation of innateimmunity was critical to the activity of LANAC adjuvants formu-lated with plasmid DNA, mice lacking the MyD88 adaptor protein(MyD88 �/�) were immunized with OVA and LANAC, andCD8� T cell responses were compared with those elicited in wild-type C57BL/6 mice. We observed that Ag-specific CD8� T cellresponses were almost completely abrogated in vaccinatedMyD88�/� mice, compared with responses in wild-type mice (Fig.3B). For example, the mean percentage of OVA8-specific CD8� Tcells in immunized MyD88 �/� mice was only 0.37% (�0.54%)following immunization with LANAC, whereas immunized wild-type mice on the same background (C57BL/6 � 129) mountedCD8� T cell responses equivalent to those of C57BL/6 mice (datanot shown). Thus, activation of innate immunity via a MyD88-dependent pathway was critical to the ability of the LANAC ad-juvant to elicit efficient CD8� T cell responses.

Cross-priming by LANAC is independent of CD4 help

We observed that immunization with OVA protein consistentlyelicited stronger CD8� T cell responses than immunization withthe same amount of peptide (see Figs. 1 and 2). The CD8� T cellresponses to immunization with protein Ag may therefore havebeen augmented by CD4 T cell help provided by MHC class IIepitopes contained within the full-length OVA protein. To assessthis possibility, CD4�/� mice on the C57BL/6 background wereimmunized with OVA and LANAC, and responses were comparedwith wild-type mice. We found that the magnitude of CD8� T cellresponses to LANAC vaccination in CD4�/� mice were not di-minished compared with responses in wild-type animals (Fig. 3C).Therefore, the ability of LANAC vaccines to cross-prime CD8� Tcell responses appeared to be independent of CD4 T cell help,although the long-term functionality of CD8� T cells elicited inCD4�/� mice was not assessed in these experiments.

Liposome-DNA complexes traffic to draining lymph nodesprimarily in CD11b� and Gr-1� cells

Targeting of vaccines to specific populations of APC can have asubstantial impact on the subsequent T cell responses that are gen-erated. Therefore, we assessed uptake of LANAC complexes byAPC in draining lymph nodes shortly after immunization. For

FIGURE 2. Comparison of cross-priming efficiency elicited by LANACvaccines and conventional vaccines. Mice (4/group) were immunized twicei.p. with 5 �g of OVA in LANAC adjuvant, then spleen cells were col-lected and stained with Kb-OVA8 tetramer, followed by staining for cellsurface determinants, as described in Materials and Methods. A, Repre-sentative FACS plot of spleen cells from a control mouse (left panel) anda mouse immunized twice with LANAC containing 5 �g of OVA (rightpanel). Total CD8� T cells were gated for analysis (after excluding MHCclass II� cells), and Kb-OVA8� cells were plotted vs CD44� expression bytotal CD8� T cells. This result is representative of �6 independent exper-iments done with OVA protein. B, Comparison of the mean percentage(�SD) of Kb-OVA8�CD8� T cells in nonimmunized control mice (4/group), as compared with mice immunized twice with bone marrow DCpulsed overnight with OVA (DC), with plasmid DNA encoding the full-length OVA cDNA (DNA), with vv-OVA, or with 5 �g of OVA inLANAC adjuvant, as described in Materials and Methods. This data waspooled from two separate experiments. �, Denotes values significantly dif-ferent (p � 0.05) when LANAC-immunized mice were compared withcontrol mice, as assessed by ANOVA and Tukey multiple meanscomparison.

7339The Journal of Immunology

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

these experiments, liposomes labeled with the fluorescent dyeBODIPY were used along with flow cytometry to track the distri-bution of LANAC complexes. Within 1 h of i.p. injection of la-beled LANAC, virtually all cells in the peritoneal cavity containedBODIPY� liposomes (data not shown). By 4 h after injection,BODIPY� cells could also be detected by flow cytometry in cellsin the mediastinal lymph nodes, but not in cells in the spleen orother lymph nodes (Fig. 4). The majority of labeled LANAC in themediastinal lymph node were present in cells that were CD11b� orGr-1� (Fig. 4B). Most (90%) BODIPY� Gr-1� cells coexpressedCD11b� (consistent with neutrophils or immature monocytes),whereas 40% of BODIY� CD11b� cells did not express Gr-1�

(most consistent with macrophages). In contrast, labeled com-plexes were relatively rare in F4/80� or CD11c� cells. These re-sults suggest that early after immunization, LANAC were taken todraining lymph nodes primarily within macrophages and neutro-phils and possibly monocytes, rather than within classical DC.However, most of the BODIPY�/CD11b�/Gr-1� macrophageswere also F4/80�, indicating that they were probably not derivedfrom resident peritoneal macrophages, which were strongly F4/80� (data not shown). The second population of CD11b�/Gr-1�

cells was most likely comprised of both neutrophils and inflam-matory monocytes (50). The presence of neutrophils in the drain-ing lymph nodes was also confirmed by cytologic examination(data not shown). The Gr-1�/CD11b� cells observed here resem-bled in some respects a recently described population of uniqueAPC present in the peritoneal cavity that are associated with Agpresentation following vaccination with aluminum hydroxide ad-juvant (51).

Additional experiments were done using BODIPY-labeled lipo-somes complexed to other TLR agonists, including poly(I:C), zy-mosan, and LPS, to assess uptake by APC in draining lymphnodes. Important differences in cell uptake were not observedwhen uptake of labeled complexes with other TLR agonists wascompared with that observed using plasmid DNA (data not

shown). Therefore, the initial uptake of LANAC appeared to bemediated primarily by the liposome component, because alteringthe TLR agonist did not affect cellular uptake. These results sug-gested that macrophages and possibly inflammatory monocytes,rather than classical DC, were the primary APC responsible fortransport of LANAC to lymph nodes soon after immunization.However, these observations do not exclude the possibility thatAgs were transferred later to other APC in the lymph nodes, or thatthe phenotype of the earliest APC transporting LANAC to lymphnodes changed over time to resemble cells more consistent withtypical DC, as has been reported previously with inflammatorymonocytes (50).

T cells elicited by LANAC vaccination are functionally activeand long-lived

Although some vaccines can generate large numbers of Ag-spe-cific T cells, in some cases these T cells are nonfunctional (52).Therefore, we conducted experiments to assess the functionality ofCD8� T cells elicited by LANAC immunization. In an in vitro51Cr-release assay using effector cells derived from spleens of im-munized mice and restimulated in vitro, high levels of peptide-specific killing were observed (data not shown). An in vivo CD8�

T cell assay was also used to directly assess cytotoxic activity.Following immunization with OVA in LANAC, very high levelsof specific CD8� T cell activity were observed in spleen cells (Fig.5A). The mean R value (ratio of nonpulsed to peptide-pulsed targetcells) for LANAC-immunized mice was 52.3 (�3.7), comparedwith 1.2 (�0.1) for nonimmunized control mice. Thus, LANACimmunization elicited CD8� T cells with functional cytolyticactivity.

The ability to generate IFN-� was also assessed in CD8� T cellselicited by LANAC adjuvant. After in vitro restimulation, spleencells from mice immunized with OVA-vaccinated mice releasedlarge quantities of IFN-� into the supernatant following overnight

FIGURE 3. Optimal cross-priming by LANAC adjuvant requires physical association of liposome-TLR agonist complexes, requires MyD88 signaling,and is independent of CD4� T cells. A, To determine whether liposomes alone or plasmid DNA alone were sufficient to cross-prime CD8� T cell responses,mice were immunized twice i.p. with 5 �g of OVA plus liposomes only (liposome), OVA plus plasmid DNA only (DNA), or with the three-part complexof liposomes, plasmid DNA, and OVA (LANAC). OVA8-specific CD8� T cell responses in spleen were quantitated using Kb-OVA8 tetramers. For theseanalyses, total CD8� T cells were gated for analysis (after excluding MHC class II� cells), the percentage of Kb-OVA8� cells was plotted vs CD44�, andrepresentative FACS plots are presented. B, To assess the role of activation of innate immunity in generating CD8� T cell responses to LANACimmunization, MyD88 �/� and wild-type mice (3/group) were immunized twice with 5 �g of OVA with LANAC adjuvant, CD8� T cell responses in thespleen were assessed using tetramers, and representative FACS data is presented. C, The role of CD4 T cells in generating CD8� T cell responses toLANAC immunization was assessed using CD4�/� mice bred on a C57BL/6 background (3 mice/group) and comparing their CD8� T cell responses tothose elicited in wild-type C57BL/6 mice. Mice were immunized twice i.p. with 5 �g of OVA in LANAC, and OVA8-specific CD8� T cell responses wereassessed using Kb-OVA8 tetramer staining. The mean (�SD) percentage of CD8� Kb-OVA8� T cells from wild-type and CD4�/� mice was plotted. �,Denotes significant differences (p � 0.01) between unvaccinated control mice and wild-type and CD4�/� mice, as assessed by ANOVA and Tukey multiplemeans comparison test.

7340 TLR AGONIST LIPOSOME ADJUVANTS

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

stimulation with 1 �M OVA8 peptide. However, when intracel-lular cytokine production was assessed, only a relatively smallpercentage (2–3%) of the total OVA8-specific CD8� T cell pop-ulation was found to produce IFN-� (Fig. 5B). It is also possiblethat the LANAC vaccine elicited IFN-� production by other cellsin addition to CD8� T cells. For example, we have shown previ-ously that liposome-DNA complexes are capable of stimulatingrelease of high levels of IFN-� by NK cells in vivo (20).

Immunization with LANAC adjuvant elicits large numbers oflong-lived memory T cells in tissues

Acute viral infections typically elicit large expansions of CD8� Tcells, but these T cells may be relatively short-lived (53, 54).Therefore, experiments were done to assess the survival and tissuedistribution of Ag-specific CD8� T cells elicited by immunizationwith LANAC. CD8� T cell responses in lung and spleen tissueswere analyzed 5 days and 3 mo after immunization with 5 �g ofOVA in LANAC. Large numbers of OVA8-specific CD8� T cellswere present in the lungs at 3 mo, although the percentage of

Ag-specific CD8� T cells had declined by �50% during this timeperiod (from 66% of total CD8� T cells to 32% of total CD8� Tcells; Fig. 5C). The phenotype of these Ag-specific CD8� T cells(CD44high, CD62Llow) was consistent with that of memory effectorCD8� T cells, which have been described as residing for longperiods of time in tissues following systemic viral infection (53).In contrast, OVA8-specific CD8� T cells were relatively rare inthe spleen at 3 mo (data not shown). Thus, LANAC vaccinationappeared to elicit large numbers of long-lived memory effectorCD8� T cells.

Adjuvants prepared with TLR9 or TLR3 agonists effectivelycross-prime CD8� T cell responses

The preceding results indicated that a TLR9 agonist (plasmidDNA) could elicit efficient cross-priming when combined with cat-ionic liposomes. To determine whether other TLR agonists com-plexed to liposomes were capable of eliciting similar responses, aseries of liposome-TLR agonist complexes were prepared and as-sessed for their ability to elicit CD8� T cell responses followingimmunization with 5 �g of OVA. The liposome-TLR agonist com-plexes evaluated included zymosan (TLR2), poly(I:C) (TLR3),LPS (TLR4), R848 (TLR7/8), and CpG oligonucleotides or plas-mid DNA (TLR9). The adjuvants were all prepared using the samerelative amounts of cationic liposome and TLR agonist. Mice wereimmunized twice i.p., and spleen, lung, and liver cells were ana-lyzed by tetramers to quantitate CD8� T cell responses. We foundthat only adjuvants prepared using TLR9 agonists (plasmid DNA,CpG oligonucleotides) or TLR3 agonists (poly(I:C)) were able toefficiently generate strong Ag-specific CD8� T cell responses (Fig.6). Each of these three adjuvants elicited large numbers of Ag-specific CD8� T cells in spleen, liver, and lung tissues followingimmunization with OVA. Of the other adjuvants evaluated, onlyliposome-zymosan complexes elicited a significant increase ( p �0.05) in OVA8-specific CD8� T cells, although the response wasstill much less than that elicited by CpG-, DNA-, and poly(I:C)-containing adjuvants. Thus, TLR3 and TLR9 agonists appeared tobe uniquely effective as vaccine adjuvants when formulated withliposomes.

Several recent reports indicate that the ability to cross-primeCD8� T cell responses is linked to induction of type I IFN pro-duction (48, 55, 56). Therefore, to determine whether a similarassociation was true with LANAC-based vaccines, we examinedthe ability of different liposome-TLR agonist complexes to induceproduction of IFN-� in vivo. We found that only liposome-TLRagonist adjuvants that contained TLR9 or TLR3 agonist elicitedsubstantial production of IFN-� in vivo (Ref. 21 and data notshown). These results are consistent with previous reports and sug-gest that the ability of vaccine adjuvants comprised of liposomesand TLR3 or TLR9 agonists to cross-prime CD8� T cell responsesmay depend in part on induction of type I IFNs.

LANAC vaccination against trp2 elicits antitumor activity inmice with established melanoma

Generation of T cell responses against shared tumor Ags such asthe endogenous melanoma Ag (trp2) is typically very difficult inmice with established tumors (57). Therefore, we assessed the abil-ity of LANAC vaccines to elicit therapeutic antitumor activity inmice with established B16 melanomas. The melanoma Ag (trp2)was used as the target for immunization. Mice with day 7 estab-lished tumors were vaccinated with the trp2 melanoma Ag weeklyfor three immunizations beginning on day 7, by either the s.c. ori.p. routes of immunization. Tumor responses were quantitated byserial tumor measurements. Vaccination with trp2 peptide (5 �g ofpeptide per mouse per immunization) in LANAC administered by

FIGURE 4. Uptake of LANAC complexes by APC in draining lymphnodes. To track the cellular uptake of LANAC by APC, complexes wereprepared using fluorescent, BODIPY-labeled liposomes to allow their de-tection by flow cytometry. Mice (3/group) were injected i.p. with labeledLANAC, and 4 h later spleen and draining lymph node tissues were col-lected for analysis of APC populations using multicolor flow cytometry, asdescribed in Materials and Methods. A, Representative FACS plot of me-diastinal lymph node cells analyzed by flow cytometry for uptake of fluo-rescently labeled LANAC by CD11b� cells (left panels) or Gr-1� cells(right panels) from an untreated control mouse (top panels) and a LANAC-immunized mouse (bottom panels), 4 h after i.p. injection. B, The mean(�SD) BODIPY� cells in different APC populations in mediastinal lymphnodes following i.p. injection of labeled LANAC was calculated and plot-ted for CD11b�, Gr-1�, B220�, F4/80�, and CD11c� cells. Similar resultswere obtained in three additional experiments.

7341The Journal of Immunology

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

the i.p. route resulted in a significant reduction in the rate of tumorgrowth, compared with control mice (Fig. 7). By comparison, im-munization of mice with trp2-pulsed DC or with liposomes aloneplus trp2 peptide did not elicit significant antitumor activity.Spleen cells from trp-2-vaccinated mice restimulated in vitro withtrp2 peptide generated cytotoxic activity and also produced IFN-�in a trp2-specific manner (data not shown). Thus, immunizationagainst a shared tumor Ag using LANAC adjuvant was capable ofeliciting therapeutic antitumor activity.

LANAC vaccination elicits protective immunity against aerosolchallenge with M. tuberculosis

The ability of LANAC vaccines to elicit protective immunity wasalso assessed in an aerosol M. tuberculosis challenge model. Forthese studies, the ESAT-6 Ag of M. tuberculosis was used, becausethis Ag has been shown previously to elicit protective immunityagainst M. tuberculosis (58–60). Mice were each immunizedtwice by the s.c. or i.p. routes with 10 �g of recombinant ESAT-6protein formulated in LANAC. The liposomal MPL adjuvant hasbeen used previously to successfully vaccinate mice against M.tuberculosis, so another group of mice was immunized withESAT-6 protein in the MPL adjuvant. Controls included mice in-jected with saline only. Mice were subjected to aerosol challenge

with virulent M. tuberculosis (Erdman strain) 3 wk after the lastvaccine, and titers in lung tissues were determined 30 days afterchallenge. We found that immunization with ESAT-6 in LANACelicited significant protection ( p � 0.05) from challenge, com-pared with nonvaccinated control mice and mice immunized withESAT-6 in MPL adjuvant (Fig. 8). Thus, immunization withLANAC adjuvant and a recombinant protein Ag was able to gen-erate significant protective immunity in a rigorous tuberculosisaerosol infection model.

DiscussionThe major findings to emerge from these studies are that 1) certainTLR agonists can be combined with cationic liposomes and Ags toproduce very potent vaccines capable of eliciting both CD4 andCD8 T cell responses; 2) liposomes complexed to TLR3 and TLR9agonists are uniquely effective at cross-priming CD8� T cell re-sponses in vivo; 3) the full activity of liposome-TLR agonist ad-juvants requires their physical association with the Ag; and 4)liposome-TLR agonist adjuvants can be used to generate effectivetherapeutic antitumor immunity and protective immunity againstaerosol challenge with M. tuberculosis. These results suggests thatcoupling vaccine delivery using liposomes with activation of in-nate immunity using specific TLR agonists represents an effective

FIGURE 5. Immunization with LANAC adjuvant elicits functional CD8� T cells. Experiments were conducted to assess the functionality and long-termsurvival of CD8� T cells elicited by vaccination with LANAC. A, CD8� T cell cytolytic activity was assessed in vivo using adoptive transfer ofpeptide-pulsed and unpulsed CFSE-labeled spleen cells into naive control mice or mice vaccinated twice with OVA in LANAC (3 mice/group), as describedin Materials and Methods. Eighteen hours later, spleen cells were harvested from the adoptively transferred mice, and the relative proportions of the twopopulations of CFSE� cells were assessed. Representative FACS plots of CFSE�-transferred target cell populations (peptide-pulsed CFSEhigh and unpulsedCFSElow cells) in spleen of a control mouse (top panel) and an OVA-vaccinated mouse (bottom panel) are shown. The mean ratio (�SD) of CFSElow

(unpulsed) to CFSEhigh (peptide pulsed) -transferred cells in the spleens of OVA LANAC-vaccinated mice was significantly greater (p � 0.05) than inunvaccinated control mice 18 h after adoptive transfer of target cells (data not shown). Similar results were obtained in one additional experiment. B, Spleencells from unvaccinated control mice (left panel) and mice (4/group) immunized twice with OVA in LANAC (right panel) were analyzed by flow cytometryfor intracellular production of IFN-� following in vitro restimulation with 1 �M OVA8 peptide, as described in Materials and Methods, and a representativeFACS plot is shown. Similar results were obtained in one additional experiment. C, Spleen cells from unvaccinated mice, mice vaccinated twice with 5�g of OVA8 peptide in LANAC, and mice vaccinated with 5 �g of OVA in LANAC were restimulated in vitro with 1 �M OVA8 peptide for 18 h, andsupernatants were collected and analyzed for production of IFN-� by ELISA, as described in Materials and Methods. The mean (�SD) IFN-� concentrationwas plotted for each treatment group (4 animals/group). Mice immunized with OVA8 or OVA both produced significantly more (p � 0.05) IFN-� thanunvaccinated control mice, as denoted by �. Similar results were obtained in two additional experiments. D, The long-term survival of OVA8-specific CD8�

T cells elicited by LANAC vaccination in peripheral tissues (lung) was assessed in mice (4/group) 44 days after the first immunization and compared withthe percentage of OVA8-specific CD8� T cells present 14 days after the first immunization. The percentage of OVA-specific CD8� T cells in lung at 44days declined by approximately half compared with day 14 mice, but there were still significantly (p � 0.01) more Ag-specific CD8� T cells in lung tissuesat day 44 than in control mice, as denoted by �. Similar results were obtained in one additional experiment.

7342 TLR AGONIST LIPOSOME ADJUVANTS

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

approach to the development of nonreplicating vaccine adjuvantscapable of eliciting strong T cell responses.

Vaccine adjuvants are generally classified as either vaccine de-livery systems or as immune potentiators, depending on their pri-mary mode of action (8). Vaccine delivery systems (e.g., lipo-somes or microparticles) deliver the Ag to appropriate APC and insome cases prolong the duration of Ag presentation. Immune po-tentiators (e.g., bacterial products or cytokines) function by acti-vating innate immune responses, leading to enhanced Ag presen-tation. The effectiveness of the liposome-TLR agonist adjuvantsdescribed in this study probably results from the combination of

enhanced Ag delivery and potent stimulation of innate immunity.For example, liposomes have been previously used as vaccine ad-juvants, primarily by facilitating delivery of Ags to APC in vivo(25, 61–63). Liposomes also facilitate cross-priming by promotingentry of protein Ags into the MHC class I pathway for presentationto CD8� T cells (64). In addition, cationic liposomes have beenshown to markedly enhance the immune stimulation elicited bybacterial DNA and CpG oligonucleotides (18–20). The fact thatCD8� T cell responses were largely eliminated in MyD88�/�

mice provided additional evidence of the importance of simulta-neous Ag delivery and activation of APCs for efficient generationof T cell responses (see Fig. 3).

One of the most notable properties of the liposome-TLR agonistcomplex adjuvants, particularly those formulated with TLR3 orTLR9 agonists, was their ability to stimulate efficient cross-prim-ing in vivo. This property appeared to be independent of CD4 help(see Fig. 3). Thus, immunization with the LANAC adjuvant prob-ably activated CD40 on APC directly, thereby bypassing the re-quirement for CD4� T cells for initial generation of Ag-specificCD8� T cells (65, 66). The efficiency of cross-priming may have

FIGURE 6. Adjuvants formulated with TLR3 or TLR9 agonists elicitefficient cross-priming. Experiments were conducted to determine whetherTLR agonists other than bacterial (plasmid) DNA were capable of elicitingefficient cross-priming when complexed with liposomes. Cationic lipo-somes were complexed to the following TLR agonists: CpG oligonucleo-tides (TLR9), poly(I:C) (TLR3), LPS (TLR4), zymosan (TLR2), and R848(TLR7). Mice (4/group) were immunized twice i.p. with 5 �g of OVAprotein mixed with the indicated liposome-TLR agonist adjuvants, thenspleen, lung, and liver tissues were analyzed by flow cytometry for quan-titation of OVA8-specific CD8� T cells, as described in Materials andMethods. A, The mean (�SD) percentage of OVA8-specific CD8� T cellsin the spleens of mice immunized with liposome-TLR agonist complexeswas plotted. �, Denotes significant differences (p � 0.05) when comparedwith control mice, as assessed by ANOVA and Tukey multiple meanscomparison test. B, The mean (�SD) percentage of OVA8-specific CD8�

T cells in the lungs of mice immunized with liposome-TLR agonist com-plexes was plotted. �, Denotes significant differences (p � 0.05), comparedwith control mice, as assessed by ANOVA and Tukey multiple meanscomparison test. C, The mean (�SD) percentage of OVA8-specific CD8�

T cells in the livers of mice immunized with liposome-TLR agonist com-plexes was plotted. �, Denotes significant differences (p � 0.05) comparedwith control mice, as assessed by ANOVA and Tukey multiple meanscomparison test. The data presented in A–C represent pooled data from twoindependent experiments.

FIGURE 7. Inhibition of growth of established melanomas by LANACvaccination against trp2. The effects of trp2 vaccination on growth of es-tablished melanoma tumors was assessed in mice with day 7 establisheds.c. B16 tumors. Mice were vaccinated weekly beginning on day 7, and thetumor area was determined every 2–3 days. A, Control mice (5 animals/group) were unvaccinated, whereas another group was vaccinated with 5�g of trp2 peptide in LANAC by the s.c. route, and a third group wasimmunized with trp2 and LANAC by the i.p. route. The mean tumor area(�SE) for each group was plotted at each time point. �, Denotes significantdifferences (p � 0.05) in tumor volume for mice immunized with trp2-LANAC by the i.p. route, compared with control mice, as assessed byANOVA and Tukey multiple means comparison test. Similar results wereobtained in one additional experiment. B, Mice (5 animals/group) werevaccinated weekly with trp2 peptide using peptide-pulsed bone marrow DCor with trp2 plus liposomes only (no DNA) as described in Materials andMethods, and tumor area was determined as described above. Immuniza-tion with trp2 using bone marrow DC or liposomes alone did not elicitsignificant differences in tumor area compared with unvaccinated micewith tumors.

7343The Journal of Immunology

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

been due in part to the ability of cationic liposomes to facilitate theinteraction of TLR3 and TLR9 agonists with their endosomal re-ceptors and thereby promote immune activation (2, 67). In addi-tion, complexes of liposomes with TLR3 or TLR9 agonists stim-ulated production of type I IFNs, including IFN-�, and theassociation between induction of type I IFNs and cross-priminghas been noted previously (55, 68).

The relative inability of other TLR agonists (e.g., TLR2, TLR4,and TLR7/TLR8 agonists) complexed to cationic liposomes tofunction as effective vaccine adjuvants may have several explana-tions (see Fig. 6). For example, TLR2 and TLR4 are expressedprimarily on the cell surface and not within endosomes (2). There-fore, they may not receive a strong activating stimulus when theirligands are delivered as liposome complexes. However, failure toactivate innate immunity is probably not the only explanation, be-cause we observed that all the liposome-TLR agonist complexesevaluated in these studies were capable of activating innate im-munity when administered in vivo (data not shown). It is alsopossible that signaling via TLR2, TLR4, and TLR7/8 agonists mayhave delivered qualitatively different signals to the APCs, whichwould have affected the magnitude of the T cell responses thatwere elicited.

The liposome-TLR agonist adjuvants evaluated in this study tar-geted an unusual population of APC in draining lymph nodes,which may have played an important role in adjuvant effective-ness. At early time points after immunization with LANAC, thecomplexes were primarily associated with CD11b�/Gr-1� cellsand were rarely present in CD11c� cells (Fig. 4). Therefore, thecells responsible for much of the early trafficking of LANAC vac-cines to APC in draining lymph nodes were primarily neutrophils,macrophages, and inflammatory monocytes, rather than classicalDC (50). Inflammatory monocytes, which can differentiate into DCin lymph nodes, may therefore play a key role as APC for lipo-some-TLR agonist-based vaccine adjuvants.

The results of these studies suggest additional strategies for de-veloping more effective vaccine adjuvants based on the liposome-

TLR agonist platform. For example, other pattern recognition re-ceptors could be incorporated in the liposomal delivery system, inaddition to TLR ligands. Moreover, coupling of targeting mole-cules such as Abs to DEC-205 into the liposomes may also proveeffective for retargeting of the complexes to more classical DC, ashas been demonstrated recently (69, 70). Other applications of theadjuvant platform include mucosal immunization, where we havefound that the liposome-TLR agonist adjuvants are effective (S.Dow, unpublished data). The LANAC adjuvants may also be use-ful clinically, because we showed recently that vaccination of petdogs with refractory atopic dermatitis with allergens complexed toLANAC adjuvants was effective in reducing clinical signs andreversing some Th2 abnormalities (71). Therefore, the liposome-TLR agonist adjuvant system described in this study may be auseful addition to the list of currently available vaccine adjuvantscapable of eliciting strong T cell responses against protein andpeptide Ags.

AcknowledgmentsWe thank Barb Rose for excellent technical assistance with these studies.

DisclosuresA patent has been filed covering the adjuvant formulation described in thismanuscript, and Steven Dow is listed as one of the coinventors.

References1. Takeda, K., T. Kaisho, and S. Akira. 2003. Toll-like receptors. Annu. Rev. Im-

munol. 21: 335–376.2. Reis e Sousa, C. 2004. Toll-like receptors and dendritic cells: for whom the bug

tolls. Semin. Immunol. 16: 27–34.3. Iwasaki, A., and R. Medzhitov. 2004. Toll-like receptor control of the adaptive

immune responses. Nat. Immunol. 5: 987–995.4. Cook, D. N., D. S. Pisetsky, and D. A. Schwartz. 2004. Toll-like receptors in the

pathogenesis of human disease. Nat. Immunol. 5: 975–995.5. Akira, S., and K. Takeda. 2004. Toll-like receptor signalling. Nat. Rev. Immunol.

4: 499–511.6. Petrovsky, N., and J. C. Aguilar. 2004. Vaccine adjuvants: current state and

future trends. Immunol. Cell Biol. 82: 488–496.7. O’Hagan, D. T. 2001. Recent developments in vaccine delivery systems. Curr.

Drug Targets Infect. Disord. 1: 273–286.8. Pashine, A., N. M. Valiante, and J. B. Ulmer. 2005. Targeting the innate immune

response with improved vaccine adjuvants. Nat. Med. 11: S63–S68.9. Schijns, V. E. 2003. Mechanisms of vaccine adjuvant activity: initiation and

regulation of immune responses by vaccine adjuvants. Vaccine 21: 829–831.10. O’Hagan, D. T., and N. M. Valiante. 2003. Recent advances in the discovery and

delivery of vaccine adjuvants. Nat. Rev. Drug Discov. 2: 727–735.11. Singh, M., and D. T. O’Hagan. 2002. Recent advances in vaccine adjuvants.

Pharm. Res. 19: 715–728.12. Tudor, D., C. Dubuquoy, V. Gaboriau, F. Lefevre, B. Charley, and S. Riffault.

2005. TLR9 pathway is involved in adjuvant effects of plasmid DNA-based vac-cines. Vaccine 23: 1258–1264.

13. Evans, J. T., C. W. Cluff, D. A. Johnson, M. J. Lacy, D. H. Persing, andJ. R. Baldridge. 2003. Enhancement of antigen-specific immunity via the TLR4ligands MPL adjuvant and Ribi.529. Expert Rev. Vaccines 2: 219–229.

14. Klinman, D. M. 2003. CpG DNA as a vaccine adjuvant. Expert Rev. Vaccines 2:305–315.

15. Krieg, A. M. 2000. Immune effects and mechanisms of action of CpG motifs.Vaccine 19: 618–622.

16. Marciani, D. J. 2003. Vaccine adjuvants: role and mechanisms of action in vac-cine immunogenicity. Drug Discov. Today 8: 934–943.

17. Ishii, N., J. Fukushima, T. Kaneko, E. Okada, K. Tani, S. I. Tanaka,K. Hamajima, K. Q. Xin, S. Kawamoto, W. Koff, et al. 1997. Cationic liposomesare a strong adjuvant for a DNA vaccine of human immunodeficiency virus type1. AIDS Res. Hum. Retroviruses 13: 1421–1428.

18. Whitmore, M., S. Li, and L. Huang. 1999. LPD lipopolyplex initiates a potentcytokine response and inhibits tumor growth. Gene. Ther. 6: 1867–1875.

19. Gursel, I., M. Gursel, K. J. Ishii, and D. M. Klinman. 2001. Sterically stabilizedcationic liposomes improve the uptake and immunostimulatory activity of CpGoligonucleotides. J. Immunol. 167: 3324–3328.

20. Dow, S. W., L. G. Fradkin, D. H. Liggitt, A. P. Willson, T. D. Heath, andT. A. Potter. 1999. Lipid-DNA complexes induce potent activation of innateimmune responses and antitumor activity when administered intravenously.J. Immunol. 163: 1552–1561.

21. Sellins, K., L. Fradkin, D. Liggitt, and S. Dow. 2005. Type I Interferons potentlysuppress gene expression following gene delivery using liposome�DNA com-plexes. Mol. Ther. 12: 451–459.

22. Rocha, A., S. Ruiz, and J. M. Coll. 2002. Improvement of DNA transfection withcationic liposomes. J. Physiol. Biochem. 58: 45–56.

FIGURE 8. Generation of protective immunity in M. tuberculosis aero-sol challenge model. The effects of vaccination with LANAC-based vac-cines on generation of protective immunity to M. tuberculosis aerosol chal-lenge was assessed. C57BL/6 mice (5/group) were immunized twice, 2 wkapart, with 10 �g of recombinant M. tuberculosis ESAT-6 protein in eitherLANAC adjuvant (i.p. and s.c. routes) or MPL adjuvant (s.c. route), asdescribed in Materials and Methods. Negative controls included micemock vaccinated with PBS and mice vaccinated with LANAC adjuvantalone. Mice were challenged by low-dose aerosol challenge with M. tu-berculosis Erdman strain 4 wk after the last vaccine was administered.Thirty days after challenge, mice were sacrificed, and the number of viableM. tuberculosis organisms in lung tissues was determined. The mean log10

protection value was calculated (CFU for saline control group � CFU forvaccinated groups) and plotted. ��, Denotes significant differences (p �0.05) between LANAC-vaccinated mice and saline control group and micevaccinated with ESAT-6 in MPL adjuvant; �, denotes significant differ-ences between control and MPL-ESAT-6-vaccinated mice.

7344 TLR AGONIST LIPOSOME ADJUVANTS

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from

23. Briane, D., D. Lesage, A. Cao, R. Coudert, N. Lievre, J. L. Salzmann, andE. Taillandier. 2002. Cellular pathway of plasmids vectorized by cholesterol-based cationic liposomes. J. Histochem. Cytochem. 50: 983–991.

24. Wagner, H. 2004. The immunobiology of the TLR9 subfamily. Trends Immunol.25: 381–386.

25. Chikh, G., and M. P. Schutze-Redelmeier. 2002. Liposomal delivery of CTLepitopes to dendritic cells. Biosci. Rep. 22: 339–353.

26. Raychaudhuri, S., M. Tonks, F. Carbone, T. Ryskamp, W. J. Morrow, andN. Hanna. 1992. Induction of antigen-specific class I-restricted cytotoxic T cellsby soluble proteins in vivo. Proc. Natl. Acad. Sci. USA 89: 8308–8312.

27. Jiang, Z. H., and R. R. Koganty. 2003. Synthetic vaccines: the role of adjuvantsin immune targeting. Curr. Med. Chem. 10: 1423–1439.

28. Moingeon, P., J. Haensler, and A. Lindberg. 2001. Towards the rational design ofTh1 adjuvants. Vaccine 19: 4363–4372.

29. Takahashi, H., T. Takeshita, B. Morein, S. Putney, R. N. Germain, andJ. A. Berzofsky. 1990. Induction of CD8� cytotoxic T cells by immunizationwith purified HIV-1 envelope protein in ISCOMs. Nature 344: 873–875.

30. Villacres, M. C., S. Behboudi, T. Nikkila, K. Lovgren-Bengtsson, and B. Morein.1998. Internalization of iscom-borne antigens and presentation under MHC classI or class II restriction. Cell. Immunol. 185: 30–38.

31. Miconnet, I., S. Koenig, D. Speiser, A. Krieg, P. Guillaume, J. C. Cerottini, andP. Romero. 2002. CpG are efficient adjuvants for specific CTL induction againsttumor antigen-derived peptide. J. Immunol. 168: 1212–1218.

32. Klinman, D. M., D. Currie, I. Gursel, and D. Verthelyi. 2004. Use of CpG oli-godeoxynucleotides as immune adjuvants. Immunol. Rev. 199: 201–216.

33. Mesa, C., and L. E. Fernandez. 2004. Challenges facing adjuvants for cancerimmunotherapy. Immunol. Cell Biol. 82: 644–650.

34. Moron, G., G. Dadaglio, and C. Leclerc. 2004. New tools for antigen delivery tothe MHC class I pathway. Trends Immunol. 25: 92–97.

35. Alving, C. R. 2002. Design and selection of vaccine adjuvants: animal modelsand human trials. Vaccine 20(Suppl. 3): S56–S64.

36. Burdin, N., B. Guy, and P. Moingeon. 2004. Immunological foundations to thequest for new vaccine adjuvants. BioDrugs 18: 79–93.

37. Ulmer, J. B. 2002. Toward the development of better vaccine adjuvants. ExpertOpin. Biol. Ther. 2: 119–121.

38. Schijns, V. E., and A. Tangeras. 2005. Vaccine adjuvant technology: from the-oretical mechanisms to practical approaches. Dev. Biol. 121: 127–134.

39. Roland, K. L., S. A. Tinge, K. P. Killeen, and S. K. Kochi. 2005. Recent advancesin the development of live, attenuated bacterial vectors. Curr. Opin. Mol. Ther.7: 62–72.

40. Suzuki, Y., D. Wakita, K. Chamoto, Y. Narita, T. Tsuji, T. Takeshima, H. Gyobu,Y. Kawarada, S. Kondo, S. Akira, et al. 2004. Liposome-encapsulated CpG oli-godeoxynucleotides as a potent adjuvant for inducing type 1 innate immunity.Cancer Res. 64: 8754–8760.

41. Dileo, J., R. Banerjee, M. Whitmore, J. V. Nayak, L. D. Falo, Jr., and L. Huang.2003. Lipid-protamine-DNA-mediated antigen delivery to antigen-presentingcells results in enhanced anti-tumor immune responses. Mol. Ther. 7: 640–648.

42. Cui, Z., and L. Huang. 2005. Liposome-polycation-DNA (LPD) particle as acarrier and adjuvant for protein-based vaccines: therapeutic effect against cervicalcancer. Cancer Immunol. Immunother. 12: 1180–1190.

43. Gursel, M., S. Tunca, M. Ozkan, G. Ozcengiz, and G. Alaeddinoglu. 1999. Im-munoadjuvant action of plasmid DNA in liposomes. Vaccine 17: 1376–1383.

44. Fairman, J., L. Roche, I. Pieslak, M. Lay, S. Corson, E. Fox, C. Luong, G. Koe,B. Lemos, R. Grove, et al. 1999. Quantitative RT-PCR to evaluate in vivo ex-pression of multiple transgenes using a common intron. BioTechniques 27:566–570.

45. Templeton, N. S., D. D. Lasic, P. M. Frederik, H. H. Strey, D. D. Roberts, andG. N. Pavlakis. 1997. Improved DNA: liposome complexes for increased sys-temic delivery and gene expression. Nat. Biotechnol. 15: 647–652.

46. Inaba, K., M. Inaba, N. Romani, H. Aya, M. Deguchi, S. Ikehara, S. Muramatsu,and R. M. Steinman. 1992. Generation of large numbers of dendritic cells frommouse bone marrow cultures supplemented with granulocyte/macrophage colo-ny-stimulating factor. J. Exp. Med. 176: 1693–1702.

47. Fisher, J. H., J. Larson, C. Cool, and S. W. Dow. 2002. Lymphocyte activationin the lungs of SP-D null mice. Am. J. Respir. Cell Mol. Biol. 27: 24–33.

48. Ahonen, C. L., C. L. Doxsee, S. M. McGurran, T. R. Riter, W. F. Wade,R. J. Barth, J. P. Vasilakos, R. J. Noelle, and R. M. Kedl. 2004. Combined TLRand CD40 triggering induces potent CD8� T cell expansion with variable de-pendence on type I IFN. J. Exp. Med. 199: 775–784.

49. Meyer, A. L., C. Trollmo, F. Crawford, P. Marrack, A. C. Steere, B. T. Huber,J. Kappler, and D. A. Hafler. 2000. Direct enumeration of Borrelia-reactive CD4T cells ex vivo by using MHC class II tetramers. Proc. Natl. Acad. Sci. USA 97:11433–11438.

50. Geissmann, F., S. Jung, and D. R. Littman. 2003. Blood monocytes consist of twoprincipal subsets with distinct migratory properties. Immunity 19: 71–82.

51. Jordan, M. B., D. M. Mills, J. Kappler, P. Marrack, and J. C. Cambier. 2004.Promotion of B cell immune responses via an alum-induced myeloid cell popu-lation. Science 304: 1808–1810.

52. Lauvau, G., S. Vijh, P. Kong, T. Horng, K. Kerksiek, N. Serbina, R. A. Tuma,and E. G. Pamer. 2001. Priming of memory but not effector CD8 T cells by akilled bacterial vaccine. Science 294: 1735–1739.

53. Gourley, T. S., E. J. Wherry, D. Masopust, and R. Ahmed. 2004. Generation andmaintenance of immunological memory. Semin. Immunol. 16: 323–333.

54. Masopust, D., and R. Ahmed. 2004. Reflections on CD8 T-cell activation andmemory. Immunol. Res. 29: 151–160.

55. Beignon, A. S., M. Skoberne, and N. Bhardwaj. 2003. Type I interferons promotecross-priming: more functions for old cytokines. Nat. Immunol. 4: 939–941.

56. Cho, H. J., T. Hayashi, S. K. Datta, K. Takabayashi, J. H. Van Uden, A. Horner,M. Corr, and E. Raz. 2002. IFN-�� promote priming of antigen-specific CD8�

and CD4� T lymphocytes by immunostimulatory DNA-based vaccines. J. Im-munol. 168: 4907–4913.

57. Pardoll, D. M. 2002. Spinning molecular immunology into successful immuno-therapy. Nat. Rev. Immunol. 2: 227–238.

58. Brandt, L., M. Elhay, I. Rosenkrands, E. B. Lindblad, and P. Andersen. 2000.ESAT-6 subunit vaccination against Mycobacterium tuberculosis. Infect. Immun.68: 791–795.

59. Mollenkopf, H. J., D. Groine-Triebkorn, P. Andersen, J. Hess, andS. H. Kaufmann. 2001. Protective efficacy against tuberculosis of ESAT-6 se-creted by a live Salmonella typhimurium vaccine carrier strain and expressed bynaked DNA. Vaccine 19: 4028–4035.

60. Olsen, A. W., P. R. Hansen, A. Holm, and P. Andersen. 2000. Efficient protectionagainst Mycobacterium tuberculosis by vaccination with a single subdominantepitope from the ESAT-6 antigen. Eur. J. Immunol. 30: 1724–1732.

61. Eriksson, K., and J. Holmgren. 2002. Recent advances in mucosal vaccines andadjuvants. Curr. Opin. Immunol. 14: 666–672.

62. Felnerova, D., J. F. Viret, R. Gluck, and C. Moser. 2004. Liposomes and viro-somes as delivery systems for antigens, nucleic acids and drugs. Curr. Opin.Biotechnol. 15: 518–529.

63. Gregoriadis, G., B. McCormack, M. Obrenovic, R. Saffie, B. Zadi, and Y. Perrie.1999. Vaccine entrapment in liposomes. Methods 19: 156–162.

64. Reis e Sousa, C., and R. N. Germain. 1995. Major histocompatibility complexclass I presentation of peptides derived from soluble exogenous antigen by asubset of cells engaged in phagocytosis. J. Exp. Med. 182: 841–851.

65. Bennett, S. R., F. R. Carbone, F. Karamalis, R. A. Flavell, J. F. Miller, andW. R. Heath. 1998. Help for cytotoxic-T-cell responses is mediated by CD40signalling. Nature 393: 478–480.

66. Schoenberger, S. P., R. E. Toes, E. I. van der Voort, R. Offringa, and C. J. Melief.1998. T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40Linteractions. Nature 393: 480–483.

67. Schulz, O., S. S. Diebold, M. Chen, T. I. Naslund, M. A. Nolte, L. Alexopoulou,Y. T. Azuma, R. A. Flavell, P. Liljestrom, and C. Reis e Sousa. 2005. Toll-likereceptor 3 promotes cross-priming to virus-infected cells. Nature 433: 887–892.

68. Santini, S. M., T. Di Pucchio, C. Lapenta, S. Parlato, M. Logozzi, andF. Belardelli. 2002. The natural alliance between type I interferon and dendriticcells and its role in linking innate and adaptive immunity. J. Interferon CytokineRes. 22: 1071–1080.

69. Demangel, C., J. Zhou, A. B. Choo, G. Shoebridge, G. M. Halliday, andW. J. Britton. 2005. Single chain antibody fragments for the selective targeting ofantigens to dendritic cells. Mol. Immunol. 42: 979–885.

70. van Broekhoven, C. L., C. R. Parish, C. Demangel, W. J. Britton, and J. G. Altin.2004. Targeting dendritic cells with antigen-containing liposomes: a highly ef-fective procedure for induction of antitumor immunity and for tumor immuno-therapy. Cancer Res. 64: 4357–4365.

71. Mueller, R. S., J. Veir, K. V. Fieseler, and S. W. Dow. 2005. Use of immuno-stimulatory liposome-nucleic acid complexes in allergen-specific immunotherapyof dogs with refractory atopic dermatitis: a pilot study. Vet. Dermatol. 16: 61–68.

7345The Journal of Immunology

by guest on February 1, 2016http://w

ww

.jimm

unol.org/D

ownloaded from