Nonmethylated CG Motifs Packaged into Virus-Like Particles Induce Protective Cytotoxic T Cell Responses in the Absence of Systemic Side Effects

Tazio Storni, Christiane Ruedl, Katrin Schwarz, Reto A. Schwendener, Wolfgang A. Renner and Martin F. Bachmann
J Immunol February 1, 2004, 172 (3) 1777-1785; DOI: https://doi.org/10.4049/jimmunol.172.3.1777

Abstract

DNA rich in nonmethylated CG motifs (CpGs) greatly facilitates induction of immune responses against coadministered Ags. CpGs are therefore among the most promising adjuvants known to date. Nevertheless, CpGs are characterized by two drawbacks. They have unfavorable pharmacokinetics and may exhibit systemic side effects, including splenomegaly. We show in this study that packaging CpGs into virus-like particles (VLPs) derived from the hepatitis B core Ag or the bacteriophage Qβ is a simple and attractive method to reduce these two problems. CpGs packaged into VLPs are resistant to DNase I digestion, enhancing their stability. In addition, and in contrast to free CpGs, packaging CpGs prevents splenomegaly in mice, without affecting their immunostimulatory capacity. In fact, vaccination with CpG-loaded VLPs was able to induce high frequencies of peptide-specific CD8+ T cells (4–14%), protected from infection with recombinant vaccinia viruses, and eradicated established solid fibrosarcoma tumors. Thus, packaging CpGs into VLPs improves both their immunogenicity and pharmacodynamics.

Viruses and other pathogens such as bacteria and parasites are known to induce potent B and T cell responses. In striking contrast, isolated recombinant proteins or T cell epitopes usually fail to induce significant immune responses unless given in strong adjuvants (1, 2, 3). Such adjuvants may enhance immune responses via two mechanisms. First, they allow the formation of an Ag depot in the host, increasing and in particular prolonging the exposure of the immune system to a selected Ag. Second, many adjuvants trigger the innate immune system usually via activation of Toll-like receptors (4, 5, 6, 7). The major cell type affected under these conditions are APCs, which become activated, migrate to T cell regions of lymphoid organs, and up-regulate costimulatory ligands (8). DNA rich in nonmethylated CpG motifs (CpGs)2 is a classical example of such an adjuvant (9, 10, 11, 12, 13), which activates dendritic cells (DCs) via stimulation of Toll-like receptor 9 within endosomes (14, 15, 16, 17), leading to translocation of NF-κB into the nucleus (18, 19), inducing maturation of DCs.

CpGs are preferentially present in prokaryotic DNA and are rare in mammalian DNA (20, 21). Thus, CpGs are an invariant pattern associated with pathogens and are recognized as such by the innate immune system. However, due to the intensive cellular activation provoked by CpGs, considerable systemic side effects may occur, in particular splenomegaly or even lethal toxic shock (22, 23). The second problem of CpGs is their unfavorable pharmacokinetics, because they exhibit a short life span in vivo. In fact, CpG oligodeoxynucleotides (ODN) containing normal phosphodiesterbonds usually fail to enhance immune responses, unless givenrepeatedly (24). Therefore, CpGs are chemically modified by introducing phosphorothioate bonds to reduce this limitation. The drawbacks of these modified oligonucleotides are the very long tissue t1/2 of ∼2 days (25) and the increased nonspecific binding to various proteins and cell types (26).

To deliver CpGs in a more specific fashion to professional APCs, to reduce their potentially serious side effects, and to enhance their stability, we packaged the synthetic DNA oligonucleotides into virus-like particles (VLPs). The hepatitis B core Ag (HBcAg) and the bacteriophage Qβ capsids were used in this study as model VLPs for the packaging of CpGs. These proteins are recombinantly produced in bacteria and can efficiently self-assemble into structured VLPs. The CTL epitope p33 derived from the lymphocytic choriomeningitis virus glycoprotein (LCMV-GP) was associated to the VLPs through two different approaches: it was genetically fused to the C terminus of the HBcAg monomers (27), whereas for the Qβ VLP the p33 peptide was chemically coupled via a bifunctional linker to lysins present in the Qβ protein structure (28). In both cases, the CpG ODNs were packaged inside the VLPs through binding to capsid-internal arginine repeats. Such VLPs packaged with various CpG sequences (with phosphorothioate and also with normal phosphodiester bonds) exhibited enhanced adjuvant properties in the absence of significant side effects and induced strong immune responses in mice. When compared side by side, VLPs containing CpGs induced stronger CTL responses than recombinant vaccinia viruses or CpGs delivered together with peptide in liposomes. The induced responses protected mice from viral infections and eradicated established solid fibrosarcoma tumors.

Materials and Methods

Production of p33-VLPs

Production, purification, and characterization of recombinant hepatitis B core p33-VLPs and Qβ VLPs were previously described in detail (27, 29, 30).

Chemical coupling of p33 to Qβ VLP

Purified Qβ VLPs (1.5 mg/ml in 20 mM HEPES, 150 mM NaCl, pH 7.2) were derivatized by a 30-min incubation at room temperature with a 10-fold molar excess of succinimidyl-6-(β-maleimidopropionamido)hexanoate (Pierce, Rockford, IL). Free cross-linker was removed by extensive dialysis against 20 mM HEPES, pH 7.2. Peptide p33 was produced in a modified version with additional 3 aa (GGC) added to the C terminus (p33-GGC) (EMC Microcollections GmbH, Tübingen, Germany) to allow coupling to VLPs. Derivatized Qβ VLPs and p33-GGC (peptide at 5-fold molar excess) were then incubated for 2 h at room temperature to allow cross-linking. Free p33-GGC was removed by dialysis against 20 mM HEPES, pH 7.2, using DispoDialyser membranes with a molecular mass cutoff of 300 kDa (Spectrum Medical Industries, Rancho Dominguez, CA). Efficiency of cross-linking was analyzed by SDS-PAGE.

CpG ODN sequences

CpGs with normal phosphodiester backbone or with DNase-protected phosphorothioate bonds were synthesized by Microsynth AG (Balgach, Switzerland). The following CpG oligonucleotide sequences were used: 1668, 5′-TCCATGACGTTCCTGAATAAT-3′; 1585, 5′-GGGGTCAACGTTGAGGGGGG-3′; 1826, 5′-TCCATGACGTTCCTGACGTT-3′; G3, 5′-GGGGACGATCGTCGGGGGG-3′; G6, 5′-GGGGGGCGACGACGATCGTCGTCGGGGGGG-3′; Liang2, 5′-GAGACCCTGAACAGTTGATC-3′.

Packaging of CpGs into p33-VLPs

Packaging of CpGs into HBcAg and Qβ VLPs was performed with the same protocol, with the sole exception that HBcAg VLPs were dissolved in PBS, whereas Qβ particles were solubilized in 20 mM HEPES. Briefly, bacterial RNA present in p33-VLPs was eliminated by incubating the capsid preparations (0.5 mg/ml concentration) with RNase A (0.2 mg/ml concentration) for at least 2 h at 37°C. Packaging of CpG oligonucleotides into the p33-VLPs was achieved by adding CpGs to RNase-treated p33-VLPs to a final concentration of 120 nmol/ml for another 3 h. The p33-VLPs loaded with CpGs were treated with DNase I (200 U/ml) for 5 h at 37°C to test the enhanced stability of packaged CpG. p33-VLPs loaded with CpGs used for immunization of mice were first extensively dialyzed with a 300-kDa molecular weight cut-off membrane (Spectrum Medical Industries) to eliminate unbound oligonucleotides from p33-VLPs. Analysis of CpG packaging into p33-VLPs was done by loading 15 μg of protein on 1% agarose gels and run for 30 min at 100 V. The presence of nucleic acids was assessed by ethidium bromide staining, while VLPs were detected through Coomassie blue staining of the gels. The Gene Ruler, 1-kb DNA Ladder (MBI Fermentas GmbH, St. Leon-Rot, Germany), was used as reference.

Preparation of liposomes

Liposomes were produced, as previously described (31). Briefly, small unilamellar liposomes were generated by freeze thawing, followed by sequential filter extrusion. The liposomal composition was 200 mg/ml soy phosphatidylcholine, 25 mg/ml cholesterol, and 1.2 mg/ml dl-α-tocopherol. The dried lipid mixture was solubilized with 1 mg/ml or 50 μg/ml p33 peptide (KAVYNFATM) alone or with 100 nmol/ml CpGs (ODN1668pt), subjected to three to five freeze-thaw cycles, and repeatedly extruded through Nucleopore filters of 0.8-, 0.4-, and 0.2-μm pore size (Sterico AG, Dietikon, Switzerland). Unencapsulated peptide and CpGs were removed by dialysis. Liposome size was determined by laser light scattering (Submicron Particle Sizer Model 370; Nicomp, Santa Barbara, CA)

Assessment of antiviral immunity

To examine antiviral immunity, vaccinated C57BL/6 mice were infected i.v. 12 days after priming with 200 PFU LCMV strain WE. Five days later, spleens were isolated and LCMV titers were determined by a LCMV focus-forming assay, as described (32). Alternatively, female C57BL/6 mice were infected i.p. with 1.5 × 106 PFU or 4 × 106 PFU recombinant vaccinia virus. Five days later, ovaries were collected and the vaccinia titers were determined on BSC 40 cells, as described (33).

Cytotoxicity assay

For detection of primary ex vivo cytotoxicity, effector cell suspensions were prepared from spleens of vaccinated mice 9 days after priming. EL-4 cells were pulsed with p33 peptide (10−6 M, 2 h at 37°C in 2% FCS MEM medium) and used in a 5-h 51Cr release assay (33). Unlabeled EL-4 cells were used as a control. Radioactivity in cell culture supernatants was measured in a Cobra II counter (Canberra Packard, Downers Grove, IL). Spontaneous release was always <10%.

Assessment of antitumor immunity

Solid tumors were obtained by s.c. implanting MC57G-GP tumor cells expressing the LCMV-GP in H-2b RAG−/− mice. Tumors were sized in 0.2 × 0.2 × 0.2-cm pieces and implanted into the flank of C57BL/6 mice. Tumor growth was monitored by measuring the orthogonal diameters weekly, and the tumor size was calculated by multiplication of the two diameters. Mice were vaccinated s.c. once the tumors had reached a size of >0.64 cm2. Mice in which tumors were reaching sizes of >3–4 cm2 were sacrificed for regulatory reasons.

Statistical analysis

Unless noted, data are presented as mean ± SD. Comparisons between two groups were performed by two-tailed Student’s t test. A p value <0.05 was considered as significant and indicated with ∗ in the figures.

For the tumor experiments, the statistical significances were calculated by comparing the survival rates of treated vs untreated mice in two-tailed Student’s t tests.

Results

Recombinantly produced p33-VLPs contain mainly RNA, which can be substituted by CpG oligonucleotides

As model VLPs for packaging CpG-containing DNA oligonucleotides, we used the HBcAg and the bacteriophage Qβ capsids. The MHC class I-restricted p33 peptide derived from the glycoprotein of LCMV was genetically fused to HBcAg and, respectively, chemically coupled to Qβ via the succinimidyl-6-(β-maleimidopropionamido)hexanoate bifunctional linker. For both VLP types and epitope attachment strategies, the VLPs maintained their assembly capacity (27, 34) (data not shown).

HBcAg p33-VLPs run as single bands in Coomassie blue-stained agarose gels (Fig. 1⇓A). Staining of a parallel gel with ethidium bromide demonstrated that the particles carried nucleic acid (Fig. 1⇓A). The nucleic acids were bound to the internal DNA/RNA binding site of the particles, because deletion mutants missing the relevant arginine-rich repeats present between aa 149–183 of the HBc protein (HBcAg VLPDEL) failed to package the nucleic acids (Fig. 1⇓A). Note that HBcAg VLPDEL have a slower migration rate on agarose gels than full-length HBcAg VLPs, most probably because of the different electrical charge of the two macromolecular complexes (28). To test whether HBcAg carries RNA or DNA, native particles or proteinase K-digested particles (data not shown) were treated with RNase A or DNase I (Fig. 1⇓A). RNase treatment almost completely removed the nucleic acid signal, while DNase had no measurable effect. Thus, recombinantly produced HBcAg carries primarily RNA bound to C-terminal arginine-rich repeats, confirming recent results obtained by Riedl et al. (35). Interestingly, only RNase treatment of full-length HBcAg VLPs, but not of VLPDEL, produced a protein smear on the agarose gels: a possible explanation for this phenomenon could be that RNase-treated HBcAg p33-VLPs, which, as confirmed by electron microscopy analysis, maintained their particulate structure after enzyme digestion (data not shown), may contain different residual RNA traces and therefore possess heterogeneous electrical charges.

FIGURE 1.
FIGURE 1.

Recombinant VLPs contain bacterial RNA that can be replaced by CpGs. A, Full-length HBcAg p33-VLPs and HBcAg VLP deletion mutants, missing the RNA/DNA binding site of the particles (VLPDEL), were left untreated or treated with RNase A, DNase I, or RNase A together with DNase I for 3 h at 37°C and analyzed on agarose gels stained with Coomassie blue (upper panel) or ethidium bromide (lower panel). B, HBcAg p33-VLPs and VLPDEL as control were left untreated or treated first with RNase A for 3 h at 37°C and subsequently incubated with ODN1668pt or ODN1668po for another 3 h. C, Preparations of HBcAg p33-VLPs were treated with RNase A in the presence of ODN1668po, followed by DNase I digestion. The Gene Ruler, 1-kb DNA Ladder was used as reference (line M). One complete set of experiments of two sets is shown. Packaging of CpGs into HBcAg p33-VLPs was established as a standard method and was repeated at least 10 times. D, HBcAg p33-VLPs packaged with ODN1668pt were digested with proteinase K, and subsequently CpG oligonucleotides were obtained through phenol extraction and DNA precipitation. The CpG amount was extrapolated then by running the dissolved oligonucleotides on an agarose gel and comparing the intensity of the CpG band with a titration curve of known CpG concentrations (5, 1, 0.5, 0.1, and 0.05 nmol of ODN1668pt).

To load VLPs with oligonucleotides, particles were digested with RNase and subsequently incubated with chemically stabilized CpG-rich oligonucleotides containing phosphorothioate (pt) bonds (ODN1668pt) or regular phosphodiester (po) oligonucleotides (ODN1668po). These oligonucleotides can diffuse into the VLPs through holes of ∼2 nm diameter present in the capside structure (36, 37, 38) and replace RNA. Addition of either type of oligonucleotides largely preserved the protein and nucleic acid bands observed in agarose gels (Fig. 1⇑B). Thus, HBcAg p33-VLPs were able to bind ODN1668po and ODN1668pt. In contrast, HBcAg lacking the RNA/DNA binding site did not package CpGs (Fig. 1⇑B). To test whether ODN1668po packaged in HBcAg were protected from degradation, p33-VLPs loaded with CpG were treated with DNase I before analysis on agarose gels (Fig. 1⇑C). Although free CpG were readily digested by DNase, packaged CpG were largely protected (Fig. 1⇑C) (note that DNase enzymes are generally larger than RNases and may not be able to enter the VLP particles). To investigate the CpG concentration packaged into VLPs, HBcAg p33-VLPs packaged with ODN1668pt were digested with proteinase K, and the packaged CpG oligonucleotides were recovered by phenol extraction and subsequent DNA precipitation. The CpG content of HBcAg p33-VLPs was calculated by running the dissolved oligonucleotides on an agarose gel and comparing the intensity of the CpG band with a titration curve of known ODN1668pt concentrations. Fig. 1⇑D shows that 80 μg of packaged HBcAg p33-VLPs contain ∼0.5 nmol of oligonucleotides: This means that every VLP is loaded on average with ∼30 ODN1668pt molecules. Note that for the packaging of oligonucleotides with different CpG sequences but of the same length, similar results were obtained (data not shown).

CpGs were packaged into Qβ particles (dissolved in 20 mM HEPES) with the same protocol as used for HBcAg VLPs. Efficiency of Qβ packaging was analyzed in the same way as described in Fig. 1⇑, and comparable results were obtained (data not shown). Specifically, Qβ packaged with the ODNG3po CpGs contained ∼85 DNA molecules per VLP capside on average, whereas Qβ packaged with the ODNG6po CpGs could be loaded with ∼25 DNA molecules per VLP particle (data not shown).

HBcAg p33-VLPs are processed in vivo for Ag presentation by DCs, but not by B or T cells

In previous studies, we have shown that DCs of lymphoid (CD11c+CD8+) and myeloid (CD11c+CD8−) origin and also macrophages (Mφ) (CD11b+) can take up HBcAg p33-VLPs for Ag processing and presentation to p33-specific T cells (27, 34). In vitro and in vivo assays indicated that DCs can cross-present the p33 epitope via a TAP-dependent endosome-to-cytosol pathway and via a TAP-independent pathway in which loading of MHC class I molecules presumably occurs within the endosome. In contrast, Mφ seem to cross-present p33 exclusively via the second mechanism (direct endosomal loading).

To strengthen the hypothesis that only professional APCs are able to load p33 on MHC class I molecules, groups of wild-type mice were immunized s.c. with 20 μg HBcAg p33-VLPs or wild-type HBcAg as negative control, and draining lymph nodes were removed 16 h after priming. DC, B, and T cells were isolated by FACS cell sorting and cocultured with specific CD8+ T cells to perform a classical [3H]thymidine incorporation assay (Fig. 2⇓). As expected, only DCs, but not B and T cells, could take up HBcAg p33-VLPs and cross-present the p33 epitope to the transgenic T cells and induce their proliferation. This underscores the suitability of VLPs for the specific targeting of Ags and CpG oligonucleotides to APCs.

FIGURE 2.
FIGURE 2.

HBcAg p33-VLPs are processed in vivo for Ag presentation by DCs, but not by B or T cells. C57BL/6 mice were immunized s.c. with 20 μg of HBcAg p33-VLPs or wild-type VLPs as control. After 16 h, draining lymph nodes were removed, and DC, B cells, and T cells were isolated by FACS cell sorting through discrimination of CD11c+, CD19+, and Thy-1.2+ cells. A total of 3 × 104 DC, B cells, and T cells was cocultured with 2 × 105 transgenic p33-specific CD8+ T cells in round-bottom 96-well plates for 40 h. Cells were pulsed with [3H]thymidine during the last 10 h of incubation. Values represent averages and SDs of triplicates.

HBcAg and Qβ p33-VLPs packaged with CpGs induce strong and protective CTL responses

The immunogenicity of p33-VLPs loaded with CpGs was tested next. C57BL/6 mice were immunized s.c. with 100 μg HBcAg p33-VLPs mixed together with 20 nmol phosphothioester-stabilized ODN1668pt or packaged with ODN1668pt. Frequencies of specific CD8+ T cells were assessed 8 days later in the blood using p33-loaded MHC tetramers and were observed to be as follows: HBcAg p33-VLP alone, 0.3 ± 0.1%; HBcAg p33-VLP mixed with ODN1668pt, 2.1 ± 0.9%; HBcAg p33-VLP with ODN1668pt packaged, 4.3 ± 1.1%, whereas untreated mice had a median frequency of 0.2% (Fig. 3⇓A). The immunogenicity of VLPs packaged with CpGs was underscored in a second experiment, in which the frequencies of specific T cells induced by live recombinant vaccinia virus expressing the p33-containing LCMV glycoprotein (Vacc-G2) (1.1 ± 0.3%) were directly compared with those induced by HBcAg p33-VLPs packaged with ODN1668pt (5.4 ± 1.4%) (Fig. 3⇓B).

FIGURE 3.
FIGURE 3.

Induction of p33-specific CTLs by HBcAg and Qβ p33-VLPs packaged with CpGs. C57BL/6 mice were vaccinated with p33-VLP preparations mixed or packaged with CpGs, whereas naive, untreated mice served as controls. Seven or eight days after Ag administration, blood lymphocytes were double stained with PE-labeled p33 tetramers and FITC-coupled anti-CD8 mAbs for p33-specific CD8+ T cell detection. A, Tetramer analysis of untreated (n = 2) mice or mice immunized with HBcAg p33-VLPs (100 μg) alone (n = 4), mixed (n = 5), or packaged (n = 5) with ODN1668pt (20 or 0.6 nmol, respectively). One representative experiment of two independent experiments is shown. ∗, p < 0.05. B, Tetramer analysis of mice immunized s.c. with HBcAg p33-VLPs (100 μg) packaged with ODN1668pt (n = 6) or infected i.v. with recombinant vaccinia virus (1 × 106 PFU) (n = 3). ∗, p < 0.05. C, Frequencies of p33-specific CD8+ T cells in mice immunized with Qβ p33-VLPs (90 μg) mixed or packaged with ODNG3po or, alternatively, ODNG6po CpGs. Three mice per group were used. One representative experiment of two independent experiments is shown. ∗, p < 0.05. D, Frequencies of p33-specific CD8+ T cells in mice immunized with packaged HBcAg p33-VLPs (100 μg) or with liposomes. In this study, various preparations were tested: liposomes packaged with p33 peptide (100 or 5 μg) alone or in combination with ODN1668pt (10 or 0.5 nmol). Five mice per group were used. ∗, p < 0.05 (two-tailed Student’s t test).

To assess protective CTL responses, mice were immunized, as described above for Fig. 3⇑A. Additional groups of mice receiving s.c. 100 μg HBcAg p33-VLPs mixed with or containing packaged chemically nonprotected ODN1668po CpGs were included. Twelve days after vaccination, mice were challenged with live Vacc-G2 virus (1.5 × 106 PFU, i.p.), and viral titers were assessed in the ovaries 5 days later. Under these conditions, CTLs are exclusively responsible for antiviral protection (39, 40). Untreated HBcAg p33-VLPs and HBcAg p33-VLPs mixed with nonprotected CpG induced low-level protection (Fig. 4⇓A). In contrast, HBcAg p33-VLPs mixed with chemically protected ODN1668pt induced protection, confirming earlier results (27). Importantly, HBcAg p33-VLPs loaded with either ODN1668po or ODN1668pt also induced strong CTL protection. Thus, normal, chemically nonprotected DNA containing CpG motifs was only able to enhance CTL responses if packaged into VLPs. Protective responses against challenge infection with LCMV were also assessed, leading to comparable results (data not shown).

FIGURE 4.
FIGURE 4.

HBcAg and Qβ p33-VLPs packaged with CpG induce protective CTL responses. A, Mice were left untreated or immunized with 100 μg of HBcAg p33-VLP alone, 20 nmol ODN1668pt alone, 100 μg of p33-VLPs mixed together with 20 nmol ODN1668pt, 100 μg of p33-VLPs containing packaged ODN1668pt, 100 μg of p33-VLPs mixed together with 20 nmol ODN1668po, or 100 μg of p33-VLPs containing packaged ODN1668po. Twelve days after priming, mice were challenged i.p. with recombinant vaccinia virus expressing LCMV-GP (1.5 × 106 PFU), and viral titers were assessed in ovaries of infected mice 5 days after challenge infection. One representative experiment of two is shown. ∗, p < 0.05. B, Viral titers in mice treated s.c. with 100 μg of HBcAg p33-VLPs packaged with protected (pt) and unprotected (po) ODN1668, ODN1585, and ODN1826, and subsequently challenged with Vacc-G2 (1.5 × 106 PFU). ∗, p < 0.05. C, Viral titers in mice treated s.c. with 90 μg of Qβ p33-VLPs mixed or packaged with ODNG3po, ODNG6po, and subsequently challenged with Vacc-G2 (1.5 × 106 PFU). One representative experiment of two is shown. ∗, p < 0.05. D, Protection from a challenge infection with Vacc-G2 (4 × 106 PFU) in mice immunized s.c. with 100 μg of packaged HBcAg p33-VLPs or with liposomes. Liposomes packaged with p33 peptide (100 or 5 μg) alone or in combination with ODN1668pt (10 or 0.5 nmol) were tested. Five mice per group were used. ∗, p < 0.05 (two-tailed Student’s t test). The detection limit for A and D was of 50 PFU per ovaries; for B and C, 10 PFU.

To further compare the immunogenicity of packaged ODN1668pt with mixed ODN1668pt, groups of C57BL/6 mice were primed s.c. with 100 μg HBcAg p33-VLPs alone, packaged, or, alternatively, mixed with ODN1668pt. Nine days after immunization, primary ex vivo cytotoxicity was tested in a 5-h 51Cr release assay (Fig. 5⇓). The best response was obtained for HBcAg p33-VLPs packaged with CpGs with a median ex vivo killing of ∼30% at an E:T ratio of 90 (Fig. 5⇓C). The corresponding killing obtained by the mixed CpGs was only in the range of 15% (Fig. 5⇓B). These results confirm values obtained by p33 tetramer staining of blood lymphocytes, in which a ∼2-fold increased frequency of p33-specific CD8+ T cells was observed.

FIGURE 5.
FIGURE 5.

Primary ex vivo cytotoxicity induced by HBcAg p33-VLPs packaged with CpGs. Groups of C57BL/6 mice were s.c. primed with 100 μg of HBcAg p33-VLP given alone (A), mixed with 20 nmol ODN1668pt (B), or, alternatively, packaged with ODN1668pt (C). Nine days later, spleen cells were tested for direct ex vivo CTL activity in a 5-h 51Cr release assay on p33-pulsed (filled symbols) or on unpulsed (open symbols) EL-4 target cells at the indicated E:T cell ratios. Two dilution series of effector cells per mouse were performed. A, Two mice per group were used, whereas in B and C data from four mice per group are shown.

Although HBcAg p33-VLPs packaged with CpGs with normal phosphodiester bonds (ODN1668po) protected mice from Vacc-G2 infection, they were less potent than HBcAg p33-VLPs containing protected ODN1668pt (Fig. 4⇑A). To investigate whether these results depended on the unprotected backbone of the CpGs or were sequence specific, a vaccinia challenge assay was performed by comparing side by side three different sequences of protected and unprotected CpGs (ODN1668, ODN1585, and ODN1826) packaged into HBcAg p33-VLPs (s.c., 100 μg) (Fig. 4⇑B). All treated mice were significantly protected, whereas no relevant differences were observed between protected and unprotected CpGs, except for ODN1668po, in which again a slightly lower level of protection was observed. To further analyze the effectiveness of packaged unprotected CpGs, we packaged two different CpG sequences (ODNG3po and ODNG6po) into Qβ p33-VLPs and used these for immunization (90 μg, s.c.). CTL responses were assessed by determining numbers of p33 tetramer-positive T cells and protection against Vacc-G2 infection (Figs. 3⇑C and 4C). For both CpG sequences, significantly improved immune responses were obtained when the oligonucleotides were packaged inside the Qβ capsids. In fact, mixed CpGs induced specific CD8+ T cell frequencies lower than 3%, whereas median numbers for packaged ODNs surpassed 13% (Fig. 3⇑C). Correspondingly, complete viral protection was observed only for Qβ p33-VLPs packaged with ODNG6po or ODNG3po (Fig. 4⇑C). In conclusion, these results strengthen the hypothesis that selected packaged CpG sequences with normal phosphodiester bonds can generate strong and protective immunity.

As expected, mice immunized s.c. with 100 μg HBcAg p33-VLPs either packaged or mixed with a DNA oligonucleotide lacking CpG motifs (ODNLiang2pt) were not significantly protected from viral replication upon challenge infection with Vacc-G2 (data not shown). This indicates that the immune-stimulating activity of DNA oligonucleotides depends on the presence of CpG motifs into the sequence.

Packaging CpGs into HBcAg p33-VLPs results in better immune responses than packaging CpGs in liposomes

To compare the immunogenicity of CpGs packaged into p33-VLPs or liposomes (31, 41), groups of mice were vaccinated s.c. with 100 μg HBcAg p33-VLPs packaged with 0.6 nmol ODN1668pt or with small unilamellar liposomes (mean diameter of 120 ± 30 nm) packaged with 100 μg p33 peptide alone or together with 10 nmol ODN1668pt. Two additional groups of mice were primed with liposomes packaged with 5 μg p33 peptide together with 10 or 0.5 nmol ODN1668pt. Note that 5 μg of p33 epitope and 0.6 nmol CpGs are delivered per HBcAg p33-VLP injection; therefore, a direct comparison between VLPs and liposomes (with the same amounts of Ag and adjuvant) corresponds to a liposome preparation containing 5 μg p33 and 0.5 nmol CpGs. The induced immune responses were assessed next by measuring frequencies of p33-specific T cells and protection against infection with Vacc-G2 (4 × 106 PFU, i.p.) (Figs. 3⇑D and 4D). Packaged HBcAg p33-VLPs induced 6.8 ± 1.1% specific CD8+ T cells and protected mice most efficiently against infection with recombinant vaccinia virus, whereas liposomes packaged with p33 peptide and 10 nmol ODN1668pt generated 3.1 ± 1.2% and 2.9 ± 0.4% of specific CD8+ T cells, for the high and low p33 doses, respectively. The viral protection was intermediate, but still significant when compared with untreated controls. However, the mice treated with liposomes containing 10 nmol CpGs suffered from splenomegaly as side effect due to the high dose of CpGs used (data not shown; see also below). Liposomes packaged with 5 μg p33 together with 0.5 nmol ODN1668pt were by contrast not immunogenic. These data indicate that VLPs are better Ag and adjuvant carriers for the generation of specific T cell responses than liposomes.

A single vaccination with CpG-loaded HBcAg p33-VLPs can cure mice with established fibrosarcoma tumors

To test whether vaccination with packaged CpG may be able to eradicate established tumors, solid fibrosarcoma tumors were transplanted into C57BL/6 mice that were therapeutically vaccinated s.c. with 100 μg HBcAg p33-VLPs packaged with ODN1668pt once the tumors had reached a size of >0.64 cm2, reflecting a large tumor burden. In a first experiment, in six of eight mice, established tumors were completely eradicated and mice remained tumor free for at least 3 mo until the experiment was stopped (Fig. 6⇓A). The other two treated mice controlled the tumor initially, as indicated by a reduction in tumor size early after vaccination. However, tumors started to grow again at a later time point. RT-PCR analysis of isolated tumor tissue demonstrated absence of LCMV-GP expression (data not shown), suggesting that the tumor had escaped CTL-mediated immunosurveillance by loss of LCMV-GP expression or that the cell line, even after cultivation in selection medium was not completely monoclonal (42). In contrast, all five untreated control mice failed to eliminate the tumor. In a second experiment, 7 of 10 mice treated with HBcAg p33-VLPs packaged with ODN1668pt were able to eliminate the tumors (Fig. 6⇓B). Again, individual mice (3 of 10) after initially responding to the therapy were not able to control tumor growth in the long-term. RT-PCR analysis of tumor tissue confirmed the absence or loss of LCMV-GP expression by tumor cells in 2 of 3 mice (data not shown). In both experiments, Student’s t tests confirmed the significance of the data (p < 0.05 when comparing treated vs untreated mice). Moreover, HBcAg p33-VLPs given alone or wild-type VLPs (not containing the p33 epitope) packaged with ODN1668pt failed to noticeably inhibit tumor progression (4 of 5 and 3 of 4 mice did not control the tumor, respectively; p > 0.05 by comparing treated vs untreated mice). Note that Kawarada et al. (43) extensively investigated the efficiency of free CpGs alone for the treatment of various tumors and could show that only repeated peritumoral CpG injections (ODN1668pt) inhibited tumor growth, whereas systemic CpG application had only limited effects.

FIGURE 6.
FIGURE 6.

A and B, Packaged CpGs can eradicate established tumors. Solid fibrosarcoma tumors expressing the LCMV-GP were transplanted into C57BL/6 mice. Therapy, consisting of a single s.c. injection of 100 μg of HBcAg p33-VLPs containing packaged ODN1668pt (•), was applied once the tumors had reached a size of >0.64 cm2. A control group of mice was left untreated (○). Value of p < 0.05 (two-tailed Student’s t test) were observed by comparing treated vs untreated mice. Mice in which tumors were reaching sizes of >3–4 cm2 were sacrificed for regulatory reasons. Note that no LCMV-GP expression could be detected by RT-PCR in the tumors of the two vaccinated mice that failed to control the tumor in experiment A and in two of three that failed therapy in experiment B. A and B, Represent two independent experiments.

Packaging of CpGs into HBcAg p33-VLPs reduces systemic toxicity induced by CpGs with phosphorothioate backbone

Induction of splenomegaly is a major side effect of CpGs with DNase-resistant phosphorothioate backbones in animal models and especially in rodents. We therefore analyzed spleen weights 12 days (Fig. 7⇓) and 6 days (data not shown) after immunization. Although mice vaccinated s.c. with 100 μg HBcAg p33-VLPs containing packaged ODN1668pt exhibited normal spleen weights, cellularity (Fig. 7⇓, A and B), and histology (Fig. 7⇓, C–F), spleens from mice immunized s.c. with 100 μg HBcAg p33-VLPs mixed with ODN1668pt (20 nmol) exhibited 4-fold increased spleen sizes, and histological analysis indicated that white pulp (dark violet-colored regions) and red pulp (regions appearing red) had largely dissolved. Thus, packaging phosphothioester-stabilized CpGs into VLPs preserved their adjuvant properties, but significantly reduced their systemic toxicity.

FIGURE 7.
FIGURE 7.

Free CpGs, but not packaged CpGs, induce splenomegaly. Mice were left untreated or immunized s.c. with 100 μg of HBcAg p33-VLPs alone, 20 nmol ODN1668pt, 100 μg of p33-VLPs mixed together with 20 nmol ODN1668pt, or 100 μg of p33-VLPs containing packaged ODN1668pt. Twelve days later, spleens were isolated, and spleen weights (A) and splenic cellularity (B) were assessed. The mean ± SD (n = 4) are shown. ∗, p < 0.05 (two-tailed Student’s t test). C–F, Formalin-fixed spleen sections of naive mice (C) or mice primed with 100 μg of p33-VLPs given alone (D) or mixed together with 20 nmol ODN1668pt (E) or containing packaged ODN1668pt (F) were stained with H&E. Similar results were obtained if spleens were analyzed 6 days after immunization (data not shown). One experiment of three similar experiments is shown for A and B.

Discussion

CpGs are among the most potent immune-stimulating agents known to date. In this study, we show that the potency of CpGs can be further increased by packaging the oligonucleotides into VLPs. Packaged CpGs are protected from degradation and therefore exhibit improved pharmacokinetics. In addition, because VLPs are primarily taken up by DCs and Mφ, the CpGs are delivered in a targeted fashion into the optimal APCs. As a consequence, strong CTL responses may be induced in mice in the absence of systemic side effects.

Although CpGs are powerful immune activators, they induce serious side effects such as systemic immune activation and splenomegaly. To minimize the toxic effects of CpGs and deliver them in a more specific way, in this study we established a simple and effective method for packaging short CpG-containing DNA oligonucleotides into HBcAg and Qβ VLPs. These VLPs have in fact two important characteristics in common, which allow for the packaging of the immunostimulatory DNA: 1) the capsids possess VLP-internal arginine repeats able to bind nucleic acids, and 2) the VLPs exhibit pores of ∼2 nm in diameter that permit the diffusion of small enzymes such as RNase A and oligonucleotides into the particles (36, 37, 38). Although CpG toxicity and other side effects in humans may be less severe than in rodents, no definitive toxicological data are available yet. Our results indicate that packaging CpGs into VLPs allows reduction of side effects in three different ways: 1) small doses of CpGs are effective at enhancing CTL responses; 2) because professional APCs are the primary cell types known to efficiently take up particles, DCs and Mφ will probably be the major cell populations exposed to the CpGs; and 3) unmodified DNA oligonucleotides may be used, which have a more favorable toxicity profile than the chemically stabilized CpGs.

This study shows that packaging CpGs into VLPs could not only reduce their side effects, but also increased their immune-stimulating effects. In particular, HBcAg p33-VLPs packaged with ODN1668pt were at least 2-fold more immunogenic than p33-VLPs mixed with CpGs, as measured by tetramer-staining and in ex vivo 51Cr release experiments. The results were even more pronounced when Qβ p33-VLPs were packaged or mixed with unmodified DNA (ODNG3po and ODNG6po). In fact, whereas free CpG coadministered with p33-VLPs generated barely detectable CTL frequencies and did not protect mice from recombinant vaccinia virus challenge, p33-VLPs packaged with CpGs induced very high frequencies of p33-specific T cells (above 10%) and fully protected mice from Vacc-G2 virus replication.

The potency of CpG-loaded VLPs for induction of CTLs was directly compared with a live vaccine, namely recombinant vaccinia virus expressing the LCMV-GP. Surprisingly, the VLPs were much more potent than the recombinant vaccinia virus, perhaps because vaccinia virus simultaneously triggers CTL responses against many epitopes, while it is unlikely that the small VLP monomers (∼20 kDa) harbor many CTL epitopes. Liposomes are also known to be an effective Ag and adjuvant delivery system for the induction of CTL responses (31, 41, 44). This prompted us to compare side by side the efficiency of p33-VLPs packaged with CpGs and liposomes encapsulated with p33 peptide and CpGs. Confirming previous studies (31, 41, 44), liposomes significantly enhanced the immunogenicity of peptides. Mice immunized with liposomes containing 100 μg p33 with 10 nmol CpGs generated ∼2–4% of p33-specific T cells and were partly protected from Vacc-G2. Nevertheless, when the amounts of Ag and adjuvant present in the liposomes were adjusted to doses typically packaged into p33-VLPs, liposomes were much less effective than VLPs. It may be possible that the size of the liposomes used (mean diameter of 120 ± 30 nm) was less suitable for efficient uptake by APCs compared with p33-VLPs. Alternatively, protein structures may be more easily taken up by DCs than lipid particulates. In addition, it cannot be excluded that liposomes may fuse not only with cellular membranes of professional APCs, but also of normal tissue cells, and therefore deliver the Ag and the adjuvant in a less specific way.

In conclusion, VLPs loaded with CpGs are able to induce potent T cell responses, reaching levels usually obtained only with live vaccines. In fact, VLPs together with CpGs exhibit the immunologically relevant feature of live viruses, which is efficient Ag presentation by activated professional APCs. VLPs in fact exhibit the proper particle size for phagocytosis, facilitating cross-presentation by professional APCs. At the same time, these APCs are activated by the CpGs liberated within the endosome. This leads to the presentation of the relevant Ags by activated APCs, allowing the induction of optimal T cell responses.

These properties may facilitate the development of human T cell vaccines. In fact, induction of strong CTL responses in humans is still a major problem for vaccination against tumors or chronic viral infections using conventional vaccination strategies. CpG packaged into epitope-carrying VLPs may be ideal candidates for generating therapeutic antiviral or tumor vaccines that have a tolerable side-effect spectrum. Because we packaged CpG into particles formed by the HBcAg, these VLPs may themselves be a promising candidate vaccine for treatment of chronic hepatitis B infection, afflicting globally >350 million people (www.who.int/inf-fs/en/fact204.html).

Acknowledgments

We thank Franziska Lechner, Edwin Meijerink, Manfred Kopf, Hans Hengartner, and Rolf Zinkernagel for helpful discussions, and Edit Horvath, Petra Jäger, and Gudrun Christiansen for excellent technical assistance.

Footnotes

  • ↵1 Address correspondence and reprint requests to Dr. Martin F. Bachmann, Cytos Biotechnology, Wagistrasse 25, 8952 Schlieren-Zürich, Switzerland. E-mail address: martin.bachmann{at}cytos.com

  • ↵2 Abbreviations used in this paper: CpG, nonmethylated CG motif; DC, dendritic cell; HBcAg, hepatitis B core Ag; LCMV, lymphocytic choriomeningitis virus; LCMV-GP, LCMV glycoprotein; Mφ, macrophage; ODN, oligodeoxynucleotide; po, phosphodiester; pt, phosphorothioate; Vacc-G2, recombinant vaccinia virus expressing glycoprotein of LCMV; VLP, virus-like particle; VLPDEL, VLP composed of HBcAg monomers in which C-terminal region rich in arginine repeats is deleted.

  • Received July 14, 2003.
  • Accepted November 17, 2003.

References

  1. ↵
    Janeway, C. A., Jr. 1989. Approaching the asymptote? Evolution and revolution in immunology. Cold Spring Harbor Symp. Quant. Biol. 54:1.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    Zinkernagel, R. M.. 1996. Immunology taught by viruses. Science 271:173.
    OpenUrlAbstract
  3. ↵
    Bachmann, M. F., R. M. Zinkernagel, A. Oxenius. 1998. Immune responses in the absence of costimulation: viruses know the trick. J. Immunol. 161:5791.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    Medzhitov, R., C. A. Janeway, Jr. 1997. Innate immunity: impact on the adaptive immune response. Curr. Opin. Immunol. 9:4.
    OpenUrlCrossRefPubMed
  5. ↵
    Medzhitov, R.. 2001. CpG DNA: security code for host defense. Nat. Immun. 2:15.
    OpenUrlCrossRef
  6. ↵
    Akira, S., K. Takeda, T. Kaisho. 2001. Toll-like receptors: critical proteins linking innate and acquired immunity. Nat. Immun. 2:675.
    OpenUrlCrossRefPubMed
  7. ↵
    Wagner, H.. 2001. Toll meets bacterial CpG-DNA. Immunity 14:499.
    OpenUrlCrossRefPubMed
  8. ↵
    Banchereau, J., R. M. Steinman. 1998. Dendritic cells and the control of immunity. Nature 392:245.
    OpenUrlCrossRefPubMed
  9. ↵
    Lipford, G. B., M. Bauer, C. Blank, R. Reiter, H. Wagner, K. Heeg. 1997. CpG-containing synthetic oligonucleotides promote B and cytotoxic T cell responses to protein antigen: a new class of vaccine adjuvants. Eur. J. Immunol. 27:2340.
    OpenUrlCrossRefPubMed
  10. ↵
    Roman, M., E. Martin-Orozco, J. S. Goodman, M. D. Nguyen, Y. Sato, A. Ronaghy, R. S. Kornbluth, D. D. Richman, D. A. Carson, E. Raz. 1997. Immunostimulatory DNA sequences function as T helper-1-promoting adjuvants. Nat. Med. 3:849.
    OpenUrlCrossRefPubMed
  11. ↵
    Davis, H. L., R. Weeratna, T. J. Waldschmidt, L. Tygrett, J. Schorr, A. M. Krieg, R. Weeranta. 1998. CpG DNA is a potent enhancer of specific immunity in mice immunized with recombinant hepatitis B surface antigen. J. Immunol. 160:870.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    Chu, R. S., O. S. Targoni, A. M. Krieg, P. V. Lehmann, C. V. Harding. 1997. CpG oligodeoxynucleotides act as adjuvants that switch on T helper 1 (Th1) immunity. J. Exp. Med. 186:1623.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Sun, S., H. Kishimoto, J. Sprent. 1998. DNA as an adjuvant: capacity of insect DNA and synthetic oligodeoxynucleotides to augment T cell responses to specific antigen. J. Exp. Med. 187:1145.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    Macfarlane, D. E., L. Manzel. 1998. Antagonism of immunostimulatory CpG-oligodeoxynucleotides by quinacrine, chloroquine, and structurally related compounds. J. Immunol. 160:1122.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    Yi, A. K., J. H. Chace, J. S. Cowdery, A. M. Krieg. 1996. IFN-γ promotes IL-6 and IgM secretion in response to CpG motifs in bacterial DNA and oligodeoxynucleotides. J. Immunol. 156:558.
    OpenUrlAbstract
  16. ↵
    Takeshita, F., C. A. Leifer, I. Gursel, K. J. Ishii, S. Takeshita, M. Gursel, D. M. Klinman. 2001. Cutting edge: role of Toll-like receptor 9 in CpG DNA-induced activation of human cells. J. Immunol. 167:3555.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    Hemmi, H., O. Takeuchi, T. Kawai, T. Kaisho, S. Sato, H. Sanjo, M. Matsumoto, K. Hoshino, H. Wagner, K. Takeda, S. Akira. 2000. A Toll-like receptor recognizes bacterial DNA. Nature 408:740.
    OpenUrlCrossRefPubMed
  18. ↵
    Hacker, H.. 2000. Signal transduction pathways activated by CpG-DNA. Curr. Top. Microbiol. Immunol. 247:77.
    OpenUrlPubMed
  19. ↵
    Akira, S., K. Hoshino, T. Kaisho. 2000. The role of Toll-like receptors and MyD88 in innate immune responses. J. Endotoxin Res. 6:383.
    OpenUrlCrossRef
  20. ↵
    Lipford, G. B., K. Heeg, H. Wagner. 1998. Bacterial DNA as immune cell activator. Trends Microbiol. 6:496.
    OpenUrlCrossRefPubMed
  21. ↵
    Krieg, A. M.. 2000. The role of CpG motifs in innate immunity. Curr. Opin. Immunol. 12:35.
    OpenUrlCrossRefPubMed
  22. ↵
    Sparwasser, T., T. Miethke, G. Lipford, K. Borschert, H. Hacker, K. Heeg, H. Wagner. 1997. Bacterial DNA causes septic shock. Nature 386:336.
    OpenUrlCrossRefPubMed
  23. ↵
    Sparwasser, T., L. Hultner, E. S. Koch, A. Luz, G. B. Lipford, H. Wagner. 1999. Immunostimulatory CpG-oligodeoxynucleotides cause extramedullary murine hemopoiesis. J. Immunol. 162:2368.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    Hartmann, G., G. J. Weiner, A. M. Krieg. 1999. CpG DNA: a potent signal for growth, activation, and maturation of human dendritic cells. Proc. Natl. Acad. Sci. USA 96:9305.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    Sands, H., L. J. Gorey-Feret, A. J. Cocuzza, F. W. Hobbs, D. Chidester, G. L. Trainor. 1994. Biodistribution and metabolism of internally 3H-labeled oligonucleotides. I. Comparison of a phosphodiester and a phosphorothioate. Mol. Pharmacol. 45:932.
    OpenUrlAbstract
  26. ↵
    Stein, C. A., Y. C. Cheng. 1993. Antisense oligonucleotides as therapeutic agents: is the bullet really magical?. Science 261:1004.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    Storni, T., F. Lechner, I. Erdmann, T. Bachi, A. Jegerlehner, T. Dumrese, T. M. Kundig, C. Ruedl, M. F. Bachmann. 2002. Critical role for activation of antigen-presenting cells in priming of cytotoxic T cell responses after vaccination with virus-like particles. J. Immunol. 168:2880.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Jegerlehner, A., A. Tissot, F. Lechner, P. Sebbel, I. Erdmann, T. Kundig, T. Bachi, T. Storni, G. Jennings, P. Pumpens, et al 2002. A molecular assembly system that renders antigens of choice highly repetitive for induction of protective B cell responses. Vaccine 20:3104.
    OpenUrlCrossRefPubMed
  29. ↵
    Kozlovska, T. M., I. Cielens, I. Vasiljeva, A. Strelnikova, A. Kazaks, A. Dislers, D. Dreilina, V. Ose, I. Gusars, P. Pumpens. 1996. RNA phage Qβ coat protein as a carrier for foreign epitopes. Intervirology 39:9.
    OpenUrlPubMed
  30. ↵
    Cielens, I., V. Ose, I. Petrovskis, A. Strelnikova, R. Renhofa, T. Kozlovska, P. Pumpens. 2000. Mutilation of RNA phage Qβ virus-like particles: from icosahedrons to rods. FEBS Lett. 482:261.
    OpenUrlCrossRefPubMed
  31. ↵
    Ludewig, B., F. Barchiesi, M. Pericin, R. M. Zinkernagel, H. Hengartner, R. A. Schwendener. 2000. In vivo antigen loading and activation of dendritic cells via a liposomal peptide vaccine mediates protective antiviral and anti-tumor immunity. Vaccine 19:23.
    OpenUrlCrossRefPubMed
  32. ↵
    Battegay, M., S. Cooper, A. Althage, J. Banziger, H. Hengartner, R. M. Zinkernagel. 1991. Quantification of lymphocytic choriomeningitis virus with an immunological focus assay in 24- or 96-well plates. J. Virol. Methods 33:191. [Published errata appear in 1991 J. Virol. Methods 35:115 and 1992 J. Virol. Methods 38:263.].
    OpenUrlCrossRefPubMed
  33. ↵
    Bachmann, M. F., T. M. Kundig, G. Freer, Y. Li, C. Y. Kang, D. H. Bishop, H. Hengartner, R. M. Zinkernagel. 1994. Induction of protective cytotoxic T cells with viral proteins. Eur. J. Immunol. 24:2228.
    OpenUrlCrossRefPubMed
  34. ↵
    Ruedl, C., T. Storni, F. Lechner, T. Bachi, M. Bachmann. 2002. Cross-presentation of virus-like particles by skin-derived CD8– dendritic cells: a dispensable role for TAP. Eur. J. Immunol. 32:818.
    OpenUrlCrossRefPubMed
  35. ↵
    Riedl, P., D. Stober, C. Oehninger, K. Melber, J. Reimann, R. Schirmbeck. 2002. Priming Th1 immunity to viral core particles is facilitated by trace amounts of RNA bound to its arginine-rich domain. J. Immunol. 168:4951.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    Conway, J. F., N. R. Watts, D. M. Belnap, N. Cheng, S. J. Stahl, P. T. Wingfield, A. C. Steven. 2003. Characterization of a conformational epitope on hepatitis B virus core antigen and quasiequivalent variations in antibody binding. J. Virol. 77:6466.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    Wynne, S. A., R. A. Crowther, A. G. Leslie. 1999. The crystal structure of the human hepatitis B virus capsid. Mol. Cell 3:771.
    OpenUrlCrossRefPubMed
  38. ↵
    Crowther, R. A., N. A. Kiselev, B. Bottcher, J. A. Berriman, G. P. Borisova, V. Ose, P. Pumpens. 1994. Three-dimensional structure of hepatitis B virus core particles determined by electron cryomicroscopy. Cell 77:943.
    OpenUrlCrossRefPubMed
  39. ↵
    Kagi, D., B. Ledermann, K. Burki, P. Seiler, B. Odermatt, K. J. Olsen, E. R. Podack, R. M. Zinkernagel, H. Hengartner. 1994. Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin-deficient mice. Nature 369:31.
    OpenUrlCrossRefPubMed
  40. ↵
    Walsh, C. M., M. Matloubian, C. C. Liu, R. Ueda, C. G. Kurahara, J. L. Christensen, M. T. Huang, J. D. Young, R. Ahmed, W. R. Clark. 1994. Immune function in mice lacking the perforin gene. Proc. Natl. Acad. Sci. USA 91:10854.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    Li, W. M., W. H. Dragowska, M. B. Bally, M. P. Schutze-Redelmeier. 2003. Effective induction of CD8+ T-cell response using CpG oligodeoxynucleotides and HER-2/neu-derived peptide co-encapsulated in liposomes. Vaccine 21:3319.
    OpenUrlCrossRefPubMed
  42. ↵
    Perez-Diez, A., P. J. Spiess, N. P. Restifo, P. Matzinger, F. M. Marincola. 2002. Intensity of the vaccine-elicited immune response determines tumor clearance. J. Immunol. 168:338.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    Kawarada, Y., R. Ganss, N. Garbi, T. Sacher, B. Arnold, G. J. Hammerling. 2001. NK− and CD8+ T cell-mediated eradication of established tumors by peritumoral injection of CpG-containing oligodeoxynucleotides. J. Immunol. 167:5247.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    Gursel, I., M. Gursel, K. J. Ishii, D. M. Klinman. 2001. Sterically stabilized cationic liposomes improve the uptake and immunostimulatory activity of CpG oligonucleotides. J. Immunol. 167:3324.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section