Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The stimulatory effects of DNA on immune cells are well described in the literature (Lipford et al., 1998), especially those associated with ODN in which the backbone has been modified to reduce molecular degradation by nucleases (Monteith et al., 1997). The degree of immune stimulation is variable depending upon the nucleic acid sequence; however, the presence of unmethylated CpG dinucleotides flanked by two 5' purines and two 3' pyrimidines induces the most potent response for PS ODN, which are molecules stabilized by substituting nonbridging oxygens with sulfur atoms to form a phosphorothioate backbone (Boggs et al., 1997). This immune stimulatory sequence (ISS) is recognized by macrophages, dendritic cells, and B cells as a danger signal indicating bacterial infection (Lipford et al., 1998; Krieg et al., 1999).

Macrophages are a critical cell type in the innate immune response. Their ability to rapidly respond to danger signals associated with pathogens provides a primary defense against infection, and the cytokines secreted by activated macrophages initiate and orchestrate the development of the adaptive immune response. Antigen presenting cells (APC), like macrophages, possess structurally conserved pattern recognition receptors capable of differentiating pathogen-associated molecular shapes, such as those described by the molecular structure of liposaccharide and peptidoglycan, from molecular forms common to the host (Medzhitov and Janeway, 2000). Bacterial DNA can be distinguished from host DNA because, in the former, CpG is present at the statistically expected frequency of 1 per 16 dinucleotides, whereas in the DNA of vertebrates CpG dimers occur at a quarter of this frequency and are also methylated at the 5' position of cytosine (Bird, 1986).

Hemmi et al. (2000) recently demonstrated that immune stimulatory CpG DNA is recognized by Toll-like receptor 9 (TLR-9). When immune cells are activated through the TLR-9 signaling pathway a strong T helper cell type 1 (Th1) response is induced, characterized by the secretion of IL-12 and IFN-, production of the antibody isotype IgG2a, and activation of cell-mediated immunity (Jakob et al., 1998). The strong Th1-biased response to CpG ODN has resulted in proposals for a number of therapeutic applications, many of which have since demonstrated proof of principle in animal models (Krieg et al., 1999). For example, CpG ODN may be effective as immune modulators of allergic responses (Kline et al., 1998; Tighe et al., 2000), where a dominant antibody (Th2) response is pathogenic. In such cases, induction of Th1 cytokines down-regulates the damaging Th2 response. Through activation of macrophages, CpG ODN also enhance the killing of intracellular pathogens that infect these cells (Walker et al., 1999), demonstrate promise as immune adjuvants when combined with weak antigens (Weiner et al., 1997; Hartmann et al., 2000), and have been shown to induce antitumor activity in part through the activation of NK cells (Dow et al., 1999).

Previously, we have described a structurally unique, lipid-based delivery system for the intravenous administration of negatively charged, nucleic acid oligomers such as antisense and ribozyme drugs. Referred to as stabilized antisense lipid particles (SALP), they enhance the circulation life time of ODN, protect their payload from degradation by nucleases, and can deliver more intact sequences to growing tumors and sites of inflammation than is possible using equivalent doses of free drug (Klimuk et al., 2000; Semple et al., 2000). It is also important to note that SALP target macrophages in vivo, in a manner similar to other liposome-like, particulate delivery systems (Juliano, 1986). Consequently, we predicted that SALP should exhibit a profound effect on the immune response if the encapsulated ODN contained an active CpG motif. In this study we have used an ODN sequence that is complementary to the initiation codon region of the human and mouse proto-oncogene c-myc, but also happens to contains a 5' CpG motif capable of inducing a strong Th1 response in mice. We demonstrate that SALP greatly enhance the activity of the ISS motif, enabling vigorous responses at ODN doses where free molecule is inactive. Moreover, because SALP fully protect ODN from nuclease degradation, they enable the use of unmodified, PO ODN as immune stimulants.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. DSPC and DODAP were purchased from Avanti Polar Lipids (Alabaster, AL), while cholesterol was from Sigma (St. Louis, MO). 1-O-(2'-(-methoxypolyethyleneglycol)succinoyl-2-N-myristoylsphingosine was synthesized by Dr. Zhao Wang (Inex Pharmaceuticals Corporation, Burnaby, BC, Canada). The ODN sequences used include a 15-mer c-myc ODN complementary to the initiation codon region of the human/mouse c-myc proto-oncogene mRNA (5'-AACGTTGAGGGGCAT-3'), referred to in the literature as LR-3280 or INX-3280; a 16-mer version of the same sequence INX-6295 with an additional 5' thymidine (5'-TAACGTTGAGGGGCAT-3'); and INX-3300, a control ODN to INX-3280 that maintains the same base composition as well as the four-guanosine sequence but disrupts the CpG motif (5'-AAGCATACGGGGTGT-3'). INX-3280 and INX-6295 were synthesized by Inex USA (Hayward, CA), while INX-3300 was purchased from Boston Biosystems, Inc. (Bedford, MA). All sequences were prepared with either PS or PO backbones. Female, 6-week-old ICR mice were obtained from Harlan Sprague-Dawley (Indianapolis, IN) and were quarantined for at least 1 week prior to use.

SALP. SALP composed of DSPC/cholesterol/DODAP/1-O-(2'-(-methoxypolyethyleneglycol)succinoyl-2-N-myristoylsphingosine (20:45:25:10, molar ratio) and encapsulated ODN were prepared as previously described (Semple et al., 2000). For PS ODN, 300 mM citrate buffer was used to dissolve the ODN, whereas for PO ODN the solubilizing buffer was 20 mM citrate, pH 4.0. Briefly, the lipid mixture dissolved in ethanol was added to the citrate solution of ODN (3.33 mg/ml) to give a final ethanol concentration of 40% (v/v). The mixture was freeze-thawed five times and extruded through two stacked 100-nm pore-sized filters using an extruder (Lipex Biomembranes, Vancouver, Canada) to make a homogeneous population of vesicles approximately 100 nm in diameter (Hope et al., 1985). The resulting vesicles were dialyzed for 2 h against citrate buffer to remove the ethanol then overnight against a 500-fold volume of HBS (150 mM NaCl, 20 mM HEPES, pH 7.5) to raise the external pH and neutralize DODAP located in the outer monolayer of the particles. Unencapsulated ODN was removed from the preparation by anion exchange chromatography using DEAE-Sepharose CL-6B. The final ODN-to-lipid ratio was calculated by measuring ODN concentration by A₂₆₀ and lipid content by phosphate assay (Bartlett and Lewis, 1970), assuming that the lipid mixture consisted of 20 mole percent DSPC. Because the phosphate on the ODN backbone would interfere with the lipid analysis, samples were first subjected to a Bligh and Dyer lipid extraction (Bligh and Dyer, 1959) followed by three water/methanol washes to remove residual ODN. The ODN-to-lipid ratio was typically 0.15 to 0.20 (w/w). Vesicle size, as determined by quasi-elastic light scattering using an NICOMP Submicron particle sizer (model 370), was approximately 120 ± 50 nm.

Free Monomer and Quadruplex ODN Preparation. ODN that contain at least four or more contiguous guanosine bases tend to combine to form G-tetrads or quadruplex structures (Jin et al., 1992). When INX-3280 is hydrated from a lyophilized powder with HBS at room temperature it exists as a mixture of quadruplexes and monomers in the mole ratio of 70:30 as determined by HPLC. It is simple to convert the mixture all to the monomer form by incubating it for 30 min at 65°C. The monomer content after this treatment is typically >99% by HPLC, consequently the monomer form was used throughout the study. Samples used for injections were prepared fresh daily.

HPLC Analysis. The proportion of INX-3280 in monomeric and quadruplex form was quantified by size exclusion chromatography-HPLC using a Pharmacia Superdex 75HR 10/30 gel filtration column equilibrated with a buffer containing 25 mM sodium phosphate, 0.25 mM Na₂EDTA, pH 7.5. The column was operated at a flow rate 0.75 ml/min for 30 min at room temperature and peaks were monitored at 260 nm.

Plasma Collection and Cytokine Analysis. ICR mice (7 weeks old at the start of the experiment) were injected intravenously with 0.2 ml of sample in HBS. At various times, the mice were killed by a terminal dose of anesthetic [3.2% (v/v) ketamine/0.8% (v/v) xylazine] and blood collected in Vacutainer tubes containing EDTA. The blood was centrifuged (500g for 10 min at 4°C) to pellet the blood cells and the plasma was isolated and frozen at 20°C until assayed. Plasma concentrations of IL-2, IL-4, IL-10, IL-12, IFN-, MCP-1, and TNF- were determined using commercial enzyme-linked immunosorbent assay kits (PharMingen, San Diego, CA). Some variability in the kinetics and magnitude (~2-fold) of cytokine induction between experiments was observed. Consequently, a single benchmark formulation (SALP INX-6295-PS) was included in all experiments as an internally consistent control group. Differences observed in the absolute levels of cytokines induced between experiments are probably due to variability in the response of the ICR mice from different litters. All data points represent the average from four mice ± 1 S.E.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Two CpG-containing ODN sequences were used throughout this study, a 15-mer (INX-3280) and a 16-mer (INX-6295). Both are antisense to the initiation codon of the murine and human c-myc proto-oncogene with the only difference between the two sequences being a thymidine added to the 5' end of INX-6295. However, no differences in the biological activity between the two ODN have been detected, both exhibit similar antisense activity against c-myc mRNA in vitro and the same immune (this report) and antitumor activity (data not shown). Each ODN sequence was synthesized in the PS and PO forms. In some experiments the 15-mer was used and in others the 16-mer, but because their activities were the same, and to make the text more readable, we simply refer to PS CpG ODN or PO CpG ODN. However, the code numbers are used to identify each ODN in the figure legends.

Plasma Cytokines Induced by Free and Encapsulated PS ODN. In the first experiment, ICR mice were injected i.v. with a 20-mg ODN/kg dose of PS-CpG ODN, either in free monomer form or as a SALP formulation. Cytokines common to both Th1 and Th2 pathways (IL-12, IFN-, IL-2, IL-4, IL-6, IL-10), as well as MCP-1 (a macrophage chemokine) and TNF- (an inflammatory mediator) were measured over a 24-h time course following administration. Of the Th1/Th2-associated cytokines, IL-12 and IFN- are strong promoters of Th1 responses, while IL-4 and IL-10 favor Th2. Free PS CpG ODN induced a significant increase in IL-12 concentration between 2 and 24 h after injection, with peak levels (a 20-fold increase compared with untreated mice) occurring at about 4 h (Fig. 1). MCP-1 (Fig. 1) and IL-10 (Fig. 2) levels were weakly increased 2- to 3-fold over baseline, while no significant differences were seen for IL-6 (Fig. 1), IFN- (Fig. 1), IL-2 (Fig. 2), IL-4 (Fig. 2), or TNF- (Fig. 2).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Plasma cytokines induced by free and encapsulated INX-6295-PS ODN. Seven-week-old ICR mice were injected i.v. with 20 mg/kg ODN of free INX-6295-PS ODN (), INX-6295-PS SALP (), or empty SALP (). The dose of the empty SALP (130 mg/kg) was equivalent to the lipid dose of INX-6295-PS SALP. At various times, the mice were killed by terminal anesthesia and blood drawn to isolate the plasma. Plasma levels of IL-12, MCP-1, IL-6, and IFN- were determined by enzyme-linked immunosorbent assay and the data are plotted as the average of four mice per group ± 1 S.E.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Plasma cytokines unaffected or weakly induced by free and encapsulated INX-6295-PS ODN. Seven-week-old ICR mice were injected i.v. with 20 mg/kg ODN of free INX-6295-PS ODN (), INX-6295-PS SALP (), or empty SALP (). The dose of the empty SALP (130 mg/kg) was equivalent to the lipid dose of INX-6295-PS SALP. At various times, the mice were killed by terminal anaethesia and blood drawn to isolate the plasma. Plasma levels of IL-10, TNF-, IL-4, and IL-2 were determined over a 24-h period. Data represent an average of four animals per group ± 1 S.E.

Effect of ODN Backbone on Plasma Cytokine Levels. The PS backbone is common to many antisense oligonucleotides because it is more resistant to nuclease degradation than the naturally occurring PO backbone. The PS structure remains a good substrate for RNase H and therefore target mRNA levels can still be reduced through an antisense mechanism (Fisher et al., 1993). Because of their sensitivity to nucleases, PO ODN are ineffective as antisense drugs. As expected, the PO CpG ODN is unable to induce an immune response when administered i.v. in the free form at a dose of 20 mg/kg (Fig. 3), presumably because it is degraded in the blood before it can enter macrophages (Semple et al., 2000). Previously, we have shown that PO ODN formulated in SALP are completely protected from interactions with extracellular proteins, including components of the complement cascade and nucleases, and are therefore stable in the circulation (Semple et al., 2000). This is consistent with the observation that a potent immune response can now be detected over the first 24 h following i.v. administration of PO CpG ODN SALP, with extensive stimulation of IL-12, MCP-1, IL-6, and IFN- (Fig. 3). As a comparison, mice were also treated with SALP containing PS CpG ODN in the same experiment and, as shown in Fig. 3, cytokine induction by encapsulated PO CpG ODN is even greater.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   PO ODN are more immunostimulatory than PS ODN over a 24-h time course when encapsulated in an SALP. ICR mice were injected with 20 mg/kg INX-3280-PO SALP (), INX-6295-PS SALP (), or free INX-3280-PO (), and plasma levels of IL-12, MCP-1, IL-6, and IFN- were measured over a 24-h time course. Data represent an average of four animals ± 1 S.E.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Plasma IFN- levels over an 8-day time course. Two IFN- induction phases were detected in ICR mice injected with 20 mg/kg INX-6295 PS or INX-3280-PO either in free form () or SALP-encapsulated (). Data represent an average of four animals ± 1 S.E.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Plasma IL-12 levels over an 8-day time course. ICR mice were injected with 20 mg/kg INX-6295 PS or INX-3280-PO either in free form () or SALP-encapsulated () and plasma IL-12 concentration was measured. The dotted line represent IL-12 levels in mice treated with HBS. Data represent an average of four animals ± 1 S.E.

Effect of ODN Sequence on Plasma Cytokine Levels. The ODN sequence used here exhibits antitumor activity in numerous animal models, which has been attributed to its antisense activity against c-myc mRNA (Leonetti et al., 1996; Citro et al., 1998). However, it is interesting to note that the negative control (nonantisense) sequence used in these reports maintained the contiguous four guanosines (4G) but disrupted the CpG motif. At the time these studies were conducted, the immune effects of CpG sequences were not widely known, but there was concern about potential nonantisense effects from 4G motifs, consequently the 4G run was maintained in the control sequence INX-3300 (Leonetti et al., 1996; Citro et al., 1998). Here we measured the ability of this control PS ODN to increase plasma IL-12 concentrations 4 to 8 h after administration. Elevated levels of plasma IL-12 were detected in mice treated with PS CpG ODN either as the free monomer (4 ± 1 ng/ml at 8 h) or formulated in SALP (15 ± 2 ng/ml at 8 h) compared with mice treated with HBS (0.6 + 0.1 ng/ml). However, no increase in plasma IL-12 was observed at either 4 or 8 h in mice treated with free monomer INX-3300 (0.59 ± 0.07 ng/ml) or encapsulated INX-3300 (0.7 ± 0.1 ng/ml) in which the CpG motif was corrupted, even though the 4G motif was intact.

Increased Potency of Encapsulated CpG ODN. In the studies described so far, a single dose of 20 mg of ODN/kg has been used, which was chosen because we have consistently found this to be a dose at which maximum antitumor activity is observed for free PS INX-6295 or PS INX-3280 (data not shown). In the following experiment SALP formulations of PS and PO CpG ODN were administered i.v. covering a dose range of 2 to 20 mg of ODN/kg, and cytokine levels measured at 8 h were compared with those induced by free ODN administered over a range of 10 to 60 mg of ODN/kg. The data clearly show that SALP greatly increase ODN potency with respect to macrophage activation as measured by IL-12 release (Fig. 6). Even at the lowest dose of encapsulated ODN tested (2 mg/kg) the concentration of plasma IL-12 is manyfold greater than an equivalent dose of free PS CpG ODN, reaching a plateau at levels not achieved by the free ODN, even at a dose of 60 mg/kg. Interestingly, when formulated in SALP, PO CpG ODN is even more potent than equivalent doses of encapsulated PS CpG ODN, inducing plasma IL-12 concentrations on the order of 50 ng/ml at a dose of 2 mg/kg and reaching a maximum of 90 to 100 ng/ml at 10 mg/kg, a level 4- to 5-fold greater than the PS formulation and more than 20-fold greater than the free PS sequence at the same dose. Plasma concentrations of IL-6, MCP-1, and IFN- were also measured and showed similar profiles to that seen for IL-12 in Fig. 6 (data not shown). In the free form PO CpG ODN did not exhibit any cytokine induction over the 10- to 60-mg/kg dose range.

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Effect of ODN dose on plasma cytokine induction. Mice were injected with various doses of INX-3280-PS SALP (), INX-3280-PO SALP (), free INX-3280-PS (), or free INX-3280-PO (), and plasma IL-12 concentrations were measured 8 h after injection. Data represent an average of four animals ± 1 S.E.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

In this study, we demonstrate that the immune response to a CpG-containing ODN is greatly enhanced when encapsulated in a structurally unique lipid particle (SALP) approximately 100 nm in diameter. Following i.v. administration, SALP containing PS CpG ODN significantly increased the plasma concentrations of IL-12, IFN-, IL-6, and MCP-1 compared with cytokine levels induced by the same dose of free ODN. Furthermore, SALP enabled the chemically unmodified PO CpG ODN to exert an effect that is even more potent than that measured for the same sequence with a PS backbone. These results are consistent with the hypothesis that SALP increase delivery of intact CpG motifs to macrophages and other APC where they are recognized as danger signals, causing cell activation and induction of a Th1-biased immune response.

SALP were previously developed to encapsulate anionic antisense ODN with high efficiency into well tolerated, lipid-based particles capable of delivering intact sequences to sites of disease (Klimuk et al., 1999; Semple et al., 2000). Although they exhibit many characteristics common to large unilamellar vesicles, they actually contain multiple layers of ODN/lipid and thus resemble small multilamellar vesicles (Semple et al., 2001). Consequently, despite their small size, each particle contains on the order of 2000 to 3000 ODN molecules. Liposomes are removed from the circulation by monocytes and macrophages (Juliano, 1986; Rao and Alving, 2000). This natural targeting process has been used to specifically deliver drugs to this cell population (Scherphof et al., 1987), including immunostimulatory components isolated from bacteria, the latter designed to induce cell-mediated cytotoxic reactions against tumors and intracellular infections (Killion and Fidler, 1998; Rao and Alving, 2000). Macrophages activated in this way release cytokines that in turn can stimulate the cytotoxic effects of NK cells and recruit T helper cells. The humoral immune response is also up-regulated in the presence of these cytokines through enhanced antigen presentation by APC and the presence of a milieu of activated T cells. In vivo, we observe that SALP containing fluorescently labeled ODN strongly distribute to tissue macrophages within the first 4 h similar to other liposome-based delivery systems (Killion and Fidler, 1998) and is consistent with macrophages being the primary effector cell population.

We conclude, therefore, that the increased immunogenicity of CpG-containing ODN encapsulated in lipid-based particles is largely the result of enhanced delivery to macrophages and other APC such as dendritic cells. Furthermore, we have shown previously that PO ODN encapsulated in lipid particles are protected from degradation by plasma nucleases (Semple et al., 2000), which probably explains the dramatic difference in cytokine induction observed here for free versus encapsulated PO CpG ODN (Figs. 3 and 4). Naked PO backbones are degraded within minutes following administration i.v. (Semple et al., 2000, 2001). Therefore, few intact CpG motifs would be expected to reach effector cells, consistent with the fact that in the free form PO ODN were unable to induce secretion of IL-12 (Fig. 3). In contrast, PS backbones are significantly more resistant to degradation, enabling a higher proportion of unencapsulated ISS to reach effector cells.

Our data also indicate that intracellular processing of the lipid-based carrier to expose encapsulated ODN to the TLR-9 receptors occurs rapidly because there is no apparent difference in the timing of IL-12 induction by encapsulated PO CpG ODN and unencapsulated PS CpG ODN (Fig. 1). It is also important to note that unlike ODN/cationic lipid complexes, which are normally used for intracellular delivery of DNA in vitro (Hope et al., 1998), SALP do not appear to be capable of disrupting endosomal structure to release ODN into the cytoplasm (B. Mui, unpublished observation). This supports the conclusions of others that the CpG pattern recognition receptors are located inside the endosomal pathway (Krieg et al., 1995; Hemmi et al., 2000). However, it should be noted that APC are capable of trafficking liposomes differently when they contain proteins (Rao and Alving, 2000), which results in cross-priming between the major histocompatibility complex II and I peptide presentation pathways. Therefore, APC may also be capable of releasing encapsulated ODN into the cytoplasm.

It is interesting to note that when delivery into effector cells has been accomplished, PO CpG ODN elicit a more potent response than the same dose of PS CpG ODN (Fig. 6), especially at the early time points (<24 h). This effect may be due to a greater affinity of the TLR-9 binding site for the natural PO backbone compared with the modified PS backbone. It is known, for example, that PS ODN exhibit weaker binding interactions with complementary mRNA targets and are poorer substrates for RNase H than their PO counterparts (Krieg and Stein, 1995). Consequently, one would expect the PO backbone to be a better activator of the TLR-9 signaling pathway than the corresponding PS backbone. It is also known that PS ODN undergo extensive nonspecific binding to proteins in vivo (Stein and Krieg, 1994; Krieg and Stein, 1995), which may reduce their effective, intracellular concentration compared with PO ODN.

Of the four cytokines measured in this study whose plasma concentrations were most significantly increased in response to encapsulated CpG ODN, IL-12 and IFN- are key to inducing antitumor effects and providing protection from infectious agents (Zimmermann et al., 1998; Dow et al., 1999; Golab and Zagozdzon, 1999; Walker et al., 1999). The early peak for IFN- occurs at 8 h compared with 4 h for IL-12. This is consistent with the initial release of the latter from macrophages, which in turn triggers the release of IFN- from NK cells (Chace et al., 1997). Of particular interest, however, is the appearance of a large, late induction of IFN- several days after the single bolus administration (Fig. 4). This second IFN- peak is considerably larger for encapsulated PS CpG ODN than it is for encapsulated PO CpG ODN. Smaller increases in IL-12 are detected prior to these second peaks and may indicate that the immune system has been altered or primed to respond more vigorously to this cytokine, possibly through expansion of NK cells (Bramson et al., 2000). The fact that the PS-induced late peak of IFN- is larger than the PO-induced peak might be a reflection of the increased stability of PS ODN over several days in vivo.

It has recently been reported that lipid/DNA complexes used for gene transfer in vivo also cause transient increases in the plasma concentrations of IL-12, IFN-, and TNF- as well as enhance IL-6 levels in the spleen, whereas an equivalent i.v. dose of naked plasmid DNA has no effect (Dow et al., 1999; Whitmore et al., 1999). In both studies, the antitumor activity of the lipid/DNA complex was correlated to the immunogenicity of the bacterial DNA rather than expression of the transgene. As noted earlier, the CpG-containing ODN used in the studies reported here is antisense to human (Leonetti et al., 1996; Citro et al., 1998) and murine c-myc mRNA (data not shown). An extensive literature exists showing that this ODN exhibits significant antitumor activity in a wide variety of preclinical animal models as a free PS-ODN (Leonetti et al., 1996; Calabretta and Skorski, 1997; Citro et al., 1998). We have confirmed these reports, showing that the PS ODN is efficacious in murine B16 melanoma, DoHH2 human B cell lymphoma, and murine C26 colon carcinoma animal models (M. J. Hope, S. C. Semple, S. K. Klimuk, and T. O. Harasym, unpublished data). However, there is clearly a significant immune component associated with the overall antitumor activity of this molecule. The large induction of IL-12 alone would be expected to exert a considerable antitumor effect, because this cytokine has been shown to exhibit antimetastatic and antiangiogenic activities in preclinical models (Golab and Zagozdzon, 1999). It is also interesting to note that the control sequence used in many of these studies is the same six base mismatch used here, in which the CpG motif is disrupted (Leonetti et al., 1996; Citro et al., 1998). This ODN does not show any antitumor or as demonstrated in this study, immune stimulatory activity in either the free or encapsulated form. Moreover, these data highlight the fact that the immune stimulation is CpG-dependent, while the four guanosines motif is not involved because this sequence is present in both the active and inactive sequences. Contiguous guanosine sequences can induce ODN to form higher ordered structures, such as quadruplexes, that may influence the biological behavior of ODN (Lee et al., 2000).

In conclusion, the results reported herein indicate that encapsulation of ISS in lipid-based particles greatly increases their immunogenicity and support further research into the use of SALP-containing CpG ODN in immune therapies. Because of their Th1 bias, PS CpG ODN are being clinically tested as bioresponse modifiers in oncology, adjuvants for protein-based vaccines, and as agents for protection from infection. We believe the virus-like characteristics of SALP, which include protection of the DNA payload from extracellular degradation and delivery to APC, combine to enhance the innate immune response to encapsulated CpG ISS. Furthermore, the lipid envelope represents a platform to which a variety of protein, carbohydrate, and lipid antigens can be readily associated, to form a potent vaccine particle.