Cationic lipid-DNA complexes (CLDC) are reported to be safe and effective for systemic gene delivery, particularly to the lungs. However, we observed that i.v. injection of CLDC induced immunologic effects not previously reported. We found that even very low doses of CLDC administered i.v. induced marked systemic immune activation. This response included strong up-regulation of CD69 expression on multiple cell types and systemic release of high levels of Th1 cytokines, from both lung and spleen mononuclear cells. CLDC were much more potent immune activators on a per weight basis than either LPS or poly(I:C). The remarkable potency of CLDC appeared to result from enhancement of the immune stimulatory properties of DNA, since cationic lipids alone were without immune stimulatory activity. Systemic treatment with CLDC controlled tumor growth and significantly prolonged survival times in mice with metastatic pulmonary tumors. NK cells accumulated to high levels in the lungs of CLDC-treated mice, were functionally activated, and released high levels of IFN-γ. The antitumor activity induced by CLDC injection was dependent on both NK cells and IFN-γ. Thus, DNA complexed to cationic liposomes becomes highly immunostimulatory and capable of inducing strong antitumor activity when administered systemically.

Reports of efficient in vivo pulmonary gene transfer using i.v. administered cationic lipid-DNA complexes (CLDC)3 have generated substantial interest, due in part to the apparent safety and repeated dosing capabilities of this form of gene delivery (1, 2, 3, 4, 5). Intravenous injection of CLDC results in preferential transfection of pulmonary vascular endothelial cells (2, 4). Adenovirus-mediated gene delivery, though extremely efficient, has been plagued by the inherent immunogenicity of adenoviral vectors, and the immune response to the adenoviral vector generally precludes repeated gene administration (6, 7). Systemic gene delivery using lipid-DNA complexes has not been previously associated with systemic immunologic sequellae, though there has been mention of toxicity (3, 4). Local administration of CLDC directly into the lung via intratracheal instillation was recently reported to induce cellular infiltration and production of the cytokines IL-12 and IFN-γ in lung tissues (8, 9).

Observations made during systemic gene delivery studies using CLDC prompted us to more closely examine the immunologic effects of systemically administered CLDC. We observed that i.v. injection of CLDC, even using noncoding plasmid DNA vectors, consistently induced strong antitumor effects in mice with experimental lung tumor metastases. We also observed adverse effects in CLDC-injected mice that resembled those induced by immune activation (depression, piloerection, dehydration) in the absence of any remarkable pulmonary pathology. Thus, i.v. injection of CLDC appeared to induce systemic effects suggestive of strong, nonspecific immune stimulation.

Therefore, we investigated immunologic responses to systemic CLDC administration, in particular immune activation, cytokine release, and antitumor activity. We report here that CLDC injected i.v. triggered release of high levels of IL-12 and IFN-γ, as well as accumulation and activation of NK cells in lung and spleen tissues. Thus, CLDC administered i.v. appear to be extremely potent inducers of innate immune responses. Immune activation by CLDC is also responsible for inhibiting the growth of established tumors.

Mice used in these studies were either 8- to 12-wk-old female C57BL/6 mice, 8- to 12-wk-old female BALB/c mice, or 8- to 12-wk-old ICR mice purchased from either The Jackson Laboratory (Bar Harbor, ME) or Harlan Sprague-Dawley (Indianapolis, IN). IFN-γ gene-disrupted mice, bred on a C57BL/6 background, were obtained from The Jackson Laboratory. Mice with a disrupted and nonfunctional recombinase activating gene type 2 (RAG-2 KO) were a kind gift from Dr. Andre Augustin (National Jewish). Protocols for these experiments were approved by the Institutional Animal Care and Use Committee at the National Jewish Medical and Research Center.

FBS, LPS, and poly(I:C) were purchased from Sigma (St. Louis, MO). Cell culture medium (modified Eagle’s medium) was prepared by Life Technologies (Gaithersburg, MD).

The eukaryotic expression vector PCR3.1 (Invitrogen, San Diego, CA) without a gene insert was used as a source of plasmid DNA. The plasmid was purified from Escherichia coli, as described previously, using modified alkaline lysis and polyethylene glycol precipitation (2). The endotoxin content of the plasmid DNA used in these experiments was between .04 and .25 EU per μg DNA. DNA for injection was resuspended in distilled water before use. In some experiments, plasmid DNA was further purified by column chromatography to assure LPS-free conditions. In other experiments, plasmid DNA was methylated in vitro with Sss I methylase (New England Biolabs, Boston, MA) according to the manufacturer’s instructions. Mock-methylated plasmid was prepared by incubation of the plasmid in the absence of the Sss I enzyme, but in the presence of the recommended enzyme reaction solution. Plasmid DNA for methylation experiments was then subsequently purified by column chromatography (Qiagen, Chatsworth, CA) before complexation with liposomes.

Cationic lipids were prepared as multilamellar vesicles for in vivo use as described previously (10). Briefly, DOTAP (1, 2 dioleoyl-3-trimethylammonium-propoane; Avanti Polar Lipids, Alabaster, AL) and cholesterol (Sigma) were mixed in a 1:1 molar ratio, dried down in round-bottom tubes, then rehydrated in 5% dextrose solution by heating at 50°C for 6 h, as described previously (10). All experiments were done with DOTAP-cholesterol liposomes, unless otherwise noted. For some experiments, DOTMA ((N-[1-(2, 3-dioleyloxy)propyl]-N,N,N-triethylammonium) (Syntex, Palo Alto, CA) was substituted for DOTAP, and DOPE (diolyl phosphatidylethanolamine; Avanti Polar Lipids) was substituted for cholesterol. For in vivo injection, CLDC were prepared immediately before injection by gently mixing cationic lipids with plasmid DNA at a ratio of 32 nmol total lipid to 1.0 μg DNA, to a final concentration of 100 μg DNA per ml in a sterile solution of 5% dextrose in water.

Lipid-DNA complexes (100 μg/ml DNA in 5% dextrose in water) were injected i.v. via the lateral tail vein. For tumor studies, injections were repeated once 7 days after the first injection.

Early cellular activation was assessed by flow cytometric measurement of CD69 expression on T cells, B cells, monocytes, and NK cells. Single cell suspensions were prepared from spleens of mice by the NH4Cl lysis procedure. Lung mononuclear cells were prepared from lung tissues by collagenase digestion. Briefly, lung tissues were minced, then digested in 1.0 mg/ml Type 1A collagenase in complete medium, along with 10 μg/ml DNase and 100 μg/ml soybean trypsin inhibitor, for 1 h at 37°C with occasional shaking. The lung tissues were then triturated and mononuclear cells isolated by Ficoll gradient centrifugation. For each experiment, spleen and lung cells were prepared from three to four animals per treatment group. Cells were analyzed using a Becton Dickinson (Mountain View, CA) FACScalibur flow cytometer, with analysis gates set by first gating on unstained spleen lymphocytes. Between 10,000 and 30,000 gated events were analyzed for each cell type. For analysis of cell activation, three-color flow cytometric analysis was done, using anti-CD69 PE (PharMingen, San Diego, CA) to quantitate the number of CD69-positive cells. T cells were labeled with an anti-αβTCR Ab (biotin H57.597; PharMingen) plus Abs to either CD4 (FITC RM4–5; PharMingen) or CD8 (FITC 53–6.7; PharMingen); B cells were dual-labeled with anti-B220 (biotin RA3-6B2; PharMingen) and either anti-IAb (FITC 3F12.35; provided by Dr. John Freed, National Jewish) or anti-IAd (FITC 14.4.4); NK cells were dual-labeled using anti-NK cell Abs (either anti-NK 1.1 (biotin PK136; PharMingen) or DX5-biotin (PharMingen)) and anti-CD3 (FITC 2C11); macrophages were evaluated using anti-CR3 (biotin Mac-1; PharMingen) and FITC anti-IAb or anti-IAd. The mean peak channel intensity and percentage of CD69-positive cells was determined for each cell type, and the mean percentage (±SD) of CD69+ cells was plotted.

A standard 4-h 51Cr-release assay was used to quantitate cytotoxic activity present in freshly isolated lung and spleen mononuclear cells, using YAC-1 cells as targets. Briefly, effector cells from lung or spleen were added in decreasing concentrations to duplicate wells of a Linbro plate, to which was then added 5 × 103 target cells that had been previously labeled for 1 h with 51Cr. The plates were incubated at 37°C for 4 h, then supernatants from each well were harvested and the amount of radioactive 51Cr present was determined by automated gamma counter. The percentage specific lysis was calculated as: [(observed 51Cr release) − (spontaneous 51Cr release)]/[(maximum 51Cr release) − (spontaneous release)] × 100.

Mice were depleted of NK cells in vivo by a single i.p. injection of 50 μl rabbit anti-asialoGM1 antiserum (Wako BioProducts, Richmond, VA). Control animals were injected with 50 μl nonimmune rabbit serum. Treatment with the asialo GM1 Ab eliminated detectable NK cells in spleen and lung (by flow cytometry) and also eliminated cytotoxic activity in spleen cells.

Cytokine release was measured in spleen cell supernatants after either in vivo or in vitro stimulation, or in serum after in vivo injection of CLDC. For assay of cytokine release after in vivo stimulation, spleen or lung mononuclear cells were prepared from mice 6 or 24 h postinjection, then cultured at a concentration of 5 × 106 cells/ml for an additional 18 h before supernatants were harvested. For in vitro stimulation of cytokine release, spleen cells were incubated in vitro with DNA, lipid, or DNA plus lipid at a final DNA concentration of 1.0 μg/ml for 18 h, at which time the supernatants were harvested for cytokine assays. Serum was obtained by tail vein bleed at various time points postinjection. IFN-γ concentrations inserum or tissue culture supernatants were assayed using a sandwich ELISA that consisted of two mAbs (XMG1.1 and biotinylated R45G). IL-10 and IL-4 concentrations in supernatants were quantitated using specific ELISA kits (PharMingen). TNF-α, total IL-12, and IL-6 concentrations were quantitated using ELISA kits obtained from Genzyme (Boston, MA) according to the manufacturer’s instructions.

B16 (clone F10) cells were obtained from Dr. Isiah Fidler (M D Anderson, Houston, TX); MCA-205 cells were provided by Dr Jack Routes (National Jewish); CT-26 cells were provided by Dr. Nicholas Restifo (National Cancer Institute, Bethesda, MD). All cell lines were maintained at 37°C in MEM supplemented with essential and nonessential amino acids, penicillin and glutamine, and 5% FBS, and were treated periodically with ciprofloxacin (10 μg/ml) to maintain mycoplasma-free conditions.

To establish experimental pulmonary metastases, mice (four per treatment group) were injected once via the lateral tail vein with 2.5 × 105 tumor cells. Treatment with CLDC was initiated 3 days after tumor injection, and was repeated once on day 10 after tumor injection. Control mice were injected with 5% dextrose in water. Mice were sacrificed on days 17–20 after tumor injection, and the number of tumor nodules per lung was determined by manual counting, as described previously (11).

Statistical analyses were done using SAS Institute (Cary, NC) software. Lung tumor counts were evaluated for statistically significant differences between treatment groups. For experiments involving two treatment groups, Student’s t test was used for analysis, whereas the Tukey-Kramer test was used to compare differences between multiple treatment groups in a given experiment. Kaplan-Meier survival and Log-Rank analysis was used to compare survival times. Significance was determined for p < 0.05.

The effect of CLDC injection on CD69 expression by various immunologically relevant cell types was evaluated in C57BL/6 mice (Fig. 1). Mice were injected i.v. with 100 μl of CLDC solution (which delivered 10 μg total plasmid DNA), or with 5% dextrose in water. Spleen cells were harvested 24 h postinjection, and five different cell populations (CD4+/TCR+, CD8+/TCR+, NK1.1+/CD3, B220+/IAb high, and Mac-1+/IAb low) were immunostained and evaluated by three-color flow cytometry for CD69 expression. The mean peak channel and percentage of CD69+ cells (±SD) was determined for each cell population. Flow cytometric analysis demonstrated strong up-regulation of CD69 expression on CD8+ and CD4+ T cells, NK cells, B220+ B cells, and Mac-1+ cells (macrophages) following CLDC injection (Fig. 1 a). The mean percentage of CD69+ cells was significantly up-regulated (p < 0.001, compared with sham-treated mice) 24 h after CLDC injection in all five different cell populations evaluated, compared with control animals. Similar results were obtained in repeated experiments with C57BL/6 mice, as well as with other strains of mice, including BALB/c, 129, and ICR (data not shown). The up-regulation of CD69 expression was maximal by 6 h postinjection and declined nearly to baseline levels by 3 days postinjection. Thus, systemic administration of CLDC induced rapid and marked activation of multiple different immune effector cells.

FIGURE 1.

Up-regulation of CD69 expression after injection of CLDC: activation of multiple cell types, effects of DNA or lipid alone, and dose-responsiveness. a, CLDC-induced activation of different spleen cell populations (T cells (CD4+ or CD8+), NK cells (NK1.1+), B cells (B220+), and monocyte/macrophages (Mac-1+)) was assessed by flow cytometry 24 h postinjection. The mean percentage CD69+ cells (±SD) was determined for each cell type in three control mice (open bars) and three CLDC-injected mice (filled bars). b, C57BL/6 mice (three per group) were injected with 100 μl diluent (control), 10 μg plasmid DNA (DNA), 320 nmol cationic lipid alone (lipid), or with CLDC (10 μg DNA + 320 nmol lipid). Twenty-four hours later, the mean percentage CD69+, NK1.1+ spleen cells (±SD) was determined by flow cytometry and plotted for each treatment group. c, C57BL/6 mice (three per treatment group) were injected with either undiluted CLDC (containing 100 μg/ml DNA) or with CLDC that had been diluted 1:10 (10 μg/ml DNA) or 1:100 (1.0 μg/ml DNA) in 5% dextrose in water. CD69 expression by NK1.1+ spleen cells was quantitated by flow cytometry 24 h postinjection. The mean percentage of CD69+/NK1.1+ cells (±SD) was plotted for each CLDC dosage group and compared with control animals. There was a significant increase in CD69 expression (p < 0.0001) for all CLDC-treated cell types, compared with control values.

FIGURE 1.

Up-regulation of CD69 expression after injection of CLDC: activation of multiple cell types, effects of DNA or lipid alone, and dose-responsiveness. a, CLDC-induced activation of different spleen cell populations (T cells (CD4+ or CD8+), NK cells (NK1.1+), B cells (B220+), and monocyte/macrophages (Mac-1+)) was assessed by flow cytometry 24 h postinjection. The mean percentage CD69+ cells (±SD) was determined for each cell type in three control mice (open bars) and three CLDC-injected mice (filled bars). b, C57BL/6 mice (three per group) were injected with 100 μl diluent (control), 10 μg plasmid DNA (DNA), 320 nmol cationic lipid alone (lipid), or with CLDC (10 μg DNA + 320 nmol lipid). Twenty-four hours later, the mean percentage CD69+, NK1.1+ spleen cells (±SD) was determined by flow cytometry and plotted for each treatment group. c, C57BL/6 mice (three per treatment group) were injected with either undiluted CLDC (containing 100 μg/ml DNA) or with CLDC that had been diluted 1:10 (10 μg/ml DNA) or 1:100 (1.0 μg/ml DNA) in 5% dextrose in water. CD69 expression by NK1.1+ spleen cells was quantitated by flow cytometry 24 h postinjection. The mean percentage of CD69+/NK1.1+ cells (±SD) was plotted for each CLDC dosage group and compared with control animals. There was a significant increase in CD69 expression (p < 0.0001) for all CLDC-treated cell types, compared with control values.

Close modal

Bacterial DNA has been shown previously to have immune stimulatory properties, including stimulation of IL-12 and IFN-γ release and activation of NK cells and B cells (12, 13, 14, 15). Therefore, experiments were done to determine the relative contributions of plasmid DNA and cationic liposomes to the systemic immune activation induced by CLDC in vivo. C57BL/6 mice (three per group) were injected i.v. with either 10 μg DNA, 320 nmol of DOTAP-cholesterol liposomes, or with CLDC comprised of 10 μg DA plus 320 nmol liposomes. Twenty-four hours postinjection, CD69 up-regulation by NK1.1+ spleen cells was measured flow cytometrically (Fig. 1,b). Intravenous injection of CLDC induced marked up-regulation of CD69 expression on NK cells, whereas CD69 expression was unchanged after i.v. injection of DNA alone or lipid alone. Similar results were obtained for T cells, B cells, and macrophages (data not shown). These data, plus cytokine data (see Fig. 3, below), indicated clearly that the complex of DNA and cationic liposomes was much more immunostimulatory than either DNA or liposome alone.

FIGURE 3.

Methylation of plasmid DNA reduces immune stimulation by CLDC. The effect of plasmid DNA methylation on cytokine release by CLDC was assessed in vitro. CLDC were formed using untreated plasmid DNA (CLDC), plasmid DNA treated with methylase (Meth+), or with plasmid DNA incubated with methylation buffer solution, but with the methylase enzyme omitted (Mock). The CLDC were then added to triplicate wells of spleen cells to a final concentration of 1.0 μg DNA per ml. Other wells were incubated with an equivalent amount of unmodified DNA only or lipid only. After 18 h of incubation, the supernatants were harvested and assayed for IL-12 release by ELISA. The mean IL-12 concentration (±SE) was plotted for each treatment group.

FIGURE 3.

Methylation of plasmid DNA reduces immune stimulation by CLDC. The effect of plasmid DNA methylation on cytokine release by CLDC was assessed in vitro. CLDC were formed using untreated plasmid DNA (CLDC), plasmid DNA treated with methylase (Meth+), or with plasmid DNA incubated with methylation buffer solution, but with the methylase enzyme omitted (Mock). The CLDC were then added to triplicate wells of spleen cells to a final concentration of 1.0 μg DNA per ml. Other wells were incubated with an equivalent amount of unmodified DNA only or lipid only. After 18 h of incubation, the supernatants were harvested and assayed for IL-12 release by ELISA. The mean IL-12 concentration (±SE) was plotted for each treatment group.

Close modal

The dose-responsiveness of immune activation (CD69 up-regulation) by CLDC was determined in C57BL/6 mice. Mice (three per group) were injected with 100 μl of a CLDC solution containing 100, 10, or 1.0 μg/ml plasmid DNA (Fig. 1 c). Twenty-four hours later, the percentage of CD69+/NK1.1+ spleen cells was determined for each group of treated mice and compared with control mice. A dose-dependent increase in CD69 up-regulation was observed. Intravenous administration of as little as 100 ng DNA (in the form of CLDC) was sufficient to induce a significant increase (p < 0.05) in CD69 expression by NK cells. Intravenous injection of LPS and poly(I:C) also induced strong CD69 up-regulation on multiple cell types (data not shown). Injection of CLDC formulated using any of three different cationic lipid formulations (DOTAP-cholesterol, DOTAP-DOPE, or DOTMA-cholesterol, see Materials and Methods) all stimulated CD69 up-regulation to an equivalent degree (data not shown). Thus, the immune-activating properties of CLDC were not dependent on use of a specific liposome.

The effect of CLDC injection on cytokine release was assessed by culturing spleen or lung cells obtained from in vivo-treated mice 24 h postinjection (Fig. 2). Mice were injected with either plasmid DNA only (10 μg), liposomes only, or CLDC. Single cell suspensions of spleen and lung mononuclear cells were prepared (see Materials and Methods) and then cultured in vitro for an additional 18 h. Release of cytokines into the supernatants was quantitated by specific cytokine ELISA.

FIGURE 2.

Induction of IL-12 and IFN-γ release after i.v. injection of CLDC. Spleens were harvested from C57BL/6 mice (three per treatment group) 24 h after i.v. injection of lipid alone (320 nmol/mouse), plasmid DNA alone (10 μg/mouse), or CLDC, and then cultured 18 h in vitro. Release of IL-12 (a) and IFN-γ (b) into the supernatants was quantitated by ELISA, and the mean cytokine concentration (±SE) for each treatment group was plotted. Similar results were obtained in four additional experiments, including experiments using several different strains of mice. c, Serum was collected from mice (five per group) at various time points before and after injection of CLDC and assayed for IFN-γ concentration. The mean (±SE) IFN-γ concentration was plotted for each time point. d, Mice (four per group) were injected i.v. with 10 μg poly(I:C), 10 μg LPS, or CLDC containing 10 μg DNA, and spleens were collected 24 h later and assayed for cytokine release, as in a. The mean concentration of IFN-γ (±SE) was plotted for each treatment group and compared with control, sham-injected mice. Similar results were also observed for lung mononuclear cells (data not shown).

FIGURE 2.

Induction of IL-12 and IFN-γ release after i.v. injection of CLDC. Spleens were harvested from C57BL/6 mice (three per treatment group) 24 h after i.v. injection of lipid alone (320 nmol/mouse), plasmid DNA alone (10 μg/mouse), or CLDC, and then cultured 18 h in vitro. Release of IL-12 (a) and IFN-γ (b) into the supernatants was quantitated by ELISA, and the mean cytokine concentration (±SE) for each treatment group was plotted. Similar results were obtained in four additional experiments, including experiments using several different strains of mice. c, Serum was collected from mice (five per group) at various time points before and after injection of CLDC and assayed for IFN-γ concentration. The mean (±SE) IFN-γ concentration was plotted for each time point. d, Mice (four per group) were injected i.v. with 10 μg poly(I:C), 10 μg LPS, or CLDC containing 10 μg DNA, and spleens were collected 24 h later and assayed for cytokine release, as in a. The mean concentration of IFN-γ (±SE) was plotted for each treatment group and compared with control, sham-injected mice. Similar results were also observed for lung mononuclear cells (data not shown).

Close modal

Spleen cells from CLDC-injected mice spontaneously released high levels of IL-12 (Fig. 2,a) and IFN-γ (Fig. 2,b), whereas there was minimal cytokine release from spleens of control mice or mice injected with equivalent amounts of DNA only or cationic liposomes only. Similar results were observed with cultured lung mononuclear cells (data not shown). In vitro addition of CLDC (1.0 μg DNA per ml medium) to naive spleen cells in vitro also induced release of IL-12 and IFN-γ (see Fig. 5, and data not shown). Injection of CLDC also induced release of high levels of IFN-γ into the serum, with the peak of release occurring 8 h postinjection and then declining thereafter (Fig. 2 c). Spleens from CLDC-injected mice released high levels of IL-6, but equivalent levels of IL-2, TNF-α, IL-4, or IL-10, compared with control mice (data not shown). Thus, i.v. injection of CLDC elicited strong systemic release of Th1 cytokines in naive mice. The complex of DNA plus a cationic lipid was a much stronger stimulus for cytokine release than DNA or lipid alone.

FIGURE 5.

Evaluation of antitumor activity induced by injection of DNA or lipid alone, or by injection of LPS or poly(I:C), and comparison to CLDC injection. a, C57BL/6 mice (four per treatment group) with day 3-established MCA-205 lung metastases were injected i.v. twice (1 wk apart) with 10 μg DNA, 320 nmol liposomes, or CLDC comprised of both. The mice were sacrificed 7 days after the second injection, and the lung tumor burdens were quantitated, as described in Materials and Methods. The mean number of lung tumor nodules (±SE) was plotted for each treatment group. Compared with control animals, there was a significant reduction in lung tumor nodules only in CLDC-treated mice. b, C57BL/6 mice with day 3-established MCA-205 tumors were injected i.v. with 10 μg LPS, 10 μg poly(I:C), or with CLDC. The injections were repeated once 7 days later, and the mice were sacrificed 1 wk later, the lung tumor burdens were quantitated, and the mean number lung tumor nodules (±SE) was plotted for each treatment group. Compared with control animals, only animals injected with CLDC had a significant reduction in lung tumor nodules. (∗, indicates a significant reduction (p < 0.05) in lung tumor burden, compared with sham-treated control animals).

FIGURE 5.

Evaluation of antitumor activity induced by injection of DNA or lipid alone, or by injection of LPS or poly(I:C), and comparison to CLDC injection. a, C57BL/6 mice (four per treatment group) with day 3-established MCA-205 lung metastases were injected i.v. twice (1 wk apart) with 10 μg DNA, 320 nmol liposomes, or CLDC comprised of both. The mice were sacrificed 7 days after the second injection, and the lung tumor burdens were quantitated, as described in Materials and Methods. The mean number of lung tumor nodules (±SE) was plotted for each treatment group. Compared with control animals, there was a significant reduction in lung tumor nodules only in CLDC-treated mice. b, C57BL/6 mice with day 3-established MCA-205 tumors were injected i.v. with 10 μg LPS, 10 μg poly(I:C), or with CLDC. The injections were repeated once 7 days later, and the mice were sacrificed 1 wk later, the lung tumor burdens were quantitated, and the mean number lung tumor nodules (±SE) was plotted for each treatment group. Compared with control animals, only animals injected with CLDC had a significant reduction in lung tumor nodules. (∗, indicates a significant reduction (p < 0.05) in lung tumor burden, compared with sham-treated control animals).

Close modal

The cytokine-releasing potency of CLDC was compared with two other classical inducers of innate immunity, LPS and poly(I:C). Twenty-four hours after i.v. injection of equivalent low doses of LPS or poly(I:C) (10 μg per mouse), cytokine release from 18-h cultured spleen cells was measured. Compared with CLDC-injected mice, spleens from LPS- or poly(I:C)-injected mice released very low levels of either IFN-γ (data not shown) or IL-12 (Fig. 2 d). Nonetheless, these doses of LPS and poly(I:C) did induce in vivo immune activation, as revealed by pronounced CD69 up-regulation (data not shown). Thus, systemically injected CLDC were more potent activators of innate immunity on a per weight basis than LPS or poly(I:C) and induced high circulating levels of IFN-γ.

Bacterial DNA is immunogenic in mammals. The immunogenicity of bacterial DNA stems in part from the high content of unmethylated CpG motifs in bacterial DNA, compared with eukaryotic DNA (12, 13, 15). Therefore, experiments were done to determine whether the marked immunogenicity of CLDC was mediated in part by increasing the immunogenicity of the plasmid DNA component of the complex. CLDC were prepared using either unmodified plasmid DNA or plasmid DNA that had been methylated in vitro (Fig. 3). Controls included plasmid DNA incubated with the appropriate buffers but without the methylase enzyme. Naive spleen cells were incubated for 18 h with modified or unmodified CLDC, and the induction of IL-12 release was quantitated.

Compared with CLDC prepared with unmodified plasmid DNA, CLDC prepared with methylated DNA elicited only 25% as much IL-12 release from spleen cells. The level of IL-12 release triggered by methylated CLDC was reduced to the levels induced by incubation with control DNA alone. Lipid alone did not induce IL-12 release (data not shown). These results suggest that formation of the CLDC greatly amplifies the inherent immunogenicity of the plasmid DNA, possibly by increasing DNA entry into cells and their nuclei. However, we have also observed immune activation by CLDC formulated with eukaryotic DNA (S.W. Dow, unpublished observations), and a recent publication suggests that any double-stranded DNA sequence is capable of inducing immune activation (16). Thus, the immune-stimulatory properties of CLDC may involve more than just immune stimulation by bacterial DNA.

Others have reported that systemic injection of CLDC containing noncoding (empty vector) plasmid DNA can induce antitumor activity (17). Therefore, we investigated in greater detail the mechanism(s) by which systemic injection of CLDC might induce antitumor activity. Three different tumor lines (fibrosarcoma (MCA-205), melanoma (B16.F10, and colon carcinoma (CT26)) were used to assess the effect of CLDC treatment on a variety of established pulmonary metastatic tumors (Fig. 4, a–d). Mice (four per treatment group) were each injected i.v. with 2.5 × 105 tumor cells. C57BL/6 mice were injected with MCA-205 cells or B16.F10 cells, and BALB/c mice were injected with CT26 cells. Mice were treated by i.v. injection of CLDC solution beginning 3 days after tumor injection. The CLDC injection was repeated once 7 days later, and the mice were sacrificed 7–10 days after the second CLDC injection (day 17–20 posttumor injection). The lung tumor burden was quantitated by counting the number of lung tumor nodules, as described previously (11). In all three tumor models, injection of CLDC induced highly significant reductions (p < 0.0001) in the lung tumor burden, compared with control animals.

FIGURE 4.

Effect of CLDC injection on tumor burden and survival times in mice with established lung tumor metastases. Lung tumors were established in mice by tail vein injection of 2.5 × 105 tumor cells per mouse. C57BL/6 mice (four per treatment group) with established MCA-205 tumors (a) or B16.F10 tumors (b) were treated 3 days posttumor injection by i.v. injection of 100 μl CLDC or by injection of diluent (control). BALB/c mice with day 3 established CT26 tumors (c) were treated similarly. CLDC injection was repeated once 7 days after the first injection, and the mice were sacrificed 7–10 days later (17–20 days posttumor injection) and the lung tumor burden quantitated by manual counting, as described in Materials and Methods. The mean number of tumor nodules per lung (±SE) was plotted for each group of mice. Each experiment was repeated at least once, with similar results. There was a significant reduction (p < 0.0001) in lung tumor burden for treated mice in each tumor model, compared with control mice, as determined by Student’s t test. d, Mice (eight per treatment group) with day 3 established CT26 tumors were given three CLDC injections, 1 wk apart, or were sham-treated (control). The survival times for mice in each group were plotted, using a Kaplan-Meier curve. Survival times for CLDC-treated group were significantly prolonged (p = .0001) compared with control mice, as determined by Log-Rank survival analysis. (∗, indicates a significant reduction (p < 0.05) in lung tumor burden, compared with sham-treated control animals).

FIGURE 4.

Effect of CLDC injection on tumor burden and survival times in mice with established lung tumor metastases. Lung tumors were established in mice by tail vein injection of 2.5 × 105 tumor cells per mouse. C57BL/6 mice (four per treatment group) with established MCA-205 tumors (a) or B16.F10 tumors (b) were treated 3 days posttumor injection by i.v. injection of 100 μl CLDC or by injection of diluent (control). BALB/c mice with day 3 established CT26 tumors (c) were treated similarly. CLDC injection was repeated once 7 days after the first injection, and the mice were sacrificed 7–10 days later (17–20 days posttumor injection) and the lung tumor burden quantitated by manual counting, as described in Materials and Methods. The mean number of tumor nodules per lung (±SE) was plotted for each group of mice. Each experiment was repeated at least once, with similar results. There was a significant reduction (p < 0.0001) in lung tumor burden for treated mice in each tumor model, compared with control mice, as determined by Student’s t test. d, Mice (eight per treatment group) with day 3 established CT26 tumors were given three CLDC injections, 1 wk apart, or were sham-treated (control). The survival times for mice in each group were plotted, using a Kaplan-Meier curve. Survival times for CLDC-treated group were significantly prolonged (p = .0001) compared with control mice, as determined by Log-Rank survival analysis. (∗, indicates a significant reduction (p < 0.05) in lung tumor burden, compared with sham-treated control animals).

Close modal

The effect of treatment with CLDC on survival times in mice with established CT26 lung tumors was also assessed (Fig. 4 d). Mice (eight per treatment group) were treated three times with i.v. injection of CLDC on days 3, 10, and 17 posttumor injection, and then treatment was discontinued. Control mice were sham-treated with diluent. The survival time in CLDC-treated animals (mean = 35 days) was significantly increased (p = .0001) compared with control animals (mean survival = 15 days). Thus, injection of CLDC induced a sustained antitumor effect in animals with established lung tumor metastases.

The relative contributions of DNA and/or liposomes to the antitumor effects of systemic CLDC injection were determined in mice with established MCA-205 lung tumor metastases. Treatment with plasmid DNA alone or liposomes alone did not induce antitumor activity, whereas injection of an equivalent amount of CLDC induced significant (p < 0.001) antitumor activity (Fig. 5,a). We also assessed whether the nonspecific immune stimulation induced by injection of LPS or poly(I:C) could also induce antitumor activity in these experimental models. In contrast to injection of CLDC, i.v. injection of equivalent amounts of (10 μg) of either LPS or poly(I:C) did not induce antitumor activity against established MCA-205 metastases (Fig. 5 b). Thus, though LPS, poly(I:C), and CLDC were all capable of inducing nonspecific immune stimulation when injected systemically, only CLDC also exerted potent antitumor activity.

Lung mononuclear cells and spleen cells were collected from mice 3 days after i.v. injection of CLDC and analyzed by flow cytometry (Fig. 6). Flow cytometric analysis revealed a pronounced, 4-fold increase in the percentage of intrapulmonary NK cells in CLDC-injected mice, compared with sham-treated control animals (p < 0.0001). There was also an increase in intrasplenic NK cells in CLDC-injected mice, though the increase was not as large as in the lungs. Thus, CLDC administered systemically serve as a stimulus for accumulation of NK cells in various tissues, particularly the lungs.

FIGURE 6.

Accumulation of NK cells in lung tissues after CLDC injection. Three days after i.v. injection of CLDC in C57BL/6 mice (four per group), mononuclear cells were harvested from lung tissues of treated or control mice by enzymatic digestion, as described in Materials and Methods. Spleen cells were obtained from the same mice. Spleen and lung mononuclear cells were analyzed by flow cytometry for expression of NK1.1 and CD3. In a representative analysis, there was a significant increase (p < 0.0001) in the percentage of NK1.1+/CD3 cells in the lungs of a mouse injected with CLDC (b), compared with a control mouse (a). The mean percentage (±SE) of NK cells in the lungs and spleens of CLDC-treated (filled bars) and control mice (open bars) was plotted in c. Similar results were observed in at least three additional experiments.

FIGURE 6.

Accumulation of NK cells in lung tissues after CLDC injection. Three days after i.v. injection of CLDC in C57BL/6 mice (four per group), mononuclear cells were harvested from lung tissues of treated or control mice by enzymatic digestion, as described in Materials and Methods. Spleen cells were obtained from the same mice. Spleen and lung mononuclear cells were analyzed by flow cytometry for expression of NK1.1 and CD3. In a representative analysis, there was a significant increase (p < 0.0001) in the percentage of NK1.1+/CD3 cells in the lungs of a mouse injected with CLDC (b), compared with a control mouse (a). The mean percentage (±SE) of NK cells in the lungs and spleens of CLDC-treated (filled bars) and control mice (open bars) was plotted in c. Similar results were observed in at least three additional experiments.

Close modal

The role of NK cells in IFN-γ release after CLDC injection was investigated by NK cell depletion experiments. Mice were depleted of NK cells by treatment with anti-asialo GM1 antiserum. Control mice were treated with nonimmune rabbit serum. Forty-eight hours after injection of anti-asialo GM1 antiserum, mice were injected with CLDC. Twenty-four hours after CLDC injection, spleen and lung mononuclear cells were harvested, cultured in vitro, and assayed for release of IFN-γ (Fig. 7, a and b). NK cell depletion almost completely eliminated release of IFN-γ from both spleen and lung cells, compared with mice treated with non-immune rabbit serum or untreated control animals. Thus, NK cells were the major source of IFN-γ release triggered by CLDC injection.

FIGURE 7.

Effect of in vivo NK cell depletion on CLDC-induced IFN-γ release. Mice (three per treatment group) were pretreated 48 h before CLDC injection with rabbit anti-asialo GM1 antiserum (CLDC/asialo), an equivalent amount of nonimmune rabbit serum (CLDC/NRS), or were not pretreated (CLDC). Spleen and lung cells from untreated mice served as controls. Spleen (a) and lung (b) mononuclear cells were collected 24 h after injection of CLDC, then cultured for 18 additional hours, and release of IFN-γ was quantitated by ELISA. The mean IFN-γ concentration (±SE) for each treatment group was plotted.

FIGURE 7.

Effect of in vivo NK cell depletion on CLDC-induced IFN-γ release. Mice (three per treatment group) were pretreated 48 h before CLDC injection with rabbit anti-asialo GM1 antiserum (CLDC/asialo), an equivalent amount of nonimmune rabbit serum (CLDC/NRS), or were not pretreated (CLDC). Spleen and lung cells from untreated mice served as controls. Spleen (a) and lung (b) mononuclear cells were collected 24 h after injection of CLDC, then cultured for 18 additional hours, and release of IFN-γ was quantitated by ELISA. The mean IFN-γ concentration (±SE) for each treatment group was plotted.

Close modal

Spleen cells harvested 24 h after i.v. injection of CLDC in C57BL/6 exhibited high levels of cytotoxic activity, as assessed in a 4-h chromium release assay, using 51Cr-labeled YAC-1 cells as targets (Fig. 8,a). The cytotoxic activity was not MHC-restricted, as revealed using MHC-mismatched target cells (data not shown). The cytotoxic activity was markedly reduced in the spleens of NK cell-depleted mice. Similar results were observed using lung mononuclear cells as effector cells (data not shown). Furthermore, cytotoxic activity was not generated by injection of either DNA alone or lipid alone (data not shown). High levels of cytotoxic activity were also detected in spleen cells of RAG-2 KO mice injected with CLDC (Fig. 8 b). Thus, i.v. administration of CLDC elicited strong functional activation of NK cells, a response that did not require T or B cells.

FIGURE 8.

Induction of NK cell cytotoxic activity by i.v. administration of CLDC. Cytotoxic activity in spleen cells of C57BL/6 mice (three per treatment group) was quantitated 24 h after i.v. injection of CLDC (a). Mice were pretreated 48 h before injection of CLDC by i.p. injection of anti-asialo GM1 antiserum to deplete NK cells, or were pretreated with nonimmune rabbit serum (NRS). Spleen cells from CLDC-injected and control mice were assayed for cytotoxic activity against 51Cr-labeled YAC-1 target cells, as described in Materials and Methods. The mean percentage specific lysis at each effector to target cell ratio was plotted. b, Cytotoxic activity was assayed in spleen cells obtained from control (□) or CLDC-injected (•) RAG-2 gene knockout mice, 24 h postinjection.

FIGURE 8.

Induction of NK cell cytotoxic activity by i.v. administration of CLDC. Cytotoxic activity in spleen cells of C57BL/6 mice (three per treatment group) was quantitated 24 h after i.v. injection of CLDC (a). Mice were pretreated 48 h before injection of CLDC by i.p. injection of anti-asialo GM1 antiserum to deplete NK cells, or were pretreated with nonimmune rabbit serum (NRS). Spleen cells from CLDC-injected and control mice were assayed for cytotoxic activity against 51Cr-labeled YAC-1 target cells, as described in Materials and Methods. The mean percentage specific lysis at each effector to target cell ratio was plotted. b, Cytotoxic activity was assayed in spleen cells obtained from control (□) or CLDC-injected (•) RAG-2 gene knockout mice, 24 h postinjection.

Close modal

The role of NK cells in mediating tumor rejection in response to CLDC injection was investigated by depleting mice of NK cells in vivo. NK cells were depleted in BALB/c mice (four per group) 1 day after i.v. injection with CT26 cells. Control mice were treated with nonimmune rabbit serum or were untreated. On day 3 posttumor injection, the mice were injected with CLDC. The Ab depletion of NK cells and CLDC treatment was repeated once 7 days later, and the mice were sacrificed on day 17 posttumor injection and the lung tumor burden was quantitated. Depletion of NK cells reduced the antitumor activity induced by systemic injection of CLDC, such that there was no significant difference in tumor burden between CLDC-treated and control mice (Fig. 9 a). However, the tumor burden in mice treated with nonimmune rabbit serum was significantly reduced by injection of CLDC. Thus, NK cells were a key mediator of the antitumor activity induced by CLDC injection.

FIGURE 9.

NK cells and IFN-γ mediate the antitumor effects of CLDC. The effect of NK cell depletion on the tumor response to CLDC injection was evaluated in BALB/c mice (four per treatment group) with day 3-established CT26 tumors (a). Tumor-bearing mice were administered an i.p. injection of anti-asialo GM1 antiserum (CLDC/asialo) or an equivalent amount of nonimmune rabbit serum (CLDC/NRS) 24 h before each injection of CLDC, as described in Materials and Methods. A third group of untreated mice served as a control. Seven days after the second CLDC injection, mice were sacrificed, the lung tumor burden was quantitated by manual counting, and the mean number of tumor nodules per lung (±SE) was plotted. There was a significant reduction (p < 0.001) in lung tumor burden in CLDC-treated mice pretreated with normal rabbit serum, compared with control mice. By contrast, there was no reduction in lung tumor burden in CLDC-treated mice that were also injected with anti-asialo GM1 antiserum. This experiment was repeated once, with similar results. b, MCA-205 lung tumors were established in wild-type C57BL/6 mice (IFN-γ+/+) or C57BL/6 mice with a nonfunctional IFN-γ gene (IFN-γ−/−). Three days later, the mice were injected with CLDC or were sham-treated (control). The treatment was repeated 7 days later, and the mice were sacrificed and lung tumors counted 7 days after that. The mean number of lung tumor nodules (±SE) per treatment group was plotted. The number of lung tumor nodules in IFN-γ+/+ mice was significantly different from that of untreated IFN-γ+/+ control mice (p < 0.001), whereas the number of lung tumor nodules in IFN-γ−/− mice treated with CLDC was not significantly different from untreated control IFN-γ−/− mice. Similar results were obtained in one additional experiment. (*, significant reduction (p < 0.05) in lung tumor burden, compared with sham-treated control animals).

FIGURE 9.

NK cells and IFN-γ mediate the antitumor effects of CLDC. The effect of NK cell depletion on the tumor response to CLDC injection was evaluated in BALB/c mice (four per treatment group) with day 3-established CT26 tumors (a). Tumor-bearing mice were administered an i.p. injection of anti-asialo GM1 antiserum (CLDC/asialo) or an equivalent amount of nonimmune rabbit serum (CLDC/NRS) 24 h before each injection of CLDC, as described in Materials and Methods. A third group of untreated mice served as a control. Seven days after the second CLDC injection, mice were sacrificed, the lung tumor burden was quantitated by manual counting, and the mean number of tumor nodules per lung (±SE) was plotted. There was a significant reduction (p < 0.001) in lung tumor burden in CLDC-treated mice pretreated with normal rabbit serum, compared with control mice. By contrast, there was no reduction in lung tumor burden in CLDC-treated mice that were also injected with anti-asialo GM1 antiserum. This experiment was repeated once, with similar results. b, MCA-205 lung tumors were established in wild-type C57BL/6 mice (IFN-γ+/+) or C57BL/6 mice with a nonfunctional IFN-γ gene (IFN-γ−/−). Three days later, the mice were injected with CLDC or were sham-treated (control). The treatment was repeated 7 days later, and the mice were sacrificed and lung tumors counted 7 days after that. The mean number of lung tumor nodules (±SE) per treatment group was plotted. The number of lung tumor nodules in IFN-γ+/+ mice was significantly different from that of untreated IFN-γ+/+ control mice (p < 0.001), whereas the number of lung tumor nodules in IFN-γ−/− mice treated with CLDC was not significantly different from untreated control IFN-γ−/− mice. Similar results were obtained in one additional experiment. (*, significant reduction (p < 0.05) in lung tumor burden, compared with sham-treated control animals).

Close modal

Since NK cells were also the primary source of IFN-γ release in response to CLDC injection, the role of IFN-γ in mediating CLDC-induced tumor rejection was investigated. CLDC-induced antitumor activity was evaluated in C57BL/6 mice with a targeted disruption of the IFN-γ gene (IFN-γ−/−). Wild-type C57BL/6 mice (IFN-γ+/+) or IFN-γ−/− mice with day 3 established MCA-205 tumor cells were treated with either CLDC or were sham-injected (control). The tumor burden in IFN −/− mice treated with CLDC was not significantly decreased, compared with control IFN−/− mice, whereas treatment of IFN+/+ mice induced a significant reduction in tumor burden (Fig. 9 b). The IFN-γ−/− mice were therefore impaired in their ability to respond to CLDC injection and control the growth of MCA-205 lung tumors. Thus, the antitumor activity induced by systemically administered CLDC was mediated to a large degree by IFN-γ released from activated NK cells.

These studies indicate that CLDC are very immunologically active when injected systemically. Indeed, CLDC administered i.v. induced rapid, marked systemic immune activation in vivo, manifested initially by up-regulation of CD69 expression on immune effector cells (T cells, B cell, macrophages, and NK cells). Injection of CLDC also served as a potent stimulus for release of Th1 cytokines (IL-12 and IFN-γ). CLDC were particularly stimulatory for NK cells, inducing functional activation, IFN-γ release, and antitumor activity.

Bacterial DNA is immunogenic in eukaryotes. Exposure to bacterial DNA triggers innate immune responses and activation of B cell, macrophages, and NK cells (12, 13, 14, 15). B cells and macrophages are activated directly by bacterial DNA, whereas NK cell activation is reported to be mediated indirectly by IL-12 released from DNA-stimulated macrophages (12, 13, 14, 15, 18, 19). The immunogenicity of bacterial DNA is due in part to the increased purine content of bacterial DNA, which is enriched in CpG motifs relative to eukaryotic DNA (12, 13, 14, 15, 20, 21, 22). The hypomethylated state of bacterial DNA is also immunogenic (12, 22). The immunogenicity of bacterial DNA has also been shown to play a role in the effectiveness of DNA vaccines, in part through induction of Th1 cytokines (23).

One surprising finding from our study was the extreme immunogenicity of lipid-DNA complexes. The cationic liposomes by themselves were without apparent immune stimulatory activity, as was the plasmid DNA alone at the low doses employed in these studies (Figs. 1 and 2). However, the complex of the two was markedly stimulatory. Complexation to liposomes was reported previously to enhance the immunogenicity of synthetic oligonucleotides (24). The enhanced immunogenicity of complexes rela-tive to naked DNA or oligonucleotides is most likely a result of lipid-facilitated intracellular entry and nuclear translocation of DNA, inasmuch as immune activation by bacterial DNA requires DNA entry into the cell nucleus (14, 25). In vivo, cationic liposomes protect DNA from degradation and prolong circulation time, which may in turn facilitate binding to monocytes and macrophages (26).

Previous studies have demonstrated that the formation of complexes with liposomes serves to increase the immune stimulatory properties of DNA, both in vitro and in lung tissues in vivo (8, 24). Our observation that methylation of the plasmid DNA reduced immune activation by CLDC (Fig. 3) is consistent with these prior observations. Intravenous injection of high doses (mg) of plasmid DNA has been reported to induce immune stimulation and IFN-γ release (18). However, we observed immune activation at DNA doses as low as 100 ng when the DNA was complexed to a cationic liposome (Fig. 1), indicative of the degree to which lipids increased the immunogenicity of bacterial DNA.

The immunostimulatory properties of CLDC were not due to contaminating LPS, as evidenced by the failure of DNA alone to induce immune stimulation, either in vitro or in vivo (Figs. 2 and 5). Injection of equivalent amounts of purified LPS also failed to elicit the same pattern of cytokine responses or antitumor responses as CLDC (Figs. 2 and 5). In addition, CLDC formulated with highly purified DNA (virtually free of detectable LPS) induced immune stimulatory effects indistinguishable from CLDC prepared with other sources of DNA (data not shown).

A major finding from this study was that CLDC injected i.v. could induce high levels of antitumor activity. Treatment of mice with established pulmonary metastases by CLDC injection induced significant reductions in lung tumor burden and prolonged survival times (Fig. 4). The antitumor activity induced by CLDC was unique in that it could not be reproduced by injection of equivalent low doses of other nonspecific activators of innate immunity, such as LPS or poly(I:C) (Fig. 5).

NK cells were the major mediators of the antitumor effects induced by CLDC. For example, NK cells accumulated to high levels in the lungs of CLDC-injected mice (Fig. 6) became highly cytotoxic (Fig. 8) and secreted large amounts of IFN-γ (Fig. 7). Furthermore, depletion of NK cells almost completely eliminated the antitumor activity of CLDC injection (Fig. 9). It is likely that NK cell activation occurred in response to IL-12 released following CLDC injection (Fig. 2). Bacterial DNA has been shown previously to activate NK cells by triggering release of IL-12 from macrophages (14, 19). However, we cannot exclude the possibility that CLDC injection might also induce production of IL-18, which could in turn trigger release of IFN-γ and antitumor activity. It should also be noted that the NK cells that accumulated in the lungs of CLDC-treated mice (Fig. 6) were classical NK cells (NK1.1+, CD3), as opposed to the recently described natural T cells (NK1.1+/CD3+) (27).

IFN-γ also played a critical role in the antitumor effects of CLDC injection. Mice with a nonfunctional IFN-γ gene were significantly impaired in their ability to control the growth of lung tumors after treatment with CLDC (Fig. 9). Thus, NK cells activated by CLDC may have induced inhibition of tumor growth indirectly by releasing IFN-γ. Both IFN-γ and IL-12 exert multiple antitumor activities in vivo, including inhibition of tumor angio-genesis (28, 29, 30). Therefore, it is conceivable that CLDC may inhibit tumor growth in part by stimulating local release of antiangiogenic cytokines within the tumor vasculature. For example, it was demonstrated recently that CLDC preferentially bound to the neovascular endothelium of tumors following i.v. injection (31).

Our findings have important implications for systemic nonviral gene delivery using CLDC. Widespread immune activation is likely to occur when CLDC are used to deliver genes systemically to the lungs or other target organs. However, the immune response to CLDC was dose-dependent and self-limiting (Figs. 1 and 2), and could therefore be controlled by dosage adjustments. The immune response elicited by CLDC injection was not vector-specific, therefore still allowing repeated CLDC administration. The development of vector-specific immunity is, by contrast, a major drawback to repeated use of viral vectors in vivo (6, 7). Continued improvements in plasmid vectors are likely to improve the duration of gene expression following systemic gene delivery, thereby reducing the frequency of administration. Activation of innate immune responses by CLDC may also represent an unanticipated advantage for immunotherapy of cancer or allergic diseases. The Th1 cytokines elicited by CLDC may provide a strong adjuvant effect for DNA vaccines administered using CLDC. For example, the superiority of CLDC for induction of immune responses to HIV Ags using DNA vaccines was reported previously (32). The use of CLDC may prove particularly effective for systemic, therapeutic vaccination against tumor Ags, given the strong antitumor effects induced by the plasmid vector alone.

We thank Dr. David Ickle for assistance with statistical analysis.

1

This work was supported by the Megabios Corporation and by Grant AI-37905 from the National Institutes of Health.

3

Abbreviation used in this paper: CLDC, cationic lipid-DNA complex.

1
Zhu, N., D. Liggitt, Y. Liu, R. Debs.
1993
. Systemic gene expression after i.v. delivery of DNA into adult mice.
Science
261
:
209
2
Liu, Y., D. Liggitt, G. Tu, W. Zhong, K. Gaensler, R. Debs.
1995
. Cationic liposome-mediated intravenous gene delivery in mice.
J. Biol. Chem.
27
:
24864
3
Canonico, A. E., J. D. Pittman, J. T. Conary, B. O. Meyrick, K. L. Brigham.
1994
. No lung toxicity after repeated aerosol or i.v. delivery of plasmid cationic liposome complexes.
J. Appl. Physiol.
77
:
415
4
Song, Y. K., L. Feng, C. Shaoyou, L. Dexi.
1997
. Characterization of cationic liposome-mediated gene transfer in vivo by intravenous administration.
Hum. Gene Ther.
8
:
1585
5
Templeton, N., D. Lasic, P. Frederik, H. Strey, D. Roberts, G. Pavlakis.
1997
. Improved DNA:liposome complexes for increased systemic delivery and gene expression.
Nat. Biotechnol.
15
:
647
6
Yang, Y., F. A. Nunes, K. Berencsi, E. E. Furth, E. Gonczol, J. M. Wilson.
1994
. Cellular immunity to viral antigens limits E1-deleted adenoviruses for gene therapy.
Proc. Natl. Acad. Sci. USA
91
:
4407
7
Knowles, M. R., K. W. Hohneker, Z. Zhou, J. C. Olsen, T. L. Noah, P. C. Hu, M. W. Leigh, J. F. Englehardt, L. J. Edwards, K. R. Jones, R. Boucher.
1995
. A controlled study of adenovirus-vector-mediated gene transfer in the nasal epithelium of patients with cystic fibrosis.
N. Engl. J. Med.
333
:
823
8
Freimark, B. D., H. P. Blezinger, V. J. Florack, J. L. Nordstrom, S. D. Long, D. S. Deshpande, S. Nochumson, K. L. Petrak.
1998
. Cationic lipids enhance cytokine and cell influx levels in the lung following administration of plasmid:cationic lipid complexes.
J. Immunol.
160
:
4580
9
Scheule, R., J. St. George, R. Bagley, J. Marshall, J. Kaplan, G. Akita, K. Wang, E. Lee, D. Harris, C. Jiang, et al
1997
. Basis of pulmonary toxicity associated with cationic lipid-mediated gene transfer to the mammalian lung.
Hum. Gene Ther.
8
:
689
10
Liu, Y., D. Liggitt, G. Tu, W. Zhong, K. Gesler, R. Debs.
1995
. Cationic lipsome-mediated i.v. gene delivery in mice.
J. Biol. Chem.
27
:
24864
11
Wexler, H..
1966
. Accurate identification of experimental pulmonary metastases.
J. Natl. Cancer Inst.
36
:
641
12
Pisetsky, D. S..
1996
. Immune activation by bacterial DNA: a new genetic code.
Immunity
5
:
303
13
Krieg, A. M..
1996
. Lymphocyte activation by CpG dinucleotide motifs in prokaryotic DNA.
Trends Microbiol.
4
:
73
14
Stacey, K. J., M. J. Sweet, D. A. Hume.
1996
. Macrophages ingest and are activated by bacterial DNA.
J. Immunol.
157
:
2116
15
Klinman, D. M., A. Yi, S. J. Beaucage, J. Conover, A. M. Krieg.
1996
. CpG motifs present in bacterial DNA rapidly induce lymphocytes to secrete interleukin-6, interleukin-12, and interferon-γ.
Proc. Natl. Acad. Sci. USA
93
:
2879
16
Suzuki, K., A. Mori, K. J. Ishii, J. Saito, D. S. Singer, D. M. Klinman, P. R. Krause, D. K. Leonard.
1999
. Activation of target-tissue immune recognition molecules by double-stranded polynucleotides.
Proc. Natl. Acad. Sci. USA
96
:
2285
17
Lesoon-Wood, L. A., W. H. Kim, H. K. Kleinman, B. D. Weintraub, A. J. Mixson.
1995
. Systemic gene therapy with p53 reduces growth and metastases of a malignant human breast cancer in nude mice.
Hum. Gene Ther.
6
:
395
18
Cowdery, J. S., J. H. Chace, A. Yi, A. M. Krieg.
1996
. Bacterial DNA induces NK cells to produce IFN-γ in vivo and increases the toxicity of lipopolysaccharides.
J. Immunol.
156
:
4570
19
Chace, J. H., N. A. Hooker, K. L. Mildenstein, A. M. Krieg, J. S. Cowdery.
1997
. Bacterial DNA-induced NK cell IFN-γ production is dependent on macrophage secretion of IL-12.
Clin. Immunol. Immunopathol.
84
:
185
20
Yi, A., D. M. Klinman, T. L. Martin, S. Matson, A. M. Krieg.
1996
. Rapid immune activation by CpG motifs in bacterial DNA.
J. Immunol.
157
:
5394
21
Ballas, Z. K., W. L. Rasmussen, A. M. Krieg.
1996
. Induction of NK activity in murine and human cells by CpG motifs in oligodeoxynucleotides and bacterial DNA.
J. Immunol.
157
:
1840
22
Karlen, S., W. Doerfler, L. R. Cardon.
1994
. Why is CpG suppressed in the genomes of virtually all small eukaryotic viruses but not in those of large eukaryotic viruses?.
J. Virol.
68
:
2889
23
Sato, Y., M. Roman, H. Tighe, D. Lee, M. Corr, M. D. Nguyen, G. J. Silveramn, M. Lotz, D. A. Carlson, E. Raz.
1996
. Immunostimulatory DNA sequences necessary for effective intradermal gene immunization.
Science
273
:
352
24
Yamamato, T., S. Yamamato, T. Katoaka, T. Tokunaga.
1994
. Lipofection of synthetic oliodeoxynulceotide have a palindromic sequence of AACGTT to murine splenocytes enhances interferon production and natural killer activity.
Microbiol. Immunol.
38
:
831
25
Friend, D. S., D. Papahadjopoulos, R. J. Debs.
1996
. Endocytosis and intracellular processing accompanying transfection mediated by cationic liposomes.
Biochem. Biophys. Acta
1278
:
41
26
McLean, J. W., E. A. Fox, P. Baluk, P. B. Bolton, A. Haskell, R. Pearlman, G. Thurston, E. Y. Umemoto, D. M. McDonald.
1997
. Organ-specific endothelial cell uptake of cationic liposome-DNA complexes.
Am. J. Physiol.
273
:
H387
27
Cui, J., T. Shin, T. Kawano, H. Sato, E. Kondo, I. Toura, Y. Kaneko, H. Koseki, M. Kanno, M. Taniguchi.
1997
. Requirement for Va14 NKT cells in IL-12 mediated rejection of tumors.
Science
278
:
1623
28
Ruegg, C., A. Yilmaz, G. Bieler, J. Bamat, P. Chaubert, F. J. Lejeune.
1998
. Evidence for the involvement of endothelial cell integrin αVβ3 in the disruption of the tumor vasculature induced by TNF and IFN-γ.
Nat. Med.
4
:
408
29
Coughlin, C. M., K. E. Salhany, M. Wysocak, E. Aruga, H. Kurzawa, A. E. Chang, C. A. Hunter, J. C. Fox, G. Trinchieri, W. M. F. Lee.
1998
. Interleukin-12 and interleukin-18 synergistically induce murine tumor regression which involves inhibition of angiogenesis.
J. Clin. Invest.
101
:
1441
30
Coughlin, C. M., K. E. Salhany, M. S. Gee, D. C. LaTemple, S. Kotenko, X. Ma, G. Gri, M. Wysocka, J. E. Kim, L. Liu, et al
1998
. Tumor cell responses to IFN-γ affect tumorigenicity and response to IL-12 therapy and antiangiogenesis.
Immunity
9
:
25
31
Thurston, G., J. W. McClean, M. Rizen, P. Baluk, A. Haskell, T. J. Murphy, D. Hanahan, D. McDonald.
1998
. Cationic liposomes that target angiogenic endothelial cells in tumors and chronic inflammation in mice.
J. Clin. Invest.
101
:
1401
32
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 liposomes are a strong adjuvant for a DNA vaccine of human immunodeficiency virus type 1.
AIDS Res. Hum. Retroviruses
13
:
1421