Design of Gadoteridol-Loaded Cationic Liposomal Adjuvant CAF01 for MRI of Lung Deposition of Intrapulmonary Administered Particles.

Designing effective and safe tuberculosis (TB) subunit vaccines for inhalation requires identification of appropriate antigens and adjuvants and definition of the specific areas to target in the lungs. Magnetic resonance imaging (MRI) enables high spatial resolution, but real-time anatomical and functional MRI of lungs is challenging. Here, we describe the design of a novel gadoteridol-loaded cationic adjuvant formulation 01 (CAF01) for MRI-guided vaccine delivery of the clinically tested TB subunit vaccine candidate H56/CAF01. Gadoteridol-loaded CAF01 liposomes were engineered by using a quality-by-design approach to (i) increase the mechanistic understanding of formulation factors governing the loading of gadoteridol and (ii) maximize the loading of gadoteridol in CAF01, which was confirmed by cryotransmission electron microscopy. The encapsulation efficiency and loading of gadoteridol were highly dependent on the buffer pH due to strong attractive electrostatic interactions between gadoteridol and the cationic lipid component. Optimal gadoteridol loading of CAF01 liposomes showed good in vivo stability and safety upon intrapulmonary administration into mice while generating 1.5-fold MRI signal enhancement associated with approximately 30% T1 relaxation change. This formulation principle and imaging approach can potentially be used for other mucosal nanoparticle-based formulations, species, and lung pathologies, which can readily be translated for clinical use.

Application of a Quality-By-Design Approach to Optimise Lipid-Polymer Hybrid Nanoparticles Loaded with a Splice-Correction Antisense Oligonucleotide: Maximising Loading and Intracellular Delivery.

BACKGROUND: Antisense oligonucleotides (ASOs) are promising therapeutics for specific modulation of cellular RNA function. However, ASO efficacy is compromised by inefficient intracellular delivery. Lipid-polymer hybrid nanoparticles (LPNs) are attractive mediators of intracellular ASO delivery due to favorable colloidal stability and sustained release properties. METHODS: LPNs composed of cationic lipidoid 5 (L5) and poly(DL-lactic-co-glycolic acid) were studied for delivery of an ASO mediating splice correction of a luciferase gene transcript (Luc-ASO). Specific purposes were: (i) to increase the mechanistic understanding of factors determining the loading of ASO in LPNs, and (ii) to optimise the LPNs and customise them for Luc-ASO delivery in HeLa pLuc/705 cells containing an aberrant luciferase gene by using a quality-by-design approach. Critical formulation variables were linked to critical quality attributes (CQAs) using risk assessment and design of experiments, followed by delineation of an optimal operating space (OOS). RESULTS: A series of CQAs were identified based on the quality target product profile. The L5 content and L5:Luc-ASO ratio (w/w) were determined as critical formulation variables, which were optimised systematically. The optimised Luc-ASO-loaded LPNs, defined from the OOS, displayed high loading and mediated splice correction at well-tolerated, lower doses as compared to those required for reference L5-based lipoplexes, L5-modified stable nucleic acid lipid nanoparticles or LPNs modified with dioleoyltrimethylammonium propane (conventional cationic lipid). CONCLUSIONS: The optimal Luc-ASO-loaded LPNs represent a robust formulation that mediates efficient intracellular delivery of Luc-ASO. This opens new avenues for further development of LPNs as a broadly applicable technology platform for delivering nucleic acid cargos intracellularly.

Identification of Factors of Importance for Spray Drying of Small Interfering RNA-Loaded Lipidoid-Polymer Hybrid Nanoparticles for Inhalation.

BACKGROUND: With the recent approval of the first small interfering RNA (siRNA) therapeutic formulated as nanoparticles, there is increased incentive for establishing the factors of importance for the design of stable solid dosage forms of such complex nanomedicines. METHODS: The aims of this study were: (i) to identify factors of importance for the design of spray-dried siRNA-loaded lipidoid-poly(DL-lactic-co-glycolic acid) hybrid nanoparticles (LPNs), and (ii) to evaluate their influence on the resulting powders by using a quality-by-design approach. Critical formulation and process parameters were linked to critical quality attributes (CQAs) using design of experiments, and an optimal operating space (OOS) was identified. RESULTS: A series of CQAs were identified based on the quality target product profile. The loading (ratio of LPNs to the total solid content) and the feedstock concentration were determined as critical parameters, which were optimized systematically. Mannitol was chosen as stabilizing excipient due to the low water content of the resulting powders. The loading negatively affected the colloidal stability of the LPNs, whereas feedstock concentration correlated positively with the powder particle size. The optimal mannitol-based solid formulation, defined from the OOS, displayed a loading of 5% (w/w), mass median aerodynamic diameter of 3.3 ± 0.2 µm, yield of 60.6 ± 6.6%, and a size ratio of 1.15 ± 0.03. Dispersed micro-embedded LPNs had preserved physicochemical characteristics as well as in vitro siRNA release profile and gene silencing, as compared to non-spray-dried LPNs. CONCLUSION: The optimal solid dosage forms represent robust formulations suitable for higher scale-up manufacturing.

Temperature-Induced Self-Assembly of the Group B Streptococcus (GBS) Fusion Antigen GBS-NN.

Group B Streptococcus (GBS) is a leading cause of serious bacterial neonatal infections worldwide, which provides an unmet medical need for a globally effective vaccine. The recombinant GBS fusion antigen GBS-NN contains the N-terminal regions of the GBS Rib and Alpha C proteins. It shows promising immunogenicity eliciting protective immunity in mice and encouraging results in early human clinical trials. Understanding the physical stability of GBS-NN containing conformational B-cell epitopes is crucial to ensure optimal vaccine stability, efficacy, and safety. We initially discovered that GBS-NN is prone to form higher-order structures at elevated temperatures. We therefore investigated the self-assembly behavior of GBS-NN and characterized the higher-order conformational structures as a function of temperature. In the native state, GBS-NN exists as a monomer and has a secondary structure containing α-helix and ß-sheet. Langmuir studies demonstrated that the native protein is highly surface-active and forms a monolayer film at the air-water interface because of its amphipathic properties. The conformational stability of GBS-NN was measured as a function of temperature. GBS-NN has an unusual thermal behavior with a phase transition of approximately 61 °C, which is not accompanied by any major changes in the secondary structure. However, the antigen showed irreversible self-assembly as a function of temperature into higher-order structures with a hydrodynamic diameter of approximately 100 nm. Cryo-transmission electron microscopy analyses demonstrated that these self-assemblies consist of vesicular, ring-like structures with a hollow aqueous interior. Therefore, GBS-NN is a physically stable monomeric protein but is prone to temperature-induced self-assembly above 61 °C.

Adjuvants Based on Synthetic Mycobacterial Cord Factor Analogues: Biophysical Properties of Neat Glycolipids and Nanoself-Assemblies with DDA.

Synthetic mycobacterial cord factor analogues, e.g., trehalose 6,6'-dibehenate (TDB), are highly promising adjuvants due to their strong immunopotentiating capabilities, but their biophysical properties have remained poorly characterized. Here, we report the synthesis of an array of synthetic TDB analogues varying in acyl chain length, degree of acylation, and headgroup display, which was subjected to biophysical characterization of neat nondispersed self-assembled nanostructures in excess buffer and as aqueous dispersions with cationic dimethyldioctadecylammonium (DDA) bromide. The array comprised trehalose mono- (TMX) and diester (TDX) analogues with symmetrically shortened acyl chains [denoted by X: arachidate (A), stearate (S), palmitate (P), myristate (Myr), and laurate (L)] and an analogue with a short hydrophilic polyethylene glycol (PEG) linker inserted between the trehalose headgroup of TDS and the acyl chains (PEG-TDS). All dispersions were liposomes, but in contrast to the colloidally stable and highly cationic TDX-containing liposomes, the zeta-potential was significantly reduced for DDA/TMX and DDA/PEG-TDS liposomes, suggesting a charge-shielding effect, which compromises the colloidal stability. An increased d-spacing was observed for the lamellar phase of neat TDB analogues in excess buffer (TDS < TMS < PEG-TDS), confirming that the charge shielding is caused by an extended molecular configuration of the more flexible headgroup. Differential scanning calorimetry showed highly cooperative phase transitions for all tested dispersions albeit the monoesters destabilized the lipid bilayers. Langmuir experiments demonstrated that incorporation of TDXs and PEG-TDS stabilized DDA monolayers due to improved hydrogen bonding and reduced intermolecular repulsions. In conclusion, data suggest that the DDA/TDS dispersions exhibit favorable physicochemical properties rendering these DDA/TDS liposomes an attractive vaccine adjuvant, and they emphasize that not only the receptor binding and immune activation but also the biophysical properties of immunopotentiator formulations should be collectively considered when designing adjuvants with optimal safety, efficacy, and storage stability.

Elucidating the mechanisms of protein antigen adsorption to the CAF/NAF liposomal vaccine adjuvant systems: effect of charge, fluidity and antigen-to-lipid ratio.

The reverse vaccinology approach has recently resulted in the identification of promising protein antigens, which in combination with appropriate adjuvants can stimulate customized, protective immune responses. Although antigen adsorption to adjuvants influences vaccine efficacy and safety, little is generally known about how antigens and adjuvants interact at the molecular level. The aim of this study was to elucidate the mechanisms of interactions between the equally sized, but oppositely charged model protein antigens α-lactalbumin and lysozyme, and i) the clinically tested cationic liposomal adjuvant CAF01 composed of cationic dimethyldioctadecylammonium (DDA) bromide and trehalose-6,6'-dibehenate (TDB) or ii) the neutral adjuvant formulation NAF01, where DDA was replaced with zwitterionic distearoylphosphatidylcholine (DSPC). The effect of liposome charge, bilayer rigidity, isoelectric point and antigen-to-lipid ratio was investigated using dynamic light scattering, transmission electron microscopy, differential scanning calorimetry, intrinsic fluorescence and Langmuir monolayers. The net anionic α-lactalbumin adsorbed onto the cationic liposomes, while there was no measureable attractive interaction with the zwitterionic liposomes. In contrast, the net cationic lysozyme showed very little interaction with either types of liposome. Adsorption of α-lactalbumin altered its tertiary structure, affected lipid membrane packing below and above the phase transition temperature, and neutralized the liposomal surface charge, resulting in reduced colloidal stability and liposome aggregation. Langmuir studies revealed that α-lactalbumin was not squeezed out of DDA monolayers upon compression, which suggests additional hydrophobic interactions. Such interactions are thus likely to affect the way vaccine antigens are presented to antigen-presenting cells, and may play an important role for the efficacy of the vaccine-induced immune response. These studies thus exemplify the importance of characterizing the molecular interactions between the vaccine antigen and adjuvant along with immunogenicity and efficacy studies.

Formulation of RNA interference-based drugs for pulmonary delivery: challenges and opportunities.

With recent advances in the field of RNAi-based therapeutics, it is possible to make any target gene 'druggable', at least in principle. The present review focuses on aspects critical for pulmonary delivery of formulations of nucleic acid-based drugs. The first part introduces the therapeutic potential of RNAi-based drugs for the treatment of lung diseases. Subsequently, we discuss opportunities for formulation-enabled pulmonary delivery of RNAi drugs in light of key physicochemical properties and physiological barriers. In the following section, an overview is included of methodologies for imparting inhalable characteristics to nucleic acid formulations. Finally, we review one of the bottlenecks in the early preclinical testing of inhalable nucleic acid-based formulations, in other words, devices suitable for pulmonary administration of powder-based formulations in rodents.

A strong adjuvant based on glycol-chitosan-coated lipid-polymer hybrid nanoparticles potentiates mucosal immune responses against the recombinant Chlamydia trachomatis fusion antigen CTH522.

Induction of mucosal immunity with vaccines is attractive for the immunological protection against pathogen entry directly at the site of infection. An example is infection with Chlamydia trachomatis (Ct), which is the most common sexually transmitted infection in the world, and there is an unmet medical need for an effective vaccine. A vaccine against Ct should elicit protective humoral and cell-mediated immune (CMI) responses in the genital tract mucosa. We previously designed an antibody- and CMI-inducing adjuvant based on poly(dl-lactic-co-glycolic acid) (PLGA) nanoparticles modified with the cationic surfactant dimethyldioctadecylammonium bromide and the immunopotentiator trehalose-6,6'-dibehenate. Here we show that immunization with these lipid-polymer hybrid nanoparticles (LPNs) coated with the mucoadhesive polymer chitosan enhances mucosal immune responses. Glycol chitosan (GC)-modified LPNs were engineered using an oil-in-water single emulsion solvent evaporation method. The nanoparticle design was optimized in a highly systematic way by using a quality-by-design approach to define the optimal operating space and to gain maximal mechanistic information about the GC coating of the LPNs. Cryo-transmission electron microscopy revealed a PLGA core coated with one or several concentric lipid bilayers. The GC coating of the surface was identified as a saturable, GC concentration-dependent increase in particle size and a reduction of the zeta-potential, and the coating layer could be compressed upon addition of salt. Increased antigen-specific mucosal immune responses were induced in the lungs and the genital tract with the optimized GC-coated LPN adjuvant upon nasal immunization of mice with the recombinant Ct fusion antigen CTH522. The mucosal responses were characterized by CTH522-specific IgG/IgA antibodies, together with CTH522-specific interferon γ-producing Th1 cells. This study demonstrates that mucosal administration of CTH522 adjuvanted with chitosan-coated LPNs represents a promising strategy to modulate the magnitude of mucosal vaccine responses.

Immunological and physical evaluation of the multistage tuberculosis subunit vaccine candidate H56/CAF01 formulated as a spray-dried powder.

Liquid vaccine dosage forms have limited stability and require refrigeration during their manufacture, distribution and storage. In contrast, solid vaccine dosage forms, produced by for example spray drying, offer improved storage stability and reduced dependence on cold-chain facilities. This is advantageous for mass immunization campaigns for global public health threats, e.g., tuberculosis (TB), and offers cheaper vaccine distribution. The multistage subunit vaccine antigen H56, which is a fusion protein of the Mycobacterium tuberculosis (Mtb) antigens Ag85B, ESAT-6, and Rv2660, has been shown to confer protective efficacy against active TB before and after Mtb exposure in preclinical models, and it is currently undergoing clinical phase 2a testing. In several studies, including a recent study comparing multiple clinically relevant vaccine adjuvants, the T helper type 1 (Th1)/Th17-inducing adjuvant CAF01 was the most efficacious adjuvant for H56 to stimulate protective immunity against Mtb. With the long-term goal of designing a thermostable and self-administrable dry powder vaccine based on H56 and CAF01 for inhalation, we compared H56 spray-dried with CAF01 with the non-spray-dried H56/CAF01 vaccine with respect to their ability to induce systemic Th1, Th17 and humoral responses after subcutaneous immunization. Here we show that spray drying of the H56/CAF01 vaccine results in preserved antigenic epitope recognition and adjuvant activity of CAF01, and the spray-dried, reconstituted vaccine induces antigen-specific Th1, Th17 and humoral immune responses, which are comparable to those stimulated by the non-spray-dried H56/CAF01 vaccine. In addition, the spray-dried and reconstituted H56/CAF01 vaccine promotes similar polyfunctional CD4+ T-cell responses as the non-spray-dried vaccine. Thus, our study provides proof-of-concept that spray drying of the subunit vaccine H56/CAF01 preserves vaccine-induced humoral and cell-mediated immune responses. These results support our ongoing efforts to develop a thermostable, dry powder-based TB vaccine.

Dual-Isotope SPECT/CT Imaging of the Tuberculosis Subunit Vaccine H56/CAF01: Induction of Strong Systemic and Mucosal IgA and T-Cell Responses in Mice Upon Subcutaneous Prime and Intrapulmonary Boost Immunization.

Pulmonary tuberculosis (TB), which is caused by Mycobacterium tuberculosis (Mtb), remains a global pandemic, despite the widespread use of the parenteral live attenuated Bacillus Calmette-Guérin (BCG) vaccine during the past decades. Mucosal administration of next generation TB vaccines has great potential, but developing a safe and efficacious mucosal vaccine is challenging. Hence, understanding the in vivo biodistribution and pharmacokinetics of mucosal vaccines is essential for shaping the desired immune response and for optimal spatiotemporal targeting of the appropriate effector cells in the lungs. A subunit vaccine consisting of the fusion antigen H56 (Ag85B-ESAT-6-Rv2660) and the liposome-based cationic adjuvant formulation (CAF01) confers efficient protection in preclinical animal models. In this study, we devise a novel immunization strategy for the H56/CAF01 vaccine, which comply with the intrapulmonary (i.pulmon.) route of immunization. We also describe a novel dual-isotope (111In/67Ga) radiolabeling approach, which enables simultaneous non-invasive and longitudinal SPECT/CT imaging and quantification of H56 and CAF01 upon parenteral prime and/or i.pulmon. boost immunization. Our results demonstrate that the vaccine is distributed evenly in the lungs, and there are pronounced differences in the pharmacokinetics of H56 and CAF01. We provide convincing evidence that the H56/CAF01 vaccine is not only well-tolerated when administered to the respiratory tract, but it also induces strong lung mucosal and systemic IgA and polyfunctional Th1 and Th17 responses after parenteral prime and i.pulmon. boost immunization. The study furthermore evaluate the application of SPECT/CT imaging for the investigation of vaccine biodistribution after parenteral and i.pulmon. immunization of mice.

Unusual Self-Assembly of the Recombinant Chlamydia trachomatis Major Outer Membrane Protein-Based Fusion Antigen CTH522 Into Protein Nanoparticles.

Sexually transmitted Chlamydia trachomatis infects more than 100 million people annually, and untreated chlamydia infections can cause severe complications. Therefore, there is an urgent need for a chlamydia vaccine. The Ctrachomatis major outer membrane protein (MOMP) is highly immunogenic but is a challenging vaccine candidate by being an integral membrane protein, and the immunogenicity depends on a correctly folded structure. We investigated the biophysical properties of the recombinant MOMP-based fusion antigen CTH522, which is tested in early human clinical trials. It consists of a truncated and cysteine-free version of MOMP fused to 4 variable domains from serovars D-G. In the native state, CTH522 did not exist as a monomer but showed an unusual self-assembly into nanoparticles with a negative zeta potential. In contrast to the ß-barrel structure of MOMP, native CTH522 contained no well-defined secondary structural elements, and no thermal transitions were measurable. Chemical unfolding resulted in monomers that upon removal of the denaturant self-assembled into higher order structures, comparable to the structure of the native protein. The conformation of CTH522 in nanoparticles is thus not entirely random and contains structural elements stabilized via denaturant-disruptable hydrophobic interactions. In conclusion, CTH522 has an unusual quaternary structure of supramolecular self-assemblies.

Influence of trehalose 6,6'-diester (TDX) chain length on the physicochemical and immunopotentiating properties of DDA/TDX liposomes.

Linking physicochemical characterization to functional properties is crucial for defining critical quality attributes during development of subunit vaccines toward optimal safety and efficacy profiles. We investigated how the trehalose 6,6'-diester (TDX) chain length influenced the physicochemical and immunopotentiating properties of the clinically tested liposomal adjuvant composed of dimethyldioctadecylammonium (DDA) bromide and analogues of trehalose-6,6'-dibehenate (TDB). TDB analogues with symmetrically shortened acyl chains [denoted X: arachidate (A), stearate (S), palmitate (P), myristate (Myr) and laurate (L)] were incorporated into DDA liposomes and characterized with respect to size, polydispersity index, charge, thermotropic phase behavior and lipid-lipid interactions. Incorporation of 11 mol% TDX into DDA liposomes significantly decreased the polydispersity index when TDA, TDS, TDP and TDMyr were incorporated, whereas both the initial size and the charge of the liposomes were unaffected. The long-term colloidal stability was only decreased when including TDL in DDA liposomes. The fatty acid length of TDX affected the phase transition of the liposomes, and for the DDA/TDP and DDA/TDS liposomes a homogeneous distribution of the lipids in the bilayer was indicated. The membrane packing was studied further by using the Langmuir monolayer technique. Incorporation of TDS improved the packing of the lipid monolayer, as compared to the other analogues, suggesting the most favorable stability. Finally, immunization of mice with the recombinant tuberculosis fusion antigen Ag85B-ESAT-6-Rv2660c (H56) and the physicochemically most optimal formulations (DDA/TDB, DDA/TDS and DDA/TDP) induced comparable T-cell responses. In conclusion, of the investigated TDB analogues, incorporation of 11 mol% TDS or TDP into DDA liposomes resulted in an adjuvant system with the most favorable physicochemical properties and an immunological profile comparable to that of DDA/TDB.

Engineering of a novel adjuvant based on lipid-polymer hybrid nanoparticles: A quality-by-design approach.

The purpose of this study was to design a novel and versatile adjuvant intended for mucosal vaccination based on biodegradable poly(DL-lactic-co-glycolic acid) (PLGA) nanoparticles (NPs) modified with the cationic surfactant dimethyldioctadecylammonium (DDA) bromide and the immunopotentiator trehalose-6,6'-dibehenate (TDB) (CAF01) to tailor humoral and cellular immunity characterized by antibodies and Th1/Th17 responses. Such responses are important for the protection against diseases caused by intracellular bacteria such as Chlamydia trachomatis and Mycobacterium tuberculosis. The hybrid NPs were engineered using an oil-in-water single emulsion method and a quality-by-design approach was adopted to define the optimal operating space (OOS). Four critical process parameters (CPPs) were identified, including the acetone concentration in the water phase, the stabilizer [polyvinylalcohol (PVA)] concentration, the lipid-to-total solid ratio, and the total concentration. The CPPs were linked to critical quality attributes consisting of the particle size, polydispersity index (PDI), zeta-potential, thermotropic phase behavior, yield and stability. A central composite face-centered design was performed followed by multiple linear regression analysis. The size, PDI, enthalpy of the phase transition and yield were successfully modeled, whereas the models for the zeta-potential and the stability were poor. Cryo-transmission electron microscopy revealed that the main structural effect on the nanoparticle architecture is caused by the use of PVA, and two different morphologies were identified: i) A PLGA core coated with one or several concentric lipid bilayers, and ii) a PLGA nanoshell encapsulating lipid membrane structures. The optimal formulation, identified from the OOS, was evaluated in vivo. The hybrid NPs induced antibody and Th1/Th17 immune responses that were similar in quality and magnitude to the response induced by DDA/TDB liposomes, showing that the adjuvant properties of DDA/TDB are maintained in the PLGA hybrid matrix. This study demonstrates the complexity of formulation design for the engineering of a hybrid lipid-polymer nanoparticle adjuvant.

Mechanism of action of lung damage caused by a nanofilm spray product.

Inhalation of waterproofing spray products has on several occasions caused lung damage, which in some cases was fatal. The present study aims to elucidate the mechanism of action of a nanofilm spray product, which has been shown to possess unusual toxic effects, including an extremely steep concentration-effect curve. The nanofilm product is intended for application on non-absorbing flooring materials and contains perfluorosiloxane as the active film-forming component. The toxicological effects and their underlying mechanisms of this product were studied using a mouse inhalation model, by in vitro techniques and by identification of the binding interaction. Inhalation of the aerosolized product gave rise to increased airway resistance in the mice, as evident from the decreased expiratory flow rate. The toxic effect of the waterproofing spray product included interaction with the pulmonary surfactants. More specifically, the active film-forming components in the spray product, perfluorinated siloxanes, inhibited the function of the lung surfactant due to non-covalent interaction with surfactant protein B, a component which is crucial for the stability and persistence of the lung surfactant film during respiration. The active film-forming component used in the present spray product is also found in several other products on the market. Hence, it may be expected that these products may have a toxicity similar to the waterproofing product studied here. Elucidation of the toxicological mechanism and identification of toxicological targets are important to perform rational and cost-effective toxicological studies. Thus, because the pulmonary surfactant system appears to be an important toxicological target for waterproofing spray products, study of surfactant inhibition could be included in toxicological assessment of this group of consumer products.

Immunity by formulation design: induction of high CD8+ T-cell responses by poly(I:C) incorporated into the CAF01 adjuvant via a double emulsion method.

The combination of nucleic acid-based Toll-like receptor (TLR)-3 or TLR9 agonists and cationic liposomes constitutes an effective vaccine adjuvant approach for eliciting CD8+ T-cell responses. However, complexing cationic liposomes and oppositely charged oligonucleotides generally results in highly unstable and heterogeneous formulations with limited clinical applicability. The aim of this study was to design, formulate, and carefully characterize a stable CD8-inducing adjuvant based on the TLR3 ligand polyinosinic-polycytidylic acid [poly(I:C)] incorporated into a cationic adjuvant system (CAF01) composed of dimethyldioctadecylammonium (DDA) and trehalose 6,6'-dibehenate (TDB). For this purpose, a modified double emulsion solvent evaporation method was investigated for complexation of high amounts of anionic poly(I:C) to gel-state DDA/TDB liposomes. Addition of a volatile, water-miscible co-solvent (ethanol) to the outer water phase enabled preparation of colloidally stable liposomes, presumably by reducing the poly(I:C)-enhanced rigidity of the lipid bilayer. Cryo-transmission electron microscopy (TEM) revealed the formation of unilamellar as well as multilamellar liposomes, the latter suggesting that poly(I:C) is intercalated between the membrane bilayers in an onion-like structure. Finally, immunization of mice with the model antigen ovalbumin (OVA) and DDA/TDB/poly(I:C) liposomes induced a remarkably strong, antigen-specific CD8+ T-cell response, which was maintained for more than two months. Importantly, whereas injection of soluble poly(I:C) led to rapid production of the pro-inflammatory cytokines tumor necrosis factor (TNF)-α and interleukin (IL)-6 in serum, administration of poly(I:C) in complex with the cationic DDA/TDB liposomes prevented this non-specific systemic pro-inflammatory response. These data emphasize the importance of improving the quality of the vaccine formulation to indeed overcome some of the major obstacles for using CD8-inducing agents such as poly(I:C) in future subunit vaccines.