Development of a New Vaccine Adjuvant System Based on the Combination of the Synthetic TLR4 Agonist FP20 and a Synthetic QS-21 Variant

Abstract

In this study, we formulated an alternative to AS01b by combining FP20, a synthetic TLR4 agonist, and QS21v, a minimal saponin adjuvant, aiming to improve the vaccine efficacy and stability. The phase transition temperature of FP20 was determined by using differential scanning calorimetry to be 43.9 °C, providing a foundation for the formulation process. The coformulation was prepared using a dry film method for even adjuvant distribution. Characterization by dynamic light scattering and nanoparticle tracking analysis revealed a uniform particle size distribution of ∼120 nm. Cryogenic electron microscopy (CryoEM) revealed nanosized interactions between FP20 and QS21v, forming stable structures that likely enhanced the antigen presentation and immune activation. These physicochemical properties contributed to a robust in vivo synergy, where the coformulation elicited significantly higher antigen-specific antibody titers compared to individual adjuvants. These findings suggest that the FP20+QS21v coformulation provides a potent, stable, and safer alternative to traditional adjuvants, enhancing both vaccine efficacy and immunogenicity.

Introduction

The rise in health crises due to outbreaks and pandemics has emphasized the significance of vaccines. With the emergence of diseases, there is an urgent need for effective vaccination strategies to halt the spread of diseases that can impact health and have economic repercussions.^1,2 The evolution of vaccination methods from inactivated viruses to modern day mRNA techniques signifies a notable shift toward safer, more precise, and robust immune responses.^3,4 Since the discovery of adjuvants, they have played a vital role in subunit vaccines by enhancing the immune response and the stability of vaccine formulations.^5,6 Adjuvants also make it possible to reduce the antigen doses and the number of booster immunizations, generate more rapid and durable immune responses, and increase the effectiveness of vaccines in poor responders. Despite their key role, few efficient adjuvants with acceptable toxicity for human use are available in licensed vaccines. For more than 70 years, alum (a mixture of diverse aluminum salts) has been the only approved adjuvant in humans.⁷ Besides aluminum salts, the other few molecules included in adjuvant systems (AS) approved for human use are the Toll-like Receptor 4 (TLR4) agonist monophosphoryl lipid A (MPLA), squalene, and the saponin natural product QS-21.⁸ MPLA is a detoxified Salmonella minnesota R595 lipid A analogue obtained by hydrolysis of the C1 phosphate and (R)-3-hydroxytetradecanoyl groups⁹ (Figure 1). The lack of the C1 phosphate group allows it to maintain its immunostimulating properties while eliminating the toxicity. Its activity, as well as that of its synthetic analogue GLA, is based on TLR4 stimulation that results in the promotion of Th1 (cellular)-biased immune response.¹⁰ MPLA is present in the formulation of AS used in human vaccines: in combination with aluminum salts (alum) in AS04, which is approved for the Cervarix vaccine against human papillomavirus (HPV), and in AS01, a liposomal formulation containing MPLA and the saponin QS-21, which has been recently approved for GSK’s malaria (Mosquirix)¹¹ and shingles (Shingrix)¹² vaccines.

Figure 1.

Chemical structures of MPLA (A), QS-21 (B), FP20 (C), and QS21v (D).

A synthetic version of MPLA is prepared through a complex chemical route consisting of about 25 reaction steps and several challenging purifications of synthesis intermediates.^6,13 Several simplified synthetic variants of MPLA have been developed such as a variety of monosaccharide-based glycolipids. Some of these derivatives have shown the ability to enhance immune responses with a similar or higher potency than MPLA, including compounds of the FP series developed by some of us¹⁴⁻¹⁶ (Figure 1). Among these, we recently developed FP20, a glucosamine-based TLR4 agonist with similar immunostimulatory potency to MPLA and easily accessible through a streamlined, sustainable 6-step synthesis, key features that have facilitated its advancement as a commercially available adjuvant candidate.¹⁷ Even though FP20 is less potent than other similar FP compounds, such as FP18, for example, in stimulating immune response,¹⁷ its improved chemical stability (phosphate group in C4 instead of C1 position of glucosamine) and efficient and scalable synthesis make this compound the best candidate for preclinical development.

QS-21 is a complex saponin natural product extracted from the bark of the tree Quillaja saponaria Molina¹⁸ (Figure 1). Despite its potent adjuvant activity and extensive clinical use, QS-21 suffers from several intrinsic shortcomings that have made difficult its approval as part of new human vaccines, with the recent, notable exceptions of the AS01 and Matrix-M adjuvant formulations licensed for use against malaria, shingles, and COVID-19.¹⁹ The high environmental footprint associated with the processes of extraction and purification of QS-21 as well as its complex, heterogeneous, and chemically unstable structure represent several key challenges in continued vaccine deployment. To address these availability and structural issues, great efforts have been made on the development of alternative approaches for accessing QS-21 (and potentially optimized analogues) beyond the laborious bark extraction, including chemical synthesis methods,²⁰ plant cell culture production,²¹ and biosynthetic strategies for heterologous expression.²² In addition to the inherent liabilities mentioned above, QS-21 is reactogenic and hemolytic and presents dose-limiting toxicity, resulting in low-grade side effects that have precluded its approval as a stand-alone adjuvant. These limitations of free QS-21 can be mitigated by formulation with cholesterol and phospholipids into liposomes (AS01,²³ and the Army Liposome Formulation Q²⁴)²⁵ and into other lipid-based complexes such as ISCOMs and Matrix-M.²⁶ A key representative example is the liposome-based adjuvant AS01 developed by GlaxoSmithKline, which combines QS-21 and MPLA and was first approved in humans as part of the Shingrix vaccine against shingles.¹² This combination of adjuvants, the so-called adjuvant systems, represents a transformative approach in vaccinology, whereby the mixture of different adjuvant molecules activating diverse immune pathways leads to a more potent, synergistic response.²⁷ This strategy has given rise to several novel adjuvant systems, such as, AS01, AS03, and AS04, each designed to target specific immunological needs.^5,28

The mixture of MPLA and QS-21 has been shown to increase vaccine immunogenicity and protection as well as the quality of the immune response due to cumulative immunological activities, both in AS01²⁷ as well as when incorporated into ISCOM nanoparticles.²⁹ This adjuvant combination creates a synergistic effect that stimulates potent innate and adaptive immune responses, with early IFN-γ production by NK and CD8⁺ T cells mediated by subcapsular macrophages via IL-18 secretion. In turn, this innate IFN-γ release promotes dendritic cell activation and Th1-biased, antigen-specific CD4⁺ T-cell responses relevant to protection.²⁷

We present here a new adjuvant mixture in which the native MPLA and QS-21 components have been replaced by synthetic, structurally simplified analogues, namely, an optimized FP20 compound and a streamlined QS-21-derived lead (QS-21 variant, QS21v) (Figure 1). This study has involved a comprehensive evaluation of the physicochemical properties of this new adjuvant combination by using several techniques: differential scanning calorimetry (DSC), dynamic light scattering (DLS), ζ-potential analysis, and cryogenic electron microscopy (CryoEM). Subsequent in vivo experiments in mice involved an ovalbumin (OVA) vaccine model incorporating our FP20/QS-21v adjuvant coformulation. The analysis of the induced antibody levels (IgG titers) after immunization revealed a synergistic, enhanced humoral response, providing insights into the efficacy of this novel adjuvant approach, which could open the door to more accessible and ecologically sustainable vaccine technologies.

Results and Discussion

Determination of the Main Phase Transition Temperature (T_m) of FP20

The transition properties of glycolipids like FP20 play a critical role in the stability and efficacy of the vaccine formulations.^30,31T_m depends on several factors such as fatty acid chain length, degree of saturation, charge, and headgroup.³²⁻³⁴T_m determines the fluidity and permeability of the liposome bilayer, as at temperatures lower than T_m, the lipids are in the gel phase, which has low fluidity and low permeability. An essential characteristic of such lipids is their phase T_m, above which they demonstrate enhanced stability.^35,36

The thermal behavior of glycolipid FP20 can be described by applying DSC. The first heating trace of FP20 shows a narrow endotherm phenomenon centered at T_m = 43.9 °C (ΔH = −14.7 J g^–1) corresponding to the phase transition of the acyl chains (Figure 2a).

Figure 2.

(a) DSC thermogram of FP20 from −30 to 55 °C recorded at 10 °C/min. (b) DSC thermal cycle of FP20 carried out at 10 °C/min: the sample is heated from −30 to 55 °C, successively cooled to −30 °C followed by a second heating run to 55 °C (the dashed line represents the first heating trace, and the solid line refers to the cooling and the second heating traces).

The reversibility of this phase transformation is demonstrated by the application of a thermal cycle carried out in the range −30 to +55 °C (Figure 2b) at 10 °C/min. During the cooling treatment, the “crystallization” of acyl chains occurs at 35 °C and the resulting “melting” of the ordered domains appears evident in the second heating trace (T_m (II run) = 41.6 °C). The enthalpy associated with this phase transition (ΔH_(II run) = −12.0 J g^–1) is lower than that measured in the first heating trace because of the high cooling rate that prevents the complete crystallization of the acyl chains. The observed T_m of 43.9 °C for FP20 is significant for several reasons as operating above this T_m is crucial for maintaining the stability of the vaccine formulation. This stability is expected to enhance the efficacy of the adjuvant, ensuring that its immunostimulatory properties are optimally preserved and delivered within the physiological environment. Additionally, this finding guides the temperature conditions for subsequent synthesis and processing steps, particularly during the sonication and formulation of the coformulations.

Formulation Studies

The lipid-based adjuvant system was initially formulated by combining FP20 and QS-21v with the amphiphile 1,2-dioleoyl-glycero-3-phosphocholine (DOPC) and cholesterol. DOPC was selected because it has the same zwitterionic unsaturated phospholipid used in commercially available AS01b. While the exact concentration of DOPC and cholesterol is not disclosed in the literature, reports suggest that the mole percent ratio of cholesterol to phospholipid in AS01 is approximately 33.7%.²⁵ In contrast, we utilized a 9:1 ratio of DOPC to cholesterol. This reduction in cholesterol was guided by previous findings that a higher cholesterol was required to mitigate the toxicity associated with the branched trisaccharide in native QS-21. Since QS21v, a simplified adjuvant lacking the branched trisaccharide, exhibits reduced toxicity,³¹ a lower cholesterol content was deemed sufficient for our formulation. Liposomes were formulated using the nitrogen-pressured extrusion system and a 0.1-μm polycarbonate membrane. The DLS analysis showed a polydispersity index (PDI) of 0.4, indicating a highly heterogeneous formulation (data not shown). Because of the deceiving results using DOPC, we then tried 1,2-dimyristoyl-glycero-3-phosphatidylcholine (DMPC), with a 14-carbon lipid chain which is similar to the 14-C fatty acid chains of FP20. The coformulation consisting of FP20 (100 μg/mL) and QS21v (500 μg/mL) in combination with DMPC (2.7 mm) and cholesterol (0.3 mM) was named Lipo-coForm.

Lipo-coForm was found to be stable, with a particle average size of 110 nm and a polydispersity index of 0.2, indicating the formation of a homogeneous formulation (Table S1). The cryoEM images of the Lipo-coform clearly show that both components FP20 and QS21v are integrated in the liposomal bilayer (Figure S1).

Mouse immunization with Lipo-coForm failed to show significant IgG responses. Indeed, we observed a reduced response compared to the one induced in mice immunized with control FP20 liposomes. The QS21v(lipos.) formulation exhibited an IgG response similar to that shown by the OVA antigen alone. Surprisingly, the FP20+QS21v combination without liposomes induced the highest activity of all the formulations tested, being superior to the antibody response observed in the control groups containing MPLA (Figure S3). These results prompted us to focus on plain vaccine formulations without additional lipids while considering the physicochemical characteristics of the combination of FP20 and QS21v based on the self-assembling properties of FP20 and QS-21v.³⁷

Thus, the formulation of both components was planned carefully, as FP20 and QS21v are not readily soluble in aqueous media due to their peculiar amphiphilic character. To form homogeneous aggregates, the widely used dry film method was used. The film was hydrated by using PBS, and the sample was incubated at 45 °C (above FP20 T_m) for 10 min. The incubated samples were sonicated at 37 Hz for 10 min. Studies on other glycolipid adjuvants, such as MPLA, have a reported T_m value of 41.7 °C, slightly lesser than for FP20.³⁸ Several studies have shown that operating above the T_m significantly improved the stability of the formulations.^35,36,39,40

We carried out the size distribution analysis of the coformulations of the FP20 and QS21v using nanoparticle tracking analysis (NTA) and DLS. Figure 3 presents the NTA results, and Table 1 tabulates the DLS results. The size distribution of the FP20, QS21v, and the coformulations was found to be uniform between the NTA and DLS methods. The uniformity of all the formulations is confirmed by the PDI as it is below 0.2. Studies have demonstrated that formulations with a PDI of 0.2 or below exhibit enhanced stability and uniformity, which is an important factor in developing efficient drug delivery systems.^41,42

Figure 3.

Size distribution and concentration profiles of FP20, QS21v and FP20+QS21v formulations.

Table 1. Average Diameter, PDI, and ζ-Potential of the FP20, QS21v, and FP20+QS21v Formulations.

formulation	diameter (nm ± SD)	PDI	ζ-potential (mV ± SD)
FP20	80 ± 2	0.15	–37 ± 3
QS21v	90 ± 3	0.20	–27 ± 1
FP20+QS21v	110 ± 6	0.17	–43 ± 1

The ζ-potential values, indicating the surface charge of all the formulations, are indicated in the Table 1. All the formulations are highly negatively charged (∼−40 mV) and suggest that they are stable in suspension. The coformulation showed a slightly more negative charge than the individual components, which could enhance particle stability by preventing aggregation through electrostatic repulsion.

As the formulations are highly negatively charged, we had problems analyzing them under the cryoEM. During the vitrification process, we coated the cryoEM grid with negatively charged ion plasmon to be able to observe it under cryoEM. We were unable to observe any formulations under the cryoEM due to the electrostatic repulsion between the grid and the formulations. We overcame this problem by coating the grid with 0.1% proline and drying for 30 min at 37 °C before loading the sample. This step prevented the sample wash-off during the vitrification step and ensured good imaging spectra. We have also collected the proline spectra alone to make sure it is not interfering with the formulations images. The proline alone has shown small background patches, which are clearly distinct from what we noticed in the formulation samples (Figure S2).

FP20 formed micellar bilayer structures, which are typically found in lipid molecules. These bilayer structures have an average size of 60 nm. The QS21v appeared to be forming rod-like structures with a tube diameter of around 5 nm. In the coformulation sample, we observed the FP20 bilayer structures interacting with the surface of the QS21v rods, forming a complex composite structure (Figure 4). The rod-like morphology in QS21v may result from its amphiphilic nature, where polar and nonpolar regions are distributed irregularly across the molecule, whereas an organized polarity in FP20 drives typical bilayer micelle formation. This structural divergence allows FP20 micelles to align peripherally along QS21v rods, creating a stabilized complex through hydrophilic–hydrophobic interactions. Such peripheral interactions not only stabilize the complex but also may enhance cellular uptake and antigen processing by APCs through direct contact with the cell membranes. FP20′s bilayer structure resembles cellular lipid layers, facilitating close contact with APC membranes, which is beneficial for endocytosis of the FP20-QS21v complex. This interaction could be important in achieving their synergistic effect as opposed to the Lipo-coform samples, which we synthesized earlier. In the Lipo-coform samples, we observed the FP20 and QS21v integrating in the liposomes but those formulations did not elicit an in vivo IgG response (Figure S3). In contrast, the peripheral interaction between FP20 and QS21v may have contributed to the observed IgG antibody induction, indicating that surface interaction could play an important role. This structured peripheral arrangement likely facilitates the dual engagement of immune receptors, known to boost dendritic cell maturation and antigen cross-presentation, leading to robust adaptive immunity. While AS01b is formulated into liposomes and many studies on the combined MPLA and native QS-21 formulations have shown synergistic activity^27,29 but key differences in structure, chemical composition and homogeneity of our adjuvants compared to the heterogeneous MPLA and QS-21 products might have contributed to this different effect. Indeed, although MPLA and FP20 are both TLR4 agonists, they exhibit significant structural differences that can influence their biological activity and formulation compatibility.

Figure 4.

CryoEM images showing the structural characteristics of all the formulations. FP20 (A) formed spherical micellar structures, while the QS21v (B) revealed a rod-like morphology. The FP20+QS21v coformulation (C) exhibited peripheral interactions between the FP20 bilayer and the QS21v rods (scale bar −100 nm).

Previous studies on FP20 showed that it activates TLR4 through a distinct pathway relative to MPLA. While MPLA triggers both MyD88-dependent and TRAM/TRIF-dependent pathways, leading to pro-inflammatory cytokine production such as TNF, IL-6, and pro-IL-1β, FP20 has been shown to specifically induce significant activation of p38 MAPK and production of IL-1β without stimulating the NF-κB (p65 subunit) or p-IRF-3 pathways.¹⁷ Similarly, even with their important structural differences in terms of truncated carbohydrate moieties and overall molecular architecture, QS21v retained native QS-21 adjuvant activity. Natural QS-21 has a complex structure comprising a quillaic acid triterpene core flanked by specific oligosaccharide chains at both sides and an elaborate acyl side chain along the periphery that is crucial for its adjuvant activity. QS21v is a streamlined synthetic variant developed to improve upon the limitations of naturally derived QS-21, such as accessibility, chemical stability, toxicity, and production scalability.⁴³ QS21v’s modifications aimed to retain the immunostimulatory properties of QS-21 while improving its toxicity and stability and streamlining the synthesis upscaling.⁴⁴ Taken together, the notable structural differences in these adjuvant derivatives compared to their parent molecules can influence how they interact with antigens and other components within the formulation, thereby affecting the overall efficacy of the vaccine. These unique structural and immunological features might have also contributed to the lack of synergistic response in the Lipo-coForm compared to the nonliposomal adjuvant formulations.

In Vivo Immunogenicity Studies

We then tested the adjuvant activity of the combined FP20+QS21v formulation in a model vaccine incorporating the prototypical OVA antigen. Mice were immunized with suboptimal concentrations (10 μg) of both adjuvants separately and in combination. As controls, we immunized groups of mice with MPLA and no adjuvant (OVA antigen only), respectively. The immunization with FP20 provided a lower IgG response to the antigen than when MPLA was used, while the IgG titers for mice immunized with QS21v alone had no detectable response at the dilutions tested (Figure 5). Notably, the animals that received the combination of FP20 and QS21v (10 μg each) responded significantly better than any of the other groups, demonstrating synergistic activity for this adjuvant mixture. To determine whether the synergistic effect was also evident when using previously established concentrations of both adjuvants administered separately, we also performed immunizations using 50 μg of QS21v. The use of a combined FP20+QS-21v formulation using high doses of QS21v did not result in increased antibody titers compared to the use of this variant alone (Figure 5). Overall, these data show that the combination of FP20 and QS21v provides a synergistic effect at suboptimal single-use doses, thus minimizing the potential secondary effects attributed to saponins in the native AS01 system.

Figure 5.

Synergistic antibody induction of coformulations of FP20 and QS21v. (A) B6 mice (5 per group) were immunized with OVA formulated with MPLA (black), FP20 (10 μg, red), QS21v (10 μg, green), FP20-QS21v coformulations (10 + 10 μg, blue) as adjuvants, or no adjuvant (OVA alone, gray). Values represent the mean ± SEM (B) Brown–Forsythe and Welch one-way analysis of variance (ANOVA) tests (with an α of 0.05) were utilized to compare the areas under each curve. *p < 0.05; **p < 0.01; ***p < 0.001. (C) Immunization of B6 mice using 50 μg of QS21v. The rest of the doses were as in A.

Conclusions

In this study, we developed a novel AS01-inspired adjuvant system by replacing MPLA with FP20 and native QS-21 with a synthetic lead variant, QS21v. Our chemical approach addresses some of the limitations affecting the parent, original adjuvants, such as the accessibility issues associated with MPLA (involving a complex chemical synthesis) and native QS-21 (requiring a laborious extraction from the bark) as well as the chemical instability and reactogenicity of the natural product saponin. The T_m of FP20 is 43.9 °C and played an important role in the stability of the formulations. The physicochemical characterization of the formulations, including FP20, QS21v, and their combinations, revealed significant insights. The mean size distribution of the FP20-QS21v coformulations, as assessed by DLS and NTA, is approximately 120 nm. The PDI is observed to be less than 0.2, indicating a narrow size distribution among the particles. These findings suggest that the coformulations possess a homogeneous particle size profile, which is critical for ensuring consistent bioavailability and therapeutic efficacy. The low PDI values further imply minimal aggregation, enhancing the stability and performance of the formulation in various applications. The ζ-potential measurements showed highly negative surface charges for the formulations, suggesting a good colloidal stability. The cryoEM imaging further confirmed the structural characteristics of the formulations: FP20 formed spherical micellar structures, QS21v exhibited rod-like morphologies, and the FP20+QS21v coformulations showed peripheral interactions between FP20 micelles and QS21v rods. These peripheral interactions may have played a crucial role in enhancing the synergistic effects of the formulations.

The immunological response observed in vivo include significantly enhanced IgG titers with the FP20+QS21v coformulation compared to the individual components. This enhancement likely stems from the complementary physicochemical properties of the adjuvants, as FP20 primarily engages TLR4, while QS21v utilizes alternative mechanisms to promote immune activation. Specifically, QS-21 is hypothesized to interact with cell surface lectins and T-cell receptors, indicating the importance of the peripheral interactions between FP20 and QS21v. By enhancement of antigen presentation and promotion of a robust immune response, these interactions could significantly improve the efficacy of the formulation, ensuring a well-coordinated link between innate and adaptive immune pathways. Ultimately, the strategic integration of FP20 and QS21v not only stabilizes the formulation but also optimizes the immunogenic potential, which may be pivotal in the development of effective vaccines.

Experimental Section

Materials

The adjuvants used in this study, FP20 and QS21v, were obtained through multistep syntheses according to published procedures.^17,43 All reagents and solvents were purchased from commercial sources and used without further purifications, unless stated otherwise.

Phase Transition Temperature (T_m) of FP20

DSC was performed to evaluate the thermal properties of FP20, focusing particularly on T_m, to obtain insights into its stability and behavior at physiological temperatures. DSC data were recorded on a Mettler Toledo Star^e DSC1 analysis system equipped with a low temperature apparatus. The experiments were run under nitrogen atmosphere in standard 40 μL Al pans from −30 to 55 °C with heating and cooling rates of 10 °C/min. FP20 sample has been analyzed after purification and rotary evaporation at 40 °C to remove the solvent.

Preparation of Coformulations

The adjuvant coformulations were prepared by dissolving FP20 (100 μg/mL) and QS21v (100 μg/mL) in methanol. This mixture was subjected to ultrasonication at a frequency of 37 kHz at a controlled temperature of 45 °C for 2 min, ensuring uniform dispersion. Following this, the methanol was evaporated by using a rotary evaporator to remove the solvent and form a thin dry film. Subsequently, PBS at 45 °C was introduced to the dried film mixture and sonicated again for 10 min at 45 °C to form the coformulations. Similar approaches have been followed for the individual adjuvants. All of the formulations were stored at 4 °C before further characterization.

Physicochemical Characterization of Coformulations

The coformulations prepared as described were characterized to determine the particle size distribution and homogeneity of the coformulations. NTA for all formulations (FP20, QS21v, FP20+QS21v) was performed using the Nanosight NS300 (Malvern Panalytical, U.K.) equipped with an sCMOS camera and a Blue488 laser. The experiment settings were kept constant for all of the samples to ensure accurate and reproducible results. Each sample was captured with a camera level of 16 for 60 s and five captures per sample were taken with the sample diluted by a factor of 1 × 10¹. The syringe pump speed was maintained at 50 turns per minute to ensure steady flow. During analysis, the detection threshold was set at 4, and the blur size was set to auto, allowing for optimal particle detection and size determination.

The average size and PDI were analyzed by DLS (Brookhaven Instruments Corporation). We performed 10 different runs for each formulation and calculated the average diameter and standard deviation. The surface charge of the formulation was analyzed in a similar way by using a ZetaPALS electrode connected to the DLS.

For cryoEM, the Quantifoil R2/2 copper grids 300 mesh were glow discharged, and 10 μL of proline (0.1%) solution was loaded onto the grid. All of the excess proline was removed with a blot paper and the grids were left for drying. The grids were loaded into an automatic plunge freezer EM GP2 (Leica systems) with a humidity of 75% at 8 °C. Four μL of each formulation of FP20, QS21v and coformulations were loaded onto the grid and incubated for 30 s followed by a 1.5 s of blotting. The samples were immediately plunged into liquid ethane and then were transferred into liquid nitrogen. The vitrified samples were transferred into a GATAN model 914 cryoEM sample holder and analyzed by using the JEM-2200 (Joel, Japan) TEM at 200 kV accelerating voltage.

In Vivo Immunogenicity Studies

Animal protocols were approved by the Animal Research Ethics Board of CIC bioGUNE and the Competent Authority (Diputación de Bizkaia) according to the guidelines of the European Union Council (Directive 2010/63/EU) and Spanish Government regulations (RD 53/2013). The Animal Facility at CIC bioGUNE is accredited by AAALAC Intl.

The in vivo efficacy of the adjuvant coformulations was assessed using 8–10 week-old C57BL/6J (B6) mice from Charles River Laboratories (Lyon, France). The animals were maintained under 12 h light/dark cycles and provided food and water ad libitum. Mice were immunized subcutaneously with an antigen (EndoFit ovalbumin, InvivoGen) combined with the prepared adjuvant coformulations. This immunization protocol involved a primary immunization (day 0) and a secondary booster immunization after 21 days. The animals were bled by intracardiac puncture at sacrifice (day 42). MPLA was administered at a dosage of 10 μg per dose as a positive control, along with no adjuvant group as a negative control. Blood was collected into serum separator tubes (BD), centrifuged at 10,000 rpm for 5 min, and the sera were recovered and stored at −80 °C until use.

Antibody titers were measured by 2-fold serial dilutions of the sera by capture enzyme-linked immuno-sorbent assay (ELISA). Ninety-six well plates (Nunc, Thermo Fisher Scientific), were coated overnight at 4 °C with 0.5 μg/mL ovalbumin in 0.2 M sodium bicarbonate buffer, pH 9.6. The plates were washed with 0.05% Tween in PBS (PBST) and blocked for 1 h with filtered in PBST containing 1% bovine serum albumin (BSA). The sera were then incubated for 1 h, followed by extensive washing and a final 45 min incubation with peroxidase-conjugated goat antimouse IgG (1/1,000; Jackson ImmunoResearch). The plates were then washed with PBST, followed by two washes with PBS. The reactions were developed using 3,3′,5,5′-tetramethylbenzidine peroxidase substrate solution (TMB, SeraCare) for 30 min, followed by 2 M sulfuric acid (stop solution). The plates were read in a BioTek Epoch microplate reader at 450 nm.

Statistical Information

All experimental results represent the mean ± the standard error of the mean (SEM) of at least three independent experiments. For ELISA experiments, means were compared by t tests (two groups) or one-way ANOVA (three or more groups). For statistical comparisons of immunization results, the areas under the ELISA titration curves were analyzed using Brown–Forsythe and Welch one-way ANOVA tests with a significance level (α) of 0.05.

Glossary

Abbreviations Used

T_m

transition temperature

DSC

differential scanning colorimetry

DLS

dynamic light scattering

NTA

nanoparticle tracking analysis

CryoEM

cryogenic electron microscopy

AS

adjuvant systems

TLR4

toll-like receptor 4

MPLA

monophosphoryl lipid A

HPV

human papillomavirus

IFN

interferon

IL

interleukin

OVA

ovalbumin

DOPC

1,2-dioleoyl-glycero-3-phosphocholine

DMPC

1,2-dimyristoyl-glycero-3-phosphatidylcholine

LPS

lipopolysaccharide

MyD88

myeloid differentiation factor 88

TRIF

TIR domain-containing adapter inducing IFN-β

TRAM

TRIF-related adapter molecule

NF-κB

NF-κB, nuclear factor kappa-light-chain enhancer of activated B cells

ELISA

enzyme-linked immuno-sorbent assay

PBS

phosphate buffer saline

PBST

tween in PBS

TMB

3,3′,5,5′-tetramethylbenzidine peroxidase substrate

NT

nontreated

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jmedchem.4c02392.

• Summary table and figures related to the liposomal formulations and their immune responses; average diameter, polydispersity index (PDI), and ζ-potential of DMPC-based liposomal formulations containing FP20 (100 μg/mL) and QS21v (500 μg/mL) (Table S1); CryoEM images of DMPC (2.7 mM) + Cholesterol (0.3 mM) liposomes, with and without FP20 and QS21v at various concentrations (Scale: 100 nm) (Figure S1); CryoEM image showing proline coating alone, displaying distinct patches separate from liposomal structures (scale: 100 nm) (Figure S2); antibody induction results from liposomal coformulations of FP20 and QS21v in B6 mice (n = 5 per group) at 21 and 42 days postimmunization (Figure S3) (PDF)

The authors declare no competing financial interest.