Skip to main content
NCBI home page
ACS Pharmacology & Translational Science logoLink to ACS Pharmacology & Translational Science
. 2024 Sep 11;7(10):3108–3118. doi: 10.1021/acsptsci.4c00306

Digitally Enabled Generic Analytical Framework Accelerating the Pace of Liquid Chromatography Method Development for Vaccine Adjuvant Formulations

Mohamed Hemida 1, Rodell C Barrientos 1,*, Caleb Kinsey 1, Nathan Kuster 1, Mayank Bhavsar 1, Armen G Beck 1, Heather Wang 1, Andrew Singh 1, Pankaj Aggarwal 1, Arthur Arcinas 1, Malini Mukherjee 1, Emmanuel Appiah-Amponsah 1, Erik L Regalado 1,*
PMCID: PMC11480893  PMID: 39421659

Abstract

graphic file with name pt4c00306_0006.jpg

The growing use of adjuvants in the fast-paced formulation of new vaccines has created an unprecedented need for meaningful analytical assays that deliver reliable quantitative data from complex adjuvant and adjuvant–antigen mixtures. Due to their complex chemical and physical properties, method development for the separation of vaccine adjuvants is considered a highly challenging and laborious task. Reversed-phase liquid chromatography (RPLC) is among the most important tests in the (bio)pharmaceutical industry for release and stability indicating measurements including adjuvant content, identity, and purity profile. However, the time constraints of developing “on-demand” robust quantitative methods prior to each change in formulation can easily lead to sample analysis becoming a bottleneck in vaccine development. Herein, a simple and efficient generic analytical framework capable of chromatographically resolving the most commonly used non-aluminum-based adjuvants across academic and industrial sectors is introduced. This was designed to seek a more proactive approach for fast-paced assay development endeavors that evolved from extensive stationary phase screening in conjunction with multifactorial in silico simulations of adjuvant retention time (RT) as a function of gradient time, temperature, organic modifier blending, and buffer concentration. The multifactorial retention models yield 3D resolution maps with excellent baseline separation of all adjuvants in a single run, which was found to be very accurate, with differences between experimental and simulated retention times of less than 1%. The analytical framework described here also includes the introduction of a more versatile approach to method development by introducing a dynamic RT database for adjuvants covering the entire library of adjuvants with broad mechanisms of action across numerous vaccine formulations with excellent linearity, accuracy, precision, and specificity. The power of this framework was also demonstrated with numerous analytical assays that can be generated rapidly from simulations guiding vaccine processes in the development of new adjuvant formulations. Analytical assay in this work covers content, purity profile by LC with diode array detector (DAD) and charged aerosol detector (CAD), and component identification by LC with mass spectrometry (MS) across complex vaccine formulations, including the use of surfactants (e.g., polysorbates) as well as their separation from adjuvant targets.

Keywords: reversed-phase liquid chromatography, vaccine adjuvant formulations, automated screening, computer-assisted optimization, charged aerosol detector

Recent years have witnessed an accelerated development of novel vaccine products facilitated by timely scientific research and technological innovations necessary to face global humanitarian emergencies.1−3 Innovations in analytical methods and technologies have also played a crucial role in streamlining vaccine development. These methods enable scientists to efficiently analyze and understand the complexity of viruses, assess their pathogenicity, and identify potential targets for vaccine development.4−7

Characterization of antibodies, large proteins, viral vectors, virus-like particles, and other macromolecules that act as active pharmaceutical ingredients can often be challenging.8−11 However, analytical assays that quantify small molecules present in vaccine products are needed throughout initial development to commercialization.12−14 Among these, adjuvants are crucial components that enhance the immune response triggered by vaccines. Adjuvants boost vaccine effectiveness, enable the use of smaller vaccine doses, and potentially expand availability.15 Typically, vaccine adjuvant formulations encompass highly diverse chemical functionalities with quite complex structures, properties, and pharmacokinetic mechanisms.16−20

In general, non-aluminum-based adjuvant formulations are typically composed of lipids, fatty acids, and glycosides with varying degrees of complexity and physicochemical diversity. In addition to being diverse in nature, most vaccine adjuvants lack ultraviolet (UV) chromophores, requiring combinations of different detectors beyond traditional UV or fluorescence.21−23 Furthermore, these molecules have a wide range of solubility and are very sensitive to pH changes.17 Vaccine adjuvants are subjected to rigorous analytical systems for quantification, identification of degradation products, impurity profiling, and stability studies. Thus, there is a growing demand for robust and fast analytical solutions to enable adjuvant formulation, development, and process analysis. These form the basis of important quality attributes for accelerating vaccine discovery and development. New analytical methods to address these challenges are rapidly surging in the field.24−29

Gas chromatography (GC) and high-performance liquid chromatography (HPLC) are commonly used separation techniques employed for lipids and adjuvants. GC can be used to characterize volatile components, whereas appropriate sample preparation is required for nonvolatile components. For example, in the analysis of monophosphoryl lipid A (MPLA), a commonly used bacterial-derived adjuvant, fatty acyl esters are first hydrolyzed and then converted to fatty acid methyl esters prior to GC analysis.30 This method is both tedious and incapable of quantifying individual components of the mixture if prior knowledge of specific fatty acid side chains is absent.30

In contrast, HPLC hyphenated with widespread detectors is currently the method of choice for the analysis of lipids and adjuvants.31−33 Reversed-phase liquid chromatography (RPLC) remains the gold standard separation mode to resolve a wide spectrum of lipids and adjuvants based on their hydrophobicity. Mass spectrometry (MS) and evaporative light scattering detection (ELSD) are often used to analyze non-UV-absorbing adjuvants in vaccine formulations. While these detectors offer some advantages, MS is generally not the best choice for quality control (QC) laboratories under the regulation of Good Manufacturing Practices (GMP), while ELSD is known for its notoriously low detection sensitivity.

Alternatively, a charged aerosol detector (CAD) is sensitive and readily available across both early and late stages of (bio)pharmaceutical development.34,35 The use of HPLC-CAD has proved to bring great value for the determination of lipid content and stability, especially those that are not UV-absorbing and have low vapor pressure.23,33 However, there are further challenges regarding the simultaneous separation of adjuvant mixtures due to their extremely wide hydrophobicity, molecular size, and polarity range. These complex multicomponent mixtures introduce significant analytical bottlenecks, both taxing current separation capabilities and delaying the development of new adjuvant formulations.

In recent years, analytical and synthetic chemists have partnered across the pharmaceutical and other industrial sectors to facilitate the development and validation of generic or more universal methods that cover multiple compound classes in a single experimental run. This has served to maximize the speed at which chemists can generate accurate and quality data to support the development of new synthetic routes, with immediate application in regulatory settings.36 In spite of this, the vaccine space is yet to benefit from these advances and currently relies on the single use of LC assays. “On-demand” creation of new analytical methods is a very inconvenient, reactive approach in fast-paced laboratories, often representing a significant obstacle and a source of delay (Figure 1a). It is evident, therefore, that a generic analytical framework that resolves the most used vaccine adjuvants in a single experimental run is missing.

Figure 1.

Figure 1

Open in a new tab

Current and proposed new approach for the LC assay development of vaccine adjuvants.

Despite the great value that such a proactive approach might have across vaccine formulations, the development of new generic or more universal chromatographic methods capable of achieving baseline resolution of multiple adjuvant targets remains a tremendous challenge.

There are numerous approaches to generate retention models that demonstrated high impact in accelerating method development for pharmaceutical analysis. Ideally, a more proactive and versatile separation concept should evolve from the integration of advanced automated stationary phase screening in conjunction with multifactorial computer-assisted modeling of adjuvant retention properties. Such a combination could yield a generic analytical framework while also enhancing method development efficiency from a dynamic built-in database that is resilient to vaccine formulation changes (Figure 1b).

To address these challenges, we herein introduce a digitally enabled generic analytical framework for immunostimulants and delivery system-type adjuvants used in vaccine formulations. With a broad selection of commonly used adjuvants in human vaccines, a dynamic RPLC database was built from in silico simulations of adjuvant retention time (RT) as a function of gradient time, temperature, and modifier concentration using the best column outcome from comprehensive stationary phase and mobile phase screening. This database can be proactively used to create new methods in silico while bypassing resource-intensive experimentation. It is dynamic, which means that new analytes, such as novel adjuvants, excipients, or even degradants, can be added to the database on demand. To the best of the authors’ knowledge, this digitally enabled generic platform assay is the first of its kind in the vaccine adjuvant space covering the widest spectrum of adjuvants used in human vaccines. The tangible power of this concept was demonstrated with real-world vaccine adjuvant formulations encountered in industry where methods that evolved from the generic analytical framework were readjusted in silico and deployed to QC laboratories.

Experimental Section

Instrumentation

Screening experiments were performed on an Agilent 1290 Infinity II UHPLC (Agilent Technologies, Santa Clara, CA), equipped with a quaternary pump (model G4204A), autosampler (model G4226A), thermostated column compartment (model G1316C) with column selector valve (model 5067-4273), and a diode array detector (DAD) module (model G7117B) with a 600 nL flow cell (model G41212-60038). The system is equipped with valve drives (model G1170A), to connect seven additional solvents to the quaternary pump. The Corona Veo RS charged aerosol detector (CAD) (Thermo Scientific, Waltham, MA) was used for the detection. High-purity nitrogen was used at 60.0 psi; nebulizer temperature was 35 °C, data acquisition frequency was 5 Hz; and power function was 1.

Chemicals and Reagents

Trifluoroacetic acid (TFA), formic acid, ammonium acetate, ammonium formate, polysorbate 80 (PS-80), and triethylamine (TEA) were purchased from Sigma-Aldrich, Inc. (St Louis, MO). Methanol and 2-propanol (HPLC grade) were obtained from Fisher (Fair Lawn, NJ). Acetic acid (99.7%) was purchased from Acros (West Chester, PA). Optima grade acetonitrile was purchased from Fisher Scientific (Fair Lawn, NJ). Deionized water produced from a Milli-Q system (model IQ-7000) was used for mobile phase preparations. 1,2-Distearoyl-sn-glycero-3-phosphocholine (DSPC) (>99%), 2-stearoyl-sn-glycero-3-phosphocholine (2-LysoPC), 18:0 DDAB dimethyldioctadecyl ammonium, 1-stearoyl-2-hydroxy-sn-glycero-3-phosphocholine (1-LysoPC), monophosphoryl 3-deacyl lipid A (3D-PHAD), monophosphoryl lipid A (MPLA-PHAD), α-galactosyl ceramide (KRN7000), d-(+)-trehalose 6,6′-dibehenate (22:0 Trehalose), and Squalene were obtained from Avanti Polar Lipids (Alabaster, AL). 1,2-Dimyristoyl-rac-glycero-3-methylpolyoxyethylene (PEG-DMG) was purchased from NOF (White Plains, NY). Cholesterol (99%) was purchased from Minakem (Dunkerque, FR). Quillaja Saponaria Fraction #21 (QS-21) was purchased from Desert King International (San Diego, CA).

Column and Mobile Phase Screening

Nine RPLC columns were used in the screening study (Table 1). All columns were used according to their manufacturer’s recommendations. Different mobile phases were also screened (Table 2). Automated screening was enabled by column and mobile phase selector valves. All columns were operated with the same column temperature, flow rate, and generic gradient: 0–0.5 min: 5.0% B, 0.5–30.0 min: 5.0–90% B, 0.5 mL/min, and 45 °C column temperature.

Table 1. UHPLC Columns Used in This Study.

  column dimension (I.D.) (mm) × length (mm) particle size (μm) stationary phase
1 Acquity CSH Phenyl Hexyl 2.1 × 50 1.7 phenyl hexyl
2 ZORBAX Eclipse Plus 2.1 × 50 1.8 C18
3 Acquity UPLC BEH 2.1 × 50 1.7 C18
4 InfinityLab Poroshell 120 2.1 × 50 1.9 phenyl hexyl
5 Acquity UPLC BEH 2.1 × 50 1.7 C8
6 Acquity UPLC HSS T3 2.1 × 50 1.8 C18
7 Acquity UPLC Protein BEH 2.1 × 50 1.7 C4
8 SB-CN RRHD 2.1 × 50 1.8 CN
9 Hypersil GOLD PFP 2.1 × 50 1.8 PFP
Open in a new tab

Table 2. Mobile Phases Used in This Study.

  mobile phase
A1 0.1% trifluoroacetic acid (TFA)
A2 0.1% acetic acid
A3 1% acetic acid and 0.08% of triethylamine (TEA, pH = 3.5)
A4 5 mM ammonium formate
A5 5 mM ammonium acetate
A6 0.1% formic acid (FA)
B1 acetonitrile (ACN)
B2 methanol (MeOH)
B3 isopropanol (IPA)
Open in a new tab

Experimental Conditions Used for Modeling

A Waters Acquity Phenyl hexyl column was chosen for in silico simulation. A 33 experimental design was employed with the column temperature (T), gradient steepness (tG), and solvent ratio of IPA to ACN in mobile phase B as factors. Each of these factors had three levels (T = 40, 50, 60 °C; tG = 15, 30, 45 min; solvent ratio = 0:100, 10:90, 20:80 of ACN to IPA (v:v) as mobile phase B). The resulting experimental data were processed using the ACD Laboratories/LC Simulator 2021.1.2 release (Version L35R41), Advanced Chemistry Development, Inc. (ACD), Toronto, Ontario, Canada (further referred to as an LC Simulator). A resolution map was obtained from LC Simulator software, indicating the optimum separation conditions listed in Figure 3. The listed references describe how LC Simulator software works.8,37−39 Adjuvant mixtures were dissolved in the relevant solvent and then mixed and diluted before LC analysis.

Figure 3.

Figure 3

Open in a new tab

Multifactorial modeling to create a generic analytical framework for lipids and vaccine adjuvants. (a) Three-dimensional resolution map generated from multifactorial optimization of gradient time (tG), column temperature (T), and solvent ratio. (b) Overlay of experimental and predicted chromatograms using conditions obtained from the model. Conditions: Phenyl hexyl column (1.7 μm, 4.6 mm × 50 mm). Mobile phase (A) 1% acetic acid and 0.08% triethylamine buffer (aq); (B) 80:20 IPA: ACN; flow rate: 0.5 mL/min; Column temperature: 50 °C. Peak assignment: (1) QS-21, (2) 18:0-LysoPC (3) PEG-DMG (4) Cholesterol, (5) DDAB, (6) DOPC, (7) DSPC, (8) Squalene, (9) KRN7000, (10) 22:0 Trehalose, (11) 3D-PHAD, (12) MPLA/PHAD.

Mass Spectrometry

A Waters Xevo G2-XS Quadrupole Time-of-Flight (QToF) mass spectrometer (Waters Corp., Milford, MA) coupled with a binary ultraperformance liquid chromatography (UPLC) system was used for mass spectrometry measurements. Source conditions were as follows: Source mode: Electrospray ionization; polarity, positive and negative, mass range, 100 to 2500m/z; Capillary voltage, 3.00 kV; Sampling cone, 40; Source offset, 80; Source temperature, 100 °C; Desolvation temperature, 250 °C; Cone gas, 50 L/h; Desolvation gas, 400 L/h.

Method Validation

The ICH quality guidelines for validation were followed for the validation of an assay for a four-component adjuvant formulation, as described in the Resultssection. Attributes evaluated include selectivity and specificity, sensitivity, linearity, precision, and accuracy of the method.

Data Analysis

All data acquisition and instrument control were performed in a ChemStation C.01.07 SR3 (Agilent Technologies, Santa Clara, CA). MassLynx V4.2 (Waters Corp., Milford, MA) was used to acquire and process the MS data. Multifactorial modeling of HPLC conditions was conducted in ACD/LC Simulator Software (ver. L35R41) from Advanced Chemistry Development, Inc. (ACD), Toronto, Ontario, Canada.

Results and Discussion

The overarching goal of this work was to create a generic analytical framework that can serve as a chromatography database of vaccine adjuvants for fast-paced method development and deployment of reliable assays. It involved the selection of a column and mobile phase combination that is suitable to resolve and elute all of the analytes. After column and mobile phase selection, a generic analytical framework with the embedded multifactorial model was built using an LC simulator. To enable a straightforward deployment in vaccine R&D settings, this platform was used with standard readily available LC coupled to DAD and CAD detection systems.

This digitally enabled generic analytical framework can serve as a dynamic method database that can be queried rapidly to create an unlimited number of assays for new vaccine formulations without the need to undergo tedious screening and optimization experiments (Figure 1b).

Column and Mobile Phase Selection

Column and mobile phase selection remains a major bottleneck in developing methods for diverse pharmaceuticals.36,40 Lipids and adjuvants used in vaccine formulations are highly complex not only in terms of their mechanisms of action but also in terms of hydrophobicity, molecular weight, stability, and solubility, thus posing significant analytical challenges. As such, in creating a generic analytical framework, it is important to identify the most suitable column and mobile phase at the outset.

In this work, automated UHPLC screening with multiple mobile phases and columns was employed to identify the column and mobile phase for downstream optimization.41 A 2-position, 6-port valve was used to enable selection between nine commercial columns and nine mobile phases. The RPLC columns used in the screening exploration were carefully selected to cover a broad range of the most common and commercially available stationary phases to comprehensively evaluate chromatographic resolution and maximize selectivity difference for better study outcomes (Table 1). The commercial columns were selected with similar dimensions spanning diverse state-of-the-art column chemistries.

Six buffers were used as aqueous mobile phases for RPLC screening (Table 2). These buffers, namely, ammonium formate, ammonium acetate, trifluoroacetic acid, formic acid, and triethylamine, were chosen because of their widespread use in adjuvant separations and compatibility with the CAD detectors.42

A total of 12 common lipids and adjuvants were screened (Table 3). These were chosen because of their widespread use in vaccines in both commercial and clinical phase development. Some compounds are endogenous in humans (e.g., cholesterol, DOPC, DSPC, squalene, etc.), but were included here because they are commonly used to formulate adjuvants in vaccine formulations.18 These compounds have a wide spectrum of hydrophobicity, molecular weight, and complex chemical properties, making them extremely challenging to measure chromatographically without a systematic and holistic method development approach (Supporting Information Table S1).

Table 3. Lipids and Adjuvants Used in This Study.

no. Name abbreviation
1 Quillaja saponaria fraction #21 QS-21
2 2-stearoyl-sn-glycero-3-phosphocholine 18:0-LysoPC
3 1,2-dimyristoyl-rac-glycero-3-methylpolyoxyethylene PEG-DMG
4 cholesterol  
5 dimethyldioctadecyl ammonium DDAB
6 1,2-dioleoyl-sn-glycero-3-phosphocholine DOPC
7 1,2-distearoyl-sn-glycero-3-phosphocholine DSPC
8 squalene  
9 α-galactosyl ceramide KRN7000
10 d-(+)-trehalose 6,6′-dibehenate 22:0 Trehalose
11 monophosphoryl 3-deacyl lipid A 3D-PHAD
12 monophosphoryl lipid A MPLA/PHAD
Open in a new tab

The screening involved a total of 162 permutations, resulting in data-rich results that require prompt interpretation. To streamline the decision-making process, an in-house data visualization dashboard was made to quickly identify which column provided the greatest number of peaks separated and best peak shape (Supporting Information Figure S1). These results indicated that the phenyl hexyl column delivered the best chromatographic performance for all 12 components that conveniently elute in the middle of the separation window by using a standard screening gradient (Figure 2). This is in contrast to C18, the most commonly used stationary phase for lipids and adjuvants,23,43 that required a longer gradient slope at high organic composition to elute highly retained analytes (9–12). An overlay of chromatograms obtained using C18 and phenyl hexyl is shown in Figure 2.

Figure 2.

Figure 2

Open in a new tab

Chromatogram (UV 210 nm and CAD traces) obtained from method screening using generic gradient conditions. (a) C18 column (1.7 μm, 4.6 mm × 50 mm I.D.) and (b) phenyl hexyl column (1.7 μm, 4.6 mm × 50 mm). Mobile phase (A) 1% acetic acid and 0.08% of triethylamine buffer (aq); (B) IPA; flow rate: 0.5 mL/min; Column temperature: 40 °C. Peak assignment: (1) QS-21, (2) 18:0-LysoPC (3) PEG-DMG (4) Cholesterol, (5) DDAB, (6) DOPC, (7) 18:0 PC (DSPC), (8) Squalene, (9) KRN7000, (10) 22:0 Trehalose, (11) 3D-PHAD, (12) MPLA/PHAD.

The results indicate that a C18 column can also be used for this work at the expense of lengthy gradient separation to ensure complete elution of hydrophobic species such as MPLA and 22:0 trehalose. It is also noteworthy that using a C18 column, some of the adjuvants could coelute that are otherwise resolved in the phenyl hexyl column, for example, compounds 10 and 11. This suggests that phenyl hexyl offers additional selectivity to the hexyl phase potentially via pi-pi interactions induced by the phenyl ring. Phenyl hexyl columns have long alkyl spacers which enhance its steric and aromatic selectivity; thus, the column exhibits both hydrophobic and pi-pi interaction retention. Some of the adjuvants used contain pi bonds in the form of carbon–carbon double bonds or carbon–oxygen double bonds, which may encourage pi-pi interactions yielding unique selectivity not observed in typical C18 column chemistry. The extremely hydrophobic nature of a C18 stationary phase makes it a challenging column to use to build a generic analytical framework for lipids and adjuvants encountered in vaccine process development. The results also demonstrate that the use of a universal detector, CAD, is crucial as most of the lipids and adjuvants are non-UV-absorbing (Figure 2).

The screening further revealed that different mobile phase combinations can be used to construct a generic analytical framework. Previous studies have demonstrated the use of HPLC mobile phase additives such as formic acid,22,43 ammonium acetate,44 ammonium formate,45 TFA,46,47 acetic acid,48 and TEA,23 for lipids and adjuvant-related analytes. In this study, among the buffers tested (Table 2), 0.1% TFA, 1% acetic acid with 0.08% TEA, and 0.1% acetic acid delivered acceptable elution profiles, peak shape, and resolution for the greatest number of adjuvant peaks in the mixture. Moving forward, the buffer containing 1% acetic acid with 0.08% TEA was selected for further optimization. The TEA was selected due to its volatility which makes it a good choice for CAD detection.23,42 It also improves the peak shapes of cationic lipids possessing amine groups that may protonate under acidic conditions.23 TEA improves the peak shape of cationic lipids by preventing secondary interactions of the amine functional groups and silica-based stationary phases.23 It is noteworthy that cationic lipids are becoming increasingly relevant in liposomal formulations for vaccines.49

Creation of a Generic Platform Assay for Vaccine Adjuvants

The creation of a generic analytical framework involves the use of computer-assisted optimization to map the separation landscape of the target adjuvants. This has been a widely used tool in developing assays that are deployed across both academic and industrial sectors.50 Next, using the selected column and mobile phase, a suitable experimental design was conceptualized. Column temperature (T), gradient steepness (tG), and solvent composition are the key parameters in an HPLC method that may influence the retention and resolution of a sample mixture. Using LC Simulator software, a multifactorial model was built to describe the retention behavior of the lipids and adjuvants. A total of 27 experiments (3 × 3 × 3 permutations) were used to build a 3D multifactorial model. The model consisted of three factors (T, tG, solvent ratio) each of which had three levels (T = 40, 50, 60 °C; tG = 15, 30, 45 min; solvent ratio = 0:100, 10:90, 20:80 of ACN to IPA (v:v)). The solvent ratio here represents the percent (volume %) of ACN in the IPA solvent used as mobile phase B. The inclusion of the solvent ratio as a factor enables the versatility to optimize solvent elution strength, something that is greatly needed to tune the elution of highly retained species.

Using experimental injections, an empirical model was constructed. Figure 3a shows a snapshot of the 3D resolution map generated by the LC Simulator. The resolution map is a heatmap representation of the resolution values of critical pairs for a given chromatogram as a function of the three parameters (T, tG, solvent ratio). The model is plotted as a spectrum of colors from blue to red to visualize the areas with the lowest and highest resolution, respectively. The desired assay condition should lie within the orange-red region of the resolution map, where peaks of interest are separated. It also shows a correlation coefficient of >0.99, suggesting an excellent fit between the experimental and predicted retention time values. This model can be used within the bounds of the experimental limits used to construct the model (T = 40, 50, 60 °C; tG = 15, 30, 45 min; solvent ratio = 0:100, 10:90, 20:80 of ACN to IPA (v:v)). This model was tested by generating methods to separate the 12 components and experimentally verifying the retention time. Results showed that the predicted and experimental retention times matched with a highly desirable <1% retention time difference (Figure 3b). This demonstrates the immense potential for the utilization of the multicolumn and multiple mobile phase screening system with software-based method development as a framework for the streamlined RPLC analysis of adjuvants in vaccine formulation development.

Evolution of Methods from the Generic Platform Assay

The goal of this work is to construct a digitally enabled generic analytical framework that can serve as a database of common lipids and adjuvants, allowing the fast-paced development of methods to circumvent the tedious lab-bench screening and optimization. Using this framework, new methods can be developed rapidly for any given formulation of lipids and adjuvants that are in the model. To demonstrate this, a method for a four-component adjuvant mixture (formulation X) was developed using the in silico generic framework (Figure 4a). The components in this adjuvant mixture are cholesterol (4), DOPC (6), 3D-PHAD (11), and MPLA (12), which are commercially available and used in similar ASO1b and ALFQ formulations.17,20,51 This final method uses a mobile phase A of 1% acetic acid with 0.08% TEA and a mobile phase B blend of ACN and IPA at a solvent ratio of 90:10 (ACN: IPA). The column temperature of 45 °C and a runtime of 15 min delivered baseline separation for this mixture with resolution to spare. Compared with the generic assay that is 30 min long for all 12 adjuvants, the method for these four components was tailored to 15 min. Because column temperature and solvent ratio are incorporated into the model, optimization of these parameters can also be performed in silico, thus significantly easing the experimental burden of optimizing them. Finally, this in silico-derived method was tested experimentally, and it was confirmed that simulated and actual retention times were in close agreement (Figure 4a). This showcases the versatility and value proposition that this generic analytical framework brings to the field in creating shorter RPLC assays by adjusting the conditions tailored to the analytes of interest while bypassing the lengthy and laborious experimentation.

Figure 4.

Figure 4

Open in a new tab

Development of methods leveraging the generic analytical framework. (a) 3D resolution map and chromatogram of the optimized method for formulation X evolved from the generic analytical framework. (b) 3D resolution map and chromatogram of the optimized method for formulation X modified by the addition of PS-80 (formulation Y). Conditions: Phenyl hexyl column (1.7 μm, 4.6 mm × 50 mm). Mobile phase (A) 1% acetic acid and 0.08% of triethylamine buffer (aq); (B) 80:20 IPA: ACN; flow rate: 0.5 mL/min; Column temperature: 40 °C. Peak assignments: (4) Cholesterol, (6) DOPC, and (12) MPLA/PHAD.

Vaccine development is a highly dynamic process, where formulations could change drastically at a fast pace. This sudden change may include the addition or removal of active components or inactive components (excipients) in the existing formulation. In instances like this, analytical laboratories must respond promptly to develop and deploy assays for a new formulation. Developing analytical methods at this stage can be rate-limiting in the clinical path of the new formulation. Failure to develop and deploy methods swiftly can cause a significant delay in the development of a vaccine candidate.

As an example, when the four-component formulation X described above (Cholesterol, DOPC, 3D-PHAD, and MPLA) underwent a sudden modification of the composition by adding a surfactant (PS-80) to yield formulation Y (Cholesterol, DOPC, 3D-PHAD, MPLA, and PS-80), the method previously developed was no longer suitable due to the interference of the PS-80 peaks with the target adjuvant components. PS-80 is a well-known surfactant added to vaccines and biopharmaceuticals.52 It is a polydisperse molecule composed of fatty acid esters of polyoxyethylene sorbitan that appears as multiple peaks in RPLC, which could interfere with the target analytes.53 Because of the addition of PS-80 to formulation X, a new method would be required to ensure that the lipids and adjuvant peaks are free from interference from PS-80.

Developing a method from scratch is a laborious and time-consuming process. In this work, the generic analytical framework described above was leveraged to come up with a new method for the formulation containing PS-80. Since the original model did not include PS-80, experimental injections using the same aforementioned 3 × 3 × 3 modeling conditions were performed to acquire retention time values for the PS-80 peaks. RPLC analysis of PS-80 showed five main peaks, which were then incorporated into the model labeled as peaks A to E (Figure 4b). This new multifactorial model was subsequently used to obtain the optimum method parameters for the formulation Y. The new assay operates at 40 °C, a runtime of 15 min, and a mobile phase B composition of 80:20 (ACN: IPA). The difference between predicted and experimental retention times obtained using the new model was <1% illustrating the power of modern in silico separation technologies.

The cases presented here demonstrate the flexibility of the model by allowing the addition of new components, which result in the creation of a digital database of analytes. The same approach can be used to develop assays for novel formulations composed of different lipids, adjuvants, excipients, and even degradation products of each of the active components. In theory, this approach can yield an unlimited number of methods within minutes for any anticipated formulation change while circumventing laborious bench work and ensuring a sustainable analytical method lifecycle.

Application of the Workflow to Develop MS-Compatible Assays

One caveat of using TEA as an additive is its potential to cause persistent memory effect in mass spectrometers which makes this a less desirable MS-compatible buffer for small-molecule analysis.54 While the mobile phase described above uses a non-MS-compatible buffer additive, the same experimental workflow can be used as a blueprint to construct a generic analytical framework for MS-compatible buffers. To illustrate, 10 mM ammonium acetate (aq) and 10 mM ammonium acetate in ACN:IPA blends were used as mobile phases A and B, respectively, to develop a model (Figure 5a). To demonstrate the use of this new model, a method was developed to quantify the lipid components of an adjuvant formulation containing DOPC, Cholesterol, 3D-PHAD, and MPLA. These adjuvant components of the formulation are outlined in order of increasing hydrophobicity, such as Cholesterol (4), DOPC (6), 3D-PHAD (11), and MPLA (12) (Figure 5a). Using this method hyphenated with MS, the identity of the peaks was assigned via electrospray ionization. MPLA as [M – H]− (m/z 1744.28), 3D-PHAD was detected as [M – H]− (m/z 1518.08), DOPC as [M + H]+ (m/z 786.61), and Cholesterol as [M-H2O+H+] (m/z 369.35) (Figure 5b–e).

Figure 5.

Figure 5

Open in a new tab

Development of an MS-compatible method using the generic analytical framework. (a) Resolution map and resulting chromatogram of the MS-compatible method to separate Cholesterol, DOPC, 3D-PHAD, and MPLA. (b) Mass spectrum of the peak (4). (c) Mass spectrum of peak (6). (d) Mass spectrum of peak (11). (e) Mass spectrum of peak (12). Conditions: Phenyl hexyl column (2.7 μm, 3.0 mm × 50 mm I.D.). Mobile phase (A) 10 mM ammonium acetate; Mobile phase (B), 10 mM ammonium acetate in 90:10 (ACN: IPA); flow rate: 0.6 mL/min; Column temperature: 40 °C. Peak assignment: (4) Cholesterol, (6) DOPC, (12) 3D-PHAD and (12) MPLA/PHAD.

Method Validation and Application to Vaccine Formulation Development

Method validation is a crucial requirement prior to the assay deployment in highly regulated pharmaceutical environments. Herein, we demonstrated that methods derived from the proposed generic model can pass the rigorous validation criteria and can be seamlessly deployed as routine adjuvant assays. Important method attributes such as accuracy, linearity, dynamic range, repeatability (intra-assay precision), interassay precision, and specificity were all assessed for the MS-compatible UHPLC method described above and its application to vaccine adjuvant formulations. The phase-appropriate method partial validation results are summarized in Table 4. The accuracy of this method was evaluated by the standard addition of each component in the formulation. Interassay precision was determined by evaluating the same sample between three separate runs on three different days. The interassay precision results for all lipids were between 1% and 3% (n = 6), while the intra-assay precision of 11 injections was less than 1% for all lipid adjuvant components (n = 11). It is worth noting that the CAD response is typically nonlinear, and a quadratic function is often the best fit to build calibration models. Method linearity was assessed using a quadratic fitting between three independent injections within the concentration ranges (0.002–0.400 mg/mL). The calibration model provided acceptable linearity with a correlation coefficient (R2) range of 0.9894–0.9998 for each quadratic fit for all of the components. The method specificity evaluation revealed that the peak area of the negative control injections at the retention time of each component was <5% of the peak area of the lowest calibration standard.

Table 4. Method Partial Validation Results for an Assay to Quantify a Four-Component Adjuvant Formulation.

attribute MPLA 3D-PHAD DOPC cholesterol
regression coefficient (R2) 0.9940 0.9894 0.9998 0.9996
limit of quantification (LOQ) as mg/mL 0.0015 0.0015 0.0303 0.0076
dynamic range (mg/mL) 0.002–0.020 0.002–0.020 0.033–0.400 0.008–0.100
accuracy (%) 100–112 100–112 103–109 101–103
interassay precision (RSD, n = 6) 3% n/a 1% 3%
intra-assay precision (RSD, n = 11) 1% n/a 1% 1%
specificity Peak area of the negative control was <5% of the lowest calibration standard
Open in a new tab

This method was successfully applied to determine the concentration of DOPC and cholesterol in two batches of a stand-alone adjuvant formulation Z (similar to ASO1b of GSK), prior to any coformulation with antigen. Partial results of the quantification are reported here. The target concentrations of cholesterol and DOPC were 0.50 mg/mL and 2.000 mg/mL, respectively. In Batch 1, the concentrations measured were 2.070 mg/mL (DOPC) and 0.50 mg/mL (cholesterol). In Batch 2, the concentrations measured were 1.900 mg/mL (DOPC) and 0.46 mg/mL (cholesterol). The precision (RSD, n = 3) was 1.1% and 3.3% for cholesterol and DOPC, respectively. This demonstrates that methods evolved from a digitally enabled generic analytical framework can be deployed to quality control (QC) laboratories as reliable assays for batch release. This assay can also be used as a stability indicating method, enabling various aspects of vaccine development.

Conclusions

Vaccine adjuvant development is a highly dynamic process, where changes in the sample formulation can happen swiftly. This often results in a need to adjust the current method or develop a completely new one. The focus of this work was to introduce a new analytical framework into the vaccine space, moving toward a paradigm shift from the traditional reactive “on-demand” assay development approach to a proactive, holistic, and sustainable way of deployment. Specifically, this work advocates for the creation of a generic analytical framework at the outset, serving as a digital library of formulation components to facilitate assay development across new vaccine processes that is resilient to adjuvant formulation changes. This framework embraces sustainability in the analytical method life cycle by eliminating the need to perform a laborious and time-consuming column and mobile phase screening when sudden changes in the formulation take place. To this end, an automated column and mobile phase screening in conjunction with multifactorial modeling using ACDLabsLC Simulator was leveraged herein to yield a generic platform assay. This approach, relying on a dynamic built-in database of retention profiles for adjuvants, excipients, or even degradants, can be leveraged for tailored RPLC-UV-CAD/MS assays that are resilient to fast-paced formulation changes. The potential of this framework was demonstrated using real-world vaccine adjuvant formulations encountered in industry where methods that evolved from the generic analytical framework were readjusted in silico and deployed to QC laboratories. The framework described here can be used not only for adjuvants but also in a wide plethora of scenarios and modalities, where method adjustment is needed in response to sudden changes in sample composition. This work is a valuable contribution to the growing role of digitalization in sustainably developing and deploying new analytical assays across academic and industrial sectors. In summary, this model captures the retention behavior of the most used adjuvants in a phenyl hexyl stationary phase with TEA and acetic acid modifiers. As this field evolves, there is an opportunity to expand this model to cover all known degradants of each lipid and adjuvant, excipient, and even antigen for future work.

Acknowledgments

The authors thank the MRL Postdoctoral Research Fellow Program (Mohamed Hemida) for sponsoring this research. This work was fully funded by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ. The authors extend their gratitude to the vaccine drug product development team in Merck & Co., Inc., West Point, PA, for the adjuvant samples.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsptsci.4c00306.

  • Chemical structures, heatmap illustrates the automated screening, list of columns and mobile phases, and chromatogram for the separation of PS-80 (PDF)

The authors declare no competing financial interest.

Supplementary Material

pt4c00306_si_001.pdf (312KB, pdf)

References

  1. Cernuschi T.; Malvolti S.; Hall S.; Debruyne L.; Bak Pedersen H.; Rees H.; Cooke E. The quest for more effective vaccine markets - Opportunities, challenges, and what has changed with the SARS-CoV-2 pandemic. Vaccine 2022, 10.1016/j.vaccine.2022.07.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Barbier A. J.; Jiang A. Y.; Zhang P.; Wooster R.; Anderson D. G. The clinical progress of mRNA vaccines and immunotherapies. Nat. Biotechnol. 2022, 40 (6), 840–854. 10.1038/s41587-022-01294-2. [DOI] [PubMed] [Google Scholar]
  3. Lin M. J.; Svensson-Arvelund J.; Lubitz G. S.; Marabelle A.; Melero I.; Brown B. D.; Brody J. D. Cancer vaccines: the next immunotherapy frontier. Nat. Cancer 2022, 3 (8), 911–926. 10.1038/s43018-022-00418-6. [DOI] [PubMed] [Google Scholar]
  4. Maharjan R.; Lee J. C.; Lee K.; Han H.-K.; Kim K. H.; Jeong S. H. Recent trends and perspectives of artificial intelligence-based machine learning from discovery to manufacturing in biopharmaceutical industry. J. Pharm. Investig. 2023, 53 (6), 803–826. 10.1007/s40005-023-00637-8. [DOI] [Google Scholar]
  5. Malburet C.; Leclercq L.; Cotte J.-F.; Thiebaud J.; Bazin E.; Garinot M.; Cottet H. Size and Charge Characterization of Lipid Nanoparticles for mRNA Vaccines. Anal. Chem. 2022, 94 (11), 4677–4685. 10.1021/acs.analchem.1c04778. [DOI] [PubMed] [Google Scholar]
  6. Vanhinsbergh C. J.; Criscuolo A.; Sutton J. N.; Murphy K.; Williamson A. J. K.; Cook K.; Dickman M. J. Characterization and Sequence Mapping of Large RNA and mRNA Therapeutics Using Mass Spectrometry. Anal. Chem. 2022, 94 (20), 7339–7349. 10.1021/acs.analchem.2c00765. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Markova N.; Cairns S.; Jankevics-Jones H.; Kaszuba M.; Caputo F.; Parot J. Biophysical Characterization of Viral and Lipid-Based Vectors for Vaccines and Therapeutics with Light Scattering and Calorimetric Techniques. Vaccines 2022, 10 (1), 49. 10.3390/vaccines10010049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Haidar Ahmad I. A.; Kiffer A.; Barrientos R. C.; Losacco G. L.; Singh A.; Shchurik V.; Wang H.; Mangion I.; Regalado E. L. In Silico Method Development of Achiral and Chiral Tandem Column Reversed-phase Liquid Chromatography for Multicomponent Pharmaceutical Mixtures. Anal. Chem. 2022, 94, 4065. 10.1021/acs.analchem.1c05551. [DOI] [PubMed] [Google Scholar]
  9. Wu G.; Yu C. F.; Wang W. B.; Zhang R. J.; Li M.; Wang L. A platform method for charge heterogeneity characterization of fusion proteins by icIEF. Anal. Biochem. 2022, 638, 8. 10.1016/j.ab.2021.114505. [DOI] [PubMed] [Google Scholar]
  10. Lorbetskie B.; White T.; Creskey M.; Zhang X.; Girard M.; Tam R. Y.; Sauvé S.; Lu H. X. Selective reversed-phase high-performance liquid chromatography method for the determination of intact SARS-CoV-2 spike protein. J. Chrom. A 2022, 1680, 6. 10.1016/j.chroma.2022.463424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Dunn Z. D.; Bohman P.; Quinteros A.; Sauerborn B.; Milman F.; Patel M.; Kargupta R.; Wu S. Y.; Hornshaw M.; Barrientos R.; et al. Automated Online-Sampling Multidimensional Liquid Chromatography with Feedback-Control Capability as a Framework for Real-Time Monitoring of mAb Critical Quality Attributes in Multiple Bioreactors. Anal. Chem. 2023, 95 (49), 18130–18138. 10.1021/acs.analchem.3c03528. [DOI] [PubMed] [Google Scholar]
  12. Bennett R.; Cohen R. D.; Wang H.; Pereira T.; Haverick M. A.; Loughney J. W.; Barbacci D. C.; Pristatsky P.; Bowman A. M.; Losacco G. L.; et al. A selective plate-based assay for trace EDTA analysis via boron trifluoride-methanol derivatization UHPLC-QqQ-MS/MS enabling biologic and vaccine processes. Anal. Chem. 2022, 94 (3), 1678–1685. 10.1021/acs.analchem.1c04224. [DOI] [PubMed] [Google Scholar]
  13. Kilgore R.; Minzoni A.; Shastry S.; Smith W.; Barbieri E.; Wu Y. X.; Lebarre J. P.; Chu W. N.; O’Brien J.; Menegatti S. The downstream bioprocess toolbox for therapeutic viral vectors. J. Chrom. A 2023, 1709, 24. 10.1016/j.chroma.2023.464337. [DOI] [PubMed] [Google Scholar]
  14. Invernizzi L.; Moyo P.; Cassel J.; Isaacs F. J.; Salvino J. M.; Montaner L. J.; Tietjen I.; Maharaj V. Use of hyphenated analytical techniques to identify the bioactive constituents of Gunnera perpensa L., a South African medicinal plant, which potently inhibit SARS-CoV-2 spike glycoprotein-host ACE2 binding. Anal. Bioanal. Chem. 2022, 414 (13), 3971–3985. 10.1007/s00216-022-04041-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Mbow M. L.; De Gregorio E.; Valiante N. M.; Rappuoli R. New adjuvants for human vaccines. Curr. Opin. Immunol. 2010, 22 (3), 411–416. 10.1016/j.coi.2010.04.004. [DOI] [PubMed] [Google Scholar]
  16. Lykins W. R.; Fox C. B. Practical Considerations for Next-Generation Adjuvant Development and Translation. Pharmaceutics 2023, 15 (7), 18. 10.3390/pharmaceutics15071850. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Facciolà A.; Visalli G.; Laganà A.; Di Pietro A. An Overview of Vaccine Adjuvants: Current Evidence and Future Perspectives. Vaccines 2022, 10 (5), 26. 10.3390/vaccines10050819. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Perrie Y.; Crofts F.; Devitt A.; Griffiths H. R.; Kastner E.; Nadella V. Designing liposomal adjuvants for the next generation of vaccines. Adv. Drug Delivery Rev. 2016, 99, 85–96. 10.1016/j.addr.2015.11.005. [DOI] [PubMed] [Google Scholar]
  19. Loos C.; Coccia M.; Didierlaurent A. M.; Essaghir A.; Fallon J. K.; Lauffenburger D.; Luedemann C.; Michell A.; van der Most R.; Zhu A. L.; et al. Systems serology-based comparison of antibody effector functions induced by adjuvanted vaccines to guide vaccine design. NPJ. Vaccines 2023, 8 (1), 34. 10.1038/s41541-023-00613-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Alving C. R.; Peachman K. K.; Matyas G. R.; Rao M.; Beck Z. Army Liposome Formulation (ALF) family of vaccine adjuvants. Expert Rev. Vaccines 2020, 19 (3), 279–292. 10.1080/14760584.2020.1745636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Thomas D.; Acworth I. N.; Bailey B.; Plante M.. Direct Analysis of Multicomponent Adjuvants by HPLC with Charged Aerosol Detection, Thermo Fisher Scientific; 2016https://assets.thermofisher.com/TFS-Assets/CMD/posters/PN-70333-Multicomponent-Adjuvants-PN70333-EN.pdf (accessed 18 March, 2024).
  22. Abucayon E. G.; Barrientos R. C.; Torres O. B.; Sweeney S.; Whalen C.; Matyas G. R. A Liquid Chromatography High-Resolution Tandem Mass Spectrometry Method to Quantify QS-21 Adjuvant and Its Degradation Products in Liposomal Drug Formulations. ACS Omega 2023, 8 (23), 21016–21025. 10.1021/acsomega.3c01877. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kinsey C.; Lu T.; Deiss A.; Vuolo K.; Klein L.; Rustandi R. R.; Loughney J. W. Determination of lipid content and stability in lipid nanoparticles using ultra high-performance liquid chromatography in combination with a Corona Charged Aerosol Detector. Electrophoresis 2022, 43 (9–10), 1091–1100. 10.1002/elps.202100244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kashyap D.; Panda M.; Baral B.; Varshney N.; Sajitha R.; Bhandari V.; Parmar H. S.; Prasad A.; Jha H. C. Outer Membrane Vesicles: An Emerging Vaccine Platform. Vaccines 2022, 10 (10), 21. 10.3390/vaccines10101578. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Tsukidate T.; Hespen C. W.; Hang H. C. Small molecule modulators of immune pattern recognition receptors. RSC Chem. Biol. 2023, 4 (12), 1014–1036. 10.1039/D3CB00096F. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Chountoulesi M.; Pispas S.; Tseti I. K.; Demetzos C. Lyotropic Liquid Crystalline Nanostructures as Drug Delivery Systems and Vaccine Platforms. Pharmaceuticals 2022, 15 (4), 32. 10.3390/ph15040429. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Ariawan D.; van Eersel J.; Martin A. D.; Ke Y. D.; Ittner L. M. Recent progress in synthetic self-adjuvanting vaccine development. Biomater. Sci. 2022, 10 (15), 4037–4057. 10.1039/D2BM00061J. [DOI] [PubMed] [Google Scholar]
  28. Sun Q. S.; Liu X. L.; Li X. L. Peptidoglycan-based immunomodulation. Appl. Microbiol. Biotechnol. 2022, 106 (3), 981–993. 10.1007/s00253-022-11795-4. [DOI] [PubMed] [Google Scholar]
  29. Brai A.; Poggialini F.; Pasqualini C.; Trivisani C. I.; Vagaggini C.; Dreassi E. Progress towards Adjuvant Development: Focus on Antiviral Therapy. Int. J. Mol. Sci. 2023, 24 (11), 29. 10.3390/ijms24119225. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Masood M. A.; Blonder J.; Veenstra T. D. Quantitation of Monophosphorylated Lipid A in the Oil-in-Water Adjuvant Delivery Systems Using Transesterification and GC-MS. J. Pharm. Sci. 2017, 106 (7), 1760–1763. 10.1016/j.xphs.2017.02.034. [DOI] [PubMed] [Google Scholar]
  31. Sommer U.; Herscovitz H.; Welty F. K.; Costello C. E. LC-MS-based method for the qualitative and quantitative analysis of complex lipid mixtures. J. Lipid Res. 2006, 47 (4), 804–814. 10.1194/jlr.M500506-JLR200. [DOI] [PubMed] [Google Scholar]
  32. McLaren D. G.; Miller P. L.; Lassman M. E.; Castro-Perez J. M.; Hubbard B. K.; Roddy T. P. An ultraperformance liquid chromatography method for the normal-phase separation of lipids. Anal. Biochem. 2011, 414 (2), 266–272. 10.1016/j.ab.2011.03.009. [DOI] [PubMed] [Google Scholar]
  33. Kim K. H.; Lee J. E.; Lee J. C.; Maharjan R.; Oh H.; Lee K.; Kim N. A.; Jeong S. H. Optimization of HPLCCAD method for simultaneous analysis of different lipids in lipid nanoparticles with analytical QbD. J. Chrom. A 2023, 1709, 464375 10.1016/j.chroma.2023.464375. [DOI] [PubMed] [Google Scholar]
  34. Haidar Ahmad I. A.; Blasko A.; Wang H.; Lu T.; Mangion I.; Regalado E. L. Charged aerosol detection in early and late-stage pharmaceutical development: selection of regression models at optimum power function value. J. Chrom. A 2021, 1641, 461997 10.1016/j.chroma.2021.461997. [DOI] [PubMed] [Google Scholar]
  35. Haidar Ahmad I. A.; Blasko A.; Tam J.; Variankaval N.; Halsey H. M.; Hartman R.; Regalado E. L. Revealing the inner workings of the power function algorithm in Charged Aerosol Detection: A simple and effective approach to optimizing power function value for quantitative analysis. J. Chromatogr. A 2019, 1603, 1–7. 10.1016/j.chroma.2019.04.017. [DOI] [PubMed] [Google Scholar]
  36. Regalado E. L.; Haidar Ahmad I. A.; Bennett R.; D’Atri V.; Makarov A. A.; Humphrey G. R.; Mangion I.; Guillarme D. The emergence of universal chromatographic methods in the research and development of new drug substances. Acc. Chem. Res. 2019, 52 (7), 1990–2002. 10.1021/acs.accounts.9b00068. [DOI] [PubMed] [Google Scholar]; doi:
  37. Makey D. M.; Shchurik V.; Wang H.; Lhotka H. R.; Stoll D. R.; Vazhentsev A.; Mangion I.; Regalado E. L.; Ahmad I. A. H. Mapping the Separation Landscape in Two-Dimensional Liquid Chromatography: Blueprints for Efficient Analysis and Purification of Pharmaceuticals Enabled by Computer-Assisted Modeling. Anal. Chem. 2021, 93 (2), 964–972. 10.1021/acs.analchem.0c03680. [DOI] [PubMed] [Google Scholar]
  38. Haidar Ahmad I. A.; Shchurik V.; Nowak T.; Mann B. F.; Regalado E. L. Introducing Multifactorial Peak Crossover in Analytical and Preparative Chromatography via Computer-Assisted Modeling. Anal. Chem. 2020, 92 (19), 13443–13451. 10.1021/acs.analchem.0c02807. [DOI] [PubMed] [Google Scholar]
  39. Barrientos R. C.; Losacco G. L.; Azizi M.; Wang H.; Nguyen A. N.; Shchurik V.; Singh A.; Richardson D.; Mangion I.; Guillarme D.; et al. Automated Hydrophobic Interaction Chromatography Screening Combined with In Silico Optimization as a Framework for Nondenaturing Analysis and Purification of Biopharmaceuticals. Anal. Chem. 2022, 94 (49), 17131–17141. 10.1021/acs.analchem.2c03453. [DOI] [PubMed] [Google Scholar]
  40. Wang H.; Herderschee H. R.; Bennett R.; Potapenko M.; Pickens C. J.; Mann B. F.; Haidar Ahmad I. A.; Regalado E. L. Introducing online multicolumn two-dimensional liquid chromatography screening for facile selection of stationary and mobile phase conditions in both dimensions. J. Chrom. A 2020, 1622, 460895 10.1016/j.chroma.2020.460895. [DOI] [PubMed] [Google Scholar]
  41. Haidar Ahmad I. A.; Chen W.; Halsey H. M.; Klapars A.; Limanto J.; Pirrone G. F.; Nowak T.; Bennett R.; Hartman R.; Makarov A. A.; et al. Multi-column ultra-high performance liquid chromatography screening with chaotropic agents and computer-assisted separation modeling enables process development of new drug substances. Analyst 2019, 144, 2872–2880. 10.1039/C8AN02499E. [DOI] [PubMed] [Google Scholar]
  42. Gamache P. H.; Kaufman S. L. Principles of Charged Aerosol Detection. Charged Aerosol Detection Liquid Chromatography Related Separation Techniques 2017, 1–65. 10.1002/9781119390725.ch1. [DOI] [Google Scholar]
  43. Bobbala S.; McDowell A.; Hook S. Quantitation of the immunological adjuvants, monophosphoryl lipid A and Quil A in poly (lactic-co-glycolic acid) nanoparticles using high performance liquid chromatography with evaporative light scattering detection. J. Chromatogr. B 2015, 975, 45–51. 10.1016/j.jchromb.2014.11.006. [DOI] [PubMed] [Google Scholar]
  44. Fauland A.; Köfeler H.; Trötzmüller M.; Knopf A.; Hartler J.; Eberl A.; Chitraju C.; Lankmayr E.; Spener F. A comprehensive method for lipid profiling by liquid chromatography-ion cyclotron resonance mass spectrometry. J. Lipid Res. 2011, 52 (12), 2314–2322. 10.1194/jlr.D016550. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Narváez-Rivas M.; Zhang Q. Comprehensive untargeted lipidomic analysis using core–shell C30 particle column and high field orbitrap mass spectrometer. J. Chromatography A 2016, 1440, 123–134. 10.1016/j.chroma.2016.02.054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Jeschek D.; Lhota G.; Wallner J.; Vorauer-Uhl K. A versatile, quantitative analytical method for pharmaceutical relevant lipids in drug delivery systems. J. Pharm. Biomed. Anal. 2016, 119, 37–44. 10.1016/j.jpba.2015.11.020. [DOI] [PubMed] [Google Scholar]
  47. Mousli Y.; Brachet M.; Chain J. L.; Ferey L. A rapid and quantitative reversed-phase HPLC-DAD/ELSD method for lipids involved in nanoparticle formulations. J. Pharm. Biomed. Anal. 2022, 220, 115011 10.1016/j.jpba.2022.115011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Plante M.; Bailey B.; Acworth I. N.. Analysis of Lipids by HPLC-CAD, 2015https://assets.thermofisher.com/TFS-Assets/CMD/Application-Notes/AN-71759-HPLC-Lipids-Corona-ultra-AN71759-EN.pdf (accessed 18 March 2024).
  49. Hou X.; Zaks T.; Langer R.; Dong Y. Lipid nanoparticles for mRNA delivery. Nat. Rev. Mater. 2021, 6 (12), 1078–1094. 10.1038/s41578-021-00358-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Hemida M.; Haidar Ahmad I. A.; Barrientos R. C.; Regalado E. L. Computer-assisted multifactorial method development for the streamlined separation and analysis of multicomponent mixtures in (Bio)pharmaceutical settings. Anal. Chim. Acta 2023, 1293, 342178 10.1016/j.aca.2023.342178. [DOI] [PubMed] [Google Scholar]
  51. Alving C. R.; Peachman K. K.; Rao M.; Reed S. G. Adjuvants for human vaccines. Curr. Opin. Immunol. 2012, 24 (3), 310–315. 10.1016/j.coi.2012.03.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Kerwin B. A. Polysorbates 20 and 80 Used in the Formulation of Protein Biotherapeutics: Structure and Degradation Pathways. J. Pharm. Sci. 2008, 97 (8), 2924–2935. 10.1002/jps.21190. [DOI] [PubMed] [Google Scholar]
  53. Zhang R.; Wang Y.; Tan L.; Zhang H. Y.; Yang M. Analysis of Polysorbate 80 and its Related Compounds by RP-HPLC with ELSD and MS Detection. J. Chromatogr. Sci. 2012, 50 (7), 598–607. 10.1093/chromsci/bms035. [DOI] [PubMed] [Google Scholar]
  54. Rütters H.; Möhring T.; Rullkötter J.; Griep-Raming J.; Metzger J. O. The persistent memory effect of triethylamine in the analysis of phospholipids by liquid chromatography/mass spectrometry. Rapid Commun. Mass Spectrom. 2000, 14 (2), 122–123. 10.1002/(SICI)1097-0231(20000130)14:2. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

pt4c00306_si_001.pdf (312KB, pdf)

Articles from ACS Pharmacology & Translational Science are provided here courtesy of American Chemical Society

RESOURCES