Spray-Dried Monoclonal Antibody Suspension for High-Concentration and Low-Viscosity Subcutaneous Injection

Spray-Dried Monoclonal Antibody Suspension for High- Concentration and Low-Viscosity Subcutaneous Injection

Chengnan Huang, Linc Chen, Lutz Franzen, Juliane Anderski, and Feng Qian*

Cite This: Mol. Pharmaceutics 2022, 19, 1505−1514 Read Online

ACCESS Metrics & More Article Recommendations ABSTRACT: Administration of highly concentrated monoclonal antibodies (mAbs) through injection is often not possible as the viscosity can be readily above 50 mPa·s when the concentration exceeds 150 mg/mL. Besides, highly concentrated mAb solutions always exhibit increased aggregation propensity and lower stability, which raise the difficulty for the successful development of highly concentrated mAb formulations. We hereby explored the possibility of suspension as another formulation form for high- concentration proteins to reduce viscosity and maintain stability. Specifically, we demonstrated that spray drying can serve as a process to prepare particles for suspension. Particles prepared from formulations with different mAb/trehalose mass ratios displayed good physical stability and antibody binding affinity, as indicated by circular dichroism, fluorescence spectroscopy, and surface plasmon resonance (SPR)-based bioassay analyses. During spray drying, a surface tension-dominated enrichment of mAb on the particle surface was observed, but this did not show a significant negative impact on mAb stability. Spray-dried particles were subsequently suspended into benzyl benzoate, and the resulting suspension showed good stability and a lower viscosity when compared to its counterpart solution. Furthermore, mAbs recovered from the suspension maintained their conformational structure.

Our study demonstrated that the suspension displayed low viscosity and good physical stability, so it may offer novel opportunities for the preparation of highly concentrated protein formulations.

KEYWORDS: monoclonal antibody, spray drying, surface enrichment, suspension, viscosity, structural stability

1. INTRODUCTION Monoclonal antibodies have increasingly become a vital class of therapeutic entity, among small molecules, biologics, and cell/gene therapies, for the treatment of a variety of diseases, including cancer, autoimmune diseases, and other human disorders, due to their high affinity and specificity as well as outstanding pharmacokinetic profile.1−4 Intravenous injection or infusion is the most widely used approach to deliver monoclonal antibodies of low concentrations to meet the dose requirement.3 Considering patient’s compliance and special application scenario, e.g., subcutaneous and intravitreous injection, where only a very limited volume is allowed to administrate, high-concentration formulations are more favored.5,6 However, highly concentrated monoclonal antibody (mAb) formulations often display undesirable properties, such as high viscosity, high aggregation propensity, low stability, etc., due to the complicated molecular interactions.7 These properties hinder the development of high-concentration formulations for the unmet clinical needs.

Great efforts were made in the last years to reduce the viscosity and improve the stability of highly concentrated mAb formulations. Protein engineering is applied to disrupt molecular interactions and harmonize solution behavior;8,9 for example, Geoghegan and colleagues9 reported that

Received:

January 15, 2022 Revised:

April 2, 2022 Accepted:

April 4, 2022 Published: April 13, 2022 Article pubs.acs.org/molecularpharmaceutics

© 2022 American Chemical Society 1505 https://doi.org/10.1021/acs.molpharmaceut.2c00039

Mol. Pharmaceutics 2022, 19, 1505−1514 Downloaded via TSINGHUA UNIV on November 4, 2022 at 12:42:13 (UTC).

See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. mutation of amino acids relating to IgG’s surface hydro- phobicity can significantly reduce viscosity. Formulation optimization can also decrease the viscosity to a certain degree if the excipients are selected suitably. Investigations were widely conducted on the effects of buffering agents, sugars, amino acids, and salts on the viscosity of monoclonal antibodies,10−15 among which sugars were demonstrated to increase significantly the viscosity of high-concentration mAb solutions and Arg·HCl, caffeine, His·HCl, and NaCl can also serve as viscosity reducers in some cases. Protein suspension is another formulation form that would reduce the viscosity differently. Researchers are trying to crystalize mAbs and suspend the crystals in aqueous media.16,17 Yang and colleagues17 demonstrated that the viscosity of crystal formulations can be lower than that of solution formulations.

More directly, protein particles were suspended into a nonaqueous solvent to achieve a highly concentrated suspension with low viscosity and good stability. Knepp and colleagues18 obtained long-term stability of proteins at elevated temperatures when they suspended plasma-derived factor IX (57 kDa) and interferon (19 kDa) into perfluorodecalin or a mixture of perfluorotributylamine and perfluorodecalin. Miller and colleagues19 achieved a low-viscosity (400 mg/mL) lysozyme suspension and proposed that the increased viscosity of the suspension to the solvent was dominantly caused by the excluded volume but not electroviscous effects, solvation of the particles, or deviation of the particle shape from a sphere. The study of monoclonal antibody suspensions20 also confirmed the viscosity reduction effects but lacked stability examination and used pharmaceutically unacceptable organic solvents, such as ethanol, toluene, methanol, etc., which left the gaps for in vivo applications.

The objective of this study was to produce a low-viscosity, highly concentrated mAb suspension with a pharmaceutically acceptable solvent, benzyl benzoate, and to describe the effects of processes including spray drying, suspending in an organic solvent, and recovering from the suspension on both physical stability and binding affinity of model antibody BM1. Different formulations of BM1/trehalose by mass were spray-dried to test the minimum trehalose fraction for stabilizing BM1 during spray drying. Meanwhile, we investigated the mechanism of surface enrichment of BM1 on spray-dried particles by a model system with different molecular interactions. BM1 powder was subsequently suspended in benzyl benzoate to make a 200 mg/ mL suspension. The viscosity of the suspension was measured and compared with that of the solution control. Besides, the stability of BM1 recovered from the suspension was evaluated by circular dichroism, fluorescence spectroscopy, and size- exclusion high-performance liquid chromatography (SEC- HPLC).

2. MATERIALS AND METHODS 2.1. Materials. The IgG2-type antibody BM1 was kindly provided by Bayer AG (Wuppertal, Germany). Reagents used in this study were of analytical grade and obtained from commercial vendors: sodium dihydrogen phosphate and disodium hydrogen phosphate (Xilong Chemical, Guangdong,

China), trehalose (Macklin Biochemical, Shanghai, China), and benzyl benzoate (J&K Chemical, Beijing, China). Milli-Q water (resistivity 15.0 MΩ·cm at 25 °C; Merck Millipore,

Billerica, MA) was used in all experiments.

2.2. Methods. 2.2.1. Solution Preparation. All buffers used in this study were prepared to have a certain pH by adjusting the addition of sodium dihydrogen phosphate and disodium hydrogen phosphate and then filtered through 0.22 μm nitrocellulose membranes (Millex, Merck Millipore Ltd.,

Ireland) before usage. Buffer exchange of BM1 was done by dialysis using a dialysis membrane of a molecular weight cutoff of 14 kDa (Union Carbide Co.) or a Slide-A-Lyzer dialysis cassette of a molecular weight cutoffof 10 kDa (Thermo

Scientific). The BM1 solution after spray drying was prepared through reconstitution from spray-dried powders by Milli-Q water to the concentration of BM1 before spray drying and then diluted to a concentration of interest for measurements by corresponding buffers. BM1 recovered from the suspension was prepared by adding 0.5 mL of suspension to 4.5 mL of water and subsequently collecting the supernatant by centrifugation.

2.2.2. Concentration Measurement. All concentrations of

BM1 in the solution except that recovered from the suspension were determined with a UV A280 of NanoDrop 2000 (Thermo Scientific). The Micro bicinchoninic acid (BCA) protein assay (Beyotime Biotechnology, Shanghai, China) was introduced to measure the concentration of BM1 recovered from the suspension as the trace of benzyl benzoate can interfere with the UV A280 measurement. For BCA measure- ments, each sample was measured in triplicate in a 96-well plate and the absorbance was measured at 562 nm using a microplate reader (SpectraMax Gemini XPS/EM Microplate

Readers, Molecular Devices). The control BM1 was measured by both UV A280 and BCA assays for normalization so that all results can be presented in the UV A280 form.

2.2.3. Measurement of Surface Tension. The measurement of surface tension was described in our previous study.21

Briefly, the Wilhelmy plate method was applied to measure the surface tension of BM1 solutions at different concentrations (0.001, 0.01, 0.1, 1, 5 mg/mL). The apparatus was validated by measuring the surface tension of water (72−73 mN/m). Each sample was measured in triplicate.

2.2.4. Determination of the Interaction Parameter (kD).

The interaction parameter was obtained through curve-fitting in terms of the dependence of the diffusion coefficient on protein concentration. Briefly, 30 mg/mL BM1 was first prepared in 50 mM phosphate buffer at different pH conditions and then diluted to 28, 26, 24, 22, 20, 18, 16, 14,

8, 6, 5, 4, 3, 2, and 1 mg/mL with the corresponding buffer.

Each BM1 sample (80 μL) was prepared in a quartzlike cuvette and applied to a Zetasizer Nano ZS90 dynamic light scattering (DLS) machine (Malvern Instruments Ltd., Worcestershire,

U.K.) to determine the diffusion coefficient (Dm). The BM1 sample was equilibrated for 120 s before measurements, and then, three measurements were conducted with each of 13 runs. The averaged Dm value was plotted against the protein concentration, where the self-diffusion coefficient (Ds) and the interaction parameter (kD) were derived using linear fitting as described by the following equation

D D k C (1 ) m s D = + (1) where C is the protein concentration (g/mL).

2.2.5. Spray-Dried Particle Preparation. A BUCHI B290 laboratory spray dryer was used to generate BM1 particles. The aspirator was set to 100%, equal to a gas flow of 583 L/min.

Other parameters are set as follows: an inlet temperature of 90

°C, which corresponds to an outlet temperature of 50−55 °C, a spray gas rate of 12.4 L/min, and a pump rate of 16%.

Molecular Pharmaceutics pubs.acs.org/molecularpharmaceutics

Article https://doi.org/10.1021/acs.molpharmaceut.2c00039

Mol. Pharmaceutics 2022, 19, 1505−1514 1506 2.2.6. Scanning Electron Microscopy (SEM). The size and morphology of spray-dried BM1 particles were analyzed using a scanning electron microscope (FEI Quanta 200, Nether- lands) operating at an excitation voltage of 15 kV. Samples were mounted onto a copper stage and coated with gold by sputtering for 60 s before observation.

2.2.7. X-ray Photoelectron Spectroscopy (XPS). XPS was employed to detect the surface elemental compositions of spray-dried BM1 particles. Details of parameter settings are described in our previous study.21 The fraction of each component on the surface was calculated as follows from the chemical formulas and the characteristic elements of BM1 (C6422H9946N1688O1997S50), trehalose (C12H22O11), and phos- phate (NaH2PO4)

M M M M M M M M M P% /(6422 12 1688 4 1997 11 50 )

NaH PO protein trehalose protein NaH PO protein trehalose protein

NaH PO 2 4 2 4 2 4 = + + + + + + + M M M M M M M M

M N% 1688 /(6422 12 1688 4 1997 11 50 ) protein protein trehalose protein

NaH PO protein trehalose protein NaH PO 2 4 2 4 = +

+ + + + + + M M M M M M M M M M M O% (4 1997 11 ) /(6422

12 1688 4 1997 11 50 ) NaH PO protein trehalose protein trehalose protein

NaH PO protein trehalose protein NaH PO 2 4 2 4 2 4

= + + + + + + + + + M M M M M M M M M M C% (6422 12

) /(6422 12 1688 4 1997 11 50 ) protein trehalose protein trehalose protein

NaH PO protein trehalose protein NaH PO 2 4 2 4 = +

+ + + + + + + M M N% P% 1688 protein NaH PO 2 4 = M

M M M N% O% 1688 4 1997 11 protein NaH PO protein trehalose

2 4 = + + M M 11 1688 6752 1997 protein trehalose N %

O % P % O % N % O % = − − M M 1688 P% N% NaH PO protein

2 4 = 2.2.8. Circular Dichroism (CD). CD measurements were performed using a Chirascan Plus CD spectrometer (Applied

Photophysics, U.K.). BM1 samples were loaded in 1.0 mm path-length quartz cuvettes for CD experiments at 25 °C. The far-UV (200−260 nm) experiments were conducted on protein samples with concentrations of 0.2 mg/mL. The CD spectrometer instrument was set with a 50 nm/min scan rate, a 0.5 nm bandwidth, and a 1 s integration time. All samples were recorded three times. The solvent spectrum obtained under the same conditions was subtracted to obtain the actual sample spectrum. The spectra were averaged and smoothened using Chirascan software.

2.2.9. Fluorescence (FL) Spectroscopy. All fluorescence experiments were performed on a Dual-FLTM spectropho- tometer system (Horiba Scientific, Edison, NJ). The spectra were measured at an excitation wavelength of 290 nm, which is exclusively due to the intrinsic tryptophan fluorophore. Briefly,

300 μL of buffer solution or 2 mg/mL BM1 solution was placed in a 2.0 mm path-length quartz cuvette. The cuvette was then inserted into the cuvette holder, and the fluorescence emission was recorded immediately; each sample was analyzed in triplicate.

2.2.10. Viscosity Measurement. Viscosity measurement of the highly concentrated protein solution using a microVISC (Rheosense Inc., CA), coupled with a Type A chipset (14HA05100550), was described in detail by our group.12

Before each measurement, water was used as a reference liquid to ensure the accuracy of the rheometer. Average viscosity values based on 3−5 measurements were recorded.

2.2.11. Rheology Measurement. Due to the size limitation of the Type A chipset of microVISC, rheologies of suspensions and a 200 mg/mL BM1 solution were measured using the parallel-plate mode of a HAAKE MARS Rheometer (Thermo

Scientific). Approximately, 1 mL of sample was placed on the plate, and the excess sample was removed before performing a measurement. Thirty measurements between shear stresses of

5 and 100 Pa were performed after the equipment was equilibrated at a room temperature of 22 °C.

2.2.12. Size-Exclusion Chromatography (SEC). A Prom- inence-i9 (LC-2030, Shimadzu, Japan) high-performance liquid chromatography system, coupled with BioCore SEC- 300 (300 mm × 7.8 mm × 5 μm, Nanochrom, Jiangsu), was used to detect the monomer of BM1 and aggregation formation. Each sample (50 μL) was loaded onto the column and eluted at a flow rate of 0.6 mL/min with a mobile phase of

150 mM sodium phosphate at pH 6.8 and room temperature of around 25 °C. The protein concentration was measured by

UV absorbance at 280 nm. The area under curve of peaks in the chromatogram was used to calculate the amount of monomers and aggregates of mAbs.

2.2.13. Surface Plasmon Resonance (SPR)-Based Analysis of Binding Affinity. Surface plasmon resonance (SPR) is a biosensor technology to prove untagged protein−protein interactions in real-time. To perform the analysis, an antihistidine antibody was attached to the surface of a sensor chip. The immobilized antihistidine antibody captured the antigen (containing a His-tag in the nonfunctional area of the protein). Specific binding of BM1 toward the captured antigen was analyzed to confirm the relative active concentration of

BM1 before/after spray drying to reference BM1.

2.2.14. Fourier Transform Infrared (FT-IR) Spectroscopy.

FT-IR analysis was performed using an FT-IR-ATR Vertex 70 spectrometer (Bruker Optics, Ettlingen, Germany) at room temperature (∼25 °C). The region of 1300−1800 cm−1 in the spectra was baseline-corrected and normalized. The region of

1600−1700 cm−1 corresponding to amide I was extracted for analysis.

3. RESULTS AND DISCUSSIONS 3.1. Size and Morphology of Spray-Dried BM1

Particles. Spray drying is a technique for powder preparation that is used widely in food and pharmaceutical research and industry;22−26 we applied it in this study and took its advantage as a controllable process for size and morphology. The spherical shape and relatively monodispersed size make the

Molecular Pharmaceutics pubs.acs.org/molecularpharmaceutics

Article https://doi.org/10.1021/acs.molpharmaceut.2c00039

Mol. Pharmaceutics 2022, 19, 1505−1514 1507 spray-dried powders eligible for further material processing, e.g., suspending in organic solvents.27

To investigate the impacts of excipients on particle properties, e.g., size, morphology, and surface component distribution, BM1 was first dialyzed to 50 mM phosphate buffer at pH 5.8 and then formulated with trehalose in the mass ratios of 8:2, 6:4, 4:6, and 2:8. For spray drying, a low inlet temperature was adopted to reduce any risk of heat- induced damage on BM1, though this process normally lasts only for a few milliseconds,28 and a low feeding rate to remove water as much as possible. The yield for the 8:2 group was

∼75%, and it decreased as the percentage of trehalose increased. Also, spray-dried powders became sticky and can hardly be collected from the cyclone for the 2:8 group. Water residues in the first three BM1 powders were below 5% as indicated by thermogravimetric analysis (TGA) (data not shown), which were favored to avoid the Tg-lowering effect of water and to keep the system in a glass state for better immobilization of proteins.29,30

The scanning electron microscopy (SEM) images provided the size and morphology of spray-dried BM1 particles (Figure

1). The average diameter of a single particle in BM1 powders was 2−8 μm. The morphology of a single particle was found to be wrinkled, though spherical particles were favored for generating a low-viscosity suspension as suggested by the volume theory.19 The wrinkled shape was caused by a typical effect of a high-molecular-weight additive, the BM1, which altered the balance of surface-to-viscous forces controlling the droplet shape during drying.31

3.2. Surface Composition of Spray-Dried BM1 Particles. Surface enrichment or depletion of a certain component of interest was widely observed and reported in different spray-dried powder systems, e.g., milk powder,26,32−35 small-molecule drugs and polymers,21 proteins, sugars36−38 etc.

The surface composition of these powders is expected to play an essential role in their storage, handling, and final application. More specifically, for the pharmaceutical industry, surface enrichment or depletion of active pharmaceutical ingredients affects their dissolution, stability, and activity.21,39

Proteins on the surface are challenged by direct contact with heat and the air−liquid/solid interface during spray drying, especially those sensitive to heat and interfaces.

X-ray photoelectron spectroscopy (XPS) is a technique that analyzes the elements along with their percentages in a 100 μm2 area with 10 nm depth40 and was utilized in this study to reveal the surface composition of spray-dried BM1 particles.

Mass fractions of BM1 in the solutions of 8:2, 6:4, and 4:6 groups were 44.9, 33.7, and 22.5%, respectively, while those on the surfaces of corresponding spray-dried particles were 55.9,

49.1, and 49.2% (Figure 2a). Proteins on the surface were around 50%, and this could be explained as follows: a protein is hundreds to thousands of times larger than a small molecule and is surrounded with trehalose/phosphate in our preformu- lation before spray drying. Molecular rearrangement and surface enrichment of proteins were observed after spray drying, but this enrichment cannot be 100% even if the particle was coated by proteins as trehalose/phosphate can still fill in the protein molecule and can be detected by XPS at a depth of

10 nm (antibody size is normally around 5 nm).

BM1 enrichment on the surface can be explained by its surface activity, molecular weight, and molecular interaction or their combinations21 (Figure 2e): (1) after spraying, large numbers of droplets formed and the resulting surface energy needed to be lowered by the surface-active BM1 (Figure 2b); (2) during the drying process, water in the outer layer evaporated due to contact with heat. This led to the formation of a concentration gradient and mass exchange, from which the inner layer of water went out to the surface and near-surface solutes went into the inner part of the droplet. Based on the

Stokes−Einstein relation, a molecule with a larger size diffuses slower than a small size molecule. In such a case, BM1 would be left on the surface during the drying process, as a result of high molecular weight, and contribute to its enrichment on the particle surface.21 Besides, molecular interactions would also affect molecule diffusion during the drying step.

To assess the effects of electrostatic interactions between protein molecules on surface enrichment of BM1 on spray- dried particles, we reformulated the 8:2 group of BM1 to phosphate buffer with pH from 5 to 8.5, across the theoretical isoelectric point, as predicted, of 8.3, to generate a model that represents different strengths of molecular interactions (Figure

2c). The six BM1 solutions that had different molecular interactions, from zero to more attractive interactions, were then spray-dried and analyzed in the same manner as before.

Due to the absence of nitrogen in the excipients and the same level of elements that existed in all solutions, the amount of nitrogen present on the surface can be used to determine the percentage of protein. To facilitate the analysis, we defined a nitrogen% change of more/less than 10% upon average as a significant change of protein%. As shown in Figure 2d, no significant variations in protein enrichment on the surface were detected by XPS, namely, there was no correlation between electrostatic interactions and surface enrichment of proteins.

We would explain that surface enrichment of BM1 on spray- dried particles was dominated by surface tension, and molecular interactions would be minor or even negligible.

3.3. Physical Stability and Binding Affinity of BM1 after Spray Drying. To evaluate the conformation stability of

BM1 after spray drying, we reconstituted the three spray-dried powders into a solution by deionized water to the concentration of BM1 before spray drying and then diluted it to a concentration of interest for CD and FL measurements by corresponding buffers. For fluorescence spectroscopy, changes of protein in terms of conformational transitions, subunit association, substrate binding, or denaturation can be detected by the emission spectra of tryptophan,41 in particular, by wavelength shift analyses. BM1 was excited by a 290 nm laser and showed an emission peak at 335 nm. As shown in

Figure 3a, there were no obvious peak shifts for all BM1 reconstitutions, when compared to the BM1 spectra before spray drying, suggesting the absence of detectable tertiary structure changes. Besides, we found that the CD spectra of

BM1 reconstitutions overlapped with those before spray

Figure 1. SEM images of spray-dried BM1 from different aqueous formulations (8:2, 6:4, and 4:6 represent the ratios of BM1 to trehalose by mass).

Molecular Pharmaceutics pubs.acs.org/molecularpharmaceutics

Article https://doi.org/10.1021/acs.molpharmaceut.2c00039

Mol. Pharmaceutics 2022, 19, 1505−1514 1508 drying, as shown in Figure 3b, suggesting the absence of detectable secondary structure alterations. A mismatch near

200 nm found in the 4:6 group in Figure 3b was possibly a result of background interference.

One interesting observation was that although BM1 was enriched on the particle surface after/during spray drying, structural damage was not detected by the widely used technologies, i.e., CD and FL. The possible explanations are as follows: (1) the spray drying process was mild and too short to induce a conformational change; (2) BM1 was insensitive to the mild heat and large air−liquid/solid interfaces due to its rigid intrinsic biophysical properties; (3) subvisible aggregates or denatured BM1 was removed by centrifugation step before

CD/FL analysis, but this should be amplified and again verified by long-term stability study; and (4) resolution of CD/FL was not enough to capture the structural alterations,42 though 8.7% of protein enriched on the surface and could be damaged (8:2 group calculated based on the model proposed by Lee’s paper28 as an example). To further evaluate the impacts of spray drying on protein stability, we subsequently characterized the BM1 by SPR, FT-IR, and SEC-HPLC. The obtained SPR data (Figure 3c) showed no significant reduction of binding affinity for BM1 after spray drying and reconstitution. FT-IR spectra (Figure 3d) showed that the major peak at around

Figure 2. (a) Mass fraction of BM1 in the solution (blank bar) and on the surface of spray-dried particles (filled bar) as determined from XPS analysis. Values 8:2, 6:4, and 4:6 of the x-axis represent the ratios of BM1 to trehalose by mass. (b) Dependence of surface tension on BM1 concentration. BM1 was prepared in different concentrations in phosphate buffer (50 mM, pH 5.8). (c) Dependence of interaction parameter on pH. BM1 in phosphate buffer (50 mM, pH was set to a range of 5.1−8.4) was diluted to concentration gradients of 30, 28, 26, 24 22, 20, 18, 16, and 14 mg/mL to determine the interaction parameter kD. (d) XPS analysis of spray-dried BM1 particles from formulations with the same BM1/ trehalose (8:2) but in different pH buffers. To facilitate the analysis, we define a nitrogen% change of more/less than 10% upon average as a

“significant change of protein%”. (e) Schematic illustration of the spray drying process for a protein−trehalose solution in phosphate buffer. Two main steps were included: spraying (or atomization) and drying (or diffusion of molecules). During the spraying step, a large surface was generated on which surface-active molecules were prone to stay to lower the surface energy from a thermodynamic consideration. For the drying step, including a constant-rate and a falling-rate drying process, the diffusivity coefficient affects the rearrangement of solutes.

Molecular Pharmaceutics pubs.acs.org/molecularpharmaceutics

Article https://doi.org/10.1021/acs.molpharmaceut.2c00039

Mol. Pharmaceutics 2022, 19, 1505−1514 1509 1638 cm−1 corresponding to the dominant β-sheet structure of mAbs did not shift, suggesting that spray drying did not cause detectable changes in the secondary structure of the mAbs.

The variation in the peak intensity is likely due to the concentration difference. We also did not detect significant monomer loss of BM1 after spray drying and storage of 5

Figure 3. Comparison of stability and affinity of unprocessed BM1 (black line or black solid bar), reconstituted BM1 (red line or red solid bar), and spray-dried BM1 powders (bue line). (a) Intrinsic fluorescence spectra, (b) far-UV CD spectra, (c) SPR-based binding affinity analysis, (d) FT-IR spectra of the amide I region, and (e) SEC-HPLC profile of BM1 reconstituted from spray-dried powders stored for 5 months.

Figure 4. Preparation of the BM1 suspension. (a) Lyophilized powder had very poor suspendability in benzyl benzoate and (b−d) suspension prepared from spray-dried BM1 and stored at room temperature for 0, 5, and 18 days.

Molecular Pharmaceutics pubs.acs.org/molecularpharmaceutics

Article https://doi.org/10.1021/acs.molpharmaceut.2c00039

Mol. Pharmaceutics 2022, 19, 1505−1514 1510 months by SEC-HPLC (Figure 3e). Together with all of the stability data, we conclude that the spray drying process did not cause detectable conformational changes in BM1 both in the solid state and in the reconstituted solution due to the mild process or the rigid biophysical properties of BM1 itself, though BM1 enrichment on the surface of spray-dried particles was observed.

3.4. BM1 Powder Showed Good Suspendability in Benzyl Benzoate. To evaluate the suspendability of BM1 powders in possible organic solvents, we first selected benzyl benzoate as a model solvent due to its low viscosity and approved use in an intramuscular injection product Proluton43 and then suspended it with spray-dried BM1 particles from the

8:2 group, which had higher drug loading. We found that the spray-dried BM1 powder can readily be suspended in benzyl benzoate to make 200 mg/mL BM1 suspension (Figure 4b), while the lyophilized group of 200 mg/mL BM1 cannot even be wetted by benzyl benzoate (Figure 4a), let alone formed a suspension, suggesting that the spray-dried BM1 powder possessed a good suspending ability. For successful delivery with highly concentrated suspensions, the settling rate, an indicator of suspension stability must be monitored, and a slow settling rate is favored to reduce agglomeration. As shown in

Figure 4b−d, the BM1 suspension was stable as only a tendency of settling from day 5 and a clear separation from day

18 were observed. The settled suspension can be easily resuspended by shaking or vortexing.

The suspension of spray-dried or lyophilized protein was tested as an alternative approach to lower the viscosity of the highly concentrated formulation and maintain long-term stability as well as extend the drug release profile.18−20,44

Researchers also tried with monoclonal antibodies,20,44 but the stability study of proteins in the suspension was not presented.

3.5. BM1 Suspension Exhibited Low Viscosity at High

Shear Rates. Highly concentrated protein solutions normally display increased viscosity due to the molecular interactions and large volume of excipients, which are used to stabilize proteins but occupy a large space of the solution and also contribute to the high viscosity.12,17,45−47 We here measured the viscosity of the BM1 solution in 50 mM phosphate buffer at different concentrations, which was prepared through concentrating using centrifugal filter units. The BM1 solution can be concentrated up to 200 mg/mL, and its viscosity displayed an exponential dependence on the concentration (Figure 5a), which is the property of a non-Newtonian solution. The viscosity increased sharply when the concen- tration exceeded 150 mg/mL due to the strengthened molecular interaction at high concentrations and the resulting deviated solution behavior from that of the ideal Newtonian solution.48,49

In suspensions, the physical properties of mAbs are different from those in solutions: particle−particle interactions would mainly exist in suspensions instead of molecular interactions; thus, a lower viscosity was expected.19 As shown in Figure 5b,

BM1 suspensions exhibited shear-thinning viscosity as

Figure 5. Viscosity measurement of the BM1 solution/suspension. (a) Dependence of viscosity on the concentration of the BM1 solution.

Different concentrations of the BM1 solution were prepared through concentrating using centrifugal filter units. (b) Dependence of viscosity on the shear rate. Pure benzyl benzoate (black square), 100 mg/mL suspension (blue square), and 200 mg/mL suspension (red square) were applied at a shear rate in the range of 100 to 9000 s−1. (c) Comparison of the viscosity between the 200 mg/mL suspension and the 200 mg/mL solution at different shear rates. (d) Comparison of the viscosity among BM1 formulations at a shear rate of 4000 s−1. Solution A was prepared through concentrating using centrifugal filter units, and solution B was prepared through reconstitution.

Molecular Pharmaceutics pubs.acs.org/molecularpharmaceutics

Article https://doi.org/10.1021/acs.molpharmaceut.2c00039

Mol. Pharmaceutics 2022, 19, 1505−1514 1511 indicated by rheology characterization, which was in line with that of other protein suspensions.19 A 200 mg/mL BM1 solution prepared through concentrating using centrifugal filter units was also applied to the rheometer to test its rheological properties for comparison (Figure 5c). The BM1 solution showed a different viscosity−shear rate relation from the BM1 suspension, suggesting different material properties or different physical interactions among molecules. The apparent viscosity of the 200 mg/mL BM1 suspension was lower than that of the

200 mg/mL BM1 solution when the shear rate was more than

500 s−1 and was only 24 mPa·s at a shear rate of 4000 s−1 produced by the 25-G needle injection.19 To take the effect of excipients into account, we prepared another 200 mg/mL BM1 solution through reconstitution of spray-dried BM1 powders, and its viscosity was 79 mPa·s (Figure 5d). This solution had a more comparable composition with that of suspension than the

BM1 solution prepared by concentrating using centrifugal filter units, during which excipients were removed.

The increased viscosity of the highly concentrated protein solution was a result of molecular interaction and excipient- occupied space. Here, we observed concentration-dependent viscosity of the BM1 solution and lower viscosity of the 200 mg/mL solution with fewer excipients than that with full excipients. In a suspension or solid state, molecular interactions will be reduced or even be mitigated, and a smaller excipient level was able to stabilize the protein.17,50 In our study, we confirmed that less excipient was possible to protect BM1 for spray drying, e.g., the 8:2 group of BM1 also exhibited comparable stability to the others (Figure 3a,b); however, this mass ratio was normally as high as 1:1 for marketed solution formulations.3 We considered that these two factors mainly contributed to the potential low viscosity of the suspension and made the suspension a promising formulation form for highly concentrated antibodies, i.e., suspension changed the molec- ular interactions and rheological properties, and fewer excipients further led to a viscosity reduction.

3.6. BM1 Recovered from the Suspension Maintained Its Physical Stability. It is challenging to prevent aggregation and activity loss of protein at high concentrations both in solutions and suspensions.18,51,52 Therefore, we performed a recovering test to compare the stability of BM1 recovered from a suspension (stored for 3 days at 4 °C, a stability study that should be acceptable if we consider the possible application scenario of protein suspension, i.e., resuspend the powder into an organic solvent before use) with that of unprocessed BM1 to evaluate the damage caused in our process. BM1 was found to be recovered readily and completely (as determined by yield, data not shown) from the suspension, and, intriguingly, to be physically stable (Figure 6). Specifically, the SEC profile (Figure 6a) showed that there were no monomer loss and aggregates in the recovered solution, and no significant changes of secondary/tertiary structures were found between

BM1 recovered from the suspension and unprocessed BM1 as indicated by the CD/FL profile (Figure 6b,c). A minor difference at 200 nm should be a result of background interference in the CD profile, and an intensity decrease but not a wavelength shift was from a change in the concentration, which was confirmed later by the BCA kit measurement.

Though the recovering test was too simple to mimic the potential manner for protein release in a specific application scenario, it revealed that such processes could be endurable for monoclonal antibodies. If we focused on the stresses involved during suspending and recovering, the solid−oil interface and oil−water interface would be challenging to BM1. During the suspension process, surface-enriched BM1 on spray-dried particles directly exposed to benzyl benzoate, the hydrophobic core of the protein would be opened and interact with the solid−oil interface, thus leading to unfolding of the protein.53

For recovering, the protein covered by the organic solvent was exposed to and dissolved in water, during which the protein encountered a large oil−liquid interface and may be damaged.54,55 In this study, no obvious structural changes were detected by the widely used technologies, i.e., SEC- HPLC, CD, and FL.

4. CONCLUSIONS Here, in this study, we reported a case indicating that the suspension can be developed as a promising formulation form for highly concentrated antibodies. Meanwhile, we discussed the potential damages to protein stability during the challenging processes including spray drying, suspending into an organic solvent, and recovering from the suspension. BM1 enrichment on the surface of the spray-dried particles was observed, and this phenomenon was considered to be dominated by the surface activity of BM1. However, no detectable structural changes and loss of activity were observed even after direct exposure to heat and interface, suggesting a

Figure 6. Stability characterization of BM1 from the suspension. (a) SEC-HPLC profile, (b) far-UV CD spectra, and (c) intrinsic fluorescence spectra at room temperature to compare the conformational stability of unprocessed BM1 (black) and BM1 recovered from the suspension (red).

Molecular Pharmaceutics pubs.acs.org/molecularpharmaceutics

Article https://doi.org/10.1021/acs.molpharmaceut.2c00039

Mol. Pharmaceutics 2022, 19, 1505−1514 1512 certain level of insensitivity of BM1 to these stresses.

Considering the drug-loading capability and excipient-driven viscosity, BM1 powder of the 8:2 mass ratio was suspended into benzyl benzoate and the resulting suspension was found to have 2.5−3.3 times lower viscosity than that of the solution counterpart. The structural characterization did not show any damage to conformational stability. This test-of-principle study provided further evidence to demonstrate the possibility of industrial and clinical applications of the protein suspension as a stable and low-viscosity, high-concentration protein for- mulation, if GMP aseptic capability to produce spray-dried protein particles with controlled powder properties could be established in the industry in the future.

■AUTHOR INFORMATION Corresponding Author Feng Qian −School of Pharmaceutical Sciences, Beijing

Advanced Innovation Center for Structural Biology, and Key

Laboratory of Bioorganic Phosphorus Chemistry & Chemical

Biology (Ministry of Education), Tsinghua University, Beijing

100084, P. R. China; orcid.org/0000-0001-7415-6997;

Email: qianfeng@tsinghua.edu.cn Authors Chengnan Huang −School of Pharmaceutical Sciences,

Beijing Advanced Innovation Center for Structural Biology, and Key Laboratory of Bioorganic Phosphorus Chemistry &

Chemical Biology (Ministry of Education), Tsinghua

University, Beijing 100084, P. R. China Linc Chen −Bayer Healthcare Co. Ltd., Beijing 100020, P. R.

China Lutz Franzen −Research & Development, Pharmaceuticals,

Bayer AG, Wuppertal 42096, Germany Juliane Anderski −Research & Development,

Pharmaceuticals, Bayer AG, Wuppertal 42096, Germany

Complete contact information is available at: https://pubs.acs.org/10.1021/acs.molpharmaceut.2c00039

Notes The authors declare no competing financial interest.

■ACKNOWLEDGMENTS The authors acknowledge Bayer AG and the National Natural

Science Foundation of China (Project Number 81773649) for providing research funding for this project. The authors also thank Zhou Tian for his support and helpful discussion.

■REFERENCES (1) Scott, A. M.; Wolchok, J. D.; Old, L. J. Antibody therapy of cancer. Nat. Rev. Cancer 2012, 12, 278−287. (2) Xu, Y.; Wang, D.; Mason, B.; Rossomando, T.; Li, N.; Liu, D.;

Cheung, J. K.; Xu, W.; Raghava, S.; Katiyar, A.; Nowak, C.; Xiang, T.;

Dong, D. D.; Sun, J.; Beck, A.; Liu, H. Structure, heterogeneity and developability assessment of therapeutic antibodies. mAbs 2019, 11,

239−264. (3) Wang, W.; Singh, S.; Zeng, D. L.; King, K.; Nema, S. Antibody structure, instability, and formulation. J. Pharm. Sci. 2007, 96, 1−26. (4) Ryman, J. T.; Meibohm, B. Pharmacokinetics of Monoclonal

Antibodies. CPT: Pharmacometrics Syst. Pharmacol. 2017, 6, 576−588. (5) Mehta, S. C.; Kelley, R. F.; Tesar, D. B. Protein conjugates and fusion proteins as ocular therapeutics. Drug Discovery Today 2019, 24,

1440−1445. (6) Nakashima, Y.; Kondo, M.; Miyahara, H.; Iwamoto, Y. Drug delivery options to increase patient adherence and satisfaction in the management of rheumatoid arthritis–focus on subcutaneous tocilizu- mab. Drug Des., Dev. Ther. 2014, 8, 913−919. (7) Yadav, S.; Laue, T. M.; Kalonia, D. S.; Singh, S. N.; Shire, S. J.

The influence of charge distribution on self-association and viscosity behavior of monoclonal antibody solutions. Mol. Pharm. 2012, 9,

791−802. (8) Chow, C. K.; Allan, B. W.; Chai, Q.; Atwell, S.; Lu, J.

Therapeutic Antibody Engineering To Improve Viscosity and Phase

Separation Guided by Crystal Structure. Mol. Pharm. 2016, 13, 915−

923. (9) Geoghegan, J. C.; Fleming, R.; Damschroder, M.; Bishop, S. M.;

Sathish, H. A.; Esfandiary, R. Mitigation of reversible self-association and viscosity in a human IgG1 monoclonal antibody by rational, structure-guided Fv engineering. mAbs 2016, 8, 941−950. (10) Whitaker, N.; Xiong, J.; Pace, S. E.; Kumar, V.; Middaugh, C.

R.; Joshi, S. B.; Volkin, D. B. A Formulation Development Approach to Identify and Select Stable Ultra-High-Concentration Monoclonal

Antibody Formulations With Reduced Viscosities. J. Pharm. Sci. 2017,

106, 3230−3241. (11) Zeng, Y.; Tran, T.; Wuthrich, P.; Naik, S.; Davagnino, J.;

Greene, D. G.; Mahoney, R. P.; Soane, D. S. Caffeine as a Viscosity

Reducer for Highly Concentrated Monoclonal Antibody Solutions. J.

Pharm. Sci. 2021, 110, 3594−3604. (12) Wang, S.; Zhang, N.; Hu, T.; Dai, W.; Feng, X.; Zhang, X.;

Qian, F. Viscosity-Lowering Effect of Amino Acids and Salts on

Highly Concentrated Solutions of Two IgG1 Monoclonal Antibodies.

Mol. Pharmaceutics 2015, 12, 4478−4487. (13) He, F.; Woods, C. E.; Litowski, J. R.; Roschen, L. A.; Gadgil, H.

S.; Razinkov, V. I.; Kerwin, B. A. Effect of sugar molecules on the viscosity of high concentration monoclonal antibody solutions.

Pharm. Res. 2011, 28, 1552−1560. (14) Borwankar, A. U.; Dear, B. J.; Twu, A.; Hung, J. J.; Dinin, A. K.;

Wilson, B. K.; Yue, J.; Maynard, J. A.; Truskett, T. M.; Johnston, K. P.

Viscosity Reduction of a Concentrated Monoclonal Antibody with

Arginine·HCl and Arginine·Glutamate. Ind. Eng. Chem. Res. 2016, 55,

11225−11234. (15) Wang, S.; Zhang, X.; Wu, G.; Tian, Z.; Qian, F. Optimization of high-concentration endostatin formulation: Harmonization of ex- cipients’ contributions on colloidal and conformational stabilities. Int.

J. Pharm. 2017, 530, 173−186. (16) Trilisky, E.; Gillespie, R.; Osslund, T. D.; Vunnum, S.

Crystallization and liquid-liquid phase separation of monoclonal antibodies and fc-fusion proteins: screening results. Biotechnol. Prog.

2011, 27, 1054−1067. (17) Yang, M. X.; Shenoy, B.; Disttler, M.; Patel, R.; McGrath, M.;

Pechenov, S.; Margolin, A. L. Crystalline monoclonal antibodies for subcutaneous delivery. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 6934−

9. (18) Knepp, V. M.; Muchnik, A.; Oldmark, S.; Kalashnikova, L.

Stability of Nonaqueous Suspension Formulations of Plasma Derived

Factor IX and Recombinant Human Alpha Interferon at Elevated

Temperatures. Pharm. Res. 1998, 15, 1090−1095. (19) Miller, M. A.; Engstrom, J. D.; Ludher, B. S.; Johnston, K. P.

Low viscosity highly concentrated injectable nonaqueous suspensions of lysozyme microparticles. Langmuir 2010, 26, 1067−1074. (20) Srinivasan, C.; Weight, A. K.; Bussemer, T.; Klibanov, A. M.

Non-aqueous suspensions of antibodies are much less viscous than equally concentrated aqueous solutions. Pharm. Res. 2013, 30, 1749−

1757. (21) Chen, Z.; Yang, K.; Huang, C.; Zhu, A.; Yu, L.; Qian, F. Surface

Enrichment and Depletion of the Active Ingredient in Spray Dried

Amorphous Solid Dispersions. Pharm. Res. 2018, 35, No. 38. (22) Bürki, K.; Jeon, I.; Arpagaus, C.; Betz, G. New insights into respirable protein powder preparation using a nano spray dryer. Int. J.

Pharm. 2011, 408, 248−256. (23) Gikanga, B.; Turok, R.; Hui, A.; Bowen, M.; Stauch, O. B.; Maa,

Y. F. Manufacturing of High-Concentration Monoclonal Antibody

Formulations via Spray Drying-the Road to Manufacturing Scale. PDA

J. Pharm. Sci. Technol. 2015, 69, 59−73.

Molecular Pharmaceutics pubs.acs.org/molecularpharmaceutics

Article https://doi.org/10.1021/acs.molpharmaceut.2c00039

Mol. Pharmaceutics 2022, 19, 1505−1514 1513 (24) Lee, S. H.; Heng, D.; Ng, W. K.; Chan, H. K.; Tan, R. B. Nano spray drying: a novel method for preparing protein nanoparticles for protein therapy. Int. J. Pharm. 2011, 403, 192−200. (25) Sollohub, K.; Cal, K. Spray drying technique: II. Current applications in pharmaceutical technology. J. Pharm. Sci. 2010, 99,

587−97. (26) Kim, E. H. J.; Chen, X. D.; Pearce, D. Surface composition of industrial spray-dried milk powders. 1. Development of surface composition during manufacture. J. Food Eng. 2009, 94, 163−168. (27) Wang, H.; Nobes, D. S.; Vehring, R. Particle Surface Roughness

Improves Colloidal Stability of Pressurized Pharmaceutical Suspen- sions. Pharm. Res. 2019, 36, No. 43. (28) Adler, M.; Lee, G. Stability and surface activity of lactate dehydrogenase in spray-dried trehalose. J. Pharm. Sci. 1999, 88, 199−

208. (29) Carrigy, N. B.; Ordoubadi, M.; Liu, Y.; Melhem, O.; Barona,

D.; Wang, H.; Milburn, L.; Ruzycki, C. A.; Finlay, W. H.; Vehring, R.

Amorphous pullulan trehalose microparticle platform for respiratory delivery. Int. J. Pharm. 2019, 563, 156−168. (30) Mensink, M. A.; Frijlink, H. W.; van der Voort Maarschalk, K.;

Hinrichs, W. L. How sugars protect proteins in the solid state and during drying (review): Mechanisms of stabilization in relation to stress conditions. Eur. J. Pharm. Biopharm. 2017, 114, 288−295. (31) Adler, M.; Unger, M.; Lee, G. Surface composition of spray- dried particles of bovine serum albumin/trehalose/surfactant. Pharm.

Res. 2000, 17, 863−870. (32) Gaiani, C.; Morand, M.; Sanchez, C.; Tehrany, E. A.; Jacquot,

M.; Schuck, P.; Jeantet, R.; Scher, J. How surface composition of high milk proteins powders is influenced by spray-drying temperature.

Colloids Surf., B 2010, 75, 377−84. (33) Foerster, M.; Gengenbach, T.; Woo, M. W.; Selomulya, C. The impact of atomization on the surface composition of spray-dried milk droplets. Colloids Surf., B 2016, 140, 460−471. (34) Kim, E. H. J.; Chen, X. D.; Pearce, D. Surface composition of industrial spray-dried milk powders. 3. Changes in the surface composition during long-term storage. J. Food Eng. 2009, 94, 182−

191. (35) Kim, E. H. J.; Chen, X. D.; Pearce, D. Surface composition of industrial spray-dried milk powders. 2. Effects of spray drying conditions on the surface composition. J. Food Eng. 2009, 94, 169−

181. (36) Wilson, N. E.; Mutukuri, T. T.; Zemlyanov, D. Y.; Taylor, L. S.;

Topp, E. M.; Zhou, Q. T. Surface Composition and Formulation

Heterogeneity of Protein Solids Produced by Spray Drying. Pharm.

Res. 2020, 37, No. 14. (37) Nuzzo, M.; Millqvist-Fureby, A.; Sloth, J.; Bergenstahl, B.

Surface Composition and Morphology of Particles Dried Individually and by Spray Drying. Drying Technol. 2015, 33, 757−767. (38) Binabaji, E.; Ma, J.; Zydney, A. L. Intermolecular Interactions and the Viscosity of Highly Concentrated Monoclonal Antibody

Solutions. Pharm. Res. 2015, 32, 3102−3109. (39) Maa, Y. F.; Nguyen, P. A.; Hsu, S. W. Spray-drying of air− liquid interface sensitive recombinant human growth hormone. J.

Pharm. Sci. 1998, 87, 152−159. (40) Van der Heide, P. X-ray Photoelectron Spectroscopy: An

Introduction to Principles and Practices; John Wiley & Sons, 2011. (41) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; Springer,

2006. (42) Greenfield, N. J. Using circular dichroism spectra to estimate protein secondary structure. Nat. Protoc. 2006, 1, 2876−2890. (43) Varma, T. R.; Morsman, J. Evaluation of the use of prolution- depot (hydroxyprogesterone hexanoate) in early pregnancy. Int. J.

Gynecol. Obstet. 1982, 20, 13−17. (44) Bowen, M.; Armstrong, N.; Maa, Y. F. Investigating high- concentration monoclonal antibody powder suspension in non- aqueous suspension vehicles for subcutaneous injection. J. Pharm.

Sci. 2012, 101, 4433−4443. (45) Lee, S. L.; Hafeman, A. E.; Debenedetti, P. G.; Pethica, B. A.;

Moore, D. J. Solid-State Stabilization of α-Chymotrypsin and Catalase with Carbohydrates. Ind. Eng. Chem. Res. 2006, 45, 5134−5147. (46) Liu, J.; Nguyen, M. D.; Andya, J. D.; Shire, S. J. Reversible self- association increases the viscosity of a concentrated monoclonal antibody in aqueous solution. J. Pharm. Sci. 2005, 94, 1928−1940. (47) Harn, N.; Allan, C.; Oliver, C.; Middaugh, C. R. Highly concentrated monoclonal antibody solutions: direct analysis of physical structure and thermal stability. J. Pharm. Sci. 2007, 96,

532−546. (48) Yearley, E. J.; Zarraga, I. E.; Shire, S. J.; Scherer, T. M.; Gokarn,

Y.; Wagner, N. J.; Liu, Y. Small-angle neutron scattering character- ization of monoclonal antibody conformations and interactions at high concentrations. Biophys. J. 2013, 105, 720−731. (49) Godfrin, P. D.; Zarraga, I. E.; Zarzar, J.; Porcar, L.; Falus, P.;

Wagner, N. J.; Liu, Y. Effect of Hierarchical Cluster Formation on the

Viscosity of Concentrated Monoclonal Antibody Formulations

Studied by Neutron Scattering. J. Phys. Chem. B 2016, 120, 278−291. (50) Carpenter, J. F.; Chang, B. S.; Garzon-Rodriguez, W.;

Randolph, T. W. Rational Design of Stable Lyophilized Protein

Formulations: Theory and Practice. In Rational Design of Stable

Protein Formulations: Theory and Practice; Carpenter, J. F.; Manning,

M. C., Eds.; Springer US: Boston, MA, 2002; pp 109−133. (51) Barnett, G. V.; Qi, W.; Amin, S.; Lewis, E. N.; Razinkov, V. I.;

Kerwin, B. A.; Liu, Y.; Roberts, C. J. Structural Changes and

Aggregation Mechanisms for Anti-Streptavidin IgG1 at Elevated

Concentration. J. Phys. Chem. B 2015, 119, 15150−15163. (52) Shire, S. J.; Shahrokh, Z.; Liu, J. Challenges in the development of high protein concentration formulations. J. Pharm. Sci. 2004, 93,

1390−1402. (53) Mitropoulos, V.; Mutze, A.; Fischer, P. Mechanical properties of protein adsorption layers at the air/water and oil/water interface: a comparison in light of the thermodynamical stability of proteins. Adv.

Colloid Interface Sci. 2014, 206, 195−206. (54) Walder, R.; Schwartz, D. K. Dynamics of protein aggregation at the oil−water interface characterized by single molecule TIRF microscopy. Soft Matter 2011, 7, 7616−7622. (55) Beverung, C. J.; Radke, C. J.; Blanch, H. W. Protein adsorption at the oil/water interface: characterization of adsorption kinetics by dynamic interfacial tension measurements. Biophys. Chem. 1999, 81,

59−80.

Molecular Pharmaceutics pubs.acs.org/molecularpharmaceutics

Article https://doi.org/10.1021/acs.molpharmaceut.2c00039

Mol. Pharmaceutics 2022, 19, 1505−1514 1514