Surface Interactions of Monoclonal Antibodies Characterized by Quartz Crystal Microbalance with Dissipation: Impact of Hydrophobicity and Protein Self-Interactions

Quartz Crystal Microbalance as a Predictive Tool for Drug-Material of Construction Interactions in Intravenous Protein Drug Administration

As a growing number of protein drug products are developed, formulation characterization is becoming important. An IgG drug product is tested at concentrations from 0.0001-0.1 mg/mL for adsorption behavior to polymer surfaces polyvinyl chloride (PVC) and polypropylene (PP) upon dilution in normal saline (NS) using quartz crystal microbalance with dissipation (QCM-D). The studies mimicked IgG antibody interaction during IV administration with polymeric surfaces within syringes, lines, and bags. Drug product was characterized with excipients, with focus on surfactant. Drug solutions were run over polymer-coated sensors to measure the adsorption behavior of the formulation with emphasis on the behavior of each of the formulation's components. Over 60 sensorgram data sets were correlated with assayed protein solution concentrations in mock NS-diluted infusions of drug product in the equivalent concentrations to QCM experiments to build a preliminary predictive model for determining fraction of drug and surfactant adsorbed and lost at the hydrophobic surface during administration. These results create a method for reliably and predictively estimating drug product adsorption behavior and protein drug dose loss on polymers at different protein drug concentrations.

Application of QCM in Peptide and Protein-Based Drug Product Development

AT-cut quartz crystals vibrating in the thickness-shear mode (TSM), especially quartz crystal resonators (QCRs), are well known as very efficient mass sensitive systems because of their sensitivity, accuracy, and biofunctionalization capacity. They are highly reliable in the measurement of the mass of deposited samples, in both gas and liquid matrices. Moreover, they offer real-time monitoring, as well as relatively low production and operation costs. These features make mass sensitive systems applicable in a wide range of different applications, including studies on protein and peptide primary packaging, formulation, and drug product manufacturing process development. This review summarizes the information on some particular implementations of quartz crystal microbalance (QCM) instruments in protein and peptide drug product development as well as their future prospects.

Interfacial Adsorption of a Monoclonal Antibody and Its Fab and Fc Fragments at the Oil/Water Interface

The physical stability of a monoclonal antibody (mAb) solution for injection in a prefilled syringe may in part depend on its behavior at the silicone oil/water interface. Here, the adsorption of a mAb (termed COE-3) and its fragment antigen-binding (Fab) and crystallizable (Fc) at the oil/water interface was measured using neutron reflection. A 1.4 ± 0.1 μm hexadecane oil film was formed on a sapphire block by a spin-freeze-thaw process, retaining its integrity upon contact with the protein solutions. Measurements revealed that adsorbed COE-3 and its Fab and Fc fragments retained their globular structure, forming layers that did not penetrate substantially into the oil phase. COE-3 and Fc were found to adsorb flat-on to the interface, with denser 45 and 42 Å inner layers, respectively, in contact with the oil and a more diffuse 17-21 Å outer layer caused by fragments adsorbing in a tilted manner. In contrast, Fab fragments formed a uniform 60 Å monolayer. Monolayers were formed under all conditions studied (10-200 ppm, using three isotopic contrasts), although changes in packing density across the COE-3 and Fc layers were observed. COE-3 had a higher affinity to the interface than either of its constituent fragments, while Fab had a lower interfacial affinity consistent with its higher net surface charge. This study extends the application of high-resolution neutron reflection measurements to the study of protein adsorption at the oil/water interface using an experimental setup mimicking the protein drug product in a siliconized prefilled syringe.

Transparent, Hydrophobic Fluorinated Ethylene Propylene Offers Rapid, Robust, and Irreversible Passive Adsorption of Diagnostic Antibodies for Sensitive Optical Biosensing

Current literature data is scarce and somehow contradictory in respect to the suitability of "nonstick" fluoropolymer surfaces for immobilization of biomolecules. We have previously shown empirically that transparent Teflon fluorinated ethylene propylene (FEP) offers rapid and sensitive optical biosensing of clinically relevant biomarkers. This study shows for the first time a comprehensive experimental analysis of passive adsorption of diagnostic IgG antibodies on actual Teflon FEP microfluidic strips. Full equilibrium isotherms and kinetics for passive adsorption were studied and modeled employing a protein titration method using hundreds of multibore microfluidic strips for a range of temperatures, pH, ionic strengths, and inner diameters, using both polyclonal and monoclonal antibody systems. Results were benchmarked against other plastic hydrophobic and glass hydrophilic capillary surfaces. For the first time, it was shown quantitatively that the hydrophobicity of fluoropolymer surfaces encourages the passive adsorption of diagnostic antibodies for biosensing and is insensitive to the temperature of incubation and to ionic buffer strength. The mass of captured antigen increased with increasing antibody surface coverage up to ∼400 ng/cm², with an optimal adsorbed antibody activity for 45-69% of full monolayer coverage, matching results of other biosensing surfaces. The equilibrium was reached fast, within 5-10 min, and surprisingly both the kinetics and equilibrium of antibody adsorption were dependent on the inner diameter of microcapillaries. This is a novel and relevant result that will generally impact on the design of miniaturized microfluidic biosensing devices. The antibody surface densities obtained with hydrophobic plastic surfaces were 2- to 4-fold lower than for a hydrophilic, glass surface, however the former presented a monolayered adsorption with a higher level of irreversibility, as shown by the adsorption and desorption rates around 1 order of magnitude smaller than for glass, which is highly desirable for biosensing with surface-coated biomolecules.

Coadsorption of a Monoclonal Antibody and Nonionic Surfactant at the SiO2/Water Interface

During the formulation of therapeutic monoclonal antibodies (mAbs), nonionic surfactants are commonly added to attenuate structural rearrangement caused by adsorption/desorption at interfaces during processing, shipping, and storage. We examined the adsorption of a mAb (COE-3) at the SiO₂/water interface in the presence of pentaethylene glycol monododecyl ether (C₁₂E₅), polysorbate 80 (PS80-20EO), and a polysorbate 80 analogue with seven ethoxylates (PS80-7EO). Spectroscopic ellipsometry was used to follow COE-3 dynamic adsorption, and neutron reflection was used to determine interfacial structure and composition. Neither PS80-20EO nor C₁₂E₅ had a notable affinity for COE-3 or the interface under the conditions studied and thus did not prevent COE-3 adsorption. In contrast, PS80-7EO did coadsorb but did not influence the dynamic process or the equilibrated amount of absorbed COE-3. Near equilibration, COE-3 underwent structural rearrangement and PS80-7EO started to bind the COE-3 interfacial layer and subsequently formed a well-defined surfactant bilayer via self-assembly. The resultant interfacial layer comprised an inner mAb layer of about 70 Å thickness and an outer surfactant layer of a further 70 Å, with distinct transitional regions across the mAb-surfactant and surfactant-bulk water boundaries. Once formed, such interfacial layers were very robust and worked to prevent further mAb adsorption, desorption, and structural rearrangement. Such robust interfacial layers could be anticipated to exist for formulated mAbs stored in type II glass vials; further research is required to understand the behavior of these layers for siliconized glass syringes.

Interfacial Adsorption of Monoclonal Antibody COE-3 at the Solid/Water Interface

Spectroscopic ellipsometry (SE) and neutron reflection (NR) data for the adsorption of a monoclonal antibody (mAb, termed COE-3, pI 8.44) at the bare SiO₂/water interface are compared here to the simulations based on Derjaguin-Landau-Verwey-Overbeek theory. COE-3 adsorption was characterized by an initial rapid increase in the surface-adsorbed amount (Γ) followed by a plateau. Only the initial rate of the increase in Γ was strongly correlated with the bulk concentration (0.002-0.2 mg/mL), with Γ at the plateau being about 2.2 mg/m² (pH 5.5). Simulations captured COE-3 adsorption at equilibrium most accurately, the point at which the outgoing flux of molecules within the adsorbed plane matched the adsorption flux. Increasing the buffer pH from 5.5 to 9 increased Γ at equilibrium to ∼3 mg/m² (0.02 mg/mL COE-3), revealing a dominant role for lateral repulsion between adsorbed mAb molecules. In contrast, increasing the buffer ionic strength (pH 6) reduced Γ, which was captured by simulations accounting for electrostatic screening by ions, in addition to mAb/SiO₂ attractive forces and lateral repulsion. NR data at the same bulk concentrations corroborated the SE data, albeit with slightly higher Γ due to longer adsorption times for data acquisition; for example, at pH 9, Γ was 3.6 mg/m² (0.02 mg/mL COE-3), equivalent to a relatively high volume fraction of 0.5. An adsorbed monolayer with a thickness of 50-52 Å was consistently determined by NR, corresponding to the short axial lengths of fragment antigen-binding and fragment crystallization and implying minimal structural perturbation. Thus, the simulations enabled a mechanistic interpretation of the experimental data of mAb adsorption at the SiO₂/water interface.

Quartz crystal microbalance as an assay to detect anti-drug antibodies for the immunogenicity assessment of therapeutic biologics

Because of their biological origins, therapeutic biologics can trigger an unwanted deleterious immune response with some patients. The immunogenicity of therapeutic biologics can affect drug efficacy and patient safety by the production of circulating anti-drug antibodies (ADA). In this study, quartz crystal microbalance (QCM) was developed as an assay to detect ADA. Etanercept (Enbrel®) was covalently grafted to dextran-modified QCM surfaces. Rabbits were immunized with etanercept to generate ADA. Results showed the QCM assay could detect purified ADA from rabbits at concentrations as low as 50 ng/mL, within the sensitivity range of ELISA. The QCM assay could also assess the ADA isotype. It was shown that the ADA were composed of the IgG isotype, but not IgM, as expected. Furthermore, it was shown that QCM surfaces that had been used to detect ADA could be regenerated in glycine-HCl solution and reused. The QCM assay was also demonstrated to detect ADA in crude serum samples. Serum was collected from the rabbits and analyzed before and after etanercept immunization. ADA were clearly detected in serum from rabbits after immunization, but not in serum before immunization. Serum from patients administered with etanercept for rheumatoid arthritis (RA) treatment was also analyzed and compared to serum from healthy donors. Sera from 10 RA patients were analyzed. Results showed one of the RA patient serum samples may have ADA present. In conclusion, QCM appears to be a viable assay to detect ADA for the immunogenicity assessment of therapeutic biologics.

Neutron Reflection Study of Surface Adsorption of Fc, Fab, and the Whole mAb

Characterizing the influence of fragment crystallization (Fc) and antigen-binding fragment (Fab) on monoclonal antibody (mAb) adsorption at the air/water interface is an important step to understanding liquid mAb drug product stability during manufacture, shipping, and storage. Here, neutron reflection is used to study the air/water adsorption of a mAb and its Fc and Fab fragments. By varying the isotopic contrast, the adsorbed amount, thickness, orientation, and immersion of the adsorbed layers could be determined unambiguously. While Fc adsorption reached saturation within the hour, its surface adsorbed amount showed little variation with bulk concentration. In contrast, Fab adsorption was slower and the adsorbed amount was concentration dependent. The much higher Fc adsorption, as compared to Fab, was linked to its lower surface charge. Time and concentration dependence of mAb adsorption was dominated by Fab behavior, although both Fab and Fc behaviors contributed to the amount of mAb adsorbed. Changing the pH from 5.5 to 8.8 did not much perturb the adsorbed amount of Fc, Fab, or mAb. However, a small decrease in adsorption was observed for the Fc over pH 8-8.8 and vice versa for the Fab and mAb, consistent with a dominant Fab behavior. As bulk concentration increased from 5 to 50 ppm, the thicknesses of the Fc layers were almost constant at 40 Å, while Fab and mAb layers increased from 45 to 50 Å. These results imply that the adsorbed mAb, Fc, and Fab all retained their globular structures and were oriented with their short axial lengths perpendicular to the interface.

Single-molecule resolution of protein dynamics on polymeric membrane surfaces: the roles of spatial and population heterogeneity

Although polymeric membranes are widely used in the purification of protein pharmaceuticals, interactions between biomolecules and membrane surfaces can lead to reduced membrane performance and damage to the product. In this study, single-molecule fluorescence microscopy provided direct observation of bovine serum albumin (BSA) and human monoclonal antibody (IgG) dynamics at the interface between aqueous buffer and polymeric membrane materials including regenerated cellulose and unmodified poly(ether sulfone) (PES) blended with either polyvinylpyrrolidone (PVP), polyvinyl acetate-co-polyvinylpyrrolidone (PVAc-PVP), or polyethylene glycol methacrylate (PEGM) before casting. These polymer surfaces were compared with model surfaces composed of hydrophilic bare fused silica and hydrophobic trimethylsilane-coated fused silica. At extremely dilute protein concentrations (10(-3)-10(-7) mg/mL), protein surface exchange was highly dynamic with protein monomers desorbing from the surface within ∼1 s after adsorption. Protein oligomers (e.g., nonspecific dimers, trimers, or larger aggregates), although less common, remained on the surface for 5 times longer than monomers. Using newly developed super-resolution methods, we could localize adsorption sites with ∼50 nm resolution and quantify the spatial heterogeneity of the various surfaces. On a small anomalous subset of the adsorption sites, proteins adsorbed preferentially and tended to reside for significantly longer times (i.e., on "strong" sites). Proteins resided for shorter times overall on surfaces that were more homogeneous and exhibited fewer strong sites (e.g., PVAc-PVP/PES). We propose that strong surface sites may nucleate protein aggregation, initiated preferentially by protein oligomers, and accelerate ultrafiltration membrane fouling. At high protein concentrations (0.3-1.0 mg/mL), fewer strong adsorption sites were observed, and surface residence times were reduced. This suggests that at high concentrations, adsorbed proteins block strong sites from further protein adsorption. Importantly, this demonstrates that strong binding sites can be modified by changing solution conditions. Membrane surfaces are intrinsically heterogeneous; by employing single-molecule techniques, we have provided a new framework for understanding protein interactions with such surfaces.

Quartz crystal microbalance study of ionic strength and pH-dependent polymer conformation and protein adsorption/desorption on PAA, PEO, and mixed PEO/PAA brushes

The conformation of polymer chains grafted on a substrate influences protein adsorption. In a previous study, adsorption/desorption of albumin was demonstrated on mixed poly(ethylene oxide) (PEO)/poly(acrylic acid) (PAA) brushes, triggered by solutions of adequate pH and ionic strength (I). In the present work, homolayers of PEO or PAA are submitted to saline solutions with pH from 3 to 9 and I from 10(-5) to 10(-1) M, and their conformation is evaluated in real time using quartz crystal microbalance with dissipation monitoring (QCM-D). Shrinkage/swelling of PAA chains and hydration and salt condensation in the brush are evidenced. The adsorption of human serum albumin (HSA) onto such brushes is also monitored in these different saline solutions, leading to a deep understanding of the influence of polymer chain conformation, modulated by pH and I, on protein adsorption. A detailed model of the conformation of PEO/PAA mixed brushes depending on pH and I is then proposed, providing a rationale for the identification of conditions for the successive adsorption and desorption of proteins on such mixed brushes. The adsorption/desorption of albumin on PEO/PAA is demonstrated using QCM-D.

Rapid analysis of antibody self-association in complex mixtures using immunogold conjugates

A key challenge in developing therapeutic antibodies is their highly variable propensities to self-associate at high antibody concentrations (>50 mg/mL) required for subcutaneous delivery. Identification of monoclonal antibodies (mAbs) in the initial discovery process that not only have high binding affinity but also have high solubility and low viscosity would simplify the development of safe and effective antibody therapeutics. Unfortunately, the low purities, small quantities and large numbers of antibody candidates during the early discovery process are incompatible with current methods of measuring antibody self-association. We report a method (affinity-capture self-interaction nanoparticle spectroscopy, AC-SINS) capable of identifying mAbs with low self-association propensity that is robust even at low mAb concentrations (5-50 μg/mL) and in the presence of cell culture media. Gold nanoparticles are coated with polyclonal antibodies specific for human antibodies, and then human mAbs are captured from dilute antibody solutions. We find that the wavelength of maximum absorbance (plasmon wavelength) of antibody-gold conjugates--which red-shifts as the distance between particles is reduced due to attractive mAb self-interactions--is well correlated with light scattering measurements conducted at several orders of magnitude higher antibody concentrations. The generality of AC-SINS makes it well suited for use in diverse settings ranging from antibody discovery to formulation development.

Surface and adsorption characteristics of three elastin-like polypeptide coatings with varying sequence lengths

The surface properties of a family of elastin-like polypeptides (ELPs), differing in molecular weight and sequence length, were investigated to understand how the nature of the polypeptide film might contribute to their thrombogenic profile. Physical adsorption of the ELPs onto Mylar increased surface wettability as the sequence length decreased while X-ray spectroscopy analysis showed an increasing amide content with sequence length. Chemical force microscopy analysis revealed that the ELP-coated surfaces displayed purely hydrophilic adhesion forces that increased as the ELP sequence length decreased. Adsorption isotherms performed using the quartz crystal microbalance with dissipation, showed that the surface coverage increased with ELP sequence length. The longer polypeptides (ELP2 and ELP4) also displayed higher specific dissipation values indicating that they established films with greater structural flexibility and associated water content than the shorter polypeptide, ELP1. Additionally, the stability of the ELP coating was lower with the shorter polypeptides. This study highlights the different surface properties of the ELP coatings as well as the dynamic nature of the ELP adsorbed layer wherein the conformational state may be an important factor contributing to their blood response.

Adsorption behavior of a human monoclonal antibody at hydrophilic and hydrophobic surfaces

One aspiration for the formulation of human monoclonal antibodies (mAb) is to reach high solution concentrations without compromising stability. Protein surface activity leading to instability is well known, but our understanding of mAb adsorption to the solid-liquid interface in relevant pH and surfactant conditions is incomplete. To investigate these conditions, we used total internal reflection fluorescence (TIRF) and neutron reflectometry (NR). The mAb tested ("mAb-1") showed highest surface loading to silica at pH 7.4 (~12 mg/m(2)), with lower surface loading at pH 5.5 (~5.5 mg/m(2), further from its pI of 8.99) and to hydrophobized silica (~2 mg/m(2)). The extent of desorption of mAb-1 from silica or hydrophobized silica was related to the relative affinity of polysorbate 20 or 80 for the same surface. mAb-1 adsorbed to silica on co-injection with polysorbate (above its critical micelle concentration) and also to silica pre-coated with polysorbate. A bilayer model was developed from NR data for mAb-1 at concentrations of 50-5000 mg/L, pH 5.5, and 50-2000 mg/L, pH 7.4. The inner mAb-1 layer was adsorbed to the SiO₂ surface at near saturation with an end-on" orientation, while the outer mAb-1 layer was sparse and molecules had a "side-on" orientation. A non-uniform triple layer was observed at 5000 mg/L, pH 7.4, suggesting mAb-1 adsorbed to the SiO₂ surface as oligomers at this concentration and pH. mAb-1 adsorbed as a sparse monolayer to hydrophobized silica, with a layer thickness increasing with bulk concentration - suggesting a near end-on orientation without observable relaxation-unfolding.