A systematic review of commercial high concentration antibody drug products approved in the US: formulation composition, dosage form design and primary packaging considerations.

Three critical aspects that define high concentration antibody products (HCAPs) are as follows: 1) formulation composition, 2) dosage form, and 3) primary packaging configuration. HCAPs have become successful in the therapeutic sector due to their unique advantage of allowing subcutaneous self-administration. Technical challenges, such as physical and chemical instability, viscosity, delivery volume limitations, and product immunogenicity, can hinder successful development and commercialization of HCAPs. Such challenges can be overcome by robust formulation and process development strategies, as well as rational selection of excipients and packaging components. We compiled and analyzed data from US Food and Drug Administration-approved and marketed HCAPs that are ≥100 mg/mL to identify trends in formulation composition and quality target product profile. This review presents our findings and discusses novel formulation and processing technologies that enable the development of improved HCAPs at ≥200 mg/mL. The observed trends can be used as a guide for further advancements in the development of HCAPs as more complex antibody-based modalities enter biologics product development.

Differential Surface Adsorption Phenomena for Conventional and Novel Surfactants Correlates with Changes in Interfacial mAb Stabilization.

Protein adsorption on surfaces can result in loss of drug product stability and efficacy during the production, storage, and administration of protein-based therapeutics. Surface-active agents (excipients) are typically added in protein formulations to prevent undesired interactions of proteins on surfaces and protein particle formation/aggregation in solution. The objective of this work is to understand the molecular-level competitive adsorption mechanism between the monoclonal antibody (mAb) and a commercially used excipient, polysorbate 80 (PS80), and a novel excipient, N-myristoyl phenylalanine-N-polyetheramine diamide (FM1000). The relative rate of adsorption of PS80 and FM1000 was studied by pendant bubble tensiometry. We find that FM1000 saturates the interface faster than PS80. Additionally, the surface-adsorbed amounts from X-ray reflectivity (XRR) measurements show that FM1000 blocks a larger percentage of interfacial area than PS80, indicating that a lower bulk FM1000 surface concentration is sufficient to prevent protein adsorption onto the air/water interface. XRR models reveal that with an increase in mAb concentration (0.5-2.5 mg/mL: IV based formulations), an increased amount of PS80 concentration (below critical micelle concentration, CMC) is required, whereas a fixed value of FM1000 concentration (above its relatively lower CMC) is sufficient to inhibit mAb adsorption, preventing mAb from co-existing with surfactants on the surface layer. With this observation, we show that the CMC of the surfactant is not the critical factor to indicate its ability to inhibit protein adsorption, especially for chemically different surfactants, PS80 and FM1000. Additionally, interface-induced aggregation studies indicate that at minimum surfactant concentration levels in protein formulations, fewer protein particles form in the presence of FM1000. Our results provide a mechanistic link between the adsorption of mAbs at the air/water interface and the aggregation induced by agitation in the presence of surfactants.

Enzymatic removal of sialic acid enables iCIEF stability monitoring of charge variants of a highly sialylated bispecific antibody.

Antibody-based therapeutic proteins have highly complex molecular structures. The final therapeutic protein product may contain a wide range of charge variants. Accurate analysis of this charge variant composition is critical to determine manufacturing process consistency and protein stability and ultimately helps to ensure that patients receive a safe and efficacious product. Here, a highly sialylated bispecific antibody (bsAb-1) challenged the ability to monitor stability by imaged capillary isoelectric focusing (iCIEF). This challenge was overcome by optimization of the iCIEF master mix buffer (adjustment of urea concentration, addition of l-arginine) and enzymatic removal of sialic acid. The method was qualified by assessing linearity, precision, LOD, LOQ, accuracy, and robustness in accordance with ICH guidance. Main species loss detectability increased up to approximately fivefold compared to the iCIEF method without desialylation when monitoring changes in stressed samples. Importantly, the results of the iCIEF method with desialylation correlated with results obtained through LC-MS tryptic peptide mapping and enabled analysis of formulation development stability samples. Finally, this analytical method shows the potential to assess low-concentration formulation development samples down to a sample concentration of 0.1 mg/ml.

Small-Scale Tools to Assess the Impact of Interfacial and Shear Stress on Biologic Drug Products.

Proper risk analysis needs to be in place to understand the susceptibility of protein to unfold and aggregate in the presence of interfacial and/or shear stress. Certain techniques, such as agitation/shaking studies, have been traditionally used to understand the impact of these stresses on the protein physical stability. However, the stresses applied in these systems are convoluted, making it difficult to define the control strategy (i.e., adjustment in process parameters to reduce foaming/bubble formation, change pump type). We have developed two small-scale tools that allow for the isolation of interfacial and shear stress, respectively. These systems, in combination with computational fluid dynamics and numerical approximations, help simulate the normal operating ranges as well as the proven acceptable ranges for different unit operations such as tangential flow filtration (TFF), mixing, and filling.

Chemical and physical instabilities in manufacturing and storage of therapeutic proteins.

Development of a robust biologic drug product is accomplished by extensive formulation and process development screening studies; however, even in the most optimal formulation, a protein can undergo spontaneous degradation during manufacture, storage, and clinical use. Chemical changes to amino acid residues, such as oxidation of methionine or tryptophan, or changes in charge such as deamidation or carbonylation, can induce conformational changes in the overall protein structure, potentially leading to changes in physical - in addition to chemical - stability. Oxidation is often caused by light exposure or the presence of metal ions or peroxides. Asparagine deamidation is more likely to occur at higher pH and/or elevated temperature. Mechanical and interfacial stresses during manufacturing can lead to physical instabilities (i.e. various forms of aggregation). A well-defined manufacturing process and effective in-process controls are essential in minimizing chemical and physical instabilities, enabling robust production and distribution of a safe and efficacious drug product. In this work, the authors provide a review of developments in these areas over the past two years, with emphasis on manufacturability of therapeutically relevant proteins and protein-based drug products.

A nondestructive and noninvasive method to determine water content in lyophilized proteins using low-field time-domain NMR.

Determining the moisture content in lyophilized solids is a fundamental step towards predicting the quality and stability of lyophilized products, but conventional methods are time-consuming, invasive, and destructive. High levels of residual moisture in a lyophilized product can lead to cake collapse, product degradation, and reduced shelf life. The aim of this study was to develop a fast, noninvasive, nondestructive, and inexpensive method for determining the moisture content in a lyophilized monoclonal antibody (mAb) formulation using benchtop low-field time-domain nuclear magnetic resonance spectroscopy.

Interfacial Stress in the Development of Biologics: Fundamental Understanding, Current Practice, and Future Perspective.

Biologic products encounter various types of interfacial stress during development, manufacturing, and clinical administration. When proteins come in contact with vapor-liquid, solid-liquid, and liquid-liquid surfaces, these interfaces can significantly impact the protein drug product quality attributes, including formation of visible particles, subvisible particles, or soluble aggregates, or changes in target protein concentration due to adsorption of the molecule to various interfaces. Protein aggregation at interfaces is often accompanied by changes in conformation, as proteins modify their higher order structure in response to interfacial stresses such as hydrophobicity, charge, and mechanical stress. Formation of aggregates may elicit immunogenicity concerns; therefore, it is important to minimize opportunities for aggregation by performing a systematic evaluation of interfacial stress throughout the product development cycle and to develop appropriate mitigation strategies. The purpose of this white paper is to provide an understanding of protein interfacial stability, explore methods to understand interfacial behavior of proteins, then describe current industry approaches to address interfacial stability concerns. Specifically, we will discuss interfacial stresses to which proteins are exposed from drug substance manufacture through clinical administration, as well as the analytical techniques used to evaluate the resulting impact on the stability of the protein. A high-level mechanistic understanding of the relationship between interfacial stress and aggregation will be introduced, as well as some novel techniques for measuring and better understanding the interfacial behavior of proteins. Finally, some best practices in the evaluation and minimization of interfacial stress will be recommended.

Pulse Proteolysis: An Orthogonal Tool for Protein Formulation Screening.

Protein formulation stability is difficult to predict a priori and generally involves long-term stability studies. It is of interest to develop an analytical method that can predict stability trends reliably. Here, pulse proteolysis was evaluated as an analytical tool to predict solution-state stability in different formulations. Four proteins formulated in different buffer and excipient compositions were subjected to urea-induced unfolding and brief enzymatic digestion ("pulse" proteolysis), and relative resistance to proteolysis was measured by microfluidics-based capillary electrophoresis-sodium dodecyl sulfate. Biophysical properties of each formulation were measured using orthogonal biophysical techniques such as differential scanning fluorimetry, differential scanning calorimetry, dynamic light scattering, circular dichroism, and fluorescence spectroscopy. Protein stability in all formulations was monitored by size exclusion chromatography on storage at 5°C and 40°C. For all 4 proteins, formulations with greater proteolytic resistance also showed higher monomer content on thermal stability. In contrast, standard biophysical techniques showed reasonable-to-no correlation with size exclusion chromatography data. The data support the use of pulse proteolysis as an orthogonal, quantitative, and predictive tool to measure protein conformational stability and rank-order formulations.

Buffer exchange path influences the stability and viscosity upon storage of a high concentration protein.

High concentration protein solutions are generally produced by spin column concentration (SCC) during early development and by tangential flow filtration (TFF) during later stages, when greater quantities of protein become available. This is based on the assumption that the protein generated by the SCC process would be fairly similar to the TFF process material. In this study, we report the case of high concentration solutions of an Fc fusion protein produced by the two processes using the same upstream drug substance (DS) with very different storage stability. The TFF and SCC batches were characterized for aggregation, viscosity, and hydrodynamic radius before and after storage at different temperatures (5°C, 25 °C, and 40 °C). Aggregation and viscosity of the solutions processed by TFF were higher than those processed by SCC upon storage at 25 °C and 40 °C for three months. Differential scanning fluorimetry (DSF) revealed differences in initial protein conformation. Upon exposure to shear stress, protein solutions showed conformational instability and increased aggregation upon storage at 35 °C. In addition, protein solution showed higher aggregation upon shearing under mixed (downstream purification process and final formulation) buffer conditions - which are more likely to be encountered during the TFF, but not SCC, process. These results were further confirmed in an independent experiment by Fourier transform-infrared (FT-IR) spectroscopy and aggregation analysis. Taken together, these data indicate that shearing the protein in intermediate, unstable buffer conditions can lead to conformational perturbation during TFF processing, which led to higher rate of aggregation and viscosity upon storage. This study highlights the importance of testing shear stress sensitivity in the transitional buffer states of the TFF process early in development to de-risk process related product instability.

(1)H, (13)C and (15)N backbone assignment of the EC-1 domain of human E-cadherin.

The Extracellular 1 (EC1) domain of E-cadherin has been shown to be important for cadherin-cadherin homophilic interactions. Cadherins are responsible for calcium-mediated cell-cell adhesion located at the adherens junction of the biological barriers (i.e., intestinal mucosa and the blood-brain barrier (BBB)). Cadherin peptides can modulate cadherin interactions to improve drug delivery through the BBB. However, the mechanism of modulating the E-cadherin interactions by cadherin peptides has not been fully elucidated. To provide a basis for subsequent examination of the structure and peptide-binding properties of the EC1 domain of human E-cadherin using solution NMR spectroscopy, the (1)H, (13)C and (15)N backbone resonance of the uniformly labeled-EC1 were assigned and the secondary structure was determined based on the chemical shift values. These resonance assignments are essential for assessing protein-ligand interactions and are reported here.

claMP Tag: a versatile inline metal-binding platform based on the metal abstraction peptide.

Molecularly targeted research and diagnostic tools are essential to advancing understanding and detection of many diseases. Metals often impart the desired functionality to these tools, and conjugation of high-affinity chelators to proteins is carried out to enable targeted delivery of the metal. This approach has been much more effective with large lanthanide series metals than smaller transition metals. Because chemical conjugation requires additional processing and purification steps and yields a heterogeneous mixture of products, inline incorporation of a peptide tag capable of metal binding is a highly preferable alternative. Development of a transition metal binding tag would provide opportunity to greatly expand metal-based analyses. The metal abstraction peptide (MAP) sequence was genetically engineered into recombinant protein to generate the claMP Tag. The effects of this tag on recombinant epidermal growth factor (EGF) protein expression, disulfide bond formation, tertiary structural integrity, and transition metal incorporation using nickel were examined to confirm the viability of utilizing the MAP sequence to generate linker-less metal conjugates.

Embedding the Ni-SOD mimetic Ni-NCC within a polypeptide sequence alters the specificity of the reaction pathway.

The unique metal abstracting peptide asparagine-cysteine-cysteine (NCC) binds nickel in a square planar 2N:2S geometry and acts as a mimic of the enzyme nickel superoxide dismutase (Ni-SOD). The Ni-NCC tripeptide complex undergoes rapid, site-specific chiral inversion to dld-NCC in the presence of oxygen. Superoxide scavenging activity increases proportionally with the degree of chiral inversion. Characterization of the NCC sequence within longer peptides with absorption, circular dichroism (CD), and magnetic CD (MCD) spectroscopies and mass spectrometry (MS) shows that the geometry of metal coordination is maintained, though the electronic properties of the complex are varied to a small extent because of bis-amide, rather than amine/amide, coordination. In addition, both Ni-tripeptide and Ni-pentapeptide complexes have charges of -2. This study demonstrates that the chiral inversion chemistry does not occur when NCC is embedded in a longer polypeptide sequence. Nonetheless, the superoxide scavenging reactivity of the embedded Ni-NCC module is similar to that of the chirally inverted tripeptide complex, which is consistent with a minor change in the reduction potential for the Ni-pentapeptide complex. Together, this suggests that the charge of the complex could affect the SOD activity as much as a change in the primary coordination sphere. In Ni-NCC and other Ni-SOD mimics, changes in chirality, superoxide scavenging activity, and oxidation of the peptide itself all depend on the presence of dioxygen or its reduced derivatives (e.g., superoxide), and the extent to which each of these distinct reactions occurs is ruled by electronic and steric effects that emenate from the organization of ligands around the metal center.

Controlling the chiral inversion reaction of the metallopeptide Ni-asparagine-cysteine-cysteine with dioxygen.

Synthetically generated metallopeptides have the potential to serve a variety of roles in biotechnology applications, but the use of such systems is often hampered by the inability to control secondary reactions. We have previously reported that the Ni(II) complex of the tripeptide LLL-asparagine-cysteine-cysteine, LLL-Ni(II)-NCC, undergoes metal-facilitated chiral inversion to dld-Ni(II)-NCC, which increases the observed superoxide scavenging activity. However, the mechanism for this process remained unexplored. Electronic absorption and circular dichroism studies of the chiral inversion reaction of Ni(II)-NCC reveal a unique dependence on dioxygen. Specifically, in the absence of dioxygen, the chiral inversion is not observed, even at elevated pH, whereas the addition of O(2) initiates this reactivity and concomitantly generates superoxide. Scavenging experiments using acetaldehyde are indicative of the formation of carbanion intermediates, demonstrating that inversion takes place by deprotonation of the alpha carbons of Asn1 and Cys3. Together, these data are consistent with the chiral inversion being dependent on the formation of a Ni(III)-NCC intermediate from Ni(II)-NCC and O(2). The data further suggest that the anionic thiolate and amide ligands in Ni(II)-NCC inhibit Cα-H deprotonation for the Ni(II) oxidation state, leading to a stable complex in the absence of O(2). Together, these results offer insights into the factors controlling reactivity in synthetic metallopeptides.

Mapping site-specific changes that affect stability of the N-terminal domain of calmodulin.

Biophysical tools have been invaluable in formulating therapeutic proteins. These tools characterize protein stability rapidly in a variety of solution conditions, but in general, the techniques lack the ability to discern site-specific information to probe how solution environment acts to stabilize or destabilize the protein. NMR spectroscopy can provide site-specific information about subtle structural changes of a protein under different conditions, enabling one to assess the mechanism of protein stabilization. In this study, NMR was employed to detect structural perturbations at individual residues as a result of altering pH and ionic strength. The N-terminal domain of calmodulin (N-CaM) was used as a model system, and the ¹H-¹5N heteronuclear single quantum coherence (HSQC) experiment was used to investigate effects of pH and ionic strength on individual residues. NMR analysis revealed that different solution conditions affect individual residues differently, even when the amino acid sequence and structure are highly similar. This study shows that addition of NMR to the formulation toolbox has the ability to extend understanding of the relationship between site-specific changes and overall protein stability.

MAPping the chiral inversion and structural transformation of a metal-tripeptide complex having ni-superoxide dismutase activity.

The metal abstraction peptide (MAP) tag is a tripeptide sequence capable of abstracting a metal ion from a chelator and binding it with extremely high affinity at neutral pH. Initial studies on the nickel-bound form of the complex demonstrate that the tripeptide asparagine-cysteine-cysteine (NCC) binds metal with 2N:2S, square planar geometry and behaves as both a structural and functional mimic of Ni superoxide dismutase (Ni-SOD). Electronic absorption, circular dichroism (CD), and magnetic CD (MCD) data collected for Ni-NCC are consistent with a diamagnetic Ni(II) center. It is apparent from the CD signal of Ni-NCC that the optical activity of the complex changes over time. Mass spectrometry data show that the mass of the complex is unchanged. Combined with the CD data, this suggests that chiral rearrangement of the complex occurs. Following incubation of the nickel-containing peptide in D(2)O and back-exchange into H(2)O, incorporation of deuterium into non-exchangeable positions is observed, indicating chiral inversion occurs at two of the α carbon atoms in the peptide. Control peptides were used to further characterize the chirality of the final nickel-peptide complex, and density functional theory (DFT) calculations were performed to validate the hypothesized position of the chiral inversions. In total, these data indicate Ni-SOD activity is increased proportionally to the degree of structural change in the complex over time. Specifically, the relationship between the change in CD signal and change in SOD activity is linear.

Novel tripeptide model of nickel superoxide dismutase.

Nickel superoxide dismutase (Ni-SOD) catalyzes the disproportionation of superoxide to molecular oxygen and hydrogen peroxide, but the overall reaction mechanism has yet to be determined. Peptide-based models of the 2N:2S nickel coordination sphere of Ni-SOD have provided some insight into the mechanism of this enzyme. Here we show that the coordination sphere of Ni-SOD can be mimicked using the tripeptide asparagine-cysteine-cysteine (NCC). NCC binds nickel with extremely high affinity at physiological pH with 2N:2S geometry, as demonstrated by electronic absorption and circular dichroism (CD) data. Like Ni-SOD, Ni-NCC has mixed amine/amide ligation that favors metal-based oxidation over ligand-based oxidation. Electronic absorption, CD, and magnetic CD (MCD) data collected for Ni-NCC are consistent with a diamagnetic Ni(II) center bound in square-planar geometry. Ni-NCC is quasi-reversibly oxidized with a midpoint potential of 0.72(2) V (vs Ag/AgCl) and breaks down superoxide in an enzyme-based assay, supporting its potential use as a model for Ni-SOD chemistry.