Characterizing and understanding the formation of cysteine conjugates and other by-products in a random, lysine-linked antibody drug conjugate.

This study describes the characterization of conjugation sites for a random, lysine conjugated 2-iminothiolane (2-IT) based antibody-drug-conjugate synthesized from an IgG1 antibody and a duocarmycin analog-based payload-linker. Of the 80 putative lysine sites, 78 were found to be conjugated via tryptic peptide mapping and LC-HRMS. Surprisingly, seven cysteine-linked conjugated peptides were also detected resulting from the conjugation of cysteine residues derived from the four inter-chain disulfide bonds during the reaction. This unexpected finding could be attributed to the free thiols of the 2-IT thiolated antibody intermediates and/or the 4-mercaptobutanamide by-product resulting from the hydrolysis of 2-IT. These free thiols could cause the four inter-chain disulfide bonds of the antibody to scramble via intra- or inter-molecular attack. The presence of only pair of non-reactive (unconjugated) lysine residues, along with the four intact intra-chain disulfide bonds, is attributed to their poor accessibility, which is consistent with solvent accessibility modeling analysis. We also discovered a major by-product derived from the hydrolysis of the amidine moiety of the N-terminus conjugate. In contrast, the amidine moiety in lysine-linked conjugates appeared stable. Based on our results, we propose plausible formation mechanisms of cysteine-linked conjugates and the hydrolysis of the N-terminus conjugate, which provide scientific insights that are beneficial to process development and drug quality control.

Investigation of anomalous charge variant profile reveals discrete pH-dependent conformations and conformation-dependent charge states within the CDR3 loop of a therapeutic mAb.

During the development of a therapeutic monoclonal antibody (mAb-1), the charge variant profile obtained by pH-gradient cation exchange chromatography (CEX) contained two main peaks, each of which exhibited a unique intrinsic fluorescence profile and demonstrated inter-convertibility upon reinjection of isolated peak fractions. Domain analysis of mAb-1 by CEX and liquid chromatography-mass spectrometry indicated that the antigen-binding fragment chromatographed as two separate peaks that had identical mass. Surface plasmon resonance binding analysis to antigen demonstrated comparable kinetics/affinity between these fractionated peaks and unfractionated starting material. Subsequent molecular modeling studies revealed that the relatively long and flexible complementarity-determining region 3 (CDR3) loop on the heavy chain could adopt two discrete pH-dependent conformations: an "open" conformation at neutral pH where the HC-CDR3 is largely solvent exposed, and a "closed" conformation at lower pH where the solvent exposure of a neighboring tryptophan in the light chain is reduced and two aspartic acid residues near the ends of the HC-CDR3 loop have atypical pKa values. The pH-dependent equilibrium between "open" and "closed" conformations of the HC-CDR3, and its proposed role in the anomalous charge variant profile of mAb-1, were supported by further CEX and hydrophobic interaction chromatography studies. This work is an example of how pH-dependent conformational changes and conformation-dependent changes to net charge can unexpectedly contribute to perceived instability and require thorough analytical, biophysical, and functional characterization during biopharmaceutical drug product development.

Cyclization of N-Terminal Glutamic Acid to pyro-Glutamic Acid Impacts Monoclonal Antibody Charge Heterogeneity Despite Its Appearance as a Neutral Transformation.

Pyroglutamic acid (pyroGlu) is commonly observed at the N-terminus of therapeutic monoclonal antibodies. Notably, the term "pyroGlu" refers to a single product that could originate from the cyclization of either an N-terminal glutamine or an N-terminal glutamic acid. This is an important and easily overlooked distinction that has major implications on the charge variant nature of a pyroGlu relative to its uncyclized form. Cyclization of an N-terminal glutamine for instance clearly produces an acidic variant with a lower isoelectric point owing to the loss of the positively charged N-terminal amine. In this report, we demonstrate that cyclization of an N-terminal glutamic acid on the other hand produces a basic variant with a higher isoelectric point contrary to the typical assumption that the simultaneous loss of the N-terminal amine and the carboxylic acid side-chain would negate the formation of a charge variant. The results of our investigation demonstrate the need to consider the relative strengths of the acidic and basic functional groups which are altered when assessing whether the product will be a charge variant. This study also adds new knowledge and experimental evidence to understand charge heterogeneity in monoclonal antibodies.

Comparison of imaged capillary isoelectric focusing and cation exchange chromatography for monitoring dextrose-mediated glycation of monoclonal antibodies in infusion solutions.

Intravenous (IV) infusion of therapeutic proteins typically involves dilution of the formulated product into infusion media such as normal saline or dextrose, 5% m/v in water. We report results from a rigorous evaluation of imaged capillary isoelectric focusing (iCIEF) for monitoring dextrose-mediated glycation of proteins in IV infusion solutions. In addition to detecting stable Amadori glycation products, iCIEF was able to detect the labile Schiff base (SB) glycation adducts since the equilibrium with free dextrose is maintained on capillary. Method parameters such as sample dilution factor and ampholyte composition (but not urea) were found to influence the observed level of SB glycation adducts. The impacts of dextrose and urea on the apparent pI values are also reported. iCIEF results were compared with results from cation exchange chromatography, which was found to preferentially detect the more stable Amadori glycation products due to the on-column decomposition of the SB adducts resulting from the separation of the protein from free dextrose which in turn altered the SB adduct- free dextrose equilibrium. These results demonstrate the need for careful consideration when selecting the analytical methodology to investigate protein sensitivity to dextrose and to monitor protein stability in dextrose-containing infusion solutions.

Improving Mass Spectral Quality of Monoclonal Antibody Middle-Up LC-MS Analysis by Shifting the Protein Charge State Distribution.

Monoclonal antibodies (mAbs) are experiencing accelerated development in the pharmaceutical industry. Utilization of middle-up LC-MS methodology can provide detailed characterization of mAbs via reduction and/or enzymatic cleavage of the mAb into smaller protein fragments. However, under typical LC-MS conditions, these fragments, especially the more heterogeneous heavy chain, can present charge state distributions (CSD) featuring a severe interference in the low mass-to-charge (m/z) region in the mass spectrum, adversely impacting spectral quality of these proteins and ultimately the deconvoluted mass spectrum. Here, we introduce a novel method to characterize protein fragments by partially reducing mAbs and using acidic mobile phases (MPs) with a trace amount of base additive. Gas-phase charge stripping occurs with the basic MP additive, causing the CSD to shift to a higher m/z region resulting in high-quality mass spectra with enhanced resolution of protein charge states. Subsequently, high-quality deconvoluted spectra and accurate mass measurement of the fragments are achieved. This method has been applied to the intact mass measurement of mAbs and antibody drug conjugates.

Drug-to-antibody determination for an antibody-drug-conjugate utilizing cathepsin B digestion coupled with reversed-phase high-pressure liquid chromatography analysis.

Antibody drug conjugates or ADCs are currently being evaluated for their effectiveness as targeted chemotherapeutic agents across the pharmaceutical industry. Due to the complexity arising from the choice of antibody, drug and linker; analytical methods for release and stability testing are required to provide a detailed understanding of both the antibody and the drug during manufacturing and storage. The ADC analyzed in this work consists of a tubulysin drug analogue that is randomly conjugated to lysine residues in a human IgG1 antibody. The drug is attached to the lysine residue through a peptidic, hydrolytically stable, cathepsin B cleavable linker. The random lysine conjugation produces a heterogeneous mixture of conjugated species with a variable drug-to-antibody ratio (DAR), therefore, the average amount of drug attached to the antibody is a critical parameter that needs to be monitored. In this work we have developed a universal method for determining DAR in ADCs that employ a cathepsin B cleavable linker. The ADC is first cleaved at the hinge region and then mildly reduced prior to treatment with the cathepsin B enzyme to release the drug from the antibody fragments. This pre-treatment allows the cathepsin B enzyme unrestricted access to the cleavage sites and ensures optimal conditions for the cathepsin B to cleave all the drug from the ADC molecule. The cleaved drug is then separated from the protein components by reversed phase high performance liquid chromatography (RP-HPLC) and quantitated using UV absorbance. This method affords superior cleavage efficiency to other methods that only employ a cathepsin digestion step as confirmed by mass spectrometry analysis. This method was shown to be accurate and precise for the quantitation of the DAR for two different random lysine conjugated ADC molecules.

Investigating the Degradation Behaviors of a Therapeutic Monoclonal Antibody Associated with pH and Buffer Species.

This study aimed in understanding the degradation behaviors of an IgG 1 subtype therapeutic monoclonal antibody A (mAb-A) associated with pH and buffer species. The information obtained in this study can augment conventional, stability-based screening paradigms by providing the direction necessary for efficient experimental design. Differential scanning calorimetry (DSC) was used for studying conformational stability. Dynamic light scattering (DLS) was utilized to generate B 22*, a modified second virial coefficient for the character of protein-protein interaction. Size-exclusion chromatography (SEC) and hydrophobic interaction chromatography (HIC) were employed to separate degradation products. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was used for determining the molecular size and liquid chromatography mass spectrometry (LC-MS) were used for identifying the sequence of the separated fragments. The results showed that both pH and buffer species played the roles in controlling the degradation behaviors of mAb-A, but the pH was more significant. In particular, pH 4.5 induced additional thermal transition peaks occurring at a low temperature compared with pH 6.5. A continual temperature-stress study illustrated that the additional thermal transition peaks related to the least stable structure and a greater fragmentation. Although mAb-A showed the comparable conformational structures and an identical amount of aggregates at time zero between the different types of buffer species at pH 6.5, the aggregation formation rate showed a buffer species-dependent discrepancy over a temperature-stress period. It was found that the levels of aggregations associated with the magnitudes of protein-protein interaction forces.

Demonstrating ß-glucan and yeast peptide clearance in biopharmaceutical downstream processes.

The use of yeast- and plant-derived hydrolysates in cell culture production processes has sparked concerns over the potential immunogenicity risk posed by ß-glucans and yeast peptides contained in these raw materials. This article utilizes a combination of in-process testing from large-scale manufacturing and scale-down spiking studies to demonstrate the clearance of ß-glucans and yeast peptides through chromatographic steps in the downstream purification process for a monoclonal antibody. ß-Glucans were found to flow through most all three modes of chromatography (Protein A, cation and anion exchange) without binding to the resins or the product. Protein A affinity chromatography was found to provide the best clearance factor. The efficacy of the resin sanitization and storage procedures to prevent carryover from one run to the next was also demonstrated. Yeast peptides were found to be metabolized during the cell culture process and were undetectable after the Protein A purification step. The data presented here serve to allay concerns about the use of hydrolysates in cell culture production. The methodology presented here provides a template to demonstrate clearance of ß-glucans and yeast peptides through chromatographic steps in downstream processing.

A simple reversed phase high-performance liquid chromatography method for polysorbate 80 quantitation in monoclonal antibody drug products.

In this paper, we discuss an improved high-performance liquid chromatography (HPLC) method for the quantitation of polysorbate 80 (polyoxyethylenesorbitan monooleate), a commonly used stabilizing excipient in therapeutic drug solutions. This method is performed by quantitation of oleic acid, a hydrolysis product of polysorbate 80. Using base hydrolysis, polysorbate 80 releases the oleic acid at a 1:1 molar ratio. The oleic acid can then be separated from other polysorbate 80 hydrolysis products and matrices using reversed phase HPLC. The oleic acid is monitored without derivatization using the absorbance at 195 nm. The method was validated and also shown to be accurate for the quantitation of polysorbate 80 in a high protein concentration monoclonal antibody drug product. For the measured polysorbate 80 concentrations, the repeatability was less than 6.2% relative standard deviation of the mean (% RSD) with the day-to-day intermediate precision being less than 8.2% RSD. The accuracy of the oleic acid quantitation averaged 94-109% in different IgG(1) and IgG(4) drug solutions with variable polysorbate 80 concentrations. In this study, polyoxyethylene, a by-product of the polysorbate 80 hydrolysis was also identified. This peak was not identified by previous methods and also increased proportionally to the polysorbate 80 concentration. We have developed and qualified a method which uses common equipment found in most laboratories and is usable over a range of monoclonal antibody subclasses and protein concentrations.

Isomerization in the CDR2 of a monoclonal antibody: Binding analysis and factors that influence the isomerization rate.

Isomerization of a monoclonal antibody is one of the common routes of protein degradation. An isomerization in the complementarity-determining region (CDR) was found previously and is investigated in depth in this work. Affinity analysis proves that the antibody with one isomerized heavy chain has lower binding. Binding constants were determined, and exhibited a slower on-rate in conjunction with a faster off-rate for this isomerization. To determine the role of the buffer on the rate of isomerization, this antibody was incubated in various matrices and the amount of isomerized antibody was determined by hydrophobic interaction chromatography (HIC). The rate was found to be dependent on the pH as well as the net negative charge of the buffer components that can act as proton acceptors. An Arrhenius plot was performed to predict the levels of isomerization and a comparison of real samples proved the model was correct. This work affirms that isomerization in the CDR of a therapeutic antibody is important to monitor and the formulation buffer plays a significant role in the rate of the isomerization.

Identification and measurement of isoaspartic acid formation in the complementarity determining region of a fully human monoclonal antibody.

Isomerization plays a key role in protein degradation. This isomerization is often difficult to detect by many protein characterization methods such as SDS-PAGE, SEC, and IEF. This work shows the identification of an isomerized aspartic acid residue in the CDR2 of the heavy chain of a fully human monoclonal antibody. This isoaspartic acid increases significantly with storage at 2-8 degrees C. Hydrophobic interaction chromatography was utilized to separate the isoaspartic variant in the intact state. Mass spectrometry including peptide mapping was employed to identify and confirm the exact location of the modification. Since this modification occurs in the complementarity determining region (CDR) it was found that binding is reduced. Therefore, three different analytical methods for regular analysis of this isomerization are evaluated. These methods include peptide mapping by LC-MS, HIC, and a protein isoaspartate methyltransferase assay. It was determined that HIC is the best method to regularly assay the level of isomerization in this monoclonal antibody.

Investigation of proteins and peptides from yeastolate and subsequent impurity testing of drug product.

Hydrolysates play an important role in modern biological production. These mixtures are mostly undefined and contain a mixture of proteins, peptides, and amino acids along with other non-amino acid-based components. Recently, there has been an interest in defining and sequencing proteins and peptides in these hydrolysates to subsequently develop an assay to ensure removal during product purification. This work investigates an ultrafiltrate of yeastolate to determine whether any protein is present. Size exclusion chromatography indicated a possible high molecular weight component (>10 kDa). This suspected high molecular weight fraction was collected and investigated. It was determined that this fraction consists of nucleic acids; and no protein was detected using sensitive modern techniques including HPLC, mass spectrometry, and SDS-PAGE. Next, five unique, yeast-specific peptides were identified, sequenced, and confirmed. Finally, an impurity assay for any residual yeast specific peptides was developed and the analytical metrics were determined including accuracy, precision, linearity, range, and limits of detection and quantitation.

Evaluation of oligosaccharide methods for carbohydrate analysis in a fully human monoclonal antibody and comparison of the results to the monosaccharide composition determination by a novel calculation.

Carbohydrates can change a drug's properties including solubility, affinity towards antigen, pharmacokinetics and pharmacodynamics. Due to this importance, carbohydrate composition is utilized as a parameter to evaluate a drug candidate's quality. In this study, the compositional monosaccharides of a drug candidate are measured by HPAEC-PAD, while the oligosaccharides are studied by HPAEC-PAD, CE-LIF and LC-MS. The advantages and limitations of these various approaches for oligosaccharide analysis are reviewed in this work. While the methods used for oligosaccharide analysis are well established we have devised a new and novel calculation for determining monosaccharide content using the relative percentages of the N-glycans. This calculation was used to evaluate the accuracy of the oligosaccharide determination methods by comparison of the N-glycan data to the experimental monosaccharide data. The results obtained from this novel calculation demonstrate that the relative abundance of carbohydrates as determined from these various approaches are consistent.

Peptide mapping of therapeutic monoclonal antibodies: improvements for increased speed and fewer artifacts.

Peptide mapping is a widely utilized technique to characterize monoclonal antibodies for the purpose of product identity and is becoming increasing important as a stability indicating assay. Many conventional peptide-mapping methods are extremely time consuming and yield a map that is wrought with processing artifact peaks such as deamidation, carbamylation, and missed cleavages. Therefore, this work examines the many common individual sample preparation steps of the peptide-mapping procedure for monoclonal antibodies including the steps of denaturing, reduction, sample cleanup, digestion, and HPLC solvent selection. Improvements in each of these steps are demonstrated that greatly help to reduce artifacts and also allow for reduction of overall sample preparation time. After evaluating the many different parameters for increased speed and fewer artifacts, the sample preparation was reduced from days to hours and the resulting peptide map is nearly free of sample or background artifacts. Therefore, this peptide map procedure and optimization scheme is an excellent tool to further examine real sample changes in a shorter amount of time.

C-terminal lysine variants in fully human monoclonal antibodies: investigation of test methods and possible causes.

The C-terminal lysine variation is commonly observed in biopharmaceutical monoclonal antibodies. This modification can be important since it is found to be sensitive to the production process. The methods commonly used to probe this charge variation, including IEF, cIEF, ion-exchange chromatography, and LC-MS, were evaluated for their ability to effectively approximate relative percentages of lysine variants. A monoclonal antibody produced in a B cell hybridoma versus a CHO cell transfectoma was examined and it was determined that the relative amount of incorporated C-terminal lysine can vary greatly between these two production schemes. Another case study is shown whereby a different monoclonal antibody is subject to some minor process changes and the extent of lysine variation also exhibits a significant difference. During these studies the different methods for determining the extent of variation were evaluated and it was determined that LC-MS after trypsin digestion provides reproducible relative percentage information and has significant advantages over other methods. The final section of this work investigates the possible origins of this modification and evidence is shown that carboxypeptidase B or another basic carboxypeptidase causes this variation.

Determination of the origin of the N-terminal pyro-glutamate variation in monoclonal antibodies using model peptides.

The purpose of this work is to determine the cause of the cyclization of the N-terminal glutamine in recombinant proteins and monoclonal antibodies. This cyclization reaction commonly occurs on the N-terminal of light and/or heavy chains of antibodies and leads to heterogeneity of the final product. Two model peptides and an antibody containing an N-terminal glutamine were used to investigate the formation of N-terminal pyro- glutamic acid under various experimental conditions and different stages of the biosynthetic process. LC-MS analysis was used to separate and quantify the N-terminal variants. Experiments prove that the cyclization reaction is spontaneous and highly dependent on temperature and buffer composition and less dependent on pH. The conditions presented in most biopharmaceutical processes accelerate the formation of this variant. The majority of the near complete conversion (>95%) of N-terminal glutamine to pyro-glutamic acid commonly observed for antibodies appears to occur inside the bioreactor with only a small contribution from purification, formulation, and analytical preparation.