Quadrupole Dalton-Based Controlled Proteolysis Method for

Subscriber access provided by UNIV OF LOUISIANA

Article

Quadrupole Dalton-Based Controlled Proteolysis Method for Characterization of Higher Order Protein Structure Xiang Cao, Shannon C Flagg, Xue Li, Naresh Chennamsetty, Gurusamy Balakrishnan, and Tapan K. Das Anal. Chem., Just Accepted Manuscript • Publication Date (Web): 27 Mar 2019 Downloaded from http://pubs.acs.org on March 27, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Quadrupole Dalton-Based Controlled Proteolysis Method for Characterization of Higher Order Protein Structure

Xiang Cao*, Shannon C. Flagg, Xue Li, Naresh Chennamsetty, Gurusamy Balakrishnan, Tapan K. Das

Biologics Methods and Analytical Development, Bristol-Myers Squibb, Hopewell, New Jersey, USA

*Corresponding Author: email: [email protected]

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 32

Abstract Higher order structure (HOS) of proteins plays a critical role in the efficacy and stability of biologic drugs. Perturbation of regional structure of proteins can affect biological activity and cause instability. Characterization of HOS has become an integral part of biologic drug development and is expected from regulatory agencies. The commonly used techniques for HOS characterization such as circular dichroism, Fourier-transform infrared, differential scanning calorimetry, intrinsic fluorescence, and hydrogen-deuterium exchange mass spectrometry have their limitations ranging from lack of sensitivity, specificity to need of high-level expertise and poor access to instrumentation due to high cost. In this study, we demonstrated a novel controlled proteolysis based LC-QDa method for the detection of HOS change. By digesting proteins directly without denaturation and reduction, the HOS information can be revealed through the digested peptides. After optimizing digestion conditions and the detection procedures, we identified thirteen signature peptides that can monitor various antibody domains for any HOS changes caused by external stress. By comparing the peptide peak areas between unknown samples and a native control sample, any regional structural changes in unknown samples can be detected. The method was subsequently applied to a wide range of forced degradation samples to demonstrate higher sensitivity compared to the near-UV CD method that is frequently used for monitoring tertiary structural changes. By further reducing the number of signature peptides to five and optimizing liquid chromatography gradient duration, a streamlined, high-throughput and controlled proteolysis method was successfully established. This method can be used to support process and formulation development, as well as potentially for stability testing.

Keywords: Peptide mapping, structure, monoclonal antibodies, molecular modeling, characterization 2 ACS Paragon Plus Environment

Page 3 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction The past two decades have seen a substantial growth of therapeutic proteins especially monoclonal antibodies (mAbs) 1-2. Antibody-based biologic drugs constitute a lion’s share of cancer therapy agents 3. A surge in regulatory approval of mAbs has been observed in recent years. From 2013 to 2017, 31 new mAbs have been introduced to market for a variety of indications in oncology, immunosciences and other therapeutic areas, creating a global market of a total of 57 mAbs by the end of 2017 4. The global mAb market is predicted to reach as high as 200 billion by 2022 4. Therapeutic proteins including mAbs possess substantial structural heterogeneity compared to small molecules, due to their complex structure and large molecular size 5-6. A diverse array of analytical technologies are employed during process development, formulation development and quality control testing for release and stability to ensure the quality of drug products. Higher order structure (HOS) of proteins, including secondary, tertiary and quaternary structure, plays a critical role in the efficacy and stability of biotherapeutics drugs 7-10.Protein unfolding and misfolding can lead to a plethora of issues to biopharmaceuticals including instability, loss of efficacy, and immunogenicity 11. Additionally, modifications in the protein primary structure including oxidation, deamidation, isomerization and glycosylation may cause of changes in HOS. The HOS of proteins can also be perturbed by physical stress during manufacturing, transportation, handling, and storage, without chemical modification in the primary sequence 7. Due to its critical role in protein activity and stability, characterization of HOS is an integral part of biologics development and expected by health agencies 12. Additionally, as development progresses, heightened HOS characterization is increasingly added in a phase-appropriate manner.



Page 4 of 32

A variety of techniques have been established in the biotechnology field for HOS characterization. Nuclear magnetic resonance (NMR) and X-ray crystallography can elucidate three-dimensional protein structure at atomic level 13-16. However, both NMR and X-ray require complex sample preparation and lengthy data analysis, rendering them impractical for routine protein structure analysis 7, 17. Spectroscopic techniques such as circular dichroism (CD) and Fourier transform infrared (FTIR) are commonly applied in biopharmaceutical development due to their ease of use 18. Raman spectroscopy is another HOS characterization tool that can be used to probe secondary and tertiary structures19. Although a selection of these techniques have been routinely adopted during formulation development, characterization and comparability studies, they suffer from the limitation of low sensitivity 20-23. A recent study reported a case in which CD and FTIR didn’t provide sufficient sensitivity to detect measurable HOS changes during forced degradation study. A panel of degraded samples of a mAb was evaluated using CD and FTIR, and their HOS was found to be largely unchanged 18. In addition, the information from CD and FTIR are derived from a sum of signal of the whole protein, and as a result subtle and local structure changes in protein domains cannot be revealed with the two methods to identify location of the HOS changes. Mass spectrometry (MS) based techniques especially hydrogendeuterium exchange mass spectrometry (HDX MS) can provide valuable information of sitespecific changes in protein structure. This technique has been used to probe protein-protein interactions, protein folding / unfolding, protein conformational dynamics and comparability studies 24-31. However, the sample preparation and data analysis for HDX MS are very timeconsuming, limiting its application in routine analytical testing. Moreover, high-end MS instruments are expensive and the MS-based methods are difficult to validate for use in quality control (QC) laboratories32. In addition to HDX MS, Musetti et al have reported a high-


Page 5 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


throughput chemoprinting platform based on affinity selection mass spectrometry to confirm the structural continuity in biologics33. This technique uses specific small-molecule ligands as molecular probes to confirm the HOS of biologics. Although the method is high-throughput and has high sensitivity, it cannot reveal specific regions of structural changes, and the requirement of high resolution MS has similar limitations as HDX MS. Therefore high-throughput and QC friendly analytical methods are in high demand for biopharmaceutical development and QC applications. Peptide mapping has been established as a powerful tool to characterize protein primary structure as well as changes in sequence such as post-translational modifications (PTMs) 5, 34-35. During sample preparation for peptide mapping, proteins are denatured and reduced before enzymatic digestion, thus information on HOS is lost in the process. Limited proteolysis, during which proteins are digested without reduction and denaturation, has been applied to extract information on protein structure and dynamics 36. Partially digested proteins derived from limited proteolysis are usually analyzed by SDS-PAGE, and chromatographic separation coupled with UV or high-end MS detectors 37-38. Since proteins are preferentially digested at exposed regions such as loops and other flexible regions, the attributes of the digestion products from limited proteolysis can provide valuable information on protein regional structures. UV detectors have been extensively used for chromatographic separations, however, the detection of complex samples such as peptides often results in impure peaks and noisy baseline. The Quadrupole Dalton (QDa) detector is an easy to use and cost-effective mass analyzer that has been used for amino acids and peptide separation 32, 39. It can be easily qualified and used in QC environment due to its robustness and ease of use.



Page 6 of 32

In this work, we demonstrated the development of a QDa-based controlled proteolysis method for the detection of HOS change for mAbs. The conditions in the method were optimized by altering the digestion time. The resulting peptides were initially identified by an Orbitrap MS as part of characterization. A selection of signature peptides was created for HOS characterization as well as to track changes in HOS. Detection was then switched to a QDa detector to detect and identify the selected signature peptides at individual channels. The method was then applied to purposefully degraded protein samples generated under relevant stress conditions commonly used in drug development. A comparison of the controlled proteolysis data with CD and in-silico molecular modeling data was also made. In the initial phase of this proofof-concept study, all of the peptides that resulted in structural changes were monitored, but subsequently detection was reduced to five key peptides for this antibody to generate a simplified, high-throughput and robust controlled proteolysis method. The streamlined method can be suitably used to support process and formulation development, as well as potentially for stability testing. Materials and methods Sample preparation All chemicals and reagents were purchased from Sigma-Aldrich (St Louis, MO) unless otherwise specified. The recombinant human monoclonal IgG4 antibody (referred to as mAb1 in this paper) used in this work was produced in-house using Chinese Hamster Ovary cells and was purified by standard antibody purification procedures including protein A affinity and ion exchange chromatography. The denatured mAb1 for method development was prepared by diluting mAb1 with 8 M urea to a final urea concentration of 6 M. After 20 min of reaction time at room temperature, the


Page 7 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


antibody solution was diluted to 10 mg/mL for enzymatic digestion. For 2, 2'-Azobis (2amidinopropane) dihydrochloride (AAPH) treatment, mAb1 sample was diluted with AAPH solution to a molar ratio of 1:100 protein: AAPH. DL-methionine was added to a final concentration of 6 mM. The solution was incubated at 37 oC for 3 days, followed by desalting using Amicon Ultra-15 30 kDa centrifugal filters (MilliporeSigma, Burlington, MA). The sample was stored at -80 oC until analysis. A control sample for AAPH treatment was prepared by replacing AAPH solution with deionized (DI) water. A deglycosylated sample was prepared by incubating mAb1 with PNGase F (New England Biolabs, Ipswich, MA) at a ratio of 10 U/µg mAb1. The solution was mixed and incubated in the dark at 37 oC for 6 hours followed by desalting using Amicon Ultra-15 30 kDa centrifugal filters and stored at -80 oC freezer. A control was prepared by replacing PNGase F with DI water. All other forced degradation samples were prepared using corresponding stress conditions: acid (pH 4.0 for 7 days at 25 oC), basic (pH 9.4 for 7 days at 25 oC), heat (40 oC for 2 months), ultraviolet A (UVA) irradiation (100 W-hr/m2), high intensity light (HIL) irradiation (0.6×106 lux-hrs) and H2O2 oxidation (mAb1 to H2O2 molar ratio of 1: 100 for 6 hrs at room temperature). Controlled proteolysis The concentration of antibody samples was measured by SoloVPE from C Technologies (Bridgewater, NJ) prior to digestion. Antibody sample was first diluted to 10 mg/mL, then 20 µL of sample was mixed with 180 µL 50 mM Tris buffer at pH 7.6. Sequencing grade modified trypsin from Promega (Madison, WI) was added to a protein:enzyme ratio of 20:1 (w/w). The mixture was incubated at 37 oC followed by aliquoting at various time intervals and quenching the reaction by adding 10 µL of formic acid. Liquid Chromatography



Page 8 of 32

Reversed-phase liquid chromatography was performed on a BEH C18 column (1.7 µm, 2.1 mm × 150 mm) from Waters (Waltham, MA) using an Acquity H-Class UPLC equipped with both a PDA and QDa detector (Waters, Waltham, MA). 20 µg of antibody sample was loaded onto the column and analyzed at a flow rate of 0.2 mL/min and column temperature of 45 oC. Mobile phase A consisted of 0.1% formic acid in water, and mobile phase B consisted of 0.1 % formic acid in acetonitrile (ThermoFisher Scientific, Waltham, MA). The column was first equilibrated with 99.8 % Buffer A for 5 min, and then antibodies were eluted with 0.2 – 15 % B for 27 min, and 15 % - 35 % B for 60 min. After washing with 80 % B for 5 min, the column was re-equilibrated with Buffer A for 15 min. UV detection was performed at 214 nm. High resolution MS was performed on an Orbitrap Velos Pro MS system (Thermo-Fisher Corp., Waltham, MA) which was operated in positive ion mode with a capillary temperature of 300 oC and source voltage of 4.25 kV. Acquisitions were performed over the range of 350 - 2000 m/z with the resolution of 30000. Data analysis was performed with Xcalibur software from Thermo. The QDa mass detector was operated in positive mode with the mass range of 350 – 1250 Da. Capillary voltage was 1.5 kV and cone voltage was 15 V. The molecular weight of each individual peptide was calculated based adjacent m/z values in its mass spectra and was compared to theoretical molecular weight to confirm peptide identity. All PDA and QDa data were analyzed by empower software from Waters.

Results and Discussion Optimization of Digestion Time for Controlled Proteolysis Enzymatic digestion of a protein requires that the enzyme interacts with the polypeptide chain(s) with a specific stereochemistry 40. Folded proteins in native state are usually very 8 ACS Paragon Plus Environment

Page 9 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


resistant to proteolysis. For this reason, peptide mapping usually requires denaturation and disulfide bond reduction before digestion. In this study, an IgG4 antibody (mAb1) was used as a model protein for method development. A purposefully denatured mAb1 sample was prepared as a control sample used in method optimization. The denatured sample was prepared by unfolding mAb1 using urea, disrupting the native structure. One critical parameter of controlled proteolysis is the reaction time for digestion. In order to determine the optimal reaction time, the native and denatured mAb1 samples were digested and quenched at five different time points. The resulting samples were analyzed by LC-UV. Figure 1a shows the overlay of chromatograms of native mAb1 samples at five digestion time points. At 30 min, only a few small peaks were observed, indicating that only a small portion of the molecule was digested. A partial digestion itself disrupts the protein native structure, which is expected to accelerate the process. As a result, more peptides were generated with increasing digestion time. The chromatograms of native mAb1 and denatured mAb1 were compared at the five time points, and 140 min was selected as the optimal reaction time for further method development. At this time point (140 min), the native mAb1 sample has sufficient sensitivity for peak quantification. The native and denatured mAb1 profiles can also be easily differentiated by comparison of chromatograms (Figure 1b). The HOS of mAb1 was highly disrupted for urea-denatured protein, creating more digestion sites for proteolysis. Therefore, the denatured mAb1 sample resulted in more peptides and higher signal intensities when compared to the native mAb1 sample. In addition to UV detection, the resulting peptides were also detected by Orbitrap MS for definitive peak identification. Fifteen peptides were identified and labeled for denatured mAb1 by comparing the molecular mass of peptides calculated from the mass spectrum with the theoretical mass (see Figure 1b). The unlabeled peaks in the chromatograms are from partial



Page 10 of 32

digestion or trypsin miscleavage. The location of the identified peptides on the molecule is listed in Table 1 and depicted in Figure 2 by in-silico molecular modeling. The digested peptides are located in different domains (VH, VL, CH1, CH2, CH3); five of these are in the CDR regions. As the peptides are widely distributed across the protein surface, these are expected to provide structural information of mAb1 in various regions. Only seven peptides for the native antibody were clearly observed in the UV chromatogram (Figure 1b), specifically, H18, H21, H22, H26, H32, H33 and H34. Interestingly all of these peptides are located in either CH2 or CH3 domains in the Fc region (Table 1 and Figure 2). The peptides from the Fab region (H3, H4, H6, H13, L1, L3, L4 and L5) were clearly observed in denatured mAb1, but their UV intensities were very low in native mAb1. Homology modeling indicates that all fifteen peptides are fully exposed to solvent, but interestingly only the peptides in the Fc region were easily digested for native mAb1. This agrees well with other modeling results that indicate the Fc region of mAbs is structurally dynamic 41-42. The CH2 domain in the Fc region of mAbs is highly flexible compared to the other domains because of its open structure. CH2 and CH3 domains also have lower melting temperatures (i.e. lower conformational stability) compared to the Fab domain for IgG4 antibodies, measured by differential scanning calorimetry 43. The Fc region including the CH2 and CH3 domains is highly exposed as well as conformationally flexible, and therefore is more prone to digestion. This explains why the major species upon digestion of native mAb1 (IgG4) are all from the Fc region. Structural perturbation of proteins during manufacturing, transportation, handling and storage is likely to occur in highly flexible and solvent accessible regions. Therefore, the regional structural changes observed in the proteolytic digestion of native protein can offer strong clues to predict vulnerable regions in the protein structure. This


Page 11 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


hypothesis is further tested and confirmed by using a comprehensive set of relevant stress conditions. LC–QDa Method Feasibility Evaluation UV detection of complex samples such as mixture of peptides often results in noisy baselines and a lack of specificity depending on the sample type. In Figure 1b, most of the peaks in the chromatogram have co-eluting species. The QDa detector has higher specificity since the detection is based on mass. In our study the QDa detector was operated in Selected Ion Recording (SIR) mode to further increase sensitivity and specificity. In this mode the QDa only records user-defined m/z values at each time point in each individual channel, while all other ions being rejected. The m/z values of the most abundant charge state for all the fifteen peptides were determined from Orbitrap mass spectra, and these values were used for setting QDa channels. The peptides H18 and H32 carry oxidation and deamidation hot spots, respectively (unpublished data from the author’s laboratories). Since all primary sequence modifications can be monitored with comprehensive peptide mapping, the two peptides were excluded from QDa detection to simplify data processing. Figure 3 shows an overlay of thirteen signature peptide channels for denatured mAb1 with each individual peptide labeled. There were no co-eluting species with these peptides since the detection is based on m/z, and the QDa sensitivity is much greater when compared to UV. By comparing the peptide peak areas between unknown samples and native mAb1, any HOS change in unknown samples can be detected. Application of Controlled Proteolysis Method on Forced Degradation Study Samples Tryptophan Oxidation The controlled proteolysis LC-QDa method described above was first applied for studying AAPH-treated mAb1 samples. Tryptophan oxidation is a common degradation pathway for



Page 12 of 32

mAbs and poses significant challenges in developing control strategy. Multiple tryptophan oxidation products, such as kynurenine (Kyn), N-formylkynurenine (NFK), dioxindolylalanine can cause color change in mAbs, which may result in failure to meet analytical specification during drug product release 44-45. In addition, Kyn and NFK can also function as photosensitizers to visible light, thereby inducing additional damages such as aggregation 46. During normal oxidizing conditions (such as by residual hydrogen peroxide), methionine and tryptophan residues are both oxidized, making it difficult to study the impact of individual degradation events on mAb biological activity. Treatment of a mAb with AAPH in combination with free methionine (to protect protein Met residues) allows selective tryptophan oxidation in the mAb 4748.

In this study, mAb1 was treated with AAPH for 3 days. The resulting sample was buffer

exchanged to 50 mM Tris buffer (pH 7.6) and analyzed by controlled proteolysis. The individual peptide peaks were integrated and the peak areas were compared for the AAPH-oxidized and native mAb1 samples. Figure 4 shows the comparison of peak areas for peptides in Fc region (a) and Fab region (b). The peak areas for the Fc region peptides in AAPH-treated mAb1 were consistent with those in native mAb1. The cause of a slight increase in peak areas of the AAPHtreated samples is not known. It is possible that large changes in the Fab structure (see below) may have caused slight perturbation in the global structure of mAb1. The peak areas for Fab region peptides in the AAPH treated sample are significantly higher than the control except for H13. In-silico modeling demonstrated that the peptides that have increased significantly (H3, H4, H6, L1, L3, L4 and L5) are close to the three Trp hot spots in the Fab region (Figure S1). The three Trp residues were oxidized by 93%, 67% and 29% respectively according to comprehensive peptide mapping data. It can be concluded that Trp oxidation induced regional structure change, and this change could be easily captured by the controlled proteolysis method


Page 13 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


described in this paper. For comparison, the Fc peptides and peptide H13 have no significant change due to lack of Trp residues in these regions. Near-UV CD spectroscopy is widely used for characterization of tertiary structure. Figure S2 shows that the changes caused by AAPH treatment can be detected by near-UV CD. The stressed sample has a higher absorption for phenylalanine (Phe) and tyrosine (Tyr) bands and less absorption for Trp bands. However, unlike in controlled proteolysis method, the near-UV data cannot identify the specific regions of higher order structural changes. Deglycosylation Glycosylation plays a critical role in antibody effector function, stability, pharmacokinetics and pharmacodynamics 49-50. The antibody used in this study has a single N-linked glycosylation site per heavy chain. Impact of the N-linked glycosylation site on HOS was evaluated by removing glycans through enzymatic digestion with PNGase F and probing the resulting samples by controlled proteolysis methods. A comparison of the peak areas of digested peptides between the deglycosylated and native control samples is shown in Figure 5. Since peptide H21 contains the glycosylation site, it was excluded from the comparison considering that the m/z of H21 will change after deglycosylation. The four peptides in the Fc region increased significantly after deglycosylation, while most of Fab region peptides remain the same with the exception of H13. Modeling data revealed that the four Fc peptides that increased after deglycosylation are in close proximity to the N-linked glycosylation site Asn297 (Figure S3). Removal of N-glycans caused local structural changes in the Fc region, and the peptides in that area are more easily digested, presumably due to increased flexibility in the region. This agrees with the reported crystal structure of IgG4, in which the Fc adopts a more open conformation in the absence of glycans 5152,

and that the deglycosylated antibodies are less resistant to proteolytic cleavage 49, 53. Peptide



Page 14 of 32

H13 is in CH1 domain, but it’s directly linked to the Fc region through the hinge region (Figure S3). So any structural changes in the Fc region caused by deglycosylation can directly affect H13 as well. Near-UV CD data showed very minor changes in the Tyr region (Figure S2) showing inadequacy of this popular method for detecting certain structural changes in protein. Similarly, other methods such as FTIR or intrinsic fluorescence spectroscopy have not been able to detect regional structural changes in deglycosylated antibodies 49. Therefore, controlled proteolysis is a powerful technique to characterize certain local structural changes that otherwise go undetected by traditional HOS techniques. Photostability During drug development, the photostability of drug substance and drug product should be studied to ensure product quality 54. Light exposure of mAbs can occur in the entire process of manufacturing, starting from production process to formulation to packaging, storage and transportation 46. ICH guideline Q1B requires that the photostability studies use both visible light and UVA stress conditions. In this study we used HIL which is in the visible light range, as well as UVA to stress mAb1. Figure 6 compares the controlled proteolysis results of the control, UVA- and HIL-stressed samples. The Fc peptides especially H26, H33 and H22 increased significantly for the HIL-stressed relative to the control samples. The same peptides showed only a slight increase during UVA exposure. With respect to the Fab peptides, both HIL- and UVAtreated samples showed significantly higher amounts of H3, H4, H6, L1, L3 and L4 peptides, and the peak increase in peak areas in the HIL sample was much greater than in the UVA sample. Photo stress such as HIL and UVA typically results in both Met and Trp oxidation. For mAb1 there are three Met oxidation hot spots in the Fc region and three Trp oxidation hot spots in the Fab region. The comprehensive peptide mapping result of the stressed samples is listed in Table


Page 15 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


S1. The modeling results demonstrated that H3, H4, H6, L1, L3 and L4 are in close proximity to the three Trp residues in the Fab region (Figure S4), supporting the notion that these regions have increased structural flexibility caused by HIL-induced oxidation. This is consistent with the experimental finding of increased peak area of the peptides around the Trp sites observed in controlled proteolysis. The peptides H22, H26 and H33 are in close proximity to the three Met residues (Figure S4), consequently these peptides have greater intensity in the HIL and UVA treated samples. Figure 6 also shows that the HIL stress induced greater structural changes in the molecule compared to UVA, as evidenced by the larger changes in peptide peak areas in the HIL-stressed sample. Interestingly, although the Met oxidation percentages are higher than Trp oxidation for both photo-stressed samples, the peak area increase for the Fc region peptides are not as significant as in the Fab region peptides (Figure 6 & Table S1). Therefore, it appears that Trp oxidation in mAb1 causes larger impact on local structure than Met oxidation. The near-UV CD data was not able to detect changes in the photo-stressed samples (Figure S2), which again demonstrated the high sensitivity and specificity of controlled proteolysis method in characterizing structural changes. Comparison of the current method with CD and HDX MS HOS characterization is critical for protein therapeutics but continues to remain challenging due to many issues including inadequacy of the analytical techniques available. Spectroscopic techniques such as CD and FTIR have been extensively used for HOS characterization in force degradation and comparability studies due to their ease of use and historical precedence of use. CD measurement over the far-UV range (180 – 260 nm) reveals information on secondary structure. In the current forced degradation studies for mAb1, no changes in the far-UV CD spectra are observed for the stressed samples, thus the far-UV CD data are not shown here. In the



Page 16 of 32

near-UV range (240 – 340 nm), CD spectrum is comprised of the contributions from aromatic amino acids including tryptophan, phenylalanine and tyrosine, and thus contains tertiary structure information with the assumption that the aromatic residues are distributed across entire structure and its CD signal is sensitive to structural changes. When there are no aromatic amino acids surrounding the structural change regions, no changes in CD spectrum are expected. Moreover, CD spectrum is not able to identify the location of the structural changes. FTIR and Raman spectral techniques also suffer from similar limitations. These limitations can be overcome by the controlled proteolysis LC-QDa method developed in this study. For example, in the photostability studies reported here, the structural changes in the stressed samples were detected by the controlled proteolysis method but not in the near-UV CD data. HDX MS is another powerful method that can provide regional structural information. In a typical HDX MS protocol, protein samples are reacted with deuterated buffer over a period of time ranging from seconds to hours. After quenching the reaction, the deuterated proteins are denatured and reduced followed by proteolytic digestion. The obtained deuterated peptides are separated by high resolution LC-MS followed by time-consuming data analysis to obtain HDX kinetic curves. Compared to HDX MS, the current controlled proteolysis method offers advantages for easier sample preparation and data processing. Additionally, the use of QDa detection reduces instrument cost and improves robustness, making the method easy to be applied in QC environment. The advantage of HDX MS is that it can detect peptide level or even residue level structural change for the whole protein peptide backbone, so it is a powerful technique for investigational studies. In summary, the controlled proteolysis LC-QDa method we demonstrated in this study can be a valuable complementary method to currently used techniques such as CD and HDX MS. It provides regional structural information with high sensitivity compared to CD,


Page 17 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and is easier to use and more cost-effective compared to HDX MS. It is noteworthy that the controlled proteolysis method uses digested peptide products to indirectly infer the HOS change. Although trypsin cleaves peptide chains at the carboxyl side of lysine or arginine both of which are very common amino acids in protein sequences, it’s still possible that in certain instances altered structural change may not change the rate of enzymatic digestion in that region. In such circumstances, orthogonal methods such as HDX MS should be explored.

Conclusions We report the development of a novel analytical method combining controlled proteolysis and LC-QDa, which is demonstrated to be a sensitive method for the detection of HOS change in proteins. By comparing with near-UV CD data for a wide selection of forced degradation samples, it is concluded that this approach demonstrates greater sensitivity and specificity than the CD method and is able detect subtle changes in regional structures. In-silico homology modeling further provided insight on how different modifications may alter regional structure and supported the experimental data. The method is expected to be applicable to other molecules with adjustments in the method development workflow. The overall method is straightforward and reproducible, and offers the benefit of high throughput. In addition, the use of a QDa mass detection instead of high resolution MS instrument enables its cost-effective use in QC environment. Considering the critical role of protein HOS in biopharmaceutical development, this method can be a valuable addition to the toolbox for characterization during process and formulation development.



Page 18 of 32

Supporting Information

Additional experimental procedure and discussion on method optimization and reproducibility evaluation. In-silico homology modeling of mAb1, near UV spectra for different mAb1 samples, comparison of peak areas of peptides between native mAb1 and other forced degradation samples, chromatograms and bar graph between native mAb1 and failed QC batch.

Acknowledgements and Disclosures We thank Drs. Jacob Bongers of Bristol-Myers Squibb Molecular and Analytical Development group for helpful discussions on this project. The authors declare no personal, financial, or non-financial conflict of interest.

References 1.

Kaplon, H.; Reichert, J. M., Antibodies to watch in 2018. MAbs 2018, 10 (2), 183-203.

2.

Singh, S.; Kumar, N.; Dwiwedi, P.; Charan, J.; Kaur, R.; Sidhu, P.; Chugh, V. K., Monoclonal

Antibodies: A Review. Curr Clin Pharmacol 2017. 3.

Weiner, G. J., Building better monoclonal antibody-based therapeutics. Nat Rev Cancer 2015, 15

(6), 361-70. 4.

Grilo, A. L.; Mantalaris, A., The Increasingly Human and Profitable Monoclonal Antibody Market.

Trends Biotechnol 2018. 5.

Beck, A.; Wagner-Rousset, E.; Ayoub, D.; Van Dorsselaer, A.; Sanglier-Cianferani, S.,

Characterization of therapeutic antibodies and related products. Anal Chem 2013, 85 (2), 715-36. 6.

Liu, H.; Gaza-Bulseco, G.; Faldu, D.; Chumsae, C.; Sun, J., Heterogeneity of monoclonal

antibodies. J Pharm Sci 2008, 97 (7), 2426-47.


Page 19 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


7.

Berkowitz, S. A.; Engen, J. R.; Mazzeo, J. R.; Jones, G. B., Analytical tools for characterizing

biopharmaceuticals and the implications for biosimilars. Nat Rev Drug Discov 2012, 11 (7), 527-40. 8.

Davies, D. R.; Metzger, H., Structural basis of antibody function. Annu Rev Immunol 1983, 1, 87-

117. 9.

Schroeder, H. W., Jr.; Cavacini, L., Structure and function of immunoglobulins. J Allergy Clin

Immunol 2010, 125 (2 Suppl 2), S41-52. 10.

Wang, W.; Singh, S.; Zeng, D. L.; King, K.; Nema, S., Antibody structure, instability, and

formulation. J Pharm Sci 2007, 96 (1), 1-26. 11.

Maas, C.; Hermeling, S.; Bouma, B.; Jiskoot, W.; Gebbink, M. F., A role for protein misfolding in

immunogenicity of biopharmaceuticals. J Biol Chem 2007, 282 (4), 2229-36. 12.

ICH Q5E Comparability of Biotechnological/Biological Products Subject to Changes in Their

Manufacturing Processes. In ICH Quality Guidelines. 13.

Gronenborn, A. M.; Clore, G. M., Protein structure determination in solution by two-dimensional

and three-dimensional nuclear magnetic resonance spectroscopy. Anal Chem 1990, 62 (1), 2-15. 14.

Ilari, A.; Savino, C., Protein structure determination by x-ray crystallography. Methods Mol Biol

2008, 452, 63-87. 15.

Parker, M. W., Protein structure from x-ray diffraction. J Biol Phys 2003, 29 (4), 341-62.

16.

Wuthrich, K., Protein structure determination in solution by nuclear magnetic resonance

spectroscopy. Science 1989, 243 (4887), 45-50. 17.

Song, Y.; Yu, D.; Mayani, M.; Mussa, N.; Li, Z. J., Monoclonal antibody higher order structure

analysis by high throughput protein conformational array. MAbs 2018, 10 (3), 397-405. 18.

Lin, J. C.; Glover, Z. K.; Sreedhara, A., Assessing the Utility of Circular Dichroism and FTIR

Spectroscopy in Monoclonal-Antibody Comparability Studies. J Pharm Sci 2015, 104 (12), 4459-4466.



19.

Page 20 of 32

Balakrishnan, G.; Barnett, G. V.; Kar, S. R.; Das, T. K., Detection and Identification of the

Vibrational Markers for the Quantification of Methionine Oxidation in Therapeutic Proteins. Anal Chem 2018, 90 (11), 6959-6966. 20.

Jiang, Y.; Li, C.; Nguyen, X.; Muzammil, S.; Towers, E.; Gabrielson, J.; Narhi, L., Qualification of

FTIR spectroscopic method for protein secondary structural analysis. J Pharm Sci 2011, 100 (11), 463141. 21.

Li, C. H.; Nguyen, X.; Narhi, L.; Chemmalil, L.; Towers, E.; Muzammil, S.; Gabrielson, J.; Jiang, Y.,

Applications of circular dichroism (CD) for structural analysis of proteins: qualification of near- and farUV CD for protein higher order structural analysis. J Pharm Sci 2011, 100 (11), 4642-54. 22.

Barnett, G. V.; Balakrishnan, G.; Chennamsetty, N.; Meengs, B.; Meyer, J.; Bongers, J.; Ludwig, R.;

Tao, L.; Das, T. K.; Leone, A.; Kar, S. R., Enhanced Precision of Circular Dichroism Spectral Measurements Permits Detection of Subtle Higher Order Structural Changes in Therapeutic Proteins. J Pharm Sci 2018, 107 (10), 2559-2569. 23.

Weiss, W. F. t.; Gabrielson, J. P.; Al-Azzam, W.; Chen, G.; Davis, D. L.; Das, T. K.; Hayes, D. B.;

Houde, D.; Singh, S. K., Technical Decision Making With Higher Order Structure Data: Perspectives on Higher Order Structure Characterization From the Biopharmaceutical Industry. J Pharm Sci 2016, 105 (12), 3465-3470. 24.

Huang, R. Y.; Chen, G., Higher order structure characterization of protein therapeutics by

hydrogen/deuterium exchange mass spectrometry. Anal Bioanal Chem 2014, 406 (26), 6541-58. 25.

Wei, H.; Mo, J.; Tao, L.; Russell, R. J.; Tymiak, A. A.; Chen, G.; Iacob, R. E.; Engen, J. R.,

Hydrogen/deuterium exchange mass spectrometry for probing higher order structure of protein therapeutics: methodology and applications. Drug Discov Today 2014, 19 (1), 95-102. 26.

Yan, X.; Maier, C. S., Hydrogen/deuterium exchange mass spectrometry. Methods Mol Biol 2009,

492, 255-71. 20 ACS Paragon Plus Environment

Page 21 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


27.

Huang, R. Y.; Iacob, R. E.; Krystek, S. R.; Jin, M.; Wei, H.; Tao, L.; Das, T. K.; Tymiak, A. A.; Engen, J.

R.; Chen, G., Characterization of Aggregation Propensity of a Human Fc-Fusion Protein Therapeutic by Hydrogen/Deuterium Exchange Mass Spectrometry. J Am Soc Mass Spectrom 2017, 28 (5), 795-802. 28.

Bou-Assaf, G. M.; Marshall, A. G., Chapter 12 - Biophysical Mass Spectrometry

for Biopharmaceutical Process Development: Focus on Hydrogen/Deuterium Exchange. In Biophysical Characterization of Proteins in Developing Biopharmaceuticals, Houde, D. J.; Berkowitz, S. A., Eds. Elsevier: Amsterdam, 2015; pp 307-339. 29.

Majumdar, R.; Middaugh, C. R.; Weis, D. D.; Volkin, D. B., Hydrogen-deuterium exchange mass

spectrometry as an emerging analytical tool for stabilization and formulation development of therapeutic monoclonal antibodies. J Pharm Sci 2015, 104 (2), 327-45. 30.

Houde, D.; Berkowitz, S. A.; Engen, J. R., The utility of hydrogen/deuterium exchange mass

spectrometry in biopharmaceutical comparability studies. J Pharm Sci 2011, 100 (6), 2071-86. 31.

Yan, Y.; Wei, H.; Fu, Y.; Jusuf, S.; Zeng, M.; Ludwig, R.; Krystek, S. R., Jr.; Chen, G.; Tao, L.; Das, T.

K., Isomerization and Oxidation in the Complementarity-Determining Regions of a Monoclonal Antibody: A Study of the Modification-Structure-Function Correlations by Hydrogen-Deuterium Exchange Mass Spectrometry. Anal Chem 2016, 88 (4), 2041-50. 32.

Xu, W.; Jimenez, R. B.; Mowery, R.; Luo, H.; Cao, M.; Agarwal, N.; Ramos, I.; Wang, X.; Wang, J., A

Quadrupole Dalton-based multi-attribute method for product characterization, process development, and quality control of therapeutic proteins. MAbs 2017, 9 (7), 1186-1196. 33.

Musetti, C.; Bean, M. F.; Quinque, G. T.; Kwiatkowski, C.; Szewczuk, L. M.; Baldoni, J.; Zajac, M.

A., High-Throughput Assessment of Structural Continuity in Biologics. Anal Chem 2018, 90 (4), 29702975. 34.

Kuster, B.; Mann, M., Identifying proteins and post-translational modifications by mass

spectrometry. Curr Opin Struct Biol 1998, 8 (3), 393-400. 21 ACS Paragon Plus Environment


35.

Page 22 of 32

Rogers, R. S.; Nightlinger, N. S.; Livingston, B.; Campbell, P.; Bailey, R.; Balland, A., Development

of a quantitative mass spectrometry multi-attribute method for characterization, quality control testing and disposition of biologics. MAbs 2015, 7 (5), 881-90. 36.

Fontana, A.; de Laureto, P. P.; Spolaore, B.; Frare, E.; Picotti, P.; Zambonin, M., Probing protein

structure by limited proteolysis. Acta Biochim Pol 2004, 51 (2), 299-321. 37.

Perrin, C.; Burkitt, W.; Perraud, X.; O'Hara, J.; Jone, C., Limited proteolysis and peptide mapping

for comparability of biopharmaceuticals: An evaluation of repeatability, intra-assay precision and capability to detect structural change. J Pharm Biomed Anal 2016, 123, 162-72. 38.

Fontana, A.; Zambonin, M.; Polverino de Laureto, P.; De Filippis, V.; Clementi, A.; Scaramella, E.,

Probing the conformational state of apomyoglobin by limited proteolysis. J Mol Biol 1997, 266 (2), 22330. 39.

Galba, J.; Michalicova, A.; Parrak, V.; Novak, M.; Kovac, A., Quantitative analysis of

phenylalanine, tyrosine, tryptophan and kynurenine in rat model for tauopathies by ultra-high performance liquid chromatography with fluorescence and mass spectrometry detection. J Pharm Biomed Anal 2016, 117, 85-90. 40.

Schechter, I.; Berger, A., On the size of the active site in proteases. I. Papain. Biochem Biophys

Res Commun 1967, 27 (2), 157-62. 41.

Remesh, S. G.; Armstrong, A. A.; Mahan, A. D.; Luo, J.; Hammel, M., Conformational Plasticity of

the Immunoglobulin Fc Domain in Solution. Structure 2018, 26 (7), 1007-1014 e2. 42.

Frank, M.; Walker, R. C.; Lanzilotta, W. N.; Prestegard, J. H.; Barb, A. W., Immunoglobulin G1 Fc

domain motions: implications for Fc engineering. J Mol Biol 2014, 426 (8), 1799-811. 43.

Garber, E.; Demarest, S. J., A broad range of Fab stabilities within a host of therapeutic IgGs.

Biochem Biophys Res Commun 2007, 355 (3), 751-7.


Page 23 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


44.

Li, Y.; Polozova, A.; Gruia, F.; Feng, J., Characterization of the degradation products of a color-

changed monoclonal antibody: tryptophan-derived chromophores. Anal Chem 2014, 86 (14), 6850-7. 45.

Song, H.; Xu, J.; Jin, M.; Huang, C.; Bongers, J.; Bai, H.; Wu, W.; Ludwig, R.; Li, Z.; Tao, L.; Das, T.

K., Investigation of Color in a Fusion Protein Using Advanced Analytical Techniques: Delineating Contributions from Oxidation Products and Process Related Impurities. Pharm Res 2016, 33 (4), 932-41. 46.

Du, C.; Barnett, G.; Borwankar, A.; Lewandowski, A.; Singh, N.; Ghose, S.; Borys, M.; Li, Z. J.,

Protection of therapeutic antibodies from visible light induced degradation: Use safe light in manufacturing and storage. Eur J Pharm Biopharm 2018, 127, 37-43. 47.

Folzer, E.; Diepold, K.; Bomans, K.; Finkler, C.; Schmidt, R.; Bulau, P.; Huwyler, J.; Mahler, H. C.;

Koulov, A. V., Selective Oxidation of Methionine and Tryptophan Residues in a Therapeutic IgG1 Molecule. J Pharm Sci 2015, 104 (9), 2824-31. 48.

Hageman, T.; Wei, H.; Kuehne, P.; Fu, J.; Ludwig, R.; Tao, L.; Leone, A.; Zocher, M.; Das, T. K.,

Impact of Tryptophan Oxidation in Complementarity-Determining Regions of Two Monoclonal Antibodies on Structure-Function Characterized by Hydrogen-Deuterium Exchange Mass Spectrometry and Surface Plasmon Resonance. Pharm Res 2018, 36 (1), 24. 49.

Zheng, K.; Bantog, C.; Bayer, R., The impact of glycosylation on monoclonal antibody

conformation and stability. MAbs 2011, 3 (6), 568-76. 50.

Liu, L., Antibody glycosylation and its impact on the pharmacokinetics and pharmacodynamics of

monoclonal antibodies and Fc-fusion proteins. J Pharm Sci 2015, 104 (6), 1866-1884. 51.

Davies, A. M.; Rispens, T.; Ooijevaar-de Heer, P.; Gould, H. J.; Jefferis, R.; Aalberse, R. C.; Sutton,

B. J., Structural determinants of unique properties of human IgG4-Fc. J Mol Biol 2014, 426 (3), 630-44. 52.

Davies, A. M.; Jefferis, R.; Sutton, B. J., Crystal structure of deglycosylated human IgG4-Fc. Mol

Immunol 2014, 62 (1), 46-53.



53.

Page 24 of 32

Raju, T. S.; Scallon, B., Fc glycans terminated with N-acetylglucosamine residues increase

antibody resistance to papain. Biotechnol Prog 2007, 23 (4), 964-71. 54.

Q1A Stability Testing of New Drug Substances and Products. In ICH Quality Guidelines.


Page 25 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure Captions Figure 1. (a) Controlled proteolysis peptide digestion of native mAb1 sample at different digestion times with UV detection at 214 nm. (b) Overlay chromatograms of controlled proteolysis digestion of native mAb1 and urea denatured mAb1 with 140 min digestion time. The peptides identified with Oribtrap MS were labeled in the chromatogram.



Page 26 of 32

Figure 2. In-silico homology modeling of mAb1 with expanded scale of Fc and Fab regions. The locations of fifteen digested peptides were labeled in the structure.

H26

Fab

H6

Fc

H3 H4

H13

H21

Glycosylation

H22

H18 H33 H32 H34

L3 L4 L5

Glycosylation

L1


Page 27 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Controlled proteolysis peptide digestion of urea denatured mAb1 at 140 min digestion time with QDa detection. The chromatograms for individual peptide channels were overlaid together.



Page 28 of 32

Figure 4. Comparison of peak areas of peptides in Fc region (a) and Fab region (b) for native mAb1 and AAPH treated mAb1. Error bars represent the standard deviation of three independent measurements for each sample


Page 29 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Comparison of peak areas of peptides in Fc region (a) and Fab region (b) for native mAb1 and deglycosylated mAb1. Error bars represent the standard deviation of three independent measurements for each sample



Page 30 of 32

Figure 6. Comparison of peak areas of peptides in Fc region (a) and Fab region (b) for native mAb1, UVA-treated mAb1 and HIL-treated mAb1. Error bars represent the standard deviation of three independent measurements for each sample


Page 31 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1 List of identified peptides and their locations on mAb1 Peptide

Domain

Peptide

Domain

H3

VH (CDR)

H32

CH3

H4

VH (CDR)

H33

CH3

H6

VH

H34

CH3

H13

CH1

L1

VL

H18

CH2

L3

VL (CDR)

H21

CH2

L4

VL (CDR)

H22

CH2

L5

VL (CDR)

H26

CH2



Page 32 of 32

TOC graphics


Quadrupole Dalton-Based Controlled Proteolysis Method for

Recommend Documents