A Generic HPLC Method for Absolute Quantification of Oxidation in

Jun 28, 2017 - We evaluated the feasibility of three strategies for absolute quantification of oxidation in the Fc region of hydrogen peroxide-stresse...
1 downloads 13 Views 1MB Size
Subscriber access provided by NEW YORK UNIV

Article

A Generic HPLC Method for Absolute Quantification of Oxidation in Monoclonal Antibodies and Fc-Fusion Proteins Using UV and MS Detection Christof Regl, Therese Wohlschlager, Johann Holzmann, and Christian G. Huber Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b01755 • Publication Date (Web): 28 Jun 2017 Downloaded from http://pubs.acs.org on June 29, 2017

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 free 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 accessible to all readers and 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.

Analytical Chemistry 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 9

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

A Generic HPLC Method for Absolute Quantification of Oxidation in Monoclonal Antibodies and Fc-Fusion Proteins Using UV and MS Detection Christof Regl,†,‡ Therese Wohlschlager,†,‡ Johann Holzmann,§,‡ and Christian G. Huber*,†,‡ †

Department of Molecular Biology, Division of Chemistry and Bioanalytics, University of Salzburg, Hellbrunner Strasse 34, 5020 Salzburg, Austria



Christian Doppler Laboratory for Innovative Tools for Biosimilar Characterization, University of Salzburg, Hellbrunner Strasse 34, 5020 Salzburg, Austria

§

Technical Development Biosimilars, Physicochemical Characterization Kundl, Novartis BTDM, Sandoz GmbH, Biochemiestrasse 10, 6250 Kundl, Austria ABSTRACT: Oxidation of biopharmaceuticals may affect their bioactivity, serum half-life, and (bio)chemical stability. The Fc domain of IgG monoclonal antibodies (mAbs) contains two methionine residues which are susceptible to oxidation. Here, we present a middle-down approach employing the cysteine protease IdeS under reducing conditions to obtain three mAb subunits of approximately 25 kDa: Fc/2, Fd’ and LC. These subunits were separated by ion-pair reversed-phase high-performance liquid chromatography (IP-RP-HPLC) and detected by UV-spectroscopy as well as Orbitrap mass spectrometry (MS), as well as MS upon allion fragmentation (AIF-MS). We evaluated the feasibility of three strategies for absolute quantification of oxidation in the Fc region of hydrogen peroxide-stressed Rituximab, using a single, commercially available software platform both for data acquisition and evaluation: UV-spectroscopy, full scan MS, and monitoring of product ions obtained by AIF-MS. UV-spectroscopy showed the lowest limits of quantification (LOQ) (0.96 ng µL–1) and featured the lowest relative process standard deviation (Vx0%) of 7.2% compared to MS and AIF-MS with LOQs of 1.24–4.32 ng µL–1 and relative process standard deviations of 9.0–14%, respectively. Our approach is generic in that it allows monitoring and quantification of oxidation in the Fc regions of fully human and humanized IgG1 mAbs as well as of Fc-fusion proteins. This is exemplified by limits of detection of 1.2%, 1.0%, and 1.2% of oxidation in drug products containing the biopharmaceuticals Rituximab, Adalimumab, and Etanercept, respectively. The presented method is an attractive alternative to conventional time-intensive peptide mapping which is prone to artificial oxidation due to extensive sample preparation.

Monoclonal antibodies (mAbs) and Fc-fusion proteins represent the fastest growing class of biopharmaceuticals and have been approved for the treatment of several major indications, including cancer, inflammatory diseases, and growth disorders.1-3 Advantages of biopharmaceuticals as therapeutic agents are their high specificity, large therapeutic impact, and long in vivo half-life.4 As large biomolecules are not amenable to chemical synthesis, therapeutic proteins are produced by recombinant protein expression technology utilizing bacterial or eukaryotic cell cultures.5 Consequently, biopharmaceuticals are inherently associated with structural variability owing to posttranslational modifications (PTMs) such as disulfide bond formation, glycosylation, amino acid cyclization, deamidation, and oxidation.6-8 These modifications require tight control of the manufacturing process in order to obtain a consistent product profile and to ensure the safety and efficacy of the drug. In-depth characterization and quantification of mAb variants during process development and drug production therefore is essential in order to obtain the desired product profile. 4,7,9,10

The Fc domain of humanized IgG1 mAbs contains two methionine residues, Met256 and Met432, which are located in loop-regions and exposed at the protein surface.11-14 Both of these residues are prone to oxidation to methionine sulfoxide during manufacturing and/or storage of the drug.15,16 Methionine oxidation has been shown to affect the structure of the Fc region17,18 which may alter FcRn binding11,18-22 and therefore impact pharmacokinetic properties. As a consequence, this protein modification may lead to faster plasma clearance and altered clinical efficacy, thus representing a shelf-life limiting factor.16,23 Identification and relative quantification of oxidation in mAbs and Fc-fusion proteins is conventionally performed via peptide mapping involving proteolytic digestion and analysis by ion-pair reversed-phase high-performance liquid chromatography (IP-RP-HPLC) coupled to mass spectrometry (MS).8,16 However, extensive sample preparation may induce artificial modifications. In that respect, reaction conditions used for tryptic digestion favor the occurrence of deamidation and oxidation.24,25 Furthermore, the context of a modification, i.e. the combination of PTMs within a certain mAb variant, is

ACS Paragon Plus Environment

1

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

not amenable by peptide mapping. Alternatively, chromatographic methods like hydrophobic interaction chromatography (HIC) or ion-exchange chromatography (IEC) combined with UV-spectroscopy are being used for rough estimation of oxidation in the intact molecule.23,26,27 Ideally, characterization of biopharmaceuticals is carried out on the intact protein level by an MS-compatible separation method like IP-RP-HPLC, combined with top-down MS.28,29 This enables simultaneous characterization and quantification of oxidation sites by UV-spectroscopy combined with MS. Nevertheless, baseline separation of intact mAb oxidation variants by HPLC, a prerequisite for quantification by UVspectroscopy, has not been accomplished to date. Furthermore, the natural isotopic distributions of oxidized and non-oxidized variants of a molecule as large as an intact mAb are too broad to be distinguishable by MS (Figure S-1) thereby preventing reliable mass spectrometric quantification without preceding separation of these variants. Characterization of mAb subunits allows a compromise between peptide analysis with its inherent limitations and the ideal but infeasible intact protein analysis. Subunits of mAbs and Fc-fusion proteins may be obtained by reduction of disulfide bonds or by partial proteolysis. The latter can be achieved using the streptococcal cysteine protease IdeS which cleaves human IgG antibodies specifically in the hinge region30. IdeS has previously been applied to study glycosylation, deamidation and oxidation of the Fc domain.31-51 Recently, Sokolowska et al. presented a fast method for quantification of oxidation upon IdeS digestion, deglycosylation with EndoS and disulfide reduction, employing IP-RP-HPLC coupled to a quadrupole time-of-flight (Q-TOF) mass spectrometer.32 Even though the method is generic for IgG mAbs, oxidized Fc variants remained unseparated, restricting quantification to mass spectrometry data. In an industrial setting, though, the robustness of UV-based HPLC quantification would be favorable. In this regard, Zhang et al. presented a method for the separation and relative quantification of oxidized and non-oxidized Fc variants by reversed-phase HPLC-UV-spectroscopy coupled with MS, but absolute quantification of Fc oxidation was not attempted.33 We here aim at developing a fast and generic method employing HPLC-UV-MS for absolute quantification of oxidation at the Fc/2 level. Subunits of mAbs or Fc-fusion proteins will be generated by parallel digestion with the proteases IdeS and Carboxypeptidase B (CpB) requiring minimum sample preparation. Chromatographic separation of the subunits by IP-RP-HPLC shall facilitate the discrimination of Fc/2 oxidation variants. Localization of the oxidation sites shall be based on diagnostic fragments obtained upon all-ion fragmentation (AIF) of Fc/2.28,29 In order to enable implementation in an industrial environment, the presented quantification strategy is primarily based on UV-spectroscopy, and corroborated by full-scan MS and AIF-MS. Finally, the universality of the method for Fc-containing biopharmaceuticals will be demonstrated by absolute quantification of oxidation in the two IgG mAbs Rituximab and Adalimumab, as well as in the Fc-fusion protein Etanercept. EXPERIMENTAL SECTION Materials. Acetonitrile (ACN, ≥ 99.9%) was purchased from VWR International (Vienna, Austria). Ammonium hexa-

Page 2 of 9

fluorophosphate (AHFP, 99.99%), tris(2-carboxyethyl) phosphine hydrochloride (TCEP, ≥ 98.0%), iodoacetamide (IAA, ≥ 99.9%), isopropanol (IPA, ≥ 99.9%), methanol (MeOH, ≥ 99.9%), trifluoroacetic acid (TFA, ≥ 99.0%), formic acid (FA, 98.0-100%), guanidine hydrochloride (GdnHCl, ≥ 99 %) and Carboxypeptidase B (CpB) from pig pancreas were obtained from Sigma-Aldrich (Vienna, Austria). Ammonium acetate (≥ 98.0%) and hydrogen peroxide (35% in H2O) were purchased from Merck (Darmstadt, Germany). Immobilized IdeS enzyme was obtained from Genovis (Lund, Sweden). Trypsin/Lys-C Mix (Mass Spec Grade, V5071) was purchased from Promega (Madison, WI, USA). For all experiments ultrapure water produced in-house by a Millipore Integral 3 from Merck/Millipore (Billerica, MA, USA) was used. Sartorius Vivaspin 500 centrifugal filters with 50 kDa cutoff were obtained from Sartorius AG (Göttingen, Germany). The biopharmaceuticals used were the commercially available monoclonal antibodies MabThera (drug product; N7025B04 exp. 02/17 respectively H0102B06 exp. 05/14, 10 mg mL–1 Rituximab) from F. Hoffmann-La Roche Ltd (Basel, Switzerland) and Humira (drug product; 26370XD15, 48.45 mg mL–1 Adalimumab) from AbbVie Inc. (Chicago, IL, USA), and the Fc-fusion therapeutic Enbrel (drug product; E88057 exp. 05/12, 50 mg mL–1 Etanercept) from Pfizer (New York, NY, USA). Sample Preparation. Rituximab, Adalimumab, and Etanercept drug product samples were diluted to a concentration of 5.0 mg mL–1 in 175 mmol.L–1 ammonium acetate, and subsequently digested with immobilized IdeS for 30 min at 20 °C applying end-over-end rotation, according to the manufacturer's protocol. Furthermore, for the analysis of Adalimumab and Etanercept, Carboxypeptidase B (CpB) was added at a substrate/enzyme ratio of 5:1 (w/w) for parallel digestion with IdeS. Subsequently, the samples were reduced in 4 mol.L-1 GdnHCl with 5 mmol.L–1 TCEP for 15 min at 60 °C. For peptide mapping, Rituximab was diluted to a concentration of 0.5 mg.mL–1 in 175 mmol.L–1 ammonium acetate. Disulfides were reduced with 5 mmol.L–1 TCEP for 15 min at 60 °C and subsequently alkylated by addition of 20 mmol L–1 IAA and incubation for 30 min at 22 °C. Digestion was performed upon addition of Trypsin-Lys-C mixture at a mAb to enzyme ratio of 10:1 and incubation for 4.0 hours at 37 °C. Calibration solutions were prepared by appropriate dilution of reduced and digested sample. Forced Oxidation of Rituximab. Chemical oxidation was induced by incubation of Rituximab (diluted to 1.0 mg mL-1 in 175 mmol.L–1 ammonium acetate) with 0.35% hydrogen peroxide for 30 min at 22 °C. The reaction was quenched by buffer exchange to 175 mmol.L–1 ammonium acetate by ultrafiltration with a 50 kDa cutoff filter. Protein concentration was determined on a nanophotometer (P330, Implen GmbH, Munich, Germany) at 280 nm assuming an extinction coefficient of 237,380 L.mol–1.cm–1. High-Performance Liquid Chromatography. Chromatographic separation of oxidation variants was carried out on an UltiMate 3000 Rapid Separation system (U3000 RSLC, Thermo Fisher Scientific, Germering, Germany) at a flow rate of 200 µL.min–1 using a MAbPac RP column (150 × 2.1 mm i.d., 4.0 µm particle size, ~1500 Å pore size, Thermo Fisher Scientific, Sunnyvale, CA, USA), operated at a temperature of

ACS Paragon Plus Environment

2

Page 3 of 9

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

Figure 1. Schematic structure of an IgG1 mAb and the subunits obtained upon IdeS digestion and disulfide bond reduction. Oxidized methionine residues are symbolized by filled orange circles. (a) Digestion of a dimeric IgG1 mAb with IdeS generates two Fc/2 and one F(ab‘)2 subunits (7). Reduction of intermolecular disulfide bonds stabilizing F(ab‘)2 yields two Fd‘ (5) and two LC (6) fragments. Furthermore, intramolecular disulfide bridges are reduced. (b) Upon digestion with IdeS and chemical reduction with TCEP, four different Fc/2 variants are obtained: doubly-oxidized (1), singly-oxidized (2 and 3) and non-oxidized (4) Fc/2. 80 °C. Mobile phase A was composed of H2O + 0.10% TFA, mobile phase B of acetonitrile + 0.10% TFA. The gradient applied was: 28.9% B for 5 min, 28.9–29.0% B in 15 min, 29.0–30.0% B in 9 min, 30.0–45.0% B in 5 min, 100% B for 5 min, and 28.9% B for 10 min. Injection was carried out using in-line split-loop mode, injection volume was 10 µL for each sample. Sample concentrations ranged from 1.8 pg µL–1 to 1060 ng µL–1 (see Supporting Information Excel file). UV detection was carried out at 214 nm with a 2.5 µL flow cell. Peptide mapping data were acquired with 5 µg of protein digest on the same system with the same mobile phases utilizing a Hypersil GOLD aQ C18 column (100 x 1.0 mm i.d., 1.9 µm particle size, 175 Å pore size, Thermo Fisher Scientific, Sunnyvale, CA, USA) at a flow rate of 60 µL.min–1 operated at a temperature of 50 °C. The gradient applied was: 2.0% B for 5 min, 5.0–10.0% B in 5 min, 10.0–40.0% B in 60 min, 100% B for 5 min, and 2.0% B for 25 min. Mass Spectrometry. Mass spectrometry was conducted on a Thermo ScientificTM Q ExactiveTM benchtop quadrupoleOrbitrap mass spectrometer equipped with an Ion MaxTM source with a heated electrospray ionization (HESI) probe, both from Thermo Fisher Scientific (Bremen, Germany), and an MXT715-000 - MX Series II Switching Valve (IDEX Health & Science LLC, Oak Harbor, WA, USA). Calibration of the instrument was conducted with Pierce™ LTQ Velos ESI Positive Ion Calibration Solution from Life Technologies (Vienna, Austria) and ammonium hexafluorophosphate. The instrument settings for analysis of mAbs and Etanercept were as follows: source heater temperature of 250 °C, spray voltage of 3.5 kV, sheath gas flow of 30 arbitrary units, auxiliary gas flow of 15 arbitrary units, capillary temperature of 320 °C, Slens RF level of 80.0, AGC target of 1e6, and a maximum injection time of 100 ms. Intact protein measurements were performed in full scan mode in a range of m/z 1,500 – 3,000 at

a resolution setting of 140,000 at m/z 200 and averaging of 10 microscans. Middle-down experiments were carried out applying AIF in the higher-energy collisional dissociation (HCD) cell at a normalized collision energy (NCE) setting of 24 with a scan range of m/z 1,000–3,000 (corresponding to 96 eV collision energy for a singly charged molecule at m/z 2,000), and a resolution setting of 140,000 at m/z 200. For peptide mapping each scan cycle consisted of a full scan at a scan range of m/z 300–2,000 and a resolution setting of 70,000, followed by 10 data-dependent HCD scans at 29 NCE at a resolution setting of 17,500. Data Evaluation. Data analysis was conducted using Chromeleon 7.2 (Thermo Fisher Scientific, Germering, Germany). Oxidation in the standard substance was determined by relative quantification of the UV areas using the Chromeleon SmartPeaks integration assistant option with the exponentially skimmed riders integration type. The chromatograms were processed with Gaussian smoothing of 7 points. For quantification based on MS data the three most intense Fc/2 glycovariants were chosen. Extracted ion current chromatograms (XICCs) of the seven most intense charge states were generated using the five most intense isotopic peaks. For quantitation of diagnostic fragments obtained by AIF, XICCs of the most intense isotopomer plus the six adjacent isotope peak (±1.5 Da) isotopes of the 2+ charge states of the two most intense diagnostic fragments for methionine oxidation, namely b34 and y45, were used. Calibration curves were generated based on the integrated peaks of five inter-day replicates per concentration covering five consecutive overlapping concentration ranges: 0.60–2.4 ng µL–1, 1.2–6.0 ng µL–1, 6.0–24.0 ng µL–1, 12.0– 60.0 ng µL–1 and 24.0–120.0 ng µL–1. Concentrations and limits of detection (LODs) as well as limits of quantification (LOQs) were calculated based on linear regression analysis. Additionally, a confidence interval

ACS Paragon Plus Environment

3

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 4 of 9

for p = 95% was determined. Linearity was checked by applying a Mandel’s fitting test to all measured concentration ranges. Deconvolution of AIF-MS fragment spectra was done with the Xtract algorithm integrated into Xcalibur 3.0 (Thermo Fisher Scientific, Sunnyvale, CA, USA). Sequence coverage of AIF-MS experiments was determined with ProSight Lite v1.3 Build 1.3.5744.16422 at a mass tolerance of 10 ppm (http://prosightlite.northwestern.edu) provided by the Kelleher Research Group (Northwestern University, Evanston, IL, USA).52 Peptide mapping data were evaluated with BioPharma Finder 1.0 (Thermo Fisher Scientific, Sunnyvale, CA, USA) with the following settings: 5 ppm mass accuracy, minimum confidence of 0.8, specificity ‘high’ for trypsin protease, Nterminal pyroglutamate formation as fixed modification and the following variable modifications: built-in N-glycan library for chinese hamster ovary cell lines, oxidation, deamidation, NH3-loss, and H2O-loss, maximum allowed number of variable modifications of 2 per peptide. RESULTS AND DISCUSSION Chromatographic Separation of mAb Subunit Oxidation Variants. Since attempts towards the separation of mAb oxidation variants by IP-RP-HPLC at the intact protein level were not successful, we implemented proteolysis with IdeS in order to obtain smaller protein entities.31-51 Digestion with

Figure 2. IP-RP-HPLC separation of Rituximab stressed for 30 min with 0.35% hydrogen peroxide upon digestion with IdeS, with disulfide bonds intact (a) or reduced (b). Chromatographic conditions are given in the experimental section. Peaks were assigned based on AIF-MS fragment spectra and are annotated corresponding to mAb subunits shown in Figure 1. IdeS generated two mAb subunits: Fc/2 and F(ab’)2 (see Figure 1a). Subsequent reduction of inter- and intramolecular disulfide bonds with TCEP yielded three antibody subunits, namely two pairs of heavy chain (HC) fragments: Fc/2 & Fd’ and a pair of light chains (LC), all comprising a molecular mass of around 25 kDa (Figure 1a). Considering two possible oxidation sites in Fc/2, we therefore expect six different non-oxidized and oxidized species of

Figure 3. Fragment ion spectrum of singly oxidized Fc/2 subunit (peak 2,3 in Fig. 2b) obtained upon AIF at 96 eV in the HCD cell. The diagnostic fragments b34 (red) for methionine residue Met256 and y45 (blue) for Met432 are annotated in the raw-spectrum (a) and position of the b34 and y45 fragments used for quantification in the Fc/2 sequence (b). antibody fragments upon IdeS and TCEP treatment, as indicated in Figure 1 a and b. The benefit of this reduction in the molecular mass from ~150 kDa to ~25 kDa of the analytes is twofold: improvement of chromatographic separation as well as feasibility of spectrum acquisition at higher resolution settings. Separation of the antibody subunits was performed with a super-macroporous, polymeric reversed-phase stationary phase developed for antibody separations53 (150 x 2.1 mm i.d. MAbPac RP, 1500 Å average pore size) and a very shallow gradient of acetonitrile in 0.10% TFA. As shown in Figure 2a, partial separation of the four differently oxidized Fc/2 species with intact intramolecular disulfide bonds was achieved. Upon reduction of disulfide bridges, baseline-separation of non-, mono-, and doubly oxidized species was accomplished (Figure 2b, corresponding spectra are shown in Figure S-2). This opens up the possibility for reliable quantification based on UV detection, while the specificity of mass spectrometric detection is able to ensure that no other potentially UVabsorbing substances are coeluting with the target analytes. Although a very shallow gradient was necessary to allow separation of the oxidized Fc/2 variants, retention times were robust with relative intra- and inter-day standard deviations in retention times ranging between 0.68 and 1.62% (see supporting Excel file). With regard to retention behavior, oxidized species, both with intact or reduced disulfides, eluted earlier in IP-RP-HPLC, indicating weaker interaction with the stationary phase compared to their non-oxidized counterparts. This is in agreement with previous reports on the separation of oxidized protein variants and may be attributed to a decrease in hydrophobicity due to incorporation of oxygen and/or a significant change in protein folding upon methionine oxidation.28,29,54

ACS Paragon Plus Environment

4

Page 5 of 9

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

Oxidation Site Assignment in Fc/2. In order to assign the oxidation sites within Fc/2, AIF was performed in the HCD cell of a Q ExactiveTM hybrid quadrupole-Orbitrap mass spectrometer.28,29 For this purpose, a mass range of m/z 1,000– 3,000 covering multiple charge states of Fc/2 was fragmented without precursor selection in the quadrupole. This strategy is favorable for molecules occurring in multiple charge states as well as glycosylation variants as it increases overall signal intensity. Figure 3 depicts a fragment ion spectrum of singlyoxidized Fc/2 (peak 2,3 in Figure 2b) obtained by AIF-MS applying a collision energy of 96 eV in the HCD cell. The fragments b34 (red) diagnostic for methionine residue Met256 and y45 (blue) diagnostic for Met432, which are present both in the non-oxidized and oxidized form, are annotated in the mass spectrum (Figure 3a). The position of the fragments in the sequence of Fc/2 can be deduced from Figure 3b. At a collision energy of 42 eV, a total of 44 sequence-specific b- and yfragments were detectable, covering 21% of the Fc/2 subunit sequence (Figure S-3). Nevertheless, additional ions were observed within the isotope patterns (Figure S-4 a, b), thus impairing correct quantification. Therefore, a relatively high collision energy of 96 eV was employed, which efficiently eliminated interfering ions through improved fragmentation at higher collision energy (Figure S-4c, d), while still providing sufficient signal intensity for confident quantification. Oxidation site assignment was achieved based on mass shifts corresponding to additional oxygen in the respective diagnostic fragments. Accordingly, peak 1 in Figure 2 was assigned as doubly-oxidized Fc/2, as indicated by +16 Da shifts in the two fragments incorporating the methionines Met256 and Met432, whereas peak 3 originated from nonoxidized Fc/2 (for corresponding XICCs see Figure S-5). As the two singly-oxidized variants coelute in peak 2,3 (Figure 2b), the diagnostic fragments b34 and y45 are present in both their non-oxidized as well their oxidized form (Figure 3a). Evaluation of Quantification Strategies for mAb Subunits. As oxidation of biopharmaceuticals is a shelf-life limiting factor,55,56 a fast, accurate, and robust method for absolute quantification of oxidation is required in quality control, process analytics, and drug product analytics. In this context we aimed at establishing a method for absolute quantification by means of UV-spectroscopy, primarily because of higher robustness and easier implementation in a good manufacturing practice (GMP) environment. In addition to UV detection, we integrated MS detection for the purpose of oxidation site as-

signment, the characterization of additional modifications such as glycosylation, and as a complementary method for analyte quantification. It was previously noted for absolute quantification of oxidation in Pegfilgrastim that, depending on the concentration range, detection by UV-spectroscopy or MS was favorable upon IP-RP-HPLC. More specifically, UVspectroscopy was preferable for high-abundant components, while MS proved to be better for lower concentration ranges.29 We here evaluated four approaches for the absolute quantification of oxidation in a hydrogen peroxide-stressed Rituximab sample: (I) UV-absorbance spectroscopy at 214 nm, (II) extracted ion current chromatograms (XICCs) as well as (III) total ion current chromatograms (TICCs) based on full scan MS, and (IV) XICCs of two diagnostic fragments generated upon AIF-MS. To cover a broad concentration range for absolute quantification, concentration-response curves for each of the four approaches were generated for five overlapping concentration ranges, starting from 0.60 ng µL–1 up to 120 ng µL–1 of Rituximab Fc/2. For this purpose, we used a non-expired Rituximab sample with a minimal amount of oxidation variants present. In total, 16 concentrations were analyzed from low to high concentration in five technical replicates within three days (see Supporting Information Excel file). To avoid possible carryover effects, four blank injections were performed between the series, revealing no Fc/2 carryover above LOD. Peak parameters in quintuplicate analysis, e.g. for the mono-oxidized Fc/2 peak of Rituximab (with reduced disulfide bonds), were very robust, as corroborated by relative standard deviations (RSD) of 0.87%, 0.75%, 1.80%, and 2.46% for retention time, peak area, peak height, and peak width at half height, respectively, which are comparable to previous studies.29,57 All calibration curves were evaluated by means of linear regression analysis and tested for linearity with Mandel’s fitting test (for calibration line parameters and linearity testing, see Supporting Information Excel file). From the calibration equations we deduced LODs in the range of 3.2-14.4 ng µL-1 and LOQs of 9.6–43.2 ng µL–1 for Fc/2 (Table 1). Relative standard deviations in peak areas (N=5) at the LOQ were 7.8% for UV detection, 8.5% for TICC, 13% for XICC, and 9.1% for AIF-MS (b34/y45) monitoring, which qualifies the UVspectroscopy mode as the most robust and reliable detection. In addition, method coefficients of variation of 7.2–13.6% demonstrate that absolute quantification is feasible with all five approaches at accuracies of better than 15%.

Table 1. Limits of Detection (LODs), Limits of Quantification (LOQs), and Statistical Parameters of the Different Quantification Approaches for Fc/2 UV TICC XICC AIF-MS (b34) AIF-MS (y45) LODa),b) [ng µL–1] a),c)

LOQ

–1

[ng µL ] –1

R² (12–60 ng µL ) RSD%

d)

0.32

0.41

0.45

1.44

1.11

0.96

1.24

1.36

4.32

3.33

0.9975

0.9956

0.9950

0.9878

0.9863

7.8

8.5

12.9

9.0

9.1

e)

Vx0% 7.2 9.4 10.3 13.6 10.5 b) Mean values from five injections carried out within 3 days. Sample volume was 10 µL in all cases; LODs calculated according to the regression line method using the slope b and the residual standard deviation sy of the 0.60–2.4 ng µL–1 calibration curves for UV, TICC, and XICC measurements and 1.2–6.0 ng µL–1 for AIF-MS measurements; c)LOQ=3 LOD; d)Relative standard deviation a)

sy /b

of peak areas at LOQ, N=5 over 3 days; e)Relative process standard deviation with Vx0 = .100%, sy residual standard deviation of x the calibration, b, slope of the calibration line, and x , mean of standard concentrations, determined with 0.60–2.4 ng µL–1 calibration curves for UV, TICC and XICC measurements and 1.2–6.0 ng µL–1 for AIF-MS measurements. 5

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 6 of 9

Table 2. Absolute Quantification of Fc/2 Oxidation Variants of Hydrogen Peroxide-Stressed Rituximab by Six Different Approaches Concentration [ng µL–1]a) Method (I) UV (II) TICC (III) XICC (IV) AIF b34(+O) (IV) AIF y45(+O) b) Peak 1 15.6 ± 0.97 20.6 ± 1.2 14.4 ± 1.4 23.9 ± 2.0 12.1 ± 2.4 49.7 ± 0.93 46.5 ± 1.2 43.5± 1.3 27.8 ± 2.0 14.6 ± 2.3 Peak 2,3b) Peak 4b) 26.6 ± 0.90 31.4 ± 1.2 26.9 ± 1.3 28.4 ± 2.0 28.5 ± 2.1 91.9 ± 0.83 98.4 ± 3.7 84.8 ± 3.1 88.8 ± 3.8f) Total Fc/2c) Mean absolute per-centage error 0 19 (n=3) 7 (n=3) 12 (n=3) [%]d) e) Met oxidized [%] 44.1 44.5 42.7 44.1 a) Mean values ± confidence interval for p=95% of 5 consecutive measurements analogous to the separation depicted in Figure 2b. b) Peak annotation corresponds to Figure 1; c)Sum of Peaks 1–4 ± pooled confidence interval for p=95%; d)

Mean absolute percentage error=

100 n

∑ni=1 

actual value-conventional true value

 %; e)Percentage of oxidized Met256 or Met432 relative to the

conventional true value concentration singly oxidized Peaks 2, 3+2*concentration doubly oxidized (Peak 1)

total number of methionines present in Fc/2; Met oxidized % = . 2*concentration total Fc/2 f) In the summation, average values for Peak 1 and Peak 4 were taken, whereas for Peak 2,3 the values for b34+O and y45+O were added. Absolute Quantification of Fc/2 Variants Obtained from Oxidatively Stressed Rituximab. As no oxidized or non-oxidized Fc/2 calibration standard for Rituximab is available, calibration curves were generated with non-expired, nonstressed drug product. Relying on comparative UVspectroscopic measurements, the use of non-oxidized drug product material has previously been proven viable for absolute quantification of protein oxidation.29 Based on the relative peak areas in the UV-trace of oxidized and non-oxidized variant of the 120 ng µL–1 calibration standard used for method evaluation, the degree of oxidation in Rituximab drug product was determined at 1.8%. The concentration of all calibration standards used for absolute quantification was corrected by this value (see Supporting Information Excel file). In order to absolutely quantify oxidation in Fc/2 of Rituximab stressed with 0.35% hydrogen peroxide, subunits were generated as described above and measured in five technical replicates. The calibration curves derived from non-stressed drug product were employed for absolute quantification of methionine oxidation in Fc/2 of the stressed Rituximab sample. Indeed, the calculated concentrations of non-, singly-, and doubly-oxidized Fc/2 matched well among the four different quantification strategies (Table 2). If we take the results of UV-spectroscopy measurements as conventional true values,29 then the XICC method (III) having 7% mean absolute percentage error clearly outperforms the other mass spectrometrybased quantification approaches (II and IV), which yielded mean average percentage errors between 12 and 19% (Table 2). Intriguingly, the measured total Fc/2 concentrations determined by all mass spectrometric methods were within ±7.5% of the value obtained by UV-spectroscopy. The detection by AIF-b34/y45 monitoring is the only approach capable of differentiating between Met256 and Met432 oxidation in Fc/2, as it is based on selective diagnostic fragment ions. Quantification of non-oxidized Fc/2 via AIF-MS b34 or y45 monitoring yielded practically identical results that are in good agreement with the values obtained by UV measurements, proving that quantification employing this mass spectrometric technique is reliable. On the other hand, the values for doubly oxidized Fc/2 obtained by b34 or y45 monitoring differ by almost a factor of two, although the measured

concentration should be identical. The obtained values imply that b34+O monitoring overestimates the concentration of double-oxidized Fc/2, while measurement of y45+O results in underestimation of the oxidation product. This clearly corroborates earlier reports that oxidation significantly impacts relative intensities of fragment ions generated by collisioninduced dissociation of oxidized or non-oxidized proteins.29 Interestingly, monitoring of fragments specific for the two methionines in Fc/2 showed an almost twofold higher abundance of Met256 oxidation as compared to Met432 oxidation (Table 2). This result is also nicely confirmed by the peak areas of the four partially separated species in the UV chromatogram shown in Figure 2a, corresponding to ratios of 19:32:23:16% (Met256&432 oxidized:Met256 oxi-

Figure 3. Chromatogram of IP-RP-HPLC separation of expired drug products of Rituximab (a), Adalimumab (b), and Etanercept (c) upon IdeS digestion and disulfide bond reduction. Adalimumab and Etanercept were additionally treated with CpB. Peak annotation corresponds to Figure 1.

ACS Paragon Plus Environment

6

Page 7 of 9

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

Table 3. Absolute Quantification of Fc/2 Oxidation Variants in Expired Rituximab, Adalimumab, and Etanercept Drug Products Using UV Detection Concentration of oxidized or non-oxidized Fc/2 speciesa) Rituximab Adalimumab Etanercept Peak 1 [ng µL–1]b) n.d.c) n.d.c) n.d.c) Peak 2,3 [ng µL–1] b) 4.3 ± 0.29 3.4 ± 0.29 2.9 ± 0.29 350.5 334.0 242.0 Total Fc/2 concentration [ng µL–1]d) Oxidation [%] 1.2 1.0 1.2 a) Mean values ±95% confidence interval of 5 consecutive measurements using UV detection; b)Peak annotation corresponding to Figure 1; c)n.d. = not detectable; d)The total Fc/2 concentration was determined by off-line UV-spectroscopy at 280 nm. dized:Met432 oxidized:non-oxidized). Finally, assessing the total percentage of oxidized methionines in Fc/2 (Table 2), we found that the values between 42.9 and 44.1% correlated very well for all four quantification approaches. Quantitative oxidation measurements based on peptide mapping yielded 50.0% oxidation of both Met256 and Met432 (Figure S-6, Table S-1), which is in good agreement with the results obtained on the level of Fc/2. The higher degree of oxidation determined by peptide mapping is most probably a consequence of oxidation occurring during extended sample preparation involving overnight digestion with trypsin.24,25 Absolute Quantification of Oxidation in Biopharmaceuticals. To demonstrate the generic applicability of our method, we absolutely quantified oxidation in three expired biopharmaceutical products, namely the IgG1 mAbs Rituximab and Adalimumab, and the Fc-fusion protein Etanercept. As UV-spectroscopy at 214 nm was not only the method exhibiting the lowest limits of detection and quantification (Table 1), but also showed the lowest RSDs in the lower concentration ranges, we pursued quantification by UV-spectroscopy. MS data were acquired in parallel to allow a more detailed characterization of oxidation and other modifications. Initial HPLC-UV-MS analyses of Adalimumab and Etanercept revealed that C-terminal lysine variants, commonly occurring in Fc-containing biopharmaceuticals, coeluted with the mono-oxidized Fc/2 species, pretending higher levels of oxidation (peak 2,3 in Figure S-7a).33 To eliminate this source of interference, Carboxypeptidase B was added to the IdeS digest, resulting in complete removal of C-terminal lysines in both Adalimumab and Etanercept (Figure S-7b). Carboxypeptidase B treatment was not required in case of Rituximab lacking these lysine variants. UV chromatograms of expired Rituximab, Adalimumab, and Etanercept obtained by the described method are displayed in Figure 3. Methionine oxidation was readily quantified by UV-spectroscopy in all three expired drug products, revealing 1.21% oxidation in Etanercept, 1.00% oxidation in Adalimumab, and 1.24% oxidation in Rituximab (Table 3). In contrast to hydrogen peroxide stressed Rituximab used for method development, we did not detect doubly-oxidized variants in the expired biopharmaceuticals. CONCLUSIONS Our combined HPLC-UV-MS method for characterization of Fc/2 variants obtained by proteolysis with IdeS under reducing conditions is generically applicable for absolute quantification of oxidation in the Fc portion of IgG1 mAbs as well as Fc-fusion proteins. The presented approach is based on chro-

matographic separation of Fc/2 oxidation variants by IP-RPHPLC combined with UV detection, allowing for absolute quantification within a concentration range from 1.8-120 ng µL–1 with RSDs below 7.8%. Thus, injection of about 2 µg of drug substance or drug product is sufficient to reliably quantify 1% of oxidized protein contamination. At higher concentrations ranging from 6.0–120 ng µL–1, quantification based on the integrated areas of total ion current chromatograms as well as extracted ion current chromatograms of full-scan mass spectra is feasible with RSDs below 13%. Moreover, AIF-MS also enables the distinction and quantification of monooxidized species modified either at Met256 or Met432 based on site-specific fragment ions. As a matter of convenience, both data acquisition and evaluation are performed using a single software platform. Finally, the assets of UV-based absolute quantification, i.e. robustness and simple method implementation, combined with the power of MS-based indepth characterization render our method especially attractive for industrial applications ranging from quality control to process and drug product analytics. ASSOCIATED CONTENT Supporting Information Supporting Information available: This material is available free of charge via the Internet at http://pubs.acs.org. AUTHOR INFORMATIONS Corresponding Author *Phone: (+) 43 662 8044 5738. Fax: (+)43 662 8044 5751. Email: [email protected] Notes. The authors declare the following competing financial interest(s): Johann Holzmann is employee of Sandoz GmbH, which provides financial support for the Christian Doppler Laboratory for Innovative Tools for Biosimilar Characterization. The salary of Therese Wohlschlager is fully funded; Christian G. Huber’s salary is partly funded by the Christian Doppler Laboratory for Biosimilar Characterization. The authors declare no other competing financial interest. ACKNOWLEDGMENTS The financial support by the Austrian Federal Ministry of Science, Research, and Economy and by a Start-up Grant of the State of Salzburg is gratefully acknowledged. We thank Ines Forstenlehner from Sandoz GmbH and Frank Steiner from Thermo Fisher Scientific for technical support as well as scientific discussions.

ACS Paragon Plus Environment

7

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

REFERENCES (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)

Walsh, G. Nat Biotechnol 2010, 28, 917-924. Beck, A.; Wurch, T.; Bailly, C.; Corvaia, N. Nat Rev Immunol 2010, 10, 345-352. Leader, B.; Baca, Q. J.; Golan, D. E. Nat Rev Drug Discov 2008, 7, 21-39. Hmiel, L. K.; Brorson, K. A.; Boyne, M. T., 2nd Anal Bioanal Chem 2015, 407, 79-94. Berlec, A.; Strukelj, B. J. Ind. Microbiol. Biotechnol. 2013, 40, 257-274. Beck, A.; Sanglier-Cianferani, S.; Van Dorsselaer, A. Anal. Chem. 2012, 84, 4637-4646. Liu, H.; Gaza-Bulseco, G.; Faldu, D.; Chumsae, C.; Sun, J. J. Pharm. Sci. 2008, 97, 2426-2447. Huang, L. J.; Chiang, C. W.; Lee, Y. W.; Wang, T. F.; Fong, C. C.; Chen, S. H. J Chromatogr B Analyt Technol Biomed Life Sci 2016, 1032, 189-197. Kuriakose, A.; Chirmule, N.; Nair, P. J Immunol Res 2016, 2016, 1298473. Shacter, E. Drug Metab Rev 2000, 32, 307-326. Gaza-Bulseco, G.; Faldu, S.; Hurkmans, K.; Chumsae, C.; Liu, H. Journal of Chromatography B 2008, 870, 55-62. Lefranc, M. P.; Pommie, C.; Kaas, Q.; Duprat, E.; Bosc, N.; Guiraudou, D.; Jean, C.; Ruiz, M.; Da Piedade, I.; Rouard, M.; Foulquier, E.; Thouvenin, V.; Lefranc, G. Dev Comp Immunol 2005, 29, 185-203. Chumsae, C.; Gaza-Bulseco, G.; Sun, J.; Liu, H. J Chromatogr B Analyt Technol Biomed Life Sci 2007, 850, 285-294. Liu, H.; Gaza-Bulseco, G.; Zhou, L. J Am Soc Mass Spectrom 2009, 20, 525-528. Vogt, W. Free Radic Biol Med 1995, 18, 93-105. Houde, D.; Kauppinen, P.; Mhatre, R.; Lyubarskaya, Y. J Chromatogr A 2006, 1123, 189-198. Liu, D.; Ren, D.; Huang, H.; Dankberg, J.; Rosenfeld, R.; Cocco, M. J.; Li, L.; Brems, D. N.; Remmele, R. L., Jr. Biochemistry 2008, 47, 5088-5100. Mo, J.; Yan, Q.; So, C. K.; Soden, T.; Lewis, M. J.; Hu, P. Anal Chem 2016. Bertolotti-Ciarlet, A.; Wang, W.; Lownes, R.; Pristatsky, P.; Fang, Y.; McKelvey, T.; Li, Y.; Li, Y.; Drummond, J.; Prueksaritanont, T.; Vlasak, J. Mol Immunol 2009, 46, 18781882. Tsuchida, D.; Yamazaki, K.; Akashi, S. Pharm Res 2016, 33, 994-1002. Wang, W.; Vlasak, J.; Li, Y.; Pristatsky, P.; Fang, Y.; Pittman, T.; Roman, J.; Wang, Y.; Prueksaritanont, T.; Ionescu, R. Mol Immunol 2011, 48, 860-866. Gao, X.; Ji, J. A.; Veeravalli, K.; Wang, Y. J.; Zhang, T.; McGreevy, W.; Zheng, K.; Kelley, R. F.; Laird, M. W.; Liu, J.; Cromwell, M. J Pharm Sci 2015, 104, 368-377. Stracke, J.; Emrich, T.; Rueger, P.; Schlothauer, T.; Kling, L.; Knaupp, A.; Hertenberger, H.; Wolfert, A.; Spick, C.; Lau, W.; Drabner, G.; Reiff, U.; Koll, H.; Papadimitriou, A. MAbs 2014, 6, 1229-1242. Chelius, D.; Rehder, D. S.; Bondarenko, P. V. Anal. Chem. 2005, 77, 6004-6011. Du, Y.; Wang, F.; May, K.; Xu, W.; Liu, H. Anal Chem 2012, 84, 6355-6360. Beck, A.; Wagner-Rousset, E.; Ayoub, D.; Van Dorsselaer, A.; Sanglier-Cianferani, S. Anal. Chem. 2013, 85, 715-736. Haverick, M.; Mengisen, S.; Shameem, M.; Ambrogelly, A. MAbs 2014, 6, 852-858. Holzmann, J.; Hausberger, A.; Rupprechter, A.; Toll, H. Anal Bioanal Chem 2013, 405, 6667-6674. Forstenlehner, I. C.; Holzmann, J.; Toll, H.; Huber, C. G. Anal Chem 2015, 87, 9336-9343.

Page 8 of 9

(30) von Pawel-Rammingen, U.; Johansson, B. P.; Bjorck, L. EMBO J 2002, 21, 1607-1615. (31) An, Y.; Zhang, Y.; Mueller, H. M.; Shameem, M.; Chen, X. MAbs 2014, 6, 879-893. (32) Sokolowska, I.; Mo, J.; Dong, J.; Lewis, M. J.; Hu, P. MAbs 2017, 0. (33) Zhang, B.; Jeong, J.; Burgess, B.; Jazayri, M.; Tang, Y.; Taylor Zhang, Y. J Chromatogr B Analyt Technol Biomed Life Sci 2016, 1032, 172-181. (34) Leblanc, Y.; Romanin, M.; Bihoreau, N.; Chevreux, G. J Chromatogr B Analyt Technol Biomed Life Sci 2014, 961, 1-4. (35) Tran, B. Q.; Barton, C.; Feng, J.; Sandjong, A.; Yoon, S. H.; Awasthi, S.; Liang, T.; Khan, M. M.; Kilgour, D. P.; Goodlett, D. R.; Goo, Y. A. J Proteomics 2016, 134, 93-101. (36) Chevreux, G.; Tilly, N.; Bihoreau, N. Anal Biochem 2011, 415, 212-214. (37) Francois, Y. N.; Biacchi, M.; Said, N.; Renard, C.; Beck, A.; Gahoual, R.; Leize-Wagner, E. Anal Chim Acta 2016, 908, 168176. (38) Hafkenscheid, L.; Bondt, A.; Scherer, H. U.; Huizinga, T. W.; Wuhrer, M.; Toes, R. E.; Rombouts, Y. Mol Cell Proteomics 2017, 16, 278-287. (39) Mittermayr, S.; Le, G. N.; Clarke, C.; Millan Martin, S.; Larkin, A. M.; O'Gorman, P.; Bones, J. J Proteome Res 2017, 16, 748762. (40) Kirley, T. L.; Greis, K. D.; Norman, A. B. Biochem Biophys Res Commun 2016, 477, 363-368. (41) Cotham, V. C.; Brodbelt, J. S. Anal Chem 2016, 88, 4004-4013. (42) Leblanc, Y.; Bihoreau, N.; Jube, M.; Andre, M. H.; Tellier, Z.; Chevreux, G. Eur J Pharm Biopharm 2016, 102, 185-190. (43) Ponniah, G.; Nowak, C.; Neill, A.; Liu, H. Anal Biochem 2017, 520, 49-57. (44) Sorensen, M.; Harmes, D. C.; Stoll, D. R.; Staples, G. O.; Fekete, S.; Guillarme, D.; Beck, A. MAbs 2016, 8, 1224-1234. (45) Novarra, S.; Grinberg, L.; Rickert, K. W.; Barnes, A.; Wilson, S.; Baca, M. MAbs 2016, 8, 1118-1125. (46) Liu, B.; Guo, H.; Zhang, J.; Xue, J.; Yang, Y.; Qin, T.; Xu, J.; Guo, Q.; Zhang, D.; Qian, W.; Li, B.; Hou, S.; Dai, J.; Guo, Y.; Wang, H. Mol Pharm 2016, 13, 2702-2710. (47) Li, Y.; Fu, T.; Liu, T.; Guo, H.; Guo, Q.; Xu, J.; Zhang, D.; Qian, W.; Dai, J.; Li, B.; Guo, Y.; Hou, S.; Wang, H. MAbs 2016, 8, 951-960. (48) Sjogren, J.; Olsson, F.; Beck, A. Analyst 2016, 141, 3114-3125. (49) Fornelli, L.; Ayoub, D.; Aizikov, K.; Beck, A.; Tsybin, Y. O. Anal Chem 2014, 86, 3005-3012. (50) Liu, T.; Guo, H.; Zhu, L.; Zheng, Y.; Xu, J.; Guo, Q.; Zhang, D.; Qian, W.; Dai, J.; Guo, Y.; Hou, S.; Wang, H. Chromatographia 2016, 79, 1491-1505. (51) Wang, B.; Gucinski, A. C.; Keire, D. A.; Buhse, L. F.; Boyne, M. T., 2nd Analyst 2013, 138, 3058-3065. (52) Fellers, R. T.; Greer, J. B.; Early, B. P.; Yu, X.; LeDuc, R. D.; Kelleher, N. L.; Thomas, P. M. Proteomics 2015, 15, 12351238. (53) Lin, S.; Zhang, T.; Wang, H.; Josephs, J.; Liu, X.; Application Note 21100: Fast Analysis of Therapeutic Monoclonal Antibody Fragments Using a Supermacroporous, ReversedPhase Chromatography Column, Thermo Scientific, 2015. (54) Toll, H.; Berger, P.; Hofmann, A.; Hildebrandt, A.; Oberacher, H.; Lenhof, H. P.; Huber, C. G. Electrophoresis 2006, 27, 2734-2746. (55) Goetze, A. M.; Schenauer, M. R.; Flynn, G. C. MAbs 2010, 2, 500-507. (56) Eon-Duval, A.; Broly, H.; Gleixner, R. Biotechnol Prog 2012, 28, 608-622. (57) Yin, Y.; Han, G.; Zhou, J.; Dillon, M.; McCarty, L.; Gavino, L.; Ellerman, D.; Spiess, C.; Sandoval, W.; Carter, P. J. MAbs 2016, 8, 1467-1476.

ACS Paragon Plus Environment

8

Page 9 of 9

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

Insert Table of Contents artwork here

ACS Paragon Plus Environment

9