Characterization of Recombinant Monoclonal Antibody Charge

May 15, 2015 - WAX-10 chromatograms from analysis of the purified fractions; deconvoluted mass spectra to show Fc-related modifications; deconvoluted ...
1 downloads 10 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Characterization of Recombinant Monoclonal Antibody Charge Variants Using OFFGEL Fractionation, Weak Anion Exchange Chromatography, and Mass Spectrometry Alyssa Neill, Christine Nowak, Rekha Patel, Gomathinayagam Ponniah, Nidia Gonzalez, Dino Miano, and Hongcheng Liu* Product Characterization, Alexion Pharmaceuticals Inc, 352 Knotter Drive, Cheshire, Connecticut 06410, United States S Supporting Information *

ABSTRACT: Recombinant monoclonal antibody charge heterogeneity has been commonly observed as multiple bands or peaks when analyzed by charge-based analytical methods such as isoelectric focusing electrophoresis and cation or anion exchange chromatography. Those charge variants have been separated by some of the above-mentioned methods and used for detailed characterization. The utility of a combination of OFFGEL fractionation and weak anion exchange chromatography to separate the charge variants of a recombinant monoclonal antibody was demonstrated in the current study. Charge variants were separated into various fractions of high purity and then analyzed thoroughly by liquid chromatography mass spectrometry. Analysis of intact molecular weights identified the presence of heavy chain leader sequence, C-terminal lysine, and C-terminal amidation. The identified modifications were further localized into different regions of the antibody from analysis of antibody fragments obtained from FabRICATOR digestion. Analysis of tryptic peptides from various fractions further confirmed the previously identified modifications in the basic variants. Asparagine deamidation and aspartate isomerization were identified in acidic fractions from analysis of tryptic peptides. Basic variants have been fully accounted for by the identified modifications. However, only a portion of the acidic variants can be explained by deamidation and isomerization, suggesting that additional modifications are yet to be identified or acidic variants are an ensemble of molecules with different structures.

H

observed basic species. However, complete understanding of acidic variants is lacking. One of the major hypotheses for the lack of understanding of acidic variants is that acidic variants are formed due to extremely low levels of Asn deamidation at multiple residues. Several methods have been commonly used to monitor charge variants of recombinant monoclonal antibodies including isoelectric focusing (IEF) gel electrophoresis,11,12,25 capillary isoelectric focusing (cIEF) electrophoresis,6,16 and cation6,11,12,14,16,20 and anion25,26 exchange chromatography. Ideally, for ease of data interpretation, charge variants should be collected using the same separation techniques as the ones used for monitoring charge variants. In this sense, fraction collection from chromatography methods can be readily achieved using the same analytical columns or using the same type of columns at a larger size. In contrast, it is almost impossible to collect fractions directly from cIEF. Fortunately, the OFFGEL fractionator, which separates proteins based on the same

eterogeneity in recombinant monoclonal antibodies (mAbs) is common and is reflected in the presence of multiple variants differing in biochemical and biophysical properties such as molecular weights, hydrophobicity, and charge.1−5 Charge variation is one of the most commonly observed and extensively characterized aspects of heterogeneity. It is routinely monitored throughout the development of recombinant monoclonal antibody therapeutics. Charge variants can be classified as acidic or basic depending on their isoelectric points (pI) relative to the main species. Charge variants with a relatively lower pI are referred to as acidic variants, while charge variants with a relatively higher pI are referred to as basic variants. Several known modifications have been reported to contribute to the formation of acidic variantssuch as the presence of sialic acid,6−10 deamidation of asparagine (Asn) residues,7,8,11−15 and glycation.8,16 On the other hand, the presence of C-terminal lysine (Lys),4,6,7,11,12 Nterminal glutamine (Gln),7,17−19 C-terminal amidation,20,21 and the formation of succinimide12,22−24 from isomerization of aspartate (Asp)residues are commonly observed modifications that form basic variants. In general, basic variants have been relatively easy to be characterized. The aforementioned basic variants forming modifications can usually account for all the © XXXX American Chemical Society

Received: March 2, 2015 Accepted: May 15, 2015

A

DOI: 10.1021/acs.analchem.5b01452 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

times in deionized water, each for 5 min. The gel was stained in 100 mL SimplyBlue Safe Stain (Life Technologies) and 10 mL 20% (w/v) sodium chloride overnight. The gel was destained in 100 mL of deionized water for 1 h. After 1 h, the deionized water was changed and 20 mL of 20% (w/v) sodium chloride was added to stabilize the stain. FabRICATOR Digestion. The samples were diluted to 0.5 mg/mL using 20 mM sodium phosphate, pH 7.0. Each vial (5000u) of FabRICATOR was reconstituted in 200 μL water. To each 100 μL of the 0.5 mg/mL sample, 1 μL of the reconstituted FabRICATOR was added. The samples were incubated at room temperature for 60 min and then analyzed by LC-MS. The digested samples were also analyzed by LC-MS after reduction using 20 mM DTT at room temperature for 30 min. LC-MS Analysis of Molecular Weights of Intact Antibody and Its Fragments. An Agilent 1100 series high performance liquid chromatography (HPLC) system, an LC/ MSD TOF mass spectrometer (Agilent, Santa Clara, CA) and a Vydac C4 (1.0 × 150 mm) column were used to measure the molecular weights of the antibody and its fragments. For intact molecular weight analysis, approximately 10 μg of each sample was injected into the column at 80% mobile phase A (0.1% TFA, 5% acetonitrile in water) and 20% mobile phase B (0.1% TFA and 80% acetonitrile in water). After 1 min, the percentage of mobile phase B was increased to 100% and remained at 100% for 10 min. The column was then equilibrated with 20% mobile phase B before the next injection. For analysis of antibody fragments, samples were loaded at 20% mobile phase B. After 5 min, the percentage of mobile phase B was increased to 95% within 30 min. The column was then washed using 95% mobile phase B and equilibrated using 20% mobile phase B prior to the next injection. Throughout the analysis, the flow-rate was set at 50 μL/min and the column temperature was set at 60 °C. The mass spectrometer was run in positive mode with a scan range of m/z 900−6000 for intact and 600−3000 for antibody fragments. A gas temperature of 350 °C and drying gas flow of 10 L/min were used for both intact antibody and its fragments. The nebulizer was set at 20 psig for intact and at 45 psig for antibody fragment analysis. The skimmer was set at 75 V for both intact and fragment analyses. The fragmentor voltage was set at 425 V for intact and at 350 V for antibody fragment analysis. LC-MS Analysis of Peptides. The recombinant monoclonal antibody and the purified fractions were denatured and reduced using 6 M guanidine hydrochloride in 20 mM Tris, pH 8.0, and 10 mM dithiothreitol (DTT) at 37 °C for 30 min. The samples were then alkylated using 25 mM iodoacetic acid at 37 °C for 30 min. The samples were then buffer exchanged into 20 mM Tris, pH 8.0 using Zeba column (Thermoscientific, Rockford, IL). Each sample was digested using trypsin at a final 1:10 (w:w) trypsin to antibody ratio at 37 °C for 4 h. The digested samples were stored at −20 °C until analysis. Tryptic peptides were analyzed using a Maxis 4 G mass spectrometer (Bruker, Billerica, MA) and an ultraperformance liquid chromatography (UPLC) system (Waters) with a Proto 200 C18 column (1.0 × 250 mm, Higgins Analytical. Inc.). The samples were loaded at 98% mobile phase A (0.1% TFA in water) and 2% mobile B (0.1% TFA in acetonitrile). After 5 min, mobile phase B was increased to 35% over 158 min and then to 60% over 20 min. The column was then washed and equilibrated. The column was heated at 45 °C and the flow-rate was set at 50 μL/min. The mass spectrometer was tuned and

mechanism as cIEF, has made fraction collection by this technique possible as demonstrated previously.27 In the current study, the feasibility of collecting recombinant monoclonal antibody charge variants using an OFFGEL Fractionator was further explored. Here, we demonstrated that fractions collected from an OFFGEL fractionator can be further purified using weak anion exchange (WAX) chromatography. Using the combination of chromatography and the OFFGEL fractionator, highly pure charge variants can be obtained for all charge variants. Characterization of basic fractions showed antibody variants with N-terminal leader sequence, C-terminal Lys, and C-terminal amidation. On the other hand, only low levels of Asn deamidation, Asp isomerization, and a species with molecular weight higher than the heavy chain were detected in acidic variants.

■

MATERIALS AND METHODS Materials. The recombinant monoclonal antibody was expressed in Chinese hamster ovary (CHO) cell line and purified at Alexion (Cheshire, CT). Acetonitrile, ammonium bicarbonate, formic acid, iodoacetic acid, and trifluoroacetic acid (TFA) were purchased from Sigma (St, Louis, MO). FabRICATOR was purchased from Genovis (Cambridge, MA). Trypsin was purchased from Worthington (Lakewood, NJ). OFFGEL Fraction Collection. An OFFGEL fractionator 3100 (Agilent, Santa Clara, CA) was used to separate the antibody charge variants. A 5 mg portion of sample per strip was focused on a ReadyStrip pH 5−8 IPG strip for 72 h at 10 °C using a pH 5.5−6.7/pH 4−7 (1:1 v:v) buffer mix. Sixteen replicate strips were run simultaneously. After focusing, fractions were collected and pooled. The pooled fractions were concentrated and buffer exchanged into 10 mM sodium phosphate, pH 7.0 using Amicon Ultra-4 centrifuge device with a molecular weight cutoff of 30 kDa (Millipore, Billerica, MA). Protein concentrations were determined using UV absorption measured using a UV spectrophotometer and the theoretical extinction coefficient calculated based on the amino acid sequence. Weak Anion Exchange Chromatography. A Waters Alliance HPLC and a weak anion exchange (WAX) column (WAX-10, 4.6 × 250 mm, Thermoscientific, Sunnyvale, CA) were used for further separation of antibody charge variants. Mobile phase A contains 20 mM Tris, pH 8.0. Mobile phase B is 20 mM Tris, 0.5 M sodium chloride, pH 8.0. Samples were injected at 72% mobile phase A and 28% mobile phase B. After 5 min, the percentage of mobile phase B was increased to 60% over 40 min. The column was washed using 100% mobile phase B and then equilibrated using 28% mobile phase B. Proteins eluted off the column were monitored using UV214 nm and UV280 nm. The flow-rate was set at 1 mL/min and the column was set at ambient temperature throughout the runs. Fraction collection was guided by UV absorption. The collected fractions were concentrated using Amicon Ultra-4 centrifuge devices with a molecular weight cutoff of 30 kDa. Isoelectric Focusing Electrophoresis. OFFGEL samples were run on a Novex pH 3−10 IEF gel to verify separation. A 3 μg portin of each fraction were mixed 1:1 (v:v) with 2× IEF pH 3−10 sample buffer (Life Technologies, Norwalk, CT) and loaded onto the gel. The gel was run at a constant 100 V for 1 h followed by a constant 200 V for 1 h and then a constant 500V for 30 min for a total run time of 2.5 h. After the run, the gel was removed from the cassette and fixed for 15 min in 12% trichloroacetic acid (TCA). After fixing, the gel was washed 3 B

DOI: 10.1021/acs.analchem.5b01452 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

agreement with the observation from IEF analysis, one major peak was observed for the acidic variants and the main band fractions, while, more than one peak was observed for the basic variant fractions (data not shown). The fractions were further purified for in-depth characterization. Charge variants in the fractions were further purified using a WAX-10 column. The purity was substantially improved. Only one band by IEF (Figure 2) and one peak by WAX

calibrated following the manufacturer’s procedure and run at the positive scan mode with m/z in the range of 150−3000. Capillary voltage was set at 4500 V. Drying gas flow was set at 10 L/min, and gas temperature was set at 220 °C. The nebulizer was set at 2.0 bar.

■

RESULTS AND DISCUSSION The workflow of the study is shown in Figure 1. In brief, charge variants of the mAb were separated by an OFFGEL fractionator

Figure 2. IEF analysis of fractions purified by OFFGEL freationator and WAX-10 chromatography. Lanes 1 and 10 are pI marker. Lane 2 is the unfractionated starting material. Lanes 3−8 correspond to fractions A, B, C, D, E, and F, respectively. Note, fractions C and D migrate to the same pI.

Figure 1. Workflow of the experiments.

into several fractions. The purity of each fraction was analyzed by IEF and WAX-10. Charge variants in each fraction were further separated by WAX-10 to obtain fractions containing only one band by IEF or one peak by WAX-10. The highly purified fractions were then thoroughly characterized using several methods. Intact molecular weights of antibody in each fraction were analyzed by LC-MS to detect modifications with substantial molecular weight difference including C-terminal Lys, C-terminal amidation, and the presence of a portion of leader sequence. LC-MS was then used to analyze the molecular weights of F(ab′)2 or Fc fragments obtained from FabRICATOR digestion of each fraction to further localize the identified modifications. LC-MS and LC-MS/MS was used to analyze tryptic peptides from the digestion of each fraction to confirm the modifications detected from the molecular weight analysis. Also it allowed for the detection of modifications with minimal molecular weight difference or no molecular weight difference. The fractions were also analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) to determine the amount of aggregates and fragments. Oligosaccharides from each fraction were released by PNGaseF digestion, labeled by reductive amidation using 2-aminobenzamide (2-AB) and then analyzed using normal-phase high performance liquid chromatography (NP-HPLC) with fluorescence detection. Analysis of OFFGEL Fractions Using IEF and WAX-10. Fractions collected from the OFFGEL fractionator were first analyzed by IEF gel electrophoresis. One major band was observed for the acidic variants and the main band fractions, while one or two major bands and additional minor bands were observed for the basic fractions (data not shown). The same fractions were also analyzed by WAX chromatography. In

(Supplemental Figure 1) were observed for each fraction. Fractions A and B contain acidic variants. Fractions C and D have the same pI by IEF, but their retention times by WAX-10 column differ slightly. Fraction C corresponds to the main peak. Fraction D corresponds to the front shoulder of the main peak. Fraction F shows one major band by IEF; however, it contains a main peak and a shoulder by WAX-10. The slightly different purities observed by IEF and WAX-10 suggest that WAX-10 either has a higher resolution than IEF or it can recognize not only the overall charge but also the charge distribution of the antibody. Identification of Modifications. In order to identify modifications forming the charge variants, fractions A−F were analyzed by LC-MS at intact, fragment, and peptide levels. Analysis at the intact levels provides information on modifications that are present at significantly higher levels with substantial molecular weight differences than the calculated molecular weights. Analysis of fragments can further localize the modifications to regions of the antibody. Ultimately, the samples were analyzed at the peptide levels to precisely localize the modifications detected by molecular weight measurement of the intact antibody and antibody fragments. In addition, modifications with a minimal molecular weight difference or no molecular weight differences such as deamidation and isomerization can only be detected at peptide levels. Intact Molecular Weight Analysis. Mass spectra from analysis of the intact antibody molecular weights are shown in Figure 3. The peak identities are summarized in Table 1. Peak C

DOI: 10.1021/acs.analchem.5b01452 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 3. Deconvoluted mass spectra from analysis of the antibody starting material and fractions A−F, as labeled in each spectrum. Identities of various peaks are summarized in Table 1.

assignments are based on the common modifications of recombinant monoclonal antibodies reported in literature and the observed molecular weights of each peak. Starting Material. As shown in Figure 3, the main peak molecular weight from analysis of the starting material is 148612 Da, which corresponds to the antibody without Cterminal Lys and with G0F on both heavy chains. The peak with the molecular weight of 148 409 Da, approximately 203 Da lower than the main peak, corresponds to the same species as the main peak but with one heavy chain with G0F lacking an

GlcNAc. The molecular weight of the shoulder on the right side of the main peak is approximately 32 Da higher than the main peak and probably indicates the presence of a trisulfide bond. The molecular weight of 148 739 Da, approximately 127 Da higher than the main peak, corresponds to the antibody with one C-terminal Lys. The molecular weight of 148 774 Da, approximately 162 Da higher than the main peak, corresponds to the antibody with one heavy chain with G0F and the other heavy chain with G1F. The peak with the molecular weight of 148 852 Da is approximately 240 Da higher than the main peak. D

DOI: 10.1021/acs.analchem.5b01452 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

amidation. C-terminal amidation has been widely observed for monoclonal antibodies.20,21,29 In addition, a peak with a molecular weight of 149 433 Da was observed, which corresponds to the addition of a portion of the leader sequence. Based on the peak profiles, fraction E contains antibody variants with one heavy chain with C-terminal Lys or one heavy chain with C-terminal amidation or with the presence of leader sequence. The major peak in fraction F has a molecular weight of 148 869 Da which is approximately 256 Da higher than the main peak in the starting material. This molecular weight corresponds to the antibody with two C-terminal Lys. Fraction F contains several small peaks with molecular weights corresponding to the antibody with partial leader sequences on the heavy chain. The peak with the molecular weight of 149 378 Da corresponds to the antibody with one C-terminal amidation and the other heavy chain with leader sequence. The peak with the molecular weight of 149 562 Da corresponds to the antibody with one C-terminal Lys and one heavy chain with leader sequence. Overall, fraction F contains antibody variants with either two C-terminal Lys, one C-terminal Lys with leader sequence, and one C-terminal amidation with leader sequence. Molecular Weight Analysis of Antibody Fragments after Digestion Using FabRICATOR. FabRICATOR is a valuable tool that cleaves antibodies into two major fragments corresponding to F(ab′)2 or Fc fragment, thus localizing the modifications to different regions.30−34 Mass spectra of the F(ab′)2 fragment are shown in Figure 4. Several peaks with the same molecular weights were observed in all fractions. The main peak with the molecular weight of 98214 Da corresponds to the F(ab′)2 fragment of the antibody with N-terminal pyroglutamate on both heavy chains. The peak right after the main peak with the molecular weight of 98 246 Da is also commonly observed in all fractions. The molecular weight of this peak is approximately 32 Da higher than the main peak, which corresponds to the presence of one trisulfide bond. The peak with the molecular weight of 98 274 Da is approximately 64 Da higher than the main peak and thus corresponds to the presence of two trisulfide bonds. The peak with the molecular weight of 98 454 Da, approximately 240 Da higher than the main peak, corresponds to the antibody with cysteinylation of two Cys residues which is also present in all fractions. From analysis of the antibody F(ab′)2 fragments, trisulfide bond and cysteinylation were localized to the F(ab′)2 fragment. This observation is in agreement with previous studies, where the trisulfide bond has been localized mainly between the light chain and the heavy chain and to a lesser extent between the two heavy chains.26,35 Several differences were also observed among different fractions. First, the relative levels of the peak with two trisulfide bonds are higher in fractions A and B. The presence of two trisulfide bonds may contribute to the peak with the molecular weight of 148 667 Da detected from LC-MS analysis of intact antibody, which is approximately 53 Da higher than the main peak in acidic fractions A and B (Figure 3A and B). The peak with the molecular weight increase of 53 Da at intact level could be due to a mixture of one and two trisulfide bonds or other modifications. Second, relatively higher levels of heterogeneity were observed for fractions A and B compared to other fractions. This is demonstrated by the presence of many poorly resolved small peaks with molecular weights higher than the main peak. Third, only fractions E and F contain heavy chain with leader sequences, which is in

Table 1. Summary of Peak Identities from Figure 3 MW (Da) 148409 148554 148612 148739 148774 148852 148869 148936 149378 149433 149562 149688

modifications

fractions

No C−K; 1HC with G0F and 1HC with G0FGIcNAc 1HC without C−K and 1HC with C-terminal amidation, 2HC with G0F No C−K, 2HC with G0F 1HC without C−K and 1HC with C−K, 2HC with G0F No C−K, 1HC with G0F, and 1HC with G1F Cysteinylation of 2 Cys residues (240 Da higher than 148612) 2HC with C−K, 2HC with G0F No C−K, 2HC with G1F or 1HC with G0F and 1HC with G2F Leader sequence (820 Da higher than 148554) Leader sequence (820 da higher than 148612) Leader sequence (820 Da higher than 148740) Leader sequence (820 Da higher than 148869)

A, B, C, D E, F A, B, C, D E A, B, C, D A,B,C,D F A, B, C, D F E F F

This molecular weight increase corresponds to cysteinylation of two cysteine residues. The peak with the molecular weight of 149 433 Da does not correspond to any known modifications. However, this peak corresponds to the molecular weight of the antibody with one heavy chain with partial leader sequence Acidic Variants. Acidic variants were enriched in fractions A and B. The peak profiles from fractions A and B are not as well-defined as the peak profile obtained from analysis of the starting material. Several poorly resolved peaks with molecular weights higher than the main peak were observed, which suggests potential modifications. The other notable difference is the absence of peaks corresponding to the antibody with Cterminal Lys. This observation is expected because antibody variants with C-terminal Lys are more basic and should be enriched in basic fractions. The peak with the molecular weight of 148 667 Da is approximately 53 Da higher than the main peak molecular weight. Modifications corresponding to this peak cannot be determined from matching molecular weight to any known modifications. This peak can also be formed due to the partial overlapping of more than one modification. Main Component. The major variants of the antibody were enriched in fractions C and D. The intact molecular weight peak profiles of fractions C and D are very similar to those obtained from analysis of the starting material, which is expected because fractions C and D account for the majority of the starting material. Fractions C and D were analyzed by RPHPLC to determine the distribution of the disulfide bond related isoforms, since it has been well-documented that IgG2 has three disulfide bond related isoforms, termed A form, A/B form and B form.28 A relatively higher amount of A/B isoform was detected in fraction C and a relatively higher amount of B was detected in fraction D (data not shown). The difference in disulfide bond linkage did play a role in the separation of fractions C and D, in addition to other potential modifications. Basic Variants. The basic variants were enriched in fractions E and F. The peak profile of fraction E is very different from the other fractions. The peak with the molecular weight of 148 741 Da is approximately 128 Da higher than the molecular weight of the main peak observed in the starting material thus it likely corresponds to the antibody with one Cterminal Lys. The peak with the molecular weight of 148 554 Da, approximately 58 Da lower than the calculated molecular weight, likely corresponds to the antibody with C-terminal E

DOI: 10.1021/acs.analchem.5b01452 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 4. Deconvoluted mass spectra of the antibody F(ab′)2 fragments generated from FabRICATOR digestion of the starting material and fractions A−F, as labeled for each spectrum.

agreement with the observation from analysis of the intact antibody. Mass spectra of the Fc portion from analysis of fractions A−F are shown in Supplemental Figure 2. Analysis of the Fc portion confirmed earlier observations. Fractions A−D contain Fc without C-terminal Lys. Fraction E contains antibody with one heavy chain without C-terminal Lys and the other heavy chain with either C-terminal Lys or C-terminal amidation. Fraction F contains antibody with C-terminal Lys on both heavy chains or one heavy chain with C-terminal Lys and the other containing C-terminal amidation. The peak without C-terminal Lys was

also observed in fraction F. This observation suggests that there is a population of antibody variants with one C-terminal Lys and one heavy chain with the leader sequence. The antibody fragments obtained from FabRICATOR digestion were also analyzed after reduction to further localize modifications to smaller fragments. For example, the leader sequence was confirmed to be associated with the heavy chain (Supplemental Figure 3). As expected, the peaks with molecular weight increases of 32 and 64 Da corresponding to the presence of one or two trisulfide bonds and also the peak with a molecular weight increase of 240 Da from cysteinylation F

DOI: 10.1021/acs.analchem.5b01452 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 5. Extracted ion current (EIC) chromatogram of the tryptic peptide with amino acid sequence of GFYPSDIAVEWESNGQPENNYK from fraction A. Similar profile was obtained from analysis of fraction B. (inset) EIC of the same peptide from fraction C. Mass spectra of doubly charged ions are shown as insets.

nonreducible covalent bond does not contribute to the formation of acidic species contained in fractions A and B. However, an extra band with a molecular weight slightly higher than the heavy chain was observed in fractions A and B, but not in fraction C. No difference in the oligosaccharide profiles was observed, suggesting oligosaccharides are not the cause of acidic variants in fractions A and B. Peptide mapping was used to further confirm the identified modifications in the basic fractions. As an example, MS and MS/MS spectra of the N-terminal heavy chain tryptic peptide containing a portion of the leader sequence are shown in Figure 6. The observed molecular weight of 2088.17 Da calculated

can no longer be detected because of reduction by DTT. Lastly, very clean spectra without much heterogeneity were observed for all the fragments from all fractions. This observation suggests that some of the heterogeneity observed for the intact antibody and F(ab′)2 fragment are due to modifications of cysteine residues including trisulfide bonds and cysteinylation that disappeared upon DTT reduction. Peptide Mapping. Fractions A−F were also analyzed by LC-MS peptide mapping with and without reduction to search for modifications with minimal or no molecular weight difference or modifications that are present at very low levels that cannot be detected from analysis of intact antibody and antibody fragments. Additionally, LC-MS peptide mapping can further confirm and localize modifications that were observed in the basic variants. Tryptic peptides were analyzed by searching for known modifications including Asn deamidation, Gln deamidation, Asp isomerization, glycation of Lys side chain and N-terminal amine, and the presence of free cysteine residues. In addition, the data was carefully evaluated for the presence of any new peaks. The only detected differences among those fractions are slightly higher levels of Asn deamidation and Asp isomerization in fractions A and B. As shown in Figure 5, four peaks were observed for the tryptic peptide with the amino acid sequence of GFYPSDIAVEWESNGQPENNYK. Peaks 1, 3, and 4 correspond to various deamidation products because of a molecular weight increase of 1 Da compared to peak 2 (Figure 5, inset). Peak 1 and 4 contain IsoAsp and Asp from deamidation of the first Asn residue and peak 3 contains Asp from deamidation of the second Asn from the peptide Nterminus.36 While similar percentages of peak 1 and 4 were observed in fractions A−C, a higher percentage of peak 3 (3.5% in fraction A, 4.4% in fraction B, 0.2% in fraction C) was observed in fractions A and B. A slightly increased level of isomerization of the Asp residue in the amino acid sequence of GFYDGYSP was also observed in fractions A and B (Supplemental Figure 4). The levels of the peptide with isoAsp were 4.5% in fraction A, 3.6% in fraction B and 0.7% in fraction C. Other than deamidation and isomerization, no other modifications were identified by LC-MS peptide mapping. Fractions A and B side-by-side with fraction C were also analyzed by SDS-PAGE (Supplemental Figure 5) to determine fragments and aggregates and NP- HPLC (Supplemental Figure 6) with fluorescence detection to determine oligosaccharide profiles. No differences in the levels of aggregates and fragments were observed in fractions A−C, indicating a

Figure 6. MS and MS/MS spectra of the peptide containing a portion of the leader sequence. The amino acid sequence of this peptide is shown on top of this figure. (A) Full scan mass spectrum of the triply charged peptide. (B) MS/MS spectrum. Matching of the major fragment ions in the spectrum to the predicted Y and B ions from the amino acid sequence confirm the identity of this peptide.

based on the triply charged ion is in agreement with the theoretical molecular weight of 2088.13 Da (Figure 6A). The correct sequence was also confirmed by MS/MS experiment, where multiple B and Y ions were observed corresponding to B and Y ions that are predicted from the amino acid sequence (Figure 6B). The presence and absence of C-terminal Lys and C-terminal amidation were also confirmed by accurate molecular weight measurement and MS/MS fragmentation (data not shown). G

DOI: 10.1021/acs.analchem.5b01452 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

■

Allison, D. E.; Wong, P. Y.; Kao, Y. H.; Quan, C.; Joshi, A.; Harris, R. J.; Motchnik, P. MAbs. 2010, 2 (6), 613−624. (9) Zhang, T.; Bourret, J.; Cano, T. J. Chromatogr. A 2011, 1218 (31), 5079−5086. (10) Alvarez, M.; Tremintin, G.; Wang, J.; Eng, M.; Kao, Y. H.; Jeong, J.; Ling, V. T.; Borisov, O. V. Anal. Biochem. 2011, 419 (1), 17− 25. (11) Perkins, M.; Theiler, R.; Lunte, S.; Jeschke, M. Pharm. Res. 2000, 17 (9), 1110−1117. (12) Harris, R. J.; Kabakoff, B.; Macchi, F. D.; Shen, F. J.; Kwong, M.; Andya, J. D.; Shire, S. J.; Bjork, N.; Totpal, K.; Chen, A. B. J. Chromatogr. B Biomed. Sci. Appl. 2001, 752 (2), 233−245. (13) Zhang, W.; Czupryn, M. J. J. Pharm. Biomed. Anal. 2003, 30 (5), 1479−1490. (14) Vlasak, J.; Bussat, M. C.; Wang, S.; Wagner-Rousset, E.; Schaefer, M.; Klinguer-Hamour, C.; Kirchmeier, M.; Corvaia, N.; Ionescu, R.; Beck, A. Anal. Biochem. 2009, 392 (2), 145−154. (15) Gandhi, S.; Ren, D.; Xiao, G.; Bondarenko, P.; Sloey, C.; Ricci, M. S.; Krishnan, S. Pharm. Res. 2012, 29 (1), 209−224. (16) Quan, C.; Alcala, E.; Petkovska, I.; Matthews, D.; Canova-Davis, E.; Taticek, R.; Ma, S. Anal. Biochem. 2008, 373 (2), 179−191. (17) Moorhouse, K. G.; Nashabeh, W.; Deveney, J.; Bjork, N. S.; Mulkerrin, M. G.; Ryskamp, T. J. Pharm. Biomed. Anal. 1997, 16 (4), 593−603. (18) Ouellette, D.; Alessandri, L.; Chin, A.; Grinnell, C.; Tarcsa, E.; Radziejewski, C.; Correia, I. Anal. Biochem. 2010, 397 (1), 37−47. (19) Xu, W.; Peng, Y.; Wang, F.; Paporello, B.; Richardson, D.; Liu, H. Anal. Biochem. 2013, 436 (1), 10−12. (20) Johnson, K. A.; Paisley-Flango, K.; Tangarone, B. S.; Porter, T. J.; Rouse, J. C. Anal. Biochem. 2007, 360 (1), 75−83. (21) Kaschak, T.; Boyd, D.; Lu, F.; Derfus, G.; Kluck, B.; Nogal, B.; Emery, C.; Summers, C.; Zheng, K.; Bayer, R.; Amanullah, A.; Yan, B. MAbs. 2011, 3 (6), 577−583. (22) Yan, B.; Steen, S.; Hambly, D.; Valliere-Douglass, J.; Vanden Bos, T.; Smallwood, S.; Yates, Z.; Arroll, T.; Han, Y.; Gadgil, H.; Latypov, R. F.; Wallace, A.; Lim, A.; Kleemann, G. R.; Wang, W.; Balland, A. J. Pharm. Sci. 2009, 98 (10), 3509−3521. (23) Chu, G. C.; Chelius, D.; Xiao, G.; Khor, H. K.; Coulibaly, S.; Bondarenko, P. V. Pharm. Res. 2007, 24 (6), 1145−1156. (24) Sreedhara, A.; Cordoba, A.; Zhu, Q.; Kwong, J.; Liu, J. Pharm. Res. 2012, 29 (1), 187−197. (25) Tsai, P. K.; Bruner, M. W.; Irwin, J. I.; Ip, C. C.; Oliver, C. N.; Nelson, R. W.; Volkin, D. B.; Middaugh, C. R. Pharm. Res. 1993, 10 (11), 1580−1586. (26) Pristatsky, P.; Cohen, S. L.; Krantz, D.; Acevedo, J.; Ionescu, R.; Vlasak, J. Anal. Chem. 2009, 81 (15), 6148−6155. (27) Meert, C. D.; Brady, L. J.; Guo, A.; Balland, A. Anal. Chem. 2010, 82 (9), 3510−3518. (28) Wypych, J.; Li, M.; Guo, A.; Zhang, Z.; Martinez, T.; Allen, M. J.; Fodor, S.; Kelner, D. N.; Flynn, G. C.; Liu, Y. D.; Bondarenko, P. V.; Ricci, M. S.; Dillon, T. M.; Balland, A. J. Biol, Chem. 2008, 283 (23), 16194−16205. (29) Tsubaki, M.; Terashima, I.; Kamata, K.; Koga, A. Int. J. Biol. Macromol. 2013, 52, 139−147. (30) Lynaugh, H.; Li, H.; Gong, B. MAbs. 2013, 5 (5), 641−645. (31) Fornelli, L.; Ayoub, D.; Aizikov, K.; Beck, A.; Tsybin, Y. O. Anal. Chem. 2014, 86 (6), 3005−3012. (32) Chevreux, G.; Tilly, N.; Bihoreau, N. Anal. Biochem. 2011, 415 (2), 212−214. (33) Beck, A.; Wagner-Rousset, E.; Ayoub, D.; Van, D. A.; SanglierCianferani, S. Anal. Chem. 2013, 85 (2), 715−736. (34) An, Y.; Zhang, Y.; Mueller, H. M.; Shameem, M.; Chen, X. MAbs. 2014, 6 (4), 879−893. (35) Gu, S.; Wen, D.; Weinreb, P. H.; Sun, Y.; Zhang, L.; Foley, S. F.; Kshirsagar, R.; Evans, D.; Mi, S.; Meier, W.; Pepinsky, R. B. Anal. Biochem. 2010, 400 (1), 89−98. (36) Gaza-Bulseco, G.; Li, B.; Bulseco, A.; Liu, H. C. Anal. Chem. 2008, 80 (24), 9491−9498.

CONCLUSIONS Charge variants were first separated by OFFGEL fractionator and then further purified by WAX-10 chromatography. The collected fractions were characterized by molecular weight measurement of intact antibody and antibody fragments generated from FabRICATOR digestion. From the molecular weight analyses, modifications in the basic variants were identified, including the presence of a partial leader sequence, C-terminal Lys, and C-terminal amidation. However, no modifications were identified from the molecular weight analyses for the acidic variants. The fractions were further analyzed by LC-MS and LC-MS/MS peptide mapping. Analysis at peptide level confirmed the modifications that were identified in the basic fractions from the molecular weight measurement. Slightly increased levels of deamidation of an Asn residue, isomerization of an Asp residue, and the presence of low level of a species with molecular weight higher than the heavy chain were the only modifications identified in the acidic variants. However, those modifications can only account for a small percentage of the acidic variants. On the other hand, basic variants can be accounted for by the identified modifications. Based on this characterization, it was hypothesized that acidic variants could contain multiple sources of modifications with minimal or no molecular weight difference and each is present at a relatively low level. It is also possible that acidic variants are an ensemble of antibodies with slightly different structures.

■

ASSOCIATED CONTENT

S Supporting Information *

WAX-10 chromatograms from analysis of the purified fractions; deconvoluted mass spectra to show Fc-related modifications; deconvoluted mass spectrum of reduced F(ab’)2 and Fc; EIC chromatogram of the triply charged peptide containing isoAsp from Asp isomerization; SDS-PAGE of acidic fractions A and B; fluorescence chromatograms of acidic fractions. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b01452.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: 203-271-8354. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Liu, H.; Gaza-Bulseco, G.; Faldu, D.; Chumsae, C.; Sun, J. J. Pharm. Sci. 2008, 97 (7), 2426−2447. (2) Vlasak, J.; Ionescu, R. Curr. Pharm. Biotechnol. 2008, 9 (6), 468− 481. (3) Du, Y.; Walsh, A.; Ehrick, R.; Xu, W.; May, K.; Liu, H. MAbs. 2012, 4 (5), 578−585. (4) Beck, A.; Bussat, M. C.; Zorn, N.; Robillard, V.; KlinguerHamour, C.; Chenu, S.; Goetsch, L.; Corvaia, N.; Van, D. A.; Haeuw, J. F. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2005, 819 (2), 203−218. (5) Beck, A.; Sanglier-Cianferani, S.; Van, D. A. Anal. Chem. 2012, 84 (11), 4637−4646. (6) Santora, L. C.; Krull, I. S.; Grant, K. Anal. Biochem. 1999, 275 (1), 98−108. (7) Lyubarskaya, Y.; Houde, D.; Woodard, J.; Murphy, D.; Mhatre, R. Anal. Biochem. 2006, 348 (1), 24−39. (8) Khawli, L. A.; Goswami, S.; Hutchinson, R.; Kwong, Z. W.; Yang, J.; Wang, X.; Yao, Z.; Sreedhara, A.; Cano, T.; Tesar, D.; Nijem, I.; H

DOI: 10.1021/acs.analchem.5b01452 Anal. Chem. XXXX, XXX, XXX−XXX