ARTICLE IN PRESS
JID: CHROMA
[m5G;February 11, 2020;13:33]
Journal of Chromatography A xxx (xxxx) xxx
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Harnessing the power of electrophoresis and chromatography: Offline coupling of reverse phase liquid chromatography-capillary zone electrophoresis-tandem mass spectrometry for peptide mapping for monoclonal antibodies Ramesh Kumar, Rohan L. Shah, Anurag S. Rathore∗ Department of Chemical Engineering, Indian Institute of Technology Delhi, New Delhi, India
a r t i c l e
i n f o
Article history: Received 3 December 2019 Revised 5 February 2020 Accepted 6 February 2020
Keywords: Reverse phase liquid chromatography Capillary zone electrophoresis Mass spectrometry Peptide mapping Monoclonal antibody
a b s t r a c t RPLC-MS/MS is the present workhorse for bottom-up proteomics-based characterization. However it suffers from limited peak capacity and challenges with respect to detection of small and hydrophilic peptides. CZE-MS/MS offers an orthogonal alternative to RPLC-MS/MS but has limited sensitivity due to low sample loading. In the present work, for the first time, offline coupling of an RPLC–CZE–MS/MS has been demonstrated for peptide mapping of a monoclonal antibody-based therapeutic product. The performance of this platform has been compared to the state-of-the-art RPLC-MS/MS and CZE-MS/MS approaches. Fifteen fractions were isolated from a mAb tryptic digest using RPLC, and each fraction was analyzed by CZE-ESI-MS/MS. Results indicate that not only the RPLC–CZE–MS/MS platform identified a larger number of distinct peptides (372) than the RPLC-MS/MS (219) and CZE-MS/MS (177) platforms, it also offered significantly superior sequence coverage (99.55% vs. 89.76% and 94.21% for the heavy chain and 98.6% vs. 92.06% and 95.79% for the light chain). Also, the distribution of amino acid residues with post translational modifications was 413 for RPLC–CZE–MS/MS, 150 for RPLC-MS/MS and 94 for CZE-MS/MS. In fact, the RPLC–CZE–MS/MS system performed better than the combined datasets from RPLC-MS/MS or CZEMS/MS. Our study shows that the traditionally used one-dimensional (1D) RPLC-MS/MS and CZE-MS/MSbased platforms may not be enough for characterization of complex molecules such as monoclonal antibodies. The proposed approach should be an essential addition to the analytical toolkit for in-depth primary structural characterization of mAbs. © 2020 Elsevier B.V. All rights reserved.
1. Introduction Reversed-phase liquid chromatography (RPLC) coupled with electrospray ionization– tandem mass spectrometry (ESI-MS/MS) is routinely used for peptide mapping in bottom-up proteomicsbased characterization [1,2]. The compatibility of RPLC with ESIMS/MS combined with its large column loading capacity offers a tool for high resolution peptide separations [3,4]. However, there are a few limitations associated to this technique including poor retention of small and hydrophilic peptides, limited peak capacity, and reduced sample throughput due to the need for reequilibration between separations [3,4]. Capillary zone electrophoresis (CZE) has long been offered as an alternative to HPLC. In its simplest mode, capillary electrophoresis ∗ Corresponding author: Department of Chemical Engineering, Indian Institute of Technology, Hauz Khas, New Delhi, 110016, India. E-mail address:
[email protected] (A.S. Rathore).
(CE) is used for separation of analytes in narrow, buffer filled capillaries under high electric fields based on their charge to size ratios [5–7]. With the introduction of improved CZE-ESI-MS/MS interfaces, it is increasingly being used for MS-based proteomics [7–14]. Currently, both RPLC and CZE are used for characterization of monoclonal antibodies (mAbs) [2,15,16]. While LC-MS/MS is considered as the gold standard in the industry, CZE-MS/MS is an emerging orthogonal alternative technique to RPLC-MS/MS [14,17,18]. Major advantages of CE over RPLC include much faster and highly efficient separation with very high peak efficiency (up to 10 0,0 0 0), low sample requirement (by one to three orders of magnitude), and efficient detection of small, basic and hydrophilic peptides [9]. Additionally, due to the orthogonal principle of separation, CZE produces complementary results when compared to RPLC, and it has been reported to perform better than RPLC, especially for mass-limited samples [7,19,20] . However, CZEMS/MS suffers from some major limitations. First, the high-speed separations and high efficiency of CE impede data acquisition
https://doi.org/10.1016/j.chroma.2020.460954 0021-9673/© 2020 Elsevier B.V. All rights reserved.
Please cite this article as: R. Kumar, R.L. Shah and A.S. Rathore, Harnessing the power of electrophoresis and chromatography: Offline coupling of reverse phase liquid chromatography-capillary zone electrophoresis-tandem mass spectrometry for peptide mapping for monoclonal antibodies, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2020.460954
JID: CHROMA 2
ARTICLE IN PRESS
[m5G;February 11, 2020;13:33]
R. Kumar, R.L. Shah and A.S. Rathore / Journal of Chromatography A xxx (xxxx) xxx
rate of the mass spectrometer. This limits the number of spectra acquired in a single CZE-MS/MS analysis. Even when operating at 10 Hz, MS may be able to acquire only 60 0 0 tandem mass spectra in ten minutes. Thus, it is difficult to analyze complex samples by CZE in a single run. Further, the limited loading capacity of CZE (few nanoliters only) compromises MS sensitivity and limits the detection of trace components in the sample [1,3,9]. Sample analysis with both online and offline pre-fractionation strategies have been reported in the literature to improve protein and peptide identifications with CE [21]. While online prefractionation limits sample loss and delivers higher sensitivity, it suffers from poor resolution due to compatibility issues between the pre-fractionation step and the following CZE [3]. On the other hand, the off-line pre-fractionation strategy allows for ease of optimization, use of commercially available instruments, lack of restrictions on separation and fractionation strategy, sample manipulation between the two dimensions, and thereby results in improved resolution [21,22]. Researchers have employed online solidphase microextraction (SPME) to fractionate a mixture of 75 yeast ribosomal proteins into 11 fractions, each of which was analyzed by CZE-MS/MS, and 66 proteins were identified [23]. In another study, an improved SPME strategy was employed by coupling transient isotachophoresis with CZE-MS/MS to study a bacterial protein mixture and 2341 peptides and 548 proteins could be identified [3]. Recently, 5700 proteoforms from E. coli proteome were identified using offline pre-fractionation based CZE-MS/MS [24]. Pre-fractionation coupled CZE-MS/MS has also been reported for highly sensitive bottom-up analysis of complex human proteomes [25,26]. Researchers have also reported achieving comparable performance from offline pre-fractionation based CZE-MS/MS than with the state-of-the-art two-dimensional (2D) LC-MS/MS [25,27]. Multidimensional protein identification technology (MudPIT) is another 2D chromatographic method that combines the strong cation exchange (SCX) and RPLC techniques in a single biphasic column, and allows 2D LC analysis of a sample on a 1D LC-MS system [28]. The technique however uses SCX in the first dimension which has limited peak capacity for tryptic peptides that get sparsely populated in the SCX chromatogram due to limited charge range of +1 to +5. Such 2D LC systems are convenient to use but offer limited orthogonality and are time consuming [29]. In another report, 2D SCX-SPE-CZE-MS/MS method with detachable sample concentrator for bottom-up proteomics analysis was reported [30]. The authors obtained significant improvement with online pH bumps but the technique has limitations of limited peak capacity in the first dimension as is evident by the excessive sample retention on the column even after first five pH bumps, huge disparity between the number of peptides eluted with different pH bumps, and limited sequence coverage of 62% on the test sample bovine serum albumin which was comparable to their single shot CZE analysis. This shows that the majority of the tryptic peptides were of similar charge that co-eluted at higher pH. In addition, the authors have used isocratic elution method which would have led to co-elution of multiple peptides. Recently, 2D LC-MS/MS-based peptide mapping of IgG1 has been reported using hydrophilic liquid interaction chromatography (HILIC), SCX, and RPLC in the first dimension and RPLC in the second dimension [31]. These techniques, however, offer limited orthogonality [29]. Another study on NIST mAb analyzed hundreds of 1D and 2D RPLC-MS/MS runs to generate a spectral library of the mAb [32]. In this study, for the first time, we have established an offline RPLC–CZE–ESI MS/MS-based platform for peptide mapping of a mAb biotherapeutic product. A stepwise approach of platform evaluation and performance comparison has been undertaken. The performance of the developed platform has been compared with the state-of-the-art RPLC-MS/MS and CZE-MS/MS platforms.
The results show a significant improvement in sequence coverage and in identification of peptides and differently modified residues with the proposed RPLC–CZE–MS/MS platform. Our study shows that the traditionally used 1D RPLC-MS/MS and CZE-MS/MS-based platforms may not be enough for characterization of complex molecules such as monoclonal antibodies. 2. Materials and methods 2.1. Materials A purified mAb sample (IgG1, 148 kDa, pI 8.5) was obtained as a gift from a leading domestic biopharmaceutical company and used in this study as the model protein. Proteomics grade trypsin was obtained from Agilent Technologies (Santa Clara, CA, US). Tris (hydroxymethyl) aminomethane (Tris), ammonium bicarbonate (NH4 HCO3 ), and urea were purchased from Sigma (St. Louis, MO, US). Dithiothreitol (DTT) was obtained from Sisco Research Laboratories (SRL, Maharashtra, India), iodoacetamide (IAM) was from Spectrochem (Maharashtra, India), acetic acid from S D Fine Chemicals Limited (SDFCL, Maharashtra, India), and hydrochloric acid (HCl) from Beckman Coulter (Brea, CA, US). Nanosep 10 K omega cut off filter from Pall Life Sciences (NY, US) and 0.22 μm nylon filter from Axiva Sichem Biotech (Delhi, India) were used. MS grade formic acid (FA), MS grade ammonium acetate, MS grade water, and MS grade acetonitrile (ACN) were purchased from Fisher Scientific (Hampton, NH, US). OptiMS Silica surface cartridge (30 μm X 91 cm) was purchased from SCIEX (Framingham, MA, US). Buffer solutions were prepared using MS grade water, filtered through the 0.22 μm filter, and degassed before use. 2.2. Methods 2.2.1. Sample preparation For digestion, mAb sample (~1 mg) was dried using CentriVap (Labconco, Kansas City, MA, US) and resuspended in 100 μl of 6 M urea in 100 mM Tris–HCl (pH 7.8) for solubilization and denaturation by gentle pipetting. About 5 μl of 200 mM DTT was added to the sample for reduction and the sample was incubated for 1hr at room temperature (RT, 25 °C), followed by addition of 20 μl of 200 mM IAM for alkylation and then by incubation for 1hr in the dark at RT. Excess IAM was quenched with the addition of 20 μl of 200 mM DTT. The sample was mixed by gentle vortex and allowed to stand at RT for 1hr. The whole volume was passed through 10 kDa cut off filter and buffer exchanged with 100 mM NH4 HCO3 . MS grade proteolytic enzyme trypsin was added (1 μg/μl) in the ratio of 1:50, and the mixture was incubated for 16 h at 37°C. The reaction was arrested by lowering the pH to less than 6 by adding 1 μl of FA. The digest was concentrated by vacuum evaporation using CentriVap and stored at −20°C until further use. 2.2.2. RPLC-MS/MS An Exion LC (SCIEX) chromatographic system was coupled to a SCIEX Triple TOF® 6600 system for RPLC-MS/MS analysis. The mAb digest was resuspended in 0.1% FA in water, and 100 ng was loaded and resolved on AdvanceBio peptide mapping C18 (4.6 × 150 mm, 2.7 μm, Agilent Technologies) column operated at 50 °C using 0.1% FA in water (buffer A) and 0.1% FA in ACN (buffer B). A gradient from 3–30% B was applied for 23 min, followed by 30–50% B in the next 5 min at a flow rate of 0.2 ml/min. Detection was performed using information dependent acquisition (IDA) mode using a Turbo Spray III ionization source. The data were acquired in technical duplicates. 2.2.3. CZE-MS/MS A CESI 80 0 0 Plus (Beckman Coulter, Brea CA, US) system along with an OptiMS silica surface cartridge (30 μm X 91 cm, SCIEX)
Please cite this article as: R. Kumar, R.L. Shah and A.S. Rathore, Harnessing the power of electrophoresis and chromatography: Offline coupling of reverse phase liquid chromatography-capillary zone electrophoresis-tandem mass spectrometry for peptide mapping for monoclonal antibodies, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2020.460954
ARTICLE IN PRESS
JID: CHROMA
[m5G;February 11, 2020;13:33]
R. Kumar, R.L. Shah and A.S. Rathore / Journal of Chromatography A xxx (xxxx) xxx
was used to conduct all CZE based experiments. Solutions of 10% and 3% acetic acid in water were employed as background electrolytes (BGE) and conductive liquid, respectively. The capillary was first cleaned by rinsing it with 0.1 N HCl at 100 psi for 5 min. Equilibration was performed using BGE in reverse direction through the conductive line at 75 psi for 3 min, followed by a forward rinse at 100 psi for 10 min through separation capillary using BGE. The mAb digest was resuspended in 50 μl of 50 mM ammonium acetate (pH 4.0) and injected at 5.0 psi for 60 s, followed by 15 s injection of BGE at 2.5 psi. The separation was performed first at 30 kV for 45 min with 1 min ramp. Voltage was decreased to 1 kV for 5 min with 5 min voltage ramp. Data were acquired in IDA mode using a SCIEX Triple TOF® 6600 system with a Nano Spray III ionization source. Data were acquired in technical duplicates. 2.2.4. RPLC–CZE–MS/MS 100 μg of digested mAb was loaded and resolved on AdvanceBio peptide mapping C18 (4.6 × 150 mm, 2.7 μm, Agilent Technologies) RP column operated at 50 °C using 0.1% FA in water (buffer A) and 0.1% FA in ACN (buffer B). A gradient from 2–100% B was applied for 63 min at a flow rate of 0.2 ml/min for elution. Detection was performed by monitoring UV absorption at 214 nm. A total of 80 fractions were collected manually every minute and pooled into 15 fractions based on the presence of peptides, confirmed by timeof-flight-mass spectrometry (TOF-MS), while trying to keep minimal disparity between the number of peptides in each pooled fraction. The pooled fractions were vacuum dried using CentriVap and stored at −80°C until further use. The dried fractions were resuspended in 30 μl of 50 mM ammonium acetate (pH 4.0) and diluted to 1:10 ratio before use in CZE. The method for fractioned samples of CZE-MS/MS analysis was the same as that for mAb whole digest. All data were acquired in technical duplicates. 2.2.5. Data acquisition and analysis Data were acquired in IDA mode with 100 ms TOF-MS survey scan, and 50 ms IDA on the top 20 ions which exceed 100 cps. Fragmentation was induced via rolling collision energy with a dynamic exclusion time of 3s and total cycle time of 1.2996s. Optimization of IDA parameters was performed to allow the duty cycle of MS to support high-speed CE separation adequately. The instrument was calibrated once every two runs to ensure maximum mass measurement accuracy. MS scan range of 20 0–160 0 m/z and MS/MS scan range of 10 0–20 0 0 m/z were used for peptide fragmentation. Raw data was processed on Protein Pilot 5.2.0 software (SCIEX, Framingham, MA, US) using Paragon method at 10% detection threshold, along with biological modifications at 1% FDR [33]. The peptide sequences were matched with the FASTA sequence of the mAb downloaded from IMGT database [34]. Only peptides with a minimum best confidence score of 99 were considered for analysis. Wherever required, qualitative analysis of MS/MS spectra was performed using PeakView 2.0 software (SCIEX, Framingham, MA, US). 3. Results and discussion In our study, the state-of-the-art RPLC system was combined with CZE and sensitive MS detection for bottom-up characterization of an IgG1 monoclonal antibody. 3.1. Evaluation of the RPLC–CZE–MS/MS method Two-dimensional separation of tryptic mAb digest was performed, using RPLC in the first dimension and CZE in the second dimension. The digested mAb was separated via an RPLC system
3
into a total of 15 fractions and analyzed via CZE-MS/MS. The electropherograms obtained with each of the techniques are presented in Supplementary Figure 1. The efficiency of fractionation was confirmed by analyzing the cumulative distribution of peptides identified in RPLC fractions and is represented in Fig. 1a. The distribution increases almost linearly, indicating efficient fractionation and absence of significant overlap between fractions [3,24]. Co-elution is a frequent limitation in LC, and CZE is known to resolve those LC peaks having more than one co-eluting peptide [21,35,36]. The efficient separation in the RPLC–CZE–MS/MS method was confirmed by plotting the detection times of common peptides between RPLC-MS/MS and RPLC–CZE–MS/MS (Fig. 1b). Two peptides were defined as common if they shared the amino acid sequence, modification(s), position in the protein sequence, and the theoretical m/z. In Fig. 1b, each dot represents a common peptide obtained between the two techniques. Dots in a straight line parallel to either x or y-axis represent the peptides that coelute in one dimension (1D) and get separated in the second dimension (2D). For instance, peptides along the red dotted line (first line from the left) in Fig. 1b elute at approximately 11 min in RPLC but are separated in CZE between 14–16 min. These observations are in accordance with the orthogonal nature of RPLC and CZE [37] and confirm the separation of the eluted peptides in the 2D. The results demonstrate high-efficiency separation offered by the 2D. Next, the number of peptides obtained in RPLC–CZE–MS/MS in a given fraction was compared with the number of peptides obtained in the corresponding elution with RPLC-MS/MS (Fig. 1c). A variable number of peptides were identified for a given fraction with both the techniques. Except for fraction number 6, the number of peptides obtained in the RPLC–CZE–MS/MS method is much higher as compared to RPLC-MS/MS alone (Supplementary Table 1). Increase in several peptides in RPLC–CZE–MS/MS demonstrates the ability of the platform to resolve limited RPLC peaks that confirm the need and applicability of the platform for mAb characterization. 3.2. Comparison of RPLC–CZE–MS/MS with RPLC–MS/MS and CZE–MS/MS An essential requirement of regulatory agencies for any recombinant protein-based therapeutic is the analysis of the primary structure of the product [38,39]. Sequence coverage, as it is called, represents the percentage of the primary structure of a target protein that is identified by a given technique. Since mAbs are produced in the biological systems, any error during translation may affect the safety and efficacy of the product [14,40]. Achieving maximum sequence coverage is always the aim of any peptide mapping-based study. Fig. 2 represents the comparison of RPLC–CZE–MS/MS with RPLC-MS/MS and CZE-MS/MS. It is evident that RPLC–CZE–MS/MS outperforms both RPLC-MS/MS and CZE-MS/MS and yields a better sequence coverage for the mAb sample under consideration. The results obtained with RPLC-MS/MS and CZE-MS/MS are shown in Figs. 2a and b, respectively. While RPLC-MS/MS covered a large part of both the heavy and light chain (represented in green), few residues remained unidentified (depicted in grey). Most of the peptides containing these unidentified residues are small, basic, and hydrophilic (Supplementary Table 2) and RPLC-MS/MS is known to be inefficient with respect to the complete detection of such peptides. On the contrary, CZE-MS/MS is known to identify them efficiently [3,9]. RPLC–CZE–MS/MS covered most of these sequences not identified by either of the two techniques (Fig. 2c). The percentage of sequence covered by each of the technique is summarized in Fig. 2d. RPLC–CZE–MS/MS covered 99.55% of the heavy
Please cite this article as: R. Kumar, R.L. Shah and A.S. Rathore, Harnessing the power of electrophoresis and chromatography: Offline coupling of reverse phase liquid chromatography-capillary zone electrophoresis-tandem mass spectrometry for peptide mapping for monoclonal antibodies, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2020.460954
JID: CHROMA 4
ARTICLE IN PRESS
[m5G;February 11, 2020;13:33]
R. Kumar, R.L. Shah and A.S. Rathore / Journal of Chromatography A xxx (xxxx) xxx
Fig. 1. Evaluation of the RPLC–CZE–MS/MS method. The efficiency of the RPLC–CZE–MS/MS method was analyzed based on fractionation strategy, orthogonality, resolution of co-eluted RPLC peptides, and efficiency of 2D separation. (a) Fractionation efficiency: the cumulative number of peptides identified by CZE-MS/MS in each of the RPLC fraction are depicted. The red diagonal represents the trendline of the peptide distribution. (b) Two-dimensional orthogonality plot: each dot represents one peptide identified in common between RPLC-MS/MS and RPLC–CZE–MS/MS. A few representative groups of peptides that co-eluted in RPLC and are separated in the second dimension by CZE are depicted by vertical red dotted lines. (c) The efficiency of 2D separation. Peptides having the same amino acid sequence but differing in modification(s) were considered distinct. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 2. Sequence coverage map. Database search results of RPLC-MS/MS (a), CZE-MS/MS/MS (b), and RPLC–CZE–MS/MS/MS (c). Residues identified with >=95% confidence are written in green, those identified with confidence >=50% but less than 95% are in yellow. Residues in red color have been identified with less than 50% confidence, and the sequence written in grey color are the ones that have not been identified. The percentage of a sequence for the heavy chain, light chain, and total (both the heavy and light chain combined) covered by each of the technique at 95% confidence is presented in (d). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
chain (HC) and 98.6% of the light chain (LC), giving a total coverage of almost 99.1% for the digested mAb sample. For HC, the coverages were 89.76% and 94.21%, and for LC, these were 92.06% and 95.79% with RPLC-MS/MS and CZE-MS/MS, respectively. Separation of LC eluted peptides in the second dimension allowed the resolution of otherwise co-eluting and low abundance peptides. Similarly, fractionation of the mAb digest by LC in the first dimension decreased the complexity of the sample allowing for better detection in the RPLC–CZE–MS/MS system compared to CZE-MS/MS alone. Moreover, improved sampling is evident by the higher number of spectra. The RPLC–CZE–MS/MS method produced 18,145 spectra compared to 1,684 and 2,007 spectra in RPLC-MS/MS and CZE-MS/MS, respectively (Fig. 3a). This corresponds to more than 10-fold improvement in mass spectra for RPLC-MS/MS and
nine-fold in case of CZE-MS/MS. Together with increased mass spectra, RPLC–CZE–MS/MS also identified more number of distinct peptides than RPLC-MS/MS or CZE-MS/MS. A peptide was classified as distinct based on the amino acid sequence of the peptide. As shown in Fig. 3a, 372 peptide sequences were obtained by RPLC–CZE–MS/MS in contrast to 219 and 177 for RPLC-MS/MS and CZE-MS/MS, respectively. RPLC–CZE–MS/MS thus led to 1.7 and 2.1 times more peptides being identified than, RPLC-MS/MS and CZE-MS/MS, respectively. A study on NIST mAb analyzed hundreds of 1D and 2D RPLC-MS/MS runs to generate a spectral library of the mAb and reported 99% sequence coverage with 1,882 peptides [32]. Our much simpler approach produced comparable results with 99.1% sequence coverage and identification of 1,202 peptides. The complete list of peptide IDs identified by each of
Please cite this article as: R. Kumar, R.L. Shah and A.S. Rathore, Harnessing the power of electrophoresis and chromatography: Offline coupling of reverse phase liquid chromatography-capillary zone electrophoresis-tandem mass spectrometry for peptide mapping for monoclonal antibodies, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2020.460954
JID: CHROMA
ARTICLE IN PRESS
[m5G;February 11, 2020;13:33]
R. Kumar, R.L. Shah and A.S. Rathore / Journal of Chromatography A xxx (xxxx) xxx
5
Fig. 3. Comparative analysis of identified peptides. The three techniques are compared via number of identified spectra and number of peptide sequences (a), and the distribution of the identified peptides (b).
Fig. 4. MS/MS spectra of “AASGFK” peptide. The spectra were analyzed with PeakView 2.0 software. Amino acids were labelled according to the mass difference between adjacent y ions.
the techniques, their modifications, and their acquisition times are given in Supplementary Table 3. A comparative analysis of the peptide sequences identified by the three techniques is presented in Fig. 3b. Of these, 120 sequences were identified by all the three techniques. The 41 peptides identified by RPLC-MS/MS alone may be because of the higher column loading, which resulted in the accumulation of low abundance peptides to the limit of MS detection. Interestingly, more than 140 sequences were identified solely with the RPLC– CZE–MS/MS technique (Supplementary Table 4). The identification of these peptides in 2D separation may be attributed to their improved resolution in multidimensional separation owing to the orthogonal nature of separation. Overall, the results obtained are similar to the previously published reports [3,41]. Fig. 4 shows the MS/MS fragmentation spectra of one of the unique peptides identified in RPLC–CZE–MS/MS. The y ion series matched considerably, as shown in the figure, thus confirming the detection of the peptide. Other physicochemical properties of peptides like isoelectric point (pI), molecular weight, and hydrophobicity were investi-
gated, and the cumulative distributions of these properties are shown in Fig. 5. In the case of isoelectric point, more than 75% of the peptides were identified by RPLC-MS/MS having pI < 7 (acidic peptides). The percentage of acidic peptides (pI < 7) detected by CZE-MS/MS and RPLC–CZE–MS/MS were approximately 58% and 59%, respectively (Fig. 5a). Thus, acidic, as well as basic peptides (pI >7), were equally represented in RPLC–CZE–MS/MS and CZE-MS/MS. The improved performance of the RPLC–CZE– MS/MS method in detection of basic peptides is in accordance with the complementary nature of CZE with RPLC and has a better ability to detect basic peptides [3,9,27,42]. The molecular size of the peptides present in the sample is another factor which decides their detection. Usually, larger peptides are well resolved and identified via RPLC-MS/MS, and CZE better represents low molecular weight peptides [27]. In the present study, RPLC–CZE– MS/MS combined the advantages of both the techniques and identified both low (<500 Da) and high molecular weight (>4500 Da) peptides (Fig. 5b). Both categories of peptides were not identified by RPLC-MS/MS. While the smallest peptide identified by RPLC-MS/MS had a molecular weight of 501 Da, CZE-MS/MS, and
Please cite this article as: R. Kumar, R.L. Shah and A.S. Rathore, Harnessing the power of electrophoresis and chromatography: Offline coupling of reverse phase liquid chromatography-capillary zone electrophoresis-tandem mass spectrometry for peptide mapping for monoclonal antibodies, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2020.460954
JID: CHROMA 6
ARTICLE IN PRESS
[m5G;February 11, 2020;13:33]
R. Kumar, R.L. Shah and A.S. Rathore / Journal of Chromatography A xxx (xxxx) xxx
Fig. 5. Physiochemical properties of the identified peptides. The peptides identified by each of the techniques were compared with respect to their pI (a), molecular weight (b), and grand average of hydropathy (GRAVY) score (c). Isoelectric point (pI) and molecular weight of the peptides were calculated using Expasy compute pI/MW tool [55]. The GRAVY score was calculated using the online GRAVY calculator [56]. Negative values of GRAVY score denote hydrophilic peptides; positive GRAVY score indicates hydrophobic peptide.
Fig. 6. MS/MS spectra for “YAM[Ox]SWVR” peptide. The spectra were analyzed with PeakView 2.0 software. Amino acids were labelled according to the mass difference between adjacent y ions.
RPLC–CZE–MS/MS identified 8 and 13 peptides, respectively, lower than 501 Da. An analysis of the hydrophobicity of the identified peptides by each of the techniques is illustrated by the grand average of hydropathy (GRAVY) score. The results are presented in Fig. 5c. While negative values of GRAVY score denote hydrophilic peptides, the positive score is for hydrophobic peptides [27,43]. The percentages of hydrophilic and hydrophobic peptides identified by RPLC-MS/MS were 61% and 39%, respectively. In the case of CZE-MS/MS, 76% of the peptides identified were hydrophilic and 24% were hydrophobic. The results are in accordance with the ability of CZE-MS/MS in separation of hydrophilic peptides [9,27]. Interestingly, the percentages of hydrophilic and hydrophobic peptides identified by RPLC– CZE–MS/MS were 69% and 31%, respectively, almost in between the respective percentages of peptides identified by RPLC-MS/MS and CZE-MS/MS individually. The list of peptides and their corresponding pI, molecular weight, and GRAVY scores are given in Supplementary Table 5. Modifications are a critical quality attribute (CQA) of mAb products and extensive characterization is mandated by the regulatory agencies [38,39]. Many of these are well known to affect the safety and efficacy of the final product [44,45]. Post-translational modifications (PTMs) are difficult to identify via RPLC-MS/MS due to co-elution, ion suppression, and a much higher abundance of unmodified peptides [44,46]. CZE is known to resolve these modified and unmodified peptides, improving their detection [46].
RPLC–CZE–MS/MS can thus adequately resolve the peptides and enhance their PTM detection. MS/MS spectra were searched to identify the modified peptides using ProteinPilot software. The representative MS/MS spectra for one of the identified modifications is depicted in Fig. 6. It is evident that even low intensity modified peptide is detected in RPLC–CZE-MS/MS, which demonstrates the ability of this method to detect even trace amounts of modifications. A total of 150, 94, and 413 PTMs were identified with RPLC-MS/MS, CZE-MS/MS, and RPLC–CZE–MS/MS, respectively (Supplementary Figure 2a). Further, most of the modifications observed in both RPLC-MS/MS and CZE-MS/MS were also observed with RPLC–CZE-MS/MS. Supplementary Figure 2b shows the numerical distribution of modifications identified in RPLCMS/MS, CZE-MS/MS, and RPLC–CZE–MS/MS. The different kinds of PTMs which were detected in all the techniques were also compared and are summarized in Supplementary Figure 2c. Since RPLC and CZE are complementary orthogonal techniques and are reported to identify distinct peptides [3,27,47], the combined data of the two techniques were analyzed to see if the simple combination of the two data sets can bypass the need of our RPLC– CZE-MS/MS method. Even after processing the combined data from RPLC-MS/MS and CZE-MS/MS of the mAb digest, the sequence coverage of HC and LC were 96.7% and 95.8% respectively (Supplementary Figure 3). This is in sharp contrast to the RPLC–CZE-MS/MS method, which covered 99.6% and 98.6% of the HC and LC, respectively (Fig. 2c).
Please cite this article as: R. Kumar, R.L. Shah and A.S. Rathore, Harnessing the power of electrophoresis and chromatography: Offline coupling of reverse phase liquid chromatography-capillary zone electrophoresis-tandem mass spectrometry for peptide mapping for monoclonal antibodies, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2020.460954
ARTICLE IN PRESS
JID: CHROMA
[m5G;February 11, 2020;13:33]
R. Kumar, R.L. Shah and A.S. Rathore / Journal of Chromatography A xxx (xxxx) xxx
4. Conclusions In this work, an RPLC–CZE-MS/MS platform has been established for a deep bottom-up analytical characterization of an IgG1 based mAb biotherapeutic product. The orthogonality of RPLC and CZE offers a significantly enhanced analytical performance as is evident from the increased sequence coverage, number of peptides, as well as PTMs. Identification of the higher number of peptides resulted in a higher confidence of sequence coverage and detection of even low abundance peptides and modifications. Additionally, RPLC–CZE–MS/MS method provided more information than even the combination of what was offered by RPLC–MS/MS and CZE–MS/MS separately. The proposed RPLC–CZE–MS/MS method is likely to be superior to the emerging 2D LC-MS/MS-based methods that offer limited orthogonality and loss of resolution due to use of high temperature and other parameters to reduce analysis time [48,49]. The multidimensional setup presented here is offline, which offers the flexibility of easy set up in any laboratory with available instruments, choice of column and fractionation strategy and better resolution compared to online methods. The RPLC– CZE–MS/MS method can be further improved by improving sample loading capacity [20], reducing the total analysis time with optimization of CZE parameters, and use of coated capillaries that offer minimal sample adsorption [50]. Sequential sample injection and array-based CZE may further increase the throughput [51–53]. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE [54] partner repository with the dataset identifier PXD017316. Funding This work was funded by the Center of Excellence for Biopharmaceutical Technologygrant (grant number BT/COE/34/SP15097/2015) from Department of Biotechnology, Ministry of Science and Technology. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. CRediT authorship contribution statement Ramesh Kumar: Methodology, Validation, Formal analysis, Investigation, Data curation, Writing - original draft, Visualization. Rohan L. Shah: Writing - review & editing. Anurag S. Rathore: Conceptualization, Methodology, Resources, Writing - review & editing, Supervision, Project administration, Funding acquisition. Acknowledgements We acknowledge the SCIEX team, Manesar, India, for providing technical support in this investigation. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.chroma.2020.460954. References [1] J.R. Lill, Introduction to biotherapeutics, in: Analytical Characterization of Biotherapeutics, 1st ed., John Wiley & Sons, 2017, pp. 1–14. [2] A.S. Rathore, I.S. Krull, S. Joshi, Analytical characterization of biotherapeutic products, part II : the analytical toolbox tools for analytical characterization, LCGC North Am. 36 (11) (2018) 814–822.
7
[3] X. Yan, L. Sun, G. Zhu, O.F. Cox, N.J. Dovichi, Over 4100 protein identifications from a xenopus laevis fertilized egg digest using reversed-phase chromatographic prefractionation followed by capillary zone electrophoresis–electrospray ionization–tandem mass spectrometry analysis, Proteomics 16 (23) (2016) 2945–2952. [4] F. Xie, R.D. Smith, Y. Shen, Advanced proteomic liquid chromatography, J. Chromatogr. A 1261 (2012) 78–90. [5] L. Suntornsuk, Capillary electrophoresis in pharmaceutical analysis : a survey on recent applications, J. Chromatogr. Sci. 45 (2007) 559–577. [6] R.L.C. Voeten, I.K. Ventouri, R. Haselberg, G.W. Somsen, Capillary electrophoresis: trends and recent advances, Anal. Chem. 90 (3) (2018) 1464–1481. [7] L. Sun, et al., Capillary zone electrophoresis for bottom-up analysis of complex proteomes, Proteomics 16 (2) (2016) 188–196. [8] J.M. Risley, C.A.G. De Jong, D.D.Y. Chen, Electrospray ionization interface development for capillary electrophoresis – mass spectrometry, in: Capillary Electrophoresis Mass Spectrometry: Principles and Applications, John Wiley & Sons, 2016, pp. 7–39. [9] Y. Li, M.M. Champion, L. Sun, P.A. DiGiuseppe Champion, R. Wojcik, N.J. Dovichi, CZE-ESI-MS/MS as an alternative proteomics platform to UPLC-ESI-MS/MS for samples of intermediate complexity, Anal. Chem. 84 (3) (2012) 1617–1622. [10] H. Mischak, J.N. Joshua, J. Coon, E.M. Weissinger, J.P. Schanstra, A.F. Dominiczak, Capillary electrophoresis–mass spectrometry as a powerful tool in biomarker discovery and clinical diagnosis: an update of recent developments, Mass Spectrom. Rev. 28 (5) (2009) 703–724. [11] G. Zhu, et al., Capillary zone electrophoresis tandem mass spectrometry detects low concentration host cell impurities in monoclonal antibodies, Electrophoresis 37 (4) (2016) 616–622. [12] W. Zhang, T. Hankemeier, R. Ramautar, Next-generation capillary electrophoresis–mass spectrometry approaches in metabolomics, Curr. Opin. Biotechnol. 43 (2017) 1–7. [13] E. Dominguez-Vega, T. De Vijlder, E.P. Romijn, G.W. Somsen, Capillary electrophoresis-tandem mass spectrometry as a highly selective tool for the compositional and site-specific assessment of multiple peptide-deamidation, Anal. Chim. Acta 982 (2017) 122–130. [14] A.S. Rathore, R. Kumar, I.S. Krull, Practical considerations of capillary electrophoresis-mass spectrometry for analysis of biotherapeutics, LCGC North. Am. 37 (6) (2019) 386–391. [15] E. Tamizi, A. Jouyban, The potential of the capillary electrophoresis techniques for quality control of biopharmaceuticals — A review, Electrophoresis 36 (2015) 831–858. [16] A.J. Reason, Structural characterization of monoclonal antibodies, BioPharm. Int. 2011 (5) (2011) s15–s22. [17] R. Gahoual, et al., Monoclonal antibodies biosimilarity assessment using transient isotachophoresis capillary zone electrophoresis-tandem mass spectrometry, MAbs 6 (6) (2014) 1464–1473. [18] O.O. Dada, Y. Zhao, N. Jaya, O. Salas-Solano, High-Resolution capillary zone electrophoresis with mass spectrometry peptide mapping of therapeutic proteins: peptide recovery and post-translational modification analysis in monoclonal antibodies and antibody-drug conjugates, Anal. Chem. 89 (21) (2017) 11236–11242. [19] M. Pioch, S.-C. Bunz, C. Neus¨uß, Capillary electrophoresis/mass spectrometry relevant to pharmaceutical and biotechnological applications, Electrophoresis 33 (11) (2012) 1517–1530. [20] D. Chen, X. Shen, L. Sun, Capillary zone electrophoresis-mass spectrometry with microliter-scale loading capacity, 140 min separation window and high peak capacity for bottom-up proteomics, Analyst 142 (12) (2017) 2118– 2127. [21] L. Ranjbar, J.P. Foley, M.C. Breadmore, Multidimensional liquid-phase separations combining both chromatography and electrophoresis-a review, Anal. Chim. Acta 950 (2016) 7–31. [22] P. Cesla, P. Dinisová, J. Fischer, Capillary electrophoresis in multidimensional separation techniques for anlaysis of natural compouds (A review), Sens. Electroanal. 6 (2011) 357–368. [23] W. Tong, A. Link, J.K. Eng, J.R. Yates III, Identification of proteins in complexes by solid-phase microextraction/multistep elution/capillary electrophoresis/tandem mass spectrometry, Anal. Chem. 71 (13) (1999) 2270–2278. [24] E.N. McCool, et al., Deep top-down proteomics using capillary zone electrophoresis-tandem mass spectrometry: identification of 5700 proteoforms from the escherichia coli proteome, Anal. Chem. 90 (9) (2018) 5529– 5533. [25] Z. Yang, X. Shen, D. Chen, L. Sun, Improved nanoflow RPLC-CZE-MS/MS system with high peak capacity and sensitivity for nanogram bottom-up proteomics, J. Proteome. Res. 18 (11) (2019) 4046–4054. [26] Z. Yang, X. Shen, D. Chen, L. Sun, Microscale reversed-phase liquid chromatography/capillary zone electrophoresis-tandem mass spectrometry for deep and highly sensitive bottom-up proteomics: identification of 7500 proteins with five micrograms of an MCF7 proteome digest, Anal. Chem. 90 (17) (2018) 10479–10486. [27] D. Chen, X. Shen, L. Sun, Strong cation exchange-reversed phase liquid chromatography-capillary zone electrophoresis-tandem mass spectrometry platform with high peak capacity for deep bottom-up proteomics, Anal. Chim. Acta 1012 (2018) 1–9. [28] D.A. Wolters, M.P. Washburn, J.R. Yates, An automated multidimensional protein identification technology for shotgun proteomics, Anal. Chem. 73 (23) (2001) 5683–5690.
Please cite this article as: R. Kumar, R.L. Shah and A.S. Rathore, Harnessing the power of electrophoresis and chromatography: Offline coupling of reverse phase liquid chromatography-capillary zone electrophoresis-tandem mass spectrometry for peptide mapping for monoclonal antibodies, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2020.460954
JID: CHROMA 8
ARTICLE IN PRESS
[m5G;February 11, 2020;13:33]
R. Kumar, R.L. Shah and A.S. Rathore / Journal of Chromatography A xxx (xxxx) xxx
[29] M. Gilar, P. Olivova, A.E. Daly, J.C. Gebler, Orthogonality of separation in two-dimensional liquid chromatography, Anal. Chem. 77 (19) (2005) 6426–6434. [30] Z. Zhang, X. Yan, L. Sun, G. Zhu, N.J. Dovichi, Detachable strong cation exchange monolith, integrated with capillary zone electrophoresis and coupled with ph gradient elution, produces improved sensitivity and numbers of peptide identifications during bottom-up analysis of complex proteomes, Anal. Chem. 87 (8) (2015) 4572–4577. [31] G. Vanhoenacker, I. Vandenheede, F. David, P. Sandra, K. Sandra, Comprehensive two-dimensional liquid chromatography of therapeutic monoclonal antibody digests, Anal. Bioanal. Chem. 407 (1) (2015) 355–366. [32] Q. Dong, et al., The NISTmAb tryptic peptide spectral library for monoclonal antibody characterization, MAbs 10 (3) (2018) 354–369. [33] I.V. Shilov, et al., The paragon algorithm, a next generation search engine that uses sequence temperature values sequence temperature values and feature probabilities to identify peptides from tandem mass spectra, Mol. Cell. Proteom. 6 (9) (2007) 1638–1655. [34] “IMGT/mAb-DB.” [Online]. Available: http://www.imgt.org/mAb-DB/doc. [Accessed: 10-May-2019]. [35] S.K.G. Grebe, R.J. Singh, LC-MS/MS in the clinical labarotory-where to from here, Clin. Biochem. Rev. 32 (2011) 5–31. [36] R. Aebersold, H.D. Morrison, Analysis of dilute peptide samples by capillary zone electrophoresis, J. Chromatogr. 516 (1990) 79–88. [37] C.R. Evans, J.W. Jorgenson, Multidimensional LC-LC and LC-CE for high-resolution separations of biological molecules, Anal. Bioanal. Chem. 378 (2004) 1952–1961. [38] ICH, ICH Q6B - specifications: test procedures and acceptance critreia for biotechnological/biological products, in: ICH Harmonised Tripartite Guideline, 1999, pp. 1–16. March. [39] B.A.J. Reason, A. Weiskopf, A.S. Rathore, Defining critical quality attributes for monoclonal antibody therapeutic products, BioPharm Int. 27 (7) (2014) 34–43. [40] S.S. Zhao, D.D.Y. Chen, Applications of capillary electrophoresis in characterizing recombinant protein therapeutics, Electrophoresis 35 (1) (2014) 96–108. [41] Y. Qu, L. Sun, Z. Zhang, N.J. Dovichi, Site-specific glycan heterogeneity characterization by hydrophilic interaction liquid chromatography solid-phase extraction, reversed-phase liquid chromatography fractionation, and capillary zone electrophoresis-electrospray ionization-tandem mass spectro, Anal. Chem. 90 (2) (2018) 1223–1233.
[42] M. Pioch, S.C. Bunz, C. Neusüß, Capillary electrophoresis/mass spectrometry relevant to pharmaceutical and biotechnological applications, Electrophoresis 33 (11) (2012) 1517–1530. [43] J. Kyte, R.F. Doolittle, A simple method for displaying the hydropathic character of a protein, J. Mol. Biol. 157 (1982) 105–132. [44] W. Li, et al., Structural elucidation of post-translational modifications in monoclonal antibodies, ACS Symp. Ser. 1201 (2015) 119–183. [45] H. Liu, et al., In vitro and in vivo modifications of recombinant and human IGG antibodies, MAbs 6 (5) (2014) 1145–1154. [46] K. Faserl, B. Sarg, P. Gruber, H.H. Lindner, Investigating capillary electrophoresis-mass spectrometry for the analysis of common post-translational modifications, Electrophoresis 39 (9–10) (2018) 1208–1215. [47] C.S. Babu, CE/MS and LC/MS Synergy, Agilent Technologies, 2013. [48] D.R. Stoll, X. Li, X. Wang, P.W. Carr, S.E.G. Porter, S.C. Rutanb, Fast, comprehensive two-dimensional liquid chromatography, J. Chromatogr. A 1168 (1–2) (2007) 3–43. [49] M. Kivilompolo, J. Pó, T. Hyötyläinen, Comprehensive two-dimensional liquid chromatography (LC × LC): a review, LCGC Eur. 24 (5) (2011) 232–243. [50] J.Q. Xia, Coated capillaries for CE-MS of therapeutic protein, in: Capillary Electrophoresis-Mass Spectrometry: Therapeutic Protein Characterization, Springer International Publishing, 2016, pp. 7–11. [51] K. Faserl, B. Sarg, L. Sola, H.H. Lindner, Enhancing proteomic throughput in capillary electrophoresis–mass spectrometry by sequential sample injection, Proteomics 17 (22) (2017) 1–13. [52] D.A. Boley, Z. Zhang, N.J. Dovichi, Multisegment injections improve peptide identification rates in capillary zone electrophoresis-based bottom-up proteomics, J. Chromatogr. A 1523 (2017) 123–126. [53] K.C. Chan, C. Liu, Q. Li, Multidimensional high performance liquid chromatography – capillary electrophoresis separation of a protein digest : an update, Electrophoresis 22 (6) (2001) 1133–1135. [54] Y. Perez-riverol, et al., The pride database and related tools and resources in 2019 : improving support for quantification data, Nucleic Acids Res. 47 (D1) (2019) D442–D450. [55] “Expasy compute pI/MW.” [Online]. Available: https://web.expasy.org/ compute_pi/. [Accessed: 23-Aug-2019]. [56] “GRAVY calculator.” [Online]. Available: http://www.gravy-calculator.de/. [Accessed: 23-Aug-2019].
Please cite this article as: R. Kumar, R.L. Shah and A.S. Rathore, Harnessing the power of electrophoresis and chromatography: Offline coupling of reverse phase liquid chromatography-capillary zone electrophoresis-tandem mass spectrometry for peptide mapping for monoclonal antibodies, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2020.460954