Characterization of a Recombinant Influenza Vaccine

Current Pharmaceutical Biotechnology, 2011, 12, 000-000

1

Characterization of a Recombinant Influenza Vaccine Candidate Using Complementary LC-MS Methods Hongwei Xie1,* , Catalin Doneanu1, Weibin Chen1, Joseph Rininger2 and Jeffery R. Mazzeo1 1

Department of Biopharmaceutical Sciences, Waters Corporation, 34 Maple Street, Milford, MA 01757; 2Protein Sciences Corporation, 1000 Research Parkway, Meriden, CT 06450, USA Abstract: Influenza vaccination is recognized as the most effective method for reducing morbidity and mortality due to seasonal influenza. To improve vaccine supply and to increase flexibility in vaccine manufacturing, cell culture-based vaccine production has emerged to overcome limitations of egg-based production. The switch of production system and the need for annual re-evaluation of vaccines for the effectiveness due to frequent viral antigenic changes call for methods for complete characterization of the hemagglutinin (HA) antigens and the final vaccine products. This study describes advanced liquid chromatography-mass spectrometry (LC-MS) methods for simultaneous identification of HA proteins and process-related impurities in a trivalent influenza candidate vaccine, comprised of purified recombinant HA (rHA) antigens produced in an insect cell-baculovirus expression vector system (BEVS). N-linked glycosylation sites and glycoforms of the three rHA proteins (corresponding to influenza A subtypes H1N1 and H3N2 and B virus, respectively) were profiled by peptide mapping using reversed-phase (RP) LC-MSE (data independent acquisition LC-MS using an alternating low and elevated collision energy scan mode). The detected site-specific glycoforms were further confirmed and quantified by hydrophilic interaction LC (HILIC)-multiple reaction monitoring (MRM) assays. LC-MSE was used to characterize the vaccine candidate, providing both protein identities and site-specific information of glycosylation and degradations on each rHA protein. HILIC-MRM methodology was used for rapid confirming and quantifying site-specific glycoforms and potential degradations on each rHA protein. These methods can contribute to the monitoring of vaccine quality especially as it pertains to product comparability studies to evaluate the impact of production process changes.

Keywords: Insect cell - baculovirus expression vector system, degradation products, glycan analysis, hemagglutinin, impurity Proteins, LC-MS, N-linked glycosylation and glycoforms, recombinant influenza vaccine, peptide mapping and protein characterization. INTRODUCTION Influenza is a viral infection of the respiratory tract by influenza A and B viruses. It affects millions of people and causes up to 500,000 deaths worldwide annually [1, 2]. Influenza vaccination is the primary prophylactic method and the principal strategy for reducing morbidity and mortality due to seasonal infection. Vaccines provide protection by immune-response neutralizing antibodies to viral hemagglutinin (HA), one of the two major functional glycoproteins on the influenza virus surface [3]. HA is responsible for attachment of the virus to sialic acid-containing host cell receptors and for membrane fusion during the infection process [4-6]. Although the functions of N-linked glycans on HA are only partly elucidated, they are believed to be potent regulators of viral pathogenicity [7] and viral life cycle [8]. Licensed vaccines for seasonal influenza contain an established quantity of HAs – a mixture of H1, H3 and B, the corresponding HA proteins of the three most common human-infected viruses – influenza A subtypes H1N1 and H3N2, and influenza B [9]. All licensed trivalent influenza vaccines are currently produced by growing the virus in *Address correspondence to this author at the Biopharmaceutical Science Department, Waters Corporation, 34 Maple Street, Milford, MA 01757, USA; Tel: 508-482-3670; Fax: 508-482-3085; E-mail: [email protected] 1389-2010/11 $58.00+.00

embryonated chicken eggs [10]. Although eggs have been an useful substrate for influenza virus propagation, the yearly development of high-growth viruses to reflect the circulating wild-type influenza strains and large-scale production are performed under considerable time pressures [11]. Eggbased production is difficult to scale up rapidly in response to an emerging need such as a pandemic. In addition, there is a concern that the supply of eggs for vaccine manufacturing may be disrupted by avian pandemic diseases. The development of alternative substrates for influenza vaccine production has been identified as a high-priority objective for security and pandemic preparedness plans. To increase vaccine manufacturing flexibility and to improve vaccine supply, there is an increasing interest in cell culture-based influenza vaccine production to overcome limitations of egg-based production system [3, 12-16]. Among them, the use of recombinant baculoviruses designed to express the influenza HA proteins to infect insect cells is especially suited for the production of influenza vaccine for a number of reasons [17]. The production system not only avoids dependence on egg supply, but has advantage for annual updating HA with virus strains in circulation with which genes can be cloned and inserted into the vector facilities to update the vaccine at regular intervals. It also avoids the need to work with potentially pathogenic live influenza viruses and the attendant biocontamination issues that would © 2011 Bentham Science Publishers

2 Current Pharmaceutical Biotechnology, 2011, Vol. 12, No. 12

be a particular concern for the development of pandemic vaccines [18]. The efficiency of protein production in the baculovirus expression system provides the opportunity to use higher, and potentially more effective, doses of vaccine [19]. The use of purified HA proteins produced from cellbased recombinant technology facilitates the application of analytical techniques previously applied for wellcharacterized biotherapeutic products to recombinant influenza vaccines. In contrast to the production of wellcharacterized biopharmaceuticals such as recombinant monoclonal antibodies, there are currently no regulatory requirements to verify the identity of each specific HA contained in the vaccine or to evaluate the overall composition of final vaccine products [11], let alone monitor important posttranslational modifications (PTMs) such as the glycosylation structures and their distribution on each antigen [20]. Currently, vaccines are characterized in terms of HA potency by single radial immunodiffusion (SRID) [21, 22] since HA antigens are of primary importance for establishing effective immunity [11]. Lot release of the vaccine includes this assay to ensure equivalent potency. Although the SRID assay verifies inclusion of influenza A (both H1 and H3 subtypes) and B virus in the vaccine, it does not distinguish between antigenic variants that could occur from influenza virus propagation in eggs. Other antigenically related proteins and potential impurities that may be present in different (or the same) vaccine preparations are neither identified nor quantified. To some extent, as discussed by Schwarzer et al. [23], influenza vaccines are not well-characterized because of their complexity. Typically, whole virus particles containing the immunogenic target molecules have to be produced and purified rather than individual recombinant proteins. In addition, initiation of an immune response while avoiding negative side effects is an extremely challenging task and depends on a variety of factors including the glycosylation of antigens. These issues cause complications in the ability to fully characterize non-recombinant vaccines and impede the systemic studies for fast, simple, and robust methods for characterization of important features of vaccines such as impurity proteins and glycosylation modifications. Better characterization may help to speed up the development of new vaccine candidates by providing more informative analytical product characterization prior to initiation of clinical trials. Clinical trials are expensive, timeconsuming and can be difficult to be established for certain populations such as young children. With vaccine development in a new production system several key questions concerning antigen quality and final composition of vaccines formulated from recombinant HA (rHA) proteins could be addressed by analytical methods. Among them, primary sequences and changes regarding PTMs such as glycosylation and potential impurities derived from cell culture are important concerns. It’s well known for recombinant glycoprotein and antibody production that selection of host cells and cell culture conditions affect the N-linked glycosylation pattern [24]. Degradation occurs in thermally labile vaccines and affects the quality of vaccines that have been subjected to improper storage and transportation [25]. In addition, allergic reactions caused by egg-derived proteins in the current vaccine production system or by the contaminant proteins de-

Xie et al.

rived from the cell culture-based production system can be properly assessed by the results from analytical characterization. Recently, efforts to develop analytical tools for more complete characterization of influenza vaccines have emerged. Getie-Kebtie et al. [11,26] successfully applied matrix-assisted laser desorption/ionization (MALDI)-tandem mass spectrometry (MS/MS) for identification of HA strains and for more complete characterization of final vaccine composition. The method was fast and convenient, but limited by the m/z range of 900-3300 for tryptic peptide monitoring. This range potentially excludes the peptides with large glycans, and tryptic peptides beyond that mass range that might provide information on sequence variants. The analysis of N-linked glycans of HA proteins from different influenza virus strains replicated from mammalian cell lines was also reported using capillary gel electrophoresis-laser induced fluorescence (CGE-LIF) [20, 23]. However, no site-specific glycosylation information for the heavily Nglycosylated HA proteins [27] was available for the licensed influenza vaccines. In addition, a number of other liquid chromatography (LC) and MS based technologies have been reported for characterizing influenza virus types and for analyzing HA antigens [28-31]. This study describes a number of LC-MS analyses of a candidate influenza vaccine formulated from purified rHA antigens that were produced in the insect cell-baculovirus expression vector system (BEVS). A recently developed peptide mapping approach [32, 33], performed with data independent acquisition LC-MS using an alternating low and elevated collision energy scan mode (LC-MSE) [34, 35], was applied to confirm the sequences of rHAs, to identify process-related impurities and to analyze site-specific glycosylation and degradations on each rHA protein. HILIC-MRM (hydrophilic interaction liquid chromatography-multiple reaction monitoring) assay was then developed to verify and quantify critical attributes of the vaccine such as site-specific glycosylation. The goal of this study was to develop advanced LC-MS methods for better characterization of recombinant influenza vaccines. MATERIALS AND EXPERIMENTS Materials Iodoacetamide (IAM), dithiothereitol (DTT), ammonium bicarbonate (NH4HCO3), ammonium formate, trifluroacetic acid (TFA), were purchased from Sigma Chemical Co. (St. Louis, MO), Sequence-grade trypsin from Promega Corp. (Madison, WI), Formic acid (FA) from EM sciences (Gibbstown, NJ), Optima-grade acetonitrile (ACN) from Fisher Scientific (Pittsburg, PA). RapiGest SF and 96-well microElution HILIC solid phase extraction (SPE) plates were from Waters Corp. (Milford, MA). A Millipore Mili-Q purification system (Bedford, MA) was used to prepare deionized water (18 M cm) for LC mobile phases and in all sample processing procedures Recombinant Trivalent Vaccine Sample The research trivalent influenza vaccine was obtained from Protein Sciences Corp. (Meriden, CT). The influenza vaccine was formulated using purified rHA proteins rH1,

LC-MS Analyses of a Recombinant Influenza Vaccine

rH3 and rB derived from the A/Brisbane/59/2007 (H1N1), A/Brisbane/16/2007 (H3N2) and B/Florida/4/2006 influenza viruses. As described by the manufacturer, all the rHAs were produced in SF+ (Spodotera frugiperda) insect cells using the baculovirus (Autographa californica) expression vector system. Each rHA was purified to >90% purity under conditions that preserve their biological activity and tertiary structure. The recombinant influenza vaccine was formulated into a PBS solution (10 mM sodium phosphate, pH 7.2, and 150 mM NaCl) containing 0.005% Tween-20. The final protein concentration was 0.27g/l as measured by BCA protein assay. The vaccine sample was stored at 2 - 8ºC before analysis. Preparation of Peptide Mixture Before digestion, the sample was denatured by heating at 80°C for 15 min, and reduced with DTT and alkylated with IAM. Tryptic digests of the vaccine (200 l) were prepared using sequence-grade trypsin (1:25 w/w) by incubating at 37°C in 50 mM NH4HCO3 solution containing 0.05% RapiGest SF (pH ~7.5) for 4 hours. Formic acid (final concentration 0.1% v/v) was used to degrade RapiGest SF and to quench tryptic enzymatic reactions. ACN was added (to 5% v/v) to help stabilize hydrophobic peptides and dissolve any precipitate due to RapiGest SF degradation. Approximately 3.7 g (in 20 l) protein tryptic digest was injected in four replicates for each LC-MS analysis. Reversed-Phase (RP) LC-MSE Peptide Mapping Experiments All LC-MSE measurements were performed on a Waters SYNAPTTM HDMS system coupled with a Waters ACQUITYTM UPLC® system for online RP LC separation. The systems were controlled by MassLynx™ 4.1 software. The peptide mixture was separated on a 2.1 x 150 mm BEH300 C18 1.7 μm column at 60 °C using a 120-min gradient (140% B) with a flow rate of 0.2 ml/min. 0.1% FA in water or ACN was used as mobile phases A and B, respectively. MS acquisition was operated in the positive ion V-mode. An alternating low collision energy (4 V) and elevated collision energy (ramping from 20 to 40 V) acquisition was used to obtain the precursor ions of peptides (MS spectrum) and their fragmentation data (MSE spectrum) without bias for MS signals, respectively. Scan time was 0.5 second (1 second total duty cycle). A capillary voltage of 3.0 kV, source temperature of 100 °C, cone voltage of 35 V, and cone gas flow of 10 L/h were maintained during the analyses. An auxiliary pump was used to spray a solution of 100 fmol/μl Glu1fibrinopeptide B (GFP) in 50/50 ACN/water containing 0.1% FA for mass accuracy (lockmass channel), at a flow rate of 20 l/min and sampling every 1 min. The system was tuned for a minimum resolution of 10,000 FWHM and calibrated using a 100 fmol/μl GFP infusion. Peptide Mapping Data Analysis The data were initially searched against a database constructed with the sequences of the three rHA proteins, the organisms Autographa californica, drosophila (as a surrogate insect for homology of Spodotera frugiperda) and yeast (culture media contained yeast hydrolysates). The resulting data-

Current Pharmaceutical Biotechnology, 2011, Vol. 12, No. 12 3

base contained 318840 protein entries. The aim was to match proteins present in the vaccine to entries in the database either by exact match or by homology using a sequence database searching engine PLGS2.4 for identification of rHA and impurity proteins [33, 35]. The search used fully tryptic cleavage rules [36] and the searching parameters were set as follows: 1) precursor monoisotopic ion intensity greater than 100 counts and mass tolerance less than 10 ppm; 2) fragment monoistopic ion intensity greater than 50 counts and mass tolerance less than 30 ppm. In order to identify small peptides with 3 or 4 amino acid residues, which have limited fragments available, a match was accepted when at least three fragment ions were identified for the peptide. The collected data were then processed by BiopharmaLynx 1.2 using fully tryptic cleavage rules. The aim of this analysis was to precisely match the known sequences of the three rHA proteins and to measure the modifications and degradations on each rHA [32]. Cysteine carbamidomethylation (+ 57.02 Da) was set as a fixed modification, while Ndeamidation (+ 0.98 Da for aspartic and isoaspartic acid products, and -17.03 Da for succinimide intermediate) and M-oxidation (+15.99 Da) as variable modifications. All potential glycan structures of glycoproteins expressed from the BEVS system [37] were set as variable modifications. The mass tolerance for both precursors and fragments was set to 30 ppm. The intensity threshold of precursors and fragments was set at 100 and 50 counts, respectively. Identified peptides were confirmed by MSE spectra. Sugar oxonium ions (such as m/z 204.1, 366.1, 168.1, 186.1, 528.2) [27] observed in MSE spectra were used to help locate and determine glycosylated peptides. HILIC-MRM Analyses All the MRM experiments were performed on a Waters XevoTM TQ MS system coupled with an ACQUITY TM UPLC® system for online separation. The aim of these experiments was to confirm and quantify site-specific glycoforms on each rHA. The HILIC LC separation was performed on a 2.1 x 150 mm, BEH glycan 1.7 m column at 40 ˚C with a flow rate of 0.2 ml/min. Mobile phases consisting of 10 mM ammonium formate in water and in 90/10 ACN/water were used as mobile phases A and B, respectively. A 60-min gradient from 90% to 75% B in 10 min, then to 50% in 50 min was applied. Data were acquired in ESI+ mode with MRM transitions using a unit mass resolution (0.75 Da FWHM) for both MS1 and MS2, and a dwell time of 5 ms. Four MRM transitions were monitored for each glycopeptide, including precursor to two ions with the backbone peptide attaching an N-acetylglucosamine (GlcNAc) or a GlcNAc plus a fucose (-GlcNAc-Fuc), and to two sugar ions 204.1 (GlcNAc) and 366.1 (GlcNAc-Fuc). The intensity sum of the two latter transitions was used for relative quantification of glycoforms on a same glycosylation site, and the two former transitions were used for further identity confirmation purpose if the masses of their fragment ions were less than 2000 Da. RESULTS AND DISCUSSION Fig. (1) presents the analytical workflow used in this study for analyses of the recombinant influenza vaccine for-


Xie et al.

Fig. (1). The analytical workflow used for the analysis of the recombinant influenza candidate vaccine.

mulated from purified rHA antigens. Limited information was available from the intact protein mass analysis due to the complexity of the sample and the heterogeneity of the HA proteins from glycosylation on multiple N-sites [27]. Characterization work was therefore concentrated at the peptide level. A peptide map was generated from the tryptic digest of the formulated vaccine with RP LC-MSE with two key goals: 1) Identify the three rHA proteins and potential impurity proteins derived from the production system by searching the collected LC-MSE data with PLGS [38] against the sequence database described in the methods section; 2) Examine the primary sequence for each rHA, including PTMs, by mining the same LC-MSE data set with a bioinformatics tool dedicated to detailed analysis of a small number of biotherapeutics (BiopharmaLynx 1.2). For a description of PLGS and BiopharmaLynx functionality see previous publications [33] and [32], respectively. Thereafter, targeted HILIC-MRM assays were developed to confirm the glycoforms detected by peptide mapping and to monitor and quantify the relative abundance of glycoforms on each glycosylation site of the rHA proteins. These assays were not only used as additional confirmation tools, but also targeted for rapid analyses of specific attributes of the formulated influenza vaccine. Identification of Proteins Table 1 lists the proteins identified from the vaccine digest by sequence database searching. As expected, those peptides identified as unique are predominantly from the three rHA antigens. Additionally, multiple proteins derived from the production system were also identified with high confidence (3 unique peptides). They include three proteins from the vector virus Autographa californica, two from drosophila (a homology of the insect Spodotera frugiperda used as expression host cells), and one from yeast (potentially a contaminant from the cell culture media). Proteins from Drosophila were incorporated in the combined database instead of Spodotera frugiperda because of the lack of protein sequences available for the latter organism. A number of impurity proteins were also detected with lower

rity proteins were also detected with lower confidence levels (2 unique peptides), including YDL182Wp-like protein, DNA replication licensing factor MCM3, KLLA0C04433p and KLLA0F13112p from yeast and GD22348, GG11723 and GG24781 from Drosophila. It is not surprising that proteins from the host cell and other contaminant proteins were identified in this research vaccine. In a recent study, Getie-Kebtie and colleagues [11] identified an envelope surface glycoprotein of baculovirus by MALDI-MS/MS-based proteomics from purified rHA of the A/New York/55/2004 prepared from BEVS system. In their study, many proteins derived from egg origin, as well as viral proteins other than HA, were also identified from an egggrown vaccine preparation. In the study described here, we identified impurity proteins derived from the virus vector, the host cell and the cell culture media in the candidate recombinant vaccine. This information can be used to guide the purification process of desired proteins during production. It is also helpful for risk evaluation of the recombinant vaccine due to impurities. Detailed Structures of rHA Proteins After determining the components in the vaccine, we next confirmed primary sequences of rHAs to check that the expected antigens were present in the formulation, profiled functional PTMs such as glycosylation on each antigen, and examined potential degradations. One of the advantages for peptide mapping with LC-MSE is that it collects fragments of all precursors with no requirements for pre-knowledge of the components and without bias on precursor intensity. The unbiased acquisition collects data that could be used for later mining certain type of information of special interest. To confirm the primary sequence and to determine PTMs and potential degradations on each of the three antigens, the same LC-MSE data set was processed with BiopharmaLynx 1.2 [32]. Primary Sequence Verification of rHA Proteins To ensure correct immune-response protection, it is important to confirm that the HA antigens in the formulated vaccine properly reflect the wild-type virus in circulation.


Table 1.


Proteins Identified from the Vaccine Tryptic Digest of the Formulated Vaccine Using the Data Acquired by RP LC-MSE and Analyzed by Sequence Database Searching with PLGS 2.4. The Drosophila Entries in the Table Indicate a Match by Homology

Protein

Organism

M.W.

Unique Peptides

rH1-H1N1A/Brisbane/59/2007

rHA

61229

45

rH3-H3N2A/Brisbane/16/2007

rHA

61814

29

rB-B/Florida/4/2006

rHA

61554

44

P17501-Major envelope glycoprotein

A. Californica

58566

16

P32651-Structural glycoprotein gp41

A. Californica

45381

7

P41678-Capsid protein p24

A. Californica

22110

3

Q05825-ATP synthase subunit beta

Spodoptera

54108

8

BI1502-GM10439

Spodoptera

36921

3

Q5KMQ8-Expressed protein (putative uncharacterized protein)

Yeast

30298

4

Proteins Identified by Homology:

High sequence coverage was obtained for all the three rHAs, with 98.6%, 98.5% and 92.9% for rB, rH1 and rH3, respectively, demonstrating that the antigens were properly produced and formulated. Unlike data dependent acquisition (DDA) LC-MS/MS and certain other techniques for peptide identification [11], peptide mapping with data independent acquisition methods like LC-MSE and BiopharmaLynx is capable of identifying co-eluting peptides [32] and unbiased for the characterization of small to large, or modified peptides. Because of these advantages, high sequence coverage is obtained for confident verification of the antigen sequences. Fig. (2) shows a typical LC-MS total ion chromatogram (TIC, Fig. 2A) of the vaccine tryptic digest and illustrates the identification of three tryptic peptides eluting at different retention times (RTs). They correspond to small, medium and large tryptic peptides rB_T44 (VDDLR, MW 616.32 Da, RT 13.49 min), rB_T17 (SGFFATMAWAVPK, MW 1141.70 Da, RT 78.09 min) and rB_T38 (GFFGAIAGFLEGGWEGMIAGWHGYTSHGAHGVAVAADLK, MW 3943.89 Da, RT 109.12 min). The software assigned peptide identities based on the recorded accurate masses of peptides. The peptide sequence can be verified by spectra reconstructed from the charge-reduced and isotope-deconvoluted MSE data (Figs. 2B-D). For example the +5, +4 and +3 charged ions in the MS spectrum of peak at 109.12 min as inserted in Fig. (2A) were converted to the singly charged ion (3944.89 m/z) for the assignment of tryptic peptide rB_T38. The MSE spectrum of the peak unambiguously confirmed the sequence of rB_T38 (Fig. 2D). Similarly, peaks at 13.49 and 78.09 min were determined to be tryptic peptides rB_T44 (Fig. 2B) and rB_T17 (Fig. 2C), respectively. The MSE spectrum for the peak at 78.09 min has a series of fragment ions with 44 Da mass differences (due to surfactant Tween-20 containing in the formulation) that are unrelated to rB_T17. However, their presence did not affect the identification of the tryptic peptide rB_T17, indicating that this methodology can be used for formulated vaccines.

Using in silico predicted unique peptides of HA from different virus strains, Getie-Kebtie and colleagues [11] reported that MALDI-MS/MS was capable of rapid characterization of virus strains. The method reported here extends beyond the detection m/z range 900-3300 of the method described by Getie-Kebtie et. al., and appears to contribute to a high sequence coverage mapping of the rHA antigens. If there are tryptic peptides containing a variant out of their defined m/z range and for modifications in tryptic peptides with large glycans (as shown below), choice for the method demonstrated here has an advantage. The presence of surfactant Tween-20 in the formulation is reflected over the RTs 60-90 min in the TIC chromatogram Fig. (2A) where there is a continually high response. The surfactant produces a series of fragment ions with 44 Da mass differences at high collision energy, as observed in the MSE spectra of the peak eluted at 78.09 min Fig. (2C). However, the fragments of contaminant surfactant appeared in the MSE spectra as unassigned ions and did not prevent the peptide identification as demonstrated above. Site-Specific Glycosylation on rHA Antigens HA proteins are heavily glycosylated with multiple Nlinked glycosylation motifs on each HA [20, 23, 27]. Considering that there is complex glycosylation in the HA antigens, site-specific glycosylation profiles at the peptide level [39, 40] would be more meaningful than released glycan profiling [20, 23]. Using all the potential glycans reported in the literature [37], we next profiled the site-specific glycosylation on each rHA antigen in the candidate trivalent influenza vaccine. It has been reported [37] that glycoproteins expressed from insect cell lines present two major types of N-linked glycans: paucimannosidic structures (Man(1-3)GlcNAc2 or Man(1-3)GlcNAc(Fuc)GlcNAc) and oligomannose structures (Man(5-9)GlcNAc2). Man represents mannose,


Xie et al.

Fig. (2). RP LC-MS total ion chromatogram (TIC) of the vaccine tryptic digest and example MSE spectra of 3 LC peaks eluted at early (13.49 min), middle (78.09 min) or later (109.12 min) in the chromatography, which were assigned by BiopharmaLynx 1.2 to be tryptic peptides rB_T44 (VDDLR), rB_T17 (SGFFATMAWAVPK) and rB_T38 (GFFGAIAGFLEGGWEGMIAGWHGYTSHGAHGVAVAADLK), respectively. A) TIC; B) MSE spectrum of rB_T44 (13.49 min, MW 616.32 Da); C) MSE spectrum of rB_T17 (78.09 min, MW 1411.70 Da); D) MSE spectrum of rB_T38 (109.12 min, 3943.89 Da). The MS spectrum for the peak at 109.12 min was also inserted in Fig. (2A). The colors in the MSE spectra represent: red – y ions; blue – b ions; green – y/b ions after losing a water or ammonia molecule; grey – unassigned ions.


Table 2.


Glycopeptides and Glycoforms Identified from the Vaccine Tryptic Digest Using the Data Acquired by RP LC-MSE and Further Confirmed by HILIC-MRM (Except for the Glycoforms of Tryptic Peptides rH3_T2, rH3_T3 and rH3_T8 Which have Multiple N-Linked Glycosylation Motifs)

Protein

Peptide

Residues

Peptide Sequence Containing -NXS/T- Motif

Peptide MW

Modification Site

Proposed Glycoforms*

rH1

T1

1-22

(-)DTICIGYHANNSTDTVDTVLEK(N)

2408.12

N11

3,3F,5,5F,7,8

T2

23-40

(K)NVTVTHSVNLLENSHNGK(L)

1961.99

N23

3F,6,8,9

T4

46-73

(K)GIAPLQLGNCSVAGWILGNPECELLISK(E)

2894.50

N54

0,3,3F, 5,8

T5

74-102

(K)ESWSYIVEKPNPENGTCYPGHFADYEELR(E)

3429.52

N87

2,3,5,6,7,8,9

T8

120-145

(K)ESSWPNHTVTGVSASCSHNGESSFYR(N)

2825.21

N125

1,1F,2,2F,3,3F

T10

154-162

(K)NGLYPNLSK(S)

1004.53

N159

0,3,3F,5,6,7,8,9

T24

278-304

(K)CQTPQGAINSSLPFQNVHPVTIGECPK(Y)

2864.39

N286

3,3F,5,8

T47

480-487

(K)NGTYDYPK(Y)

956.42

N480

3,3F,5,7,8,9

T52

504-545

(K)LESMGVYQILAIYSTVASSLVLLVSLGAISFWMCSNGSLQCR(I)

4509.28

T2

3-27

(K)LPGNDNSTATLCLGHHAVPNGTIVK(T)

2528.28

N8-N22

2F-2F

T3

28-82

(K)TITNDQIEVTNATELVQSSSTGEICDSPHQILDGENCTLIDALLGDPQCDGFQNK(K)

5889.71

N38-N63

3-3,3-3F

T8

110-141

(R)SLVASSGTLEFNNESFNWTGVTQNGTSSACIR(R)

3376.56

N122/N126 /N133

6-6, 7-7, 7-8

T10

143-150

(R)SNNSFFSR(L)

957.43

N144

0,3,3F,5,7,8,9

T13

161-173

(K)YPALNVTMPNNEK(F)

1489.72

N165

3,3F,5,6,9

T21

230-255

(R)ISIYWTIVKPGDILLINSTGNLIAPR(G)

2866.63

N246

5,6,7,8,9

T27

277-299

(K)CNSECITPNGSIPNDKPFQNVNR(I)

2546.16

N285

3F,6,7,8,9

T52

483-492

(R)NGTYDHDVYR(D)

1238.53

N483

3F,5,5F,6,8

T3

18-38

(K)TATQGEVNVTGVIPLTTTPTK(S)

2127.14

N25

3,3F,5,6,7,8,9

T8

53-80

(K)LCPDCLNCTDLDVALGRPMCVGTTPSAK(A)

2892.33

N59

0,3F

T16

137-149

(R)LGTSGSCPNATSK(S)

1221.57

N145

3,3F,5,6,7,8,9

T19

167-197

(K)NATNPLTVEVPYICTEGEDQITVWGFHSDDK(T)

3477.60

N167

3,7,8,9

T30

299-304

(K)YGGLNK(S)

650.34

N303

3,3F,5,6,7,8,9

T34

330-335

(K)LANGTK(Y)

602.34

N332

0,3,5,6,7,8,9

T51

490-497

(K)CNQTCLDR(I)

951.39

nd

T52

498-560

(R )IAAGTFNAGEFSLPTFDSLNITAASLNDDGLDNHTILLYYSTAASSLAVTLMLAIFIVYMVSR(D)

6699.38

nd

T53

561-569

(R)DNVSCSICL(-)

952.40

nd

rH3

rB

nd

*0 - peptide without attaching any glycan was identified nd - no confident glycoforms was detected n - peptide attaching a glycan containing -GlcNAC-GlcNAC-(Man)n nF - peptide attaching a glycan containing -GlcNAC[Fuc]-GlcNAC-(Man)n n-n or nF-nF - peptide with two glycans attached on two different N sites

GlcNAc is N-acetylglucosamine, and Fuc is fucose. After entering all the 11 possible glycosylation forms as variable N-linked glycosylation modifications in the data processing, we were able to identify 8, 8, and 6 glycopeptides for rH1, rH3 and rB, respectively. Each N-linked glycosylation site

may have multiple glycoforms (or a series of different glycan moieties). Table 2 summarizes the glycopeptides and glycoforms identified. Three typical examples (one from each rHA antigen in the vaccine sample) were next selected to demonstrate how RP LC-MS E separates and characterizes glyco-


peptides and glycoforms. Alternative HILIC-MRM assays were developed to confirm and quantify the detected glycoforms. Fig. (3) shows the RP LC-MSE separation and identification of unmodified and glycosylated tryptic peptide rB_T8. The glycosylated rB_T8 eluted earlier due to increased hydrophilicity upon glycosylation (Fig. 3A). The peptide sequence was confirmed by MSE spectra (Figs. 3B and 3C). The glycosylation of rB_T8 was also indicated by a series of sugar oxonium ions at m/z 138.05, 204.09, 366.14 and 528.12 in the low m/z range and larger ions of backbone peptide attached with different sugar groups at m/z 3324.48, 3470.53, 4014.78 and 4159.81 (Fig. 3C). The mass difference of 1038.36 Da between the precursors (or y28) without (Fig. 3B) and with (Fig. 3C) glycosylation corresponds to the mass of a glycan moiety (-GlcNAc(Fuc)GlcNAcMan3). Based on the MS signal intensities of singly charged precursors processed by BiopharmaLynx, the relative concentration of glycosylated rB_T8 is estimated to be ~20%, assuming the unmodified peptide and its glycosylated form having same ionization efficiency. The concentration is consistent with the quantification obtained by extracted ion chromatographic peak areas of precursor masses (Fig. 3A). Unlike rB_T8, most of the identified glycopeptides were observed fully glycosylated with multiple glycoforms. For example, no unmodified tryptic peptide rH1_T24 was identified. However, four glycoforms of rH1_T24 were chromatographically resolved and detected by RP LC-MSE (Fig.

Xie et al.

4A). The glycoform elution order correlated with the size of glycan moiety. The heavier the glycan moiety, the earlier the glycoform elutes. The most abundant glycoform with the glycan group –GlcNAc2Man3 was confirmed by MSE spectrum (see Supplemental Fig. 1A) and the glycosylation site was demonstrated to only occur on the N286 site which has the –NSS- motif instead of on the other N site (with –NVHmotif) in the tryptic peptide. This is consistent with the rule that N-linked glycosylations occur on N sites with –NXS/Tmotif [41]. Again, the oxonium sugar ions in the low m/z range and the fragment ions of backbone peptide plus sugar groups in the high m/z range further prove the identity of the glycopeptide. However, no supporting MSE spectra were available for the other three lower level glycoforms. To confirm the presence of lower-level glycoforms a HILIC-MRM assay, combining selective separation of glycopeptides by HILIC LC [39, 40] with sensitive detection by MRM, was developed by monitoring 4 selected MRM transitions during a 60-min gradient run. The four transitions include 1) precursor to oxonium ion 204.1; 2) to oxonium ion 366.1; 3) to a large fragment ion with backbone peptide containing an -GlcNAc; and 4) to another large fragment ion with backbone peptide containing a glycan moiety –GlcNAcFuc. The latter two transitions were employed to test if a glycoform belongs to X or XF structure (see Table 2 for details). X represents peptide with a glycan containing – GlcNAc-GlcNAc-(Man)n and XF for peptide with a glycan containing –GlcNAc[Fuc]-GlcNAc-(Man)n. The response of

Fig. (3). Separation, identification and quantification of unmodified and glycosylated tryptic peptide rB_T8 (LCPDCLNCTDLDVALGRPMCVGTTPSAK). A) Extracted ion chromatograms (XIC) of RP LC-MS; B) MSE spectrum of unmodified rB_T8; C) MSE spectrum of glycosylated rB_T8. The colors in the MSE spectra represent: red – y ions; blue – b ions; green – y/b ions after losing a water or ammonia molecule; grey – unassigned ions.


multiple transitions at the same RT is used to confirm the identity of a glycoform (see Supplemental Fig. 2A) for the confirmation of less abundance rH1_T24-GlcNAc2Man5). The sum of peak areas of the first two transitions was used to quantify the relative abundance of each glycoform (Fig. 4B) because these two transitions are common to all glycopeptides and glycoforms. Using a singly N-link glycosylated recombinant monoclonal antibody as a model sample (data will be present in a separate publication), we have demonstrated that the relative quantification of glycoforms achieved by this method was comparable with direct UV measurements in a HILIC-MRM/UV experiment. It also matched the quantification data indirectly obtained from 2AB labeled released glycan profiling by HILICFluorescence. The calculated relative abundance of the four glycoforms of rH1_T24 is shown in (Fig. 4). The improved separation of glycoforms by HILIC is helpful for quantifying relative abundance of site-specific glycoforms. Such cases are particularly significant when limited separation was achievable by the RP LC, such as for glycoforms of rH3_T21 as described below. Five glycoforms were detected by RP LC-MSE on tryptic peptide rH3_T21 (Fig. 5), but these glycoforms eluted in a single LC peak at 91.65 min (Fig. 5A). In addition to the MS


detection of the five glycoforms (Fig. 5B), the peptide sequence and the glycosylated modification site were confirmed by the MSE spectrum of the most abundant glycoform with a glycan moiety -GlcNAc2Man9 (see Supplemental Fig. 1B). But no MSE spectra were available for the four lower-abundance glycoforms. It is uncertain by the RP LCMSE data that the four less abundant and smaller glycoforms are native, because in-source fragmentation of the most abundant and largest glycoform may introduce these species. They were required to be further investigated and verified. All five glycoforms of rH3_T21 detected by RP LC-MSE were separated and confirmed by a HILIC-MRM assay (Fig. 5B) and Supplemental Fig. 2B). The assay also allowed for quantification of the relative concentration of the five glycoforms (Fig. 5B). It was observed that glycopeptides tend to ionize to multiple precursors with different charge states in RP ESI+ LCMS; for example +3 and +4 precursors for all glycoforms of rH3_T21 co-existed (Fig. 5B). However, in HILIC-MRM, +3 precursors were dominant for all the glycoforms of rH3_T21. Similarly observation for other glycopeptides (precursors dominant in one charge format), plus the above described common MRM transitions, helped us to rapidly develop HILIC-MRM assays to verify and quantify the glycopeptides and glycoforms detected from each rHA antigen

Fig. (4). Separation and confirmation of 4 glycoforms of N-link glycosylated tryptic peptide rH1_T24 (CQTPQGAINSSLPFQNVHPVTIGECPK). A) XIC of RP LC-MS; B) HILIC-MRM chromatogram, sum of the 2 monitored MRM transitions used for quantification (see text for details). The most abundant glycoform was confirmed by BiopharmaLynx interpreted MSE spectrum (see Supplemental Fig. 1A). The other 3 detected low-level glycoforms without supporting MSE spectra were confirmed by HILIC-MRM with multiple MRM transitions (see Supplemental Fig. 2A).


Xie et al.

Fig. (5). Separation and confirmation of 5 glycoforms of N-link glycosylated tryptic peptide rH3_T21 (ISIYWTIVKPGDILLINSTGNLIAPR). A) One chromatographic peak showing that the glycoforms were not separated in RP LC-MS; B) Accurate mass measurement indicated 5 co-existing glycoforms; C) HILIC-MRM chromatogram, sum of the 2 monitored MRM transitions used for quantification (see text for details), confirmed the 5 glycoforms by separation. The most abundant glycoform was confirmed by BiopharmaLynx interpreted MS E spectrum (see Supplemental Fig. 1B). The other 4 lower-level glycoforms without supporting MSE spectra were confirmed by HILIC-MRM with multiple MRM transitions (see Supplemental Fig. 2B). Precursors with both +3 and +4 charges were co-existed in the RP ESI+ LC-MS spectrum (Fig. 5B) for the glycoforms of this glycopeptide, but +3 charged precursors were observed dominantly in HILICMRM experiments.

in the formulated vaccine. All the glycoforms (except for rH3_T2, rH3_T3 and rH3_T8) listed in Table 2 were confirmed by a single LC run (60-min gradient). For the rH1_T1 peptide (see Table 2), the glycosylation was located to be at N11 instead of N10 by examining b/y fragment ions in the MSE spectra of its glycoforms, although both sites have an -NXS/T- motif for potential N-linked glycosylation. In a similar N-link glycosylated tryptic peptide (SQDICIGYHANNSTEQVDTIMEK) of the H5 antigen in the H5N1 virus, Blake and colleagues [27] also identified that the glycosylation occurred in the later N site instead of the former N site by LC-MS/MS analysis of the glycopeptide after de-glycosylation. However, no enough evidences were acquired in our experiments to determine the exact N site of glycans for the glycopeptides of rH3_T2, rH3_T3 and rH3_T8, which all have multiple N-linked glycosylation motifs (Table 2). For these glycopeptides, further investigations are required, such as using alternative digestion enzymes or protease [42] to separate the N-linked glycosylation sites in one tryptic peptide to different peptides.

The combination of discovery by reversed-phase LCMSE with confirmation by HILIC-MRM allows for obtaining the site-specific N-linked glycosylation profiles of HA antigens in the complex research influenza vaccine. Although no observation of O-linked glycosylation has been reported on HA antigen [27], we also tried to discover any potential Olinked glycosylation sites using the methods described above and none of them was identified from the three rHA proteins. Site-Specific Degradations Peptide mapping by LC-MSE was also applied to characterize potential degradation sites in the rHA antigens. Degradations may affect stability and efficacy of vaccines [25]. Moxidation and N-deamidation are two major degradation pathways for glycoproteins [43-45] which may occur during storage and transportation of vaccines. Using M-oxidation (+15.99 Da) and N-deamidation (+0.98 Da) as optional modifications, the LC-MSE data were processed by BiopharmaLynx for any potential degradation on each rHA antigen.


MSE is able to determine degradation type and locate the degradation sites and the intensities of precursors can be used to estimate the concentration of degradations [32, 33]. An example is shown in (Fig. 6), which presents the separation and quantification of the modified rH1_T22 (259GFGSGIINSNAPMDK-273). The oxidation site M271 and deamidation site N266 were confirmed by y and b ions in the MSE spectra, and the relative quantification was determined by the signal intensities of the precursors of M271-oxidized, N266-deamidated and unmodified rH1_T22. As expected, the M271-oxidized product eluted earlier than the unmodified rH1_T22 because M-oxidation increases the hydrophilicity of peptides, and two isoforms (isoD and D) [46] of the N266-deamidation products were observed. About 3.3% M271 oxidation and about 4.4% N266 deamidation were measured in antigen rH1. Same approach was applied for examination of other potential degradation sites in the rHA antigens. We have previously demonstrated that site-specific oxidation and deamidation at 0.5% level could be successfully identified and quantified by reversed-phase LC-MSE [33]. The quantification reproducibility was within several percent of standard deviation (SD), increasing to 10% or more for modified peptides near the limits of detection.


Secondly, HILIC-MRM was developed to confirm and monitor site-specific glycoforms that were poorly separated by RP LC. Low-abundance glycoforms without supporting MSE spectra were also confirmed by HILIC-MRM. Although RP LC is the most widely-used separation technique for peptide mapping and protein characterization, we demonstrated here that HILIC LC separation was particularly suitable for the analysis of site-specific glycoforms. The selective HILIC resolution [39, 40, 42] was helpful for determining and quantifying low-abundance glycoforms. By focusing on specific attributes to monitor, shorter LC gradients were developed for LC-MRM to confirm and quantify specific components of interest (e.g., site-specific PTMs) from the complex vaccine digest. This study opens the possibility of targeted LC-MRM assays to contribute to the monitoring of vaccine quality from batch to batch in production or after storage and transportation. In conclusion, the methods demonstrated here are suitable for the characterization of complex samples such as formulated influenza vaccines. These analytical tools can contribute to assays for a well-characterized vaccine. ACKNOWLEDGEMENTS The authors thank Drs. St. John Skilton, John C. Gebler and Martin Gilar for helpful discussion and critical review of this manuscript. SUPPLEMENTARY MATERIAL Supplementary material is available on the publishers Web site along with the published article. REFERENCES [1] [2]

Fig. (6). Separation (XIC) and quantification of unmodified, Moxidized and N-deamidated tryptic peptide rH1_T22 (GFGSGIINSNAPMDK) by RP LC-MSE and BiopharmaLnx 1.2.

[3]

[4]

CONCLUSIONS This study describes complementary LC-MS methods for characterization and monitoring of a recombinant influenza candidate vaccine formulated from purified rHA antigens. Firstly, peptide mapping with RP LC-MSE provided: 1) Identification of impurity proteins contaminant from the production processing; 2) Verification of rHA primary sequences with high (>90%) sequence coverage;

[5] [6] [7]

[8]

3) Discovery of site-specific glycosylation and glycoforms on each rHA protein; 4) Examination of potential degradations (e.g. M-oxidation and N-deamidation sites) on each rHA antigen.

[9]

Girard, M. P.; Cherian, T.; Pervikov, Y.; Kieny, M. P. A review of vaccine research and development: human acute respiratory infections. Vaccine, 2005, 23, 5708-5724. Stevens, J.; Corper, A. L.; Basler, C. F.; Taubenberger, J. K.; Palese P.; Wilson I.A. Structure of the uncleaved human H1 hemagglutinin from the extinct 1918 influenza virus. Science, 2004, 303, 1866-1870. Johansson, B. E. Immunization with influenza A virus hemagglutinin and neuraminidase produced in recombinant baculovirus results in a balanced and broadened immune response superior to conventional vaccine. Vaccine, 1999, 17, 2073-2080. Robertson, J. S.; Bootman, J. S.; Newman, R.; Oxford, J. S.; Daniels, R.S.; Webster, R.G.; Schild, G,C. Structural changes in the haemagglutinin which accompany egg adaptation of an influenza A(H1N1) virus. Virology, 1987, 160, 31-37. Daniels, R. S.; Downie, J. C.; Hay, A. J.; Knossow, M., Skehel, J.J.; Wang, M.L.; Wiley, D.C. Fusion mutants of the influenza virus hemagglutinin glycoprotein. Cell, 1985, 40, 431-439. Wiley, D. C.; Skehel, J. J. The structure and function of the hemagglutinin membrane glycoprotein of influenza virus. Annu. Rev. Biochem., 1987, 56, 365-394. Deom, C. M.; Caton, A. J.; Schulze, I. T. Host cell-mediated selection of a mutant influenza A virus that has lost a complex oligosaccharide from the tip of the hemagglutinin. Proc. Natl. Acad. Sci. USA, 1986, 83, 3771-3775. Wagner, R.; Wolff, T.; Herwig, A.; Pleschka, S.; Klenk, H. D. Interdependence of hemagglutinin glycosylation and neuraminidase as regulators of influenza virus growth: a study by reverse genetics. J. Virol., 2000, 74, 6316-6323. Williams, T. L.; Luna, L.; Guo, Z.; Cox, N. J.; Pirkle, J.L.; Donis, R.O.; Barr, J,R. Quantification of influenza virus hemagglutinins in complex mixtures using isotope dilution tandem mass spectrometry. Vaccine, 2008, 26, 2510-2520.

12 Current Pharmaceutical Biotechnology, 2011, Vol. 12, No. 12 [10]

[11] [12]

[13]

[14] [15]

[16] [17] [18]

[19]

[20]

[21] [22]

[23] [24] [25] [26] [27]

[28]

Robertson, J. S.; Cook, P.; Attwell, A. M.; Williams, S. P. Replicative advantage in tissue culture of egg-adapted influenza virus over tissue-culture derived virus: implications for vaccine manufacture. Vaccine, 1995, 13, 1583-1588. Getie-Kebtie, M.; Chen, D.; Eichelberger, M.; Alterman, M. Proteomics-based characterization of hemagglutinins in different strains of influenza virus. Proteomics Clin. Appl., 2009, 3, 979-988. Tree, J. A.; Richardson, C.; Fooks, A. R.; Clegg, J. C.; Looby, D. Comparison of large-scale mammalian cell culture systems with egg culture for the production of influenza virus A vaccine strains. Vaccine, 2001, 19, 3444-3450. Price, P. M.; Reichelderfer, C. F.; Johansson, B. E.; Kilbourne, E. D.; Acs, G. Complementation of recombinant baculoviruses by coinfection with wild-type virus facilitates production in insect larvae of antigenic proteins of hepatitis B virus and influenza virus. Proc. Natl. Acad. Sci. USA, 1989, 86, 1453-1456. Merten, O. W.; Manuguerra, J. C.; Hannoun, C.; van der Werf, S. Production of influenza virus in serum-free mammalian cell cultures. Dev. Biol. Stand., 1999, 98, 23-37. Halperin, S. A.; Smith, B.; Mabrouk, T.; Germain, M., Trépanier, P.; Hassell, T.; Treanor, J.; Gauthier, R.; Mills, E.L. Safety and immunogenicity of a trivalent, inactivated, mammalian cell culturederived influenza vaccine in healthy adults, seniors, and children. Vaccine, 2002, 20, 1240-1247. Cox, M. M. Cell-based protein vaccines for influenza. Curr. Opin. Mol. Ther., 2005, 7, 24-29. Cox, M. M. Progress on baculovirus-derived influenza vaccines. Curr. Opin. Mol. Ther., 2008, 10, 56-61. Treanor, J. J.; Wilkinson, B. E.; Masseoud, F.; Hu-Primmer, J.; Battaglia, R.; O'Brien, D.; Wolff, M.; Rabinovich, G.; Blackwelder, W.; Katz, J.M. Safety and immunogenicity of a recombinant hemagglutinin vaccine for H5 influenza in humans. Vaccine, 2001, 19, 1732-1737. Treanor, J. J.; Schiff, G. M.; Hayden, F. G.; Brady, R. C.; Hay, C.M.; Meyer, A.L.; Holden-Wiltse, J.; Liang, H.; Gilbert, A.; Cox, M. Safety and immunogenicity of a baculovirus-expressed hemagglutinin influenza vaccine: a randomized controlled trial. JAMA, 2007, 297, 1577-1582. Schwarzer, J.; Rapp, E.; Hennig, R.; Genzel, Y.; Jordan, I.; Sandig, V.; Reichl, U. Glycan analysis in cell culture-based influenza vaccine production: influence of host cell line and virus strain on the glycosylation pattern of viral hemagglutinin. Vaccine, 2009, 27, 4325-4336. Williams, M. S. Single-radial-immunodiffusion as an in vitro potency assay for human inactivated viral vaccines. Vet. Microbiol., 1993, 37, 253-262. Wood, J. M.; Mumford, J.; Schild, G. C.; Webster, R. G.; Nicholson, K. G. Single-radial-immunodiffusion potency tests of inactivated influenza vaccines for use in man and animals. Dev. Biol. Stand., 1986, 64, 169-177. Schwarzer, J.; Rapp, E.; Reichl, U. N-glycan analysis by CGE-LIF: profiling influenza A virus hemagglutinin N-glycosylation during vaccine production. Electrophoresis, 2008, 29, 4203-4214. Jenkins, N.; Parekh, R. B.; James, D. C. Getting the glycosylation right: implications for the biotechnology industry. Nat. Biotechnol., 1996, 14, 975-981. Brandau, D. T.; Jones, L. S.; Wiethoff, C. M.; Rexroad, J.; Middaugh, C. R. Thermal stability of vaccines. J. Pharm. Sci., 2003, 92, 218-231. Morrissey, B.; Downard, K. M. A proteomics approach to survey the antigenicity of the influenza virus by mass spectrometry. Proteomics, 2006, 6, 2034-2041. Blake, T. A.; Williams, T. L.; Pirkle, J. L.; Barr, J. R. Targeted Nlinked glycosylation analysis of H5N1 influenza hemagglutinin by selective sample preparation and liquid chromatography/tandem mass spectrometry. Anal. Chem., 2009, 81, 3109-3118. Garcia-Canas, V.; Lorbetskie, B.; Bertrand, D.; Cyr, T. D.; Girard, M. Selective and quantitative detection of influenza virus proteins in commercial vaccines using two-dimensional high-performance

Received: September 26, 2010

Revised: December 17, 2010

Accepted: January 11, 2011

Xie et al.

[29]

[30]

[31] [32]

[33]

[34]

[35] [36]

[37]

[38]

[39] [40]

[41] [42] [43]

[44] [45] [46]

liquid chromatography and fluorescence detection. Anal. Chem., 2007, 79, 3164-3172. Liu, N.; Song, W.; Lee, K. C.; Wang, P.; Chen, H.; Cai, Z. Identification of amino acid substitutions in avian influenza virus (H5N1) matrix protein 1 by using nanoelectrospray MS and MS/MS. J. Am. Soc. Mass Spectrom., 2009, 20, 312-320. Luna, L. G.; Williams, T. L.; Pirkle, J. L.; Barr, J. R. Ultra performance liquid chromatography isotope dilution tandem mass spectrometry for the absolute quantification of proteins and peptides. Anal. Chem., 2008, 80, 2688-2693. Schwahn, A. B.; Wong, J. W.; Downard, K. M. Subtyping of the influenza virus by high resolution mass spectrometry. Anal. Chem., 2009, 81, 3500-3506. Xie, H.; Chakraborty, A.; Ahn, J.; Yu, Y. Q.; Dakshinamoorthy, D.P.; Gilar, M.; Chen, W.; Skilton, S.J.; Mazzeo, J.R. Rapid comparison of a candidate biosimilar to an innovator monoclonal antibody with advanced liquid chromatography and mass spectrometry technologies. MAbs, 2010, 2, 379-394. Xie, H.; Gilar, M.; Gebler, J. C. Characterization of protein impurities and site-specific modifications using peptide mapping with liquid chromatography and data independent acquisition mass spectrometry. Anal. Chem., 2009, 81, 5699-5708. Silva, J. C.; Denny, R.; Dorschel, C. A.; Gorenstein, M.; Kass, I.J.; Li, G.Z.; McKenna, T.; Nold, M.J.; Richardson, K.; Young, P.; Geromanos, S. Quantitative proteomic analysis by accurate mass retention time pairs. Anal. Chem., 2005, 77, 2187-2200. Silva, J. C.; Gorenstein, M. V.; Li, G. Z.; Vissers, J. P.; Geromanos, S. J. Absolute quantification of proteins by LCMSE: a virtue of parallel MS acquisition. Mol. Cell Proteomics, 2006, 5, 144-156. Xie, H.; Griffin, T. J. Trade-off between high sensitivity and increased potential for false positive peptide sequence matches using a two-dimensional linear ion trap for tandem mass spectrometrybased proteomics. J. Proteome Res., 2006, 5, 1003-1009. Tomiya, N.; Narang, S.; Lee, Y. C.; Betenbaugh, M. J. Comparing N-glycan processing in mammalian cell lines to native and engineered lepidopteran insect cell lines. Glycoconj. J., 2004, 21, 343360. Li, G. Z.; Vissers, J. P.; Silva, J. C.; Golick, D.; Gorenstein, M.V.; Geromanos, S.J. Database searching and accounting of multiplexed precursor and product ion spectra from the data independent analysis of simple and complex peptide mixtures. Proteomics, 2009, 9, 1696-1719. Wuhrer, M.; de Boer, A. R.; Deelder, A. M. Structural glycomics using hydrophilic interaction chromatography (HILIC) with mass spectrometry. Mass Spectrom. Rev., 2009, 28, 192-206. Takegawa, Y.; Ito, H.; Keira, T.; Deguchi, K.; Nakagawa, H.; Nishimura, S. Profiling of N- and O-glycopeptides of erythropoietin by capillary zwitterionic type of hydrophilic interaction chromatography/electrospray ionization mass spectrometry. J. Sep. Sci., 2008, 31, 1585-1593. Gavel, Y.; von Heijne, G. Sequence differences between glycosylated and non-glycosylated Asn-X-Thr/Ser acceptor sites: implications for protein engineering. Protein. Eng., 1990, 3, 433-442. Zauner, G.; Koeleman, C. A.; Deelder, A. M.; Wuhrer, M. Protein glycosylation analysis by HILIC-LC-MS of Proteinase K-generated N- and O-glycopeptides. J. Sep. Sci., 2010, 33, 903-910. Vlasak, J.; Bussat, M. C.; Wang, S.; Wagner-Rousset, E.; Schaefer, M.; Klinguer-Hamour, C.; Kirchmeier, M.; Corvaïa, N.; Ionescu, R.; Beck, A. Identification and characterization of asparagine deamidation in the light chain CDR1 of a humanized IgG1 antibody. Anal. Biochem., 2009, 392, 145-154. Liu, H.; Gaza-Bulseco, G.; Xiang, T.; Chumsae, C. Structural effect of deglycosylation and methionine oxidation on a recombinant monoclonal antibody. Mol. Immunol., 2008, 45, 701-708. Kroon, D. J.; Baldwin-Ferro, A.; Lalan, P. Identification of sites of degradation in a therapeutic monoclonal antibody by peptide mapping. Pharm. Res., 1992, 9, 1386-1393. Robinson, N. E.; Robinson A. B. Molecular Clocks: Deamidation of Asparaginyl and Glutaminyl Residues in Peptides and Proteins, Athouse Press, Cave Junction, OR 2004.