Glycoforms of Immunoglobulin G Based Biopharmaceuticals Are

Subscriber access provided by The Univ of Iowa Libraries

Article

Glycoforms of immunoglobulin G based biopharmaceuticals are differentially cleaved by trypsin due to the glycoform influence on higher order structure David Falck, Bas Cornelis Jansen, Rosina Plomp, Dietmar Reusch, Markus Haberger, and Manfred Wuhrer J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.5b00573 • Publication Date (Web): 05 Aug 2015 Downloaded from http://pubs.acs.org on August 11, 2015

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

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

Page 1 of 31

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

Journal of Proteome Research

Glycoforms of immunoglobulin G based biopharmaceuticals are differentially cleaved by trypsin due to the glycoform influence on higher order structure David Falcka,*, Bas C. Jansena, Rosina Plompa, Dietmar Reuschb, Markus Habergerb, and Manfred Wuhrera,c

a

Center for Proteomics and Metabolomics, Leiden University Medical Center,

Albinusdreef 2, 2333 ZA Leiden, The Netherlands b

Pharma Biotech Development Penzberg, Roche Diagnostics GmbH, 82377 Penzberg,

Germany c

Division of BioAnalytical Chemistry, VU University Amsterdam, De Boelelaan 1083,

1081 HV Amsterdam, The Netherlands

*to whom correspondence should be addressed

ACS Paragon Plus Environment

1


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

Page 2 of 31

Abstract It has been reported that glycosylation can influence the proteolytic cleavage of proteins. A thorough investigation of this phenomenon was conducted for the serine protease trypsin which is essential in many proteomics workflows. Monoclonal and polyclonal immunoglobulin G biopharmaceuticals were employed as model substances which are highly relevant for the bioanalytical applications. Relative quantitation of glycopeptides derived from the conserved Fc-glycosylation site allowed resolution of biases on the level of individual glycan compositions. As a result, a strong preferential digestion of high mannose, hybrid, alpha2-3-sialylated and bisected glycoforms was observed over the most abundant neutral, fucosylated glycoforms. Interestingly, this bias was, to a large extend, dependent on the intact higher order structure of the antibodies and, consequently, was drastically reduced in denatured versus intact antibodies. In addition, a cleavage protocol with acidic denaturation was tested which featured reduced handson time and toxicity while showing highly comparable results to a published denaturation, reduction and alkylation based protocol.

Keywords: -

Glycoproteomics

-

glycosylation

-

Trypsin substrate specificity

-

Tryptic cleavage

-

Biopharmaceuticals

-

Immunoglobulin G

-

Monoclonal antibodies

-

Method development

-

Higher order structure

-

Proteolytic biases


2

Page 3 of 31

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


Introduction Bottom-up proteomics approaches are widely used in biological and clinical research.1-3 In addition, they are becoming an integral part of the analysis of biopharmaceuticals in discovery, development and batch control.4 At the heart of bottom-up proteomics is the cleavage of the protein of interest by proteases, most commonly the serine protease trypsin, into more readily measurable peptides.5 Therefore, a robust analysis has to incorporate an efficient and unbiased tryptic cleavage.6 Biases can be introduced by an incomplete cleavage of the target protein. This can lead to a lower sequence coverage which hinders protein identification in studies targeting multiple proteins.5 Similarly, the numerous proteoforms of a single protein, many of which arise from post-translational modifications (PTM), may introduce biases.1 In terms of quantitation, proteolytic cleavage biases can be tackled by employing isotopically-labeled internal standards (ILIS).7 However, a PTM bias could result in a bias in protein quantitation, if the proteoforms of ILIS differ in susceptibility to proteolysis. PTMs, such as glycosylation, are often addressed by defining site occupancy as well as relative quantification of glycoforms per site. Therefore, a bias in cleavage will directly translate into an apparent bias in the glycoform profile. It has been shown for trypsin

8,9

and other serine proteases

8,10

that the deglycosylated protein is more rapidly

digested than the glycosylated proteoforms. Additionally, there are hints that different glycoforms may also be preferentially digested. For example, when comparing an extensive tryptic cleavage of a monoclonal antibody with a rapid trypsin/LysC digest, Du et al. found a higher relative abundance of high mannose glycoforms with the rapid protocol.11 In another case, Deshpande et al. observed a higher ratio of afucosylated glycoforms in a tryptic peptide of the sIgA joining chain with one missed cleavage as compared to the fully digested peptide in an application of an automated data analysis software package.12 With regard to the explanation of these phenomena, steric hindrance has been suggested to cause the preferential cleavage of glycoforms with smaller glycans.9,12 Alternatively, the conformation of the glycoprotein may play a major


3


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

Page 4 of 31

role.11 Notably, the conformation of the peptide backbone of antibodies is known to be influenced by the glycan, specifically the IgG Fc N-glycan, which changes and stabilizes the CH2-domain.13,14 In order to investigate the differential cleavage of glycoforms, we measured the glycopeptides from the conserved IgG Fc glycosylation site of two biopharmaceutical preparations: 1. A marketed therapeutic IgG1 mAb and 2. Flebogamma, an intravenous IgG (IVIG) prepared from healthy donors. Both mAbs and IVIGs have a wide variety of therapeutic indications, treatment of cancer being a prominent example for mAb use15 and IVIG finding application in the treatment of primary immunodeficiency and autoimmune diseases.16 In this study, we demonstrate that different IgG glycoforms show a different susceptibility to tryptic cleavage beyond the simple difference between glycosylated and non-glycosylated proteoforms. We empirically explore these preferences using IgG1 mAb and Flebogamma as examples of recombinant and natural human antibodies, respectively. Finally, we propose an efficient denaturation and cleavage protocol which offers complete cleavage of biopharmaceutical IgG preparations into their glycopeptides with minimal hands-on time and omitting toxic reagents.


4

Page 5 of 31

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


Materials and methods Samples and chemicals An immunoglobulin 1 monoclonal antibody (IgG1 mAb) was obtained from a Chinese hamster ovary (CHO) cell line and purified by the Downstream Processing Group at Roche GmbH. Flebogamma (IVIG) was kindly provided by H. U. Scherer (Department of Rheumatology, Leiden University Medical Center, Leiden, The Netherlands). Modified and tosyl-phenylalanyl-chloromethyl-ketone(TPCK) treated, porcine trypsin was purchased from Promega (sequencing grade; Leiden, The Netherlands). Water was purified with a Purelab ultra (ELGA Labwater, Ede, The Netherlands). Acetonitrile (LC–MS grade) was from Biosolve (Valkenswaard, The Netherlands) and 2-propanol (Chromasolv grade) was from Sigma-Aldrich (Schnelldorf, Germany). All other chemicals were of analytical grade and were obtained from SigmaAldrich.

Tryptic cleavage of IgG1 mAb and Flebogamma For the experiments without prior denaturation, 100 µg of antibody were diluted in 25 mM freshly prepared ammonium bicarbonate buffer (ABC, brought to pH 7.8 with formic acid) to a final volume of 100 µL. Then, 100 µL of trypsin (enzyme:substrate ratio 1:10; 10 µg) in ABC were added. The mixture was incubated at 37°C and 20 µL samples were taken at 1 h, 2 h, 3 h, 4 h, 6 h, 9.5 h, 22 h and 26 h. These samples were quenched in 20 µL ice-cold ACN and stored at -20°C until the analysis. For denaturation with formic acid (FA), antibody stock solutions were diluted in 100 mM FA to a final volume of 100 µL (75 mM minimal final FA concentration) and mixed by vortexing. These solutions were incubated for 15 min at room temperature and then evaporated to dryness in a centrifugal vacuum concentrator (RVC2-25COplus, Martin Christ Gefriertrocknungsanlagen, Osterode am Harz, Germany) at 60°C. The dried antibodies were re-dissolved in 200 µL of trypsin (enzyme:substrate ratio 1:10; 10 µg) in ABC on a shaker (Titramax 100, Heidolph Instruments, Schwabach, Germany) for 5 min at 400 rpm. Incubation and sampling was performed as described above.


5


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

Page 6 of 31

Experiments with denaturation, reduction and alkylation followed a protocol derived from Damen et al.17 An aqueous solution containing 5.5 M urea, 270 mM ABC, 10 mM DL-dithiothreitol (DTT) and 2.4 mg/mL monoclonal IgG1 was incubated for 30 min at 60°C in order to denature and reduce the antibody. Afterwards, IAA is added to a final concentration of 23 mM. For alkylation, this mixture is incubated for 45 min at room temperature in the dark and then another 10 min at room temperature in direct light. For the tryptic cleavage the following reagents remain at the given final concentrations: A solution containing ca. 1.2 M urea, 570 mM ABC, 2.1 mM DTT, 0.5 mg/mL antibody and 0.05 mg/mL trypsin (enzyme:substrate ratio 1:10). Incubation and sampling was performed as described above with the exception of the 6 h and the 9.5 h sample which were instead taken at 5.5 h and 9 h, respectively.For detergentassisted denaturation with sodium dodecyl sulphate (SDS), 100 µg of antibody were diluted in 25 mM ABC (pH 7.8) to a volume of 90 µL. 10 µL of a 0.2% (w/w) SDS solution were added (0.02% final concentration). Afterwards, the mixture was incubated for 15 min at 60°C. Finally, 100 µL of trypsin (enzyme:substrate ratio 1:10; 10 µg) were mixed with the (partially) denatured antibody solution and incubated at 37°C.

Purification and complete cleavage of the undigested antibody fragments The three 6 h samples of IgG1 mAb from the trypsin cleavage time curves with detergent-assisted denaturation were pooled and half of it (approximately 15 µg) was freeze-dried overnight and re-suspended in water. The sample was reduced by heating for 10 min at 95°C in Laemmli buffer containing 2-mercaptoethanol. The reduced partial digest was then run on a NuPage 4-12% Bis-Tris SDS-PAGE gel (Life Technologies, Paisley,

UK),

stained

with

Coommassie

G-250

(SimplyBlue

SafeStain,

Life

Technologies) and de-stained in water. Bands at approximately 55, 45, 30 and 14 kDa were excised and cut into pieces. The gel pieces were washed with 25 mM ABC followed by dehydration with acetonitrile. Additional reduction in 10 mM DTT and 25 mM ABC at 55°C for 30 min was followed by dehydration in acetonitrile and alkylation in 55 mM iodoacetamide and 25 mM ABC for 40 min in the dark. The gel pieces were then washed in 25 mM ABC and dehydrated in acetonitrile, washed and dehydrated again and dried in a centrifugal vacuum


6

Page 7 of 31

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


concentrator (Eppendorf, Hamburg, Germany). Finally, 30 µL of 25 mM ABC containing 0.25 µg of trypsin were added to each of the samples. After an hour on ice to allow the liquid to be taken up by the gel pieces, the samples were incubated overnight at 37°C. The liquid surrounding the gel pieces was collected for nanoLC–ESI-MS analysis.

NanoLC–ESI-MS method The glycopeptides from the antibody digests were analyzed on a nanoLC–ESIMS platform described earlier.18 In short, an Ultimate 3000 RSLCnano (Thermo Fisher Scientific, Breda, NL) was hyphenated to a Maxis Impact HD quadrupole-time-of-flight mass spectrometer (q-TOF; Bruker Daltonics, Bremen, Germany) using a sheath-flow nanoESI sprayer (Agilent Technologies, Amstelveen, The Netherlands). 250 nL of the sample were injected into a flow of 25 µL/min solvent A and trapped on a Dionex Acclaim PepMap100 C18 5 mmx300 µm (Thermo Fisher Scientific, Breda, The Netherlands). Afterwards, the glycopeptides were eluted at a flow rate of 900 nL/min onto an Acentis Express C18 nanoLC column 50 mm x 75 µm with 2.7 µm fused core particles (Supelco, Bellefonte, PA) by back-flushing and separated with the following gradient: During injection 1% solvent B, 0 min at 3% B, linear to 6% B at 2 min, linear to 18% B at 4.5 min, linear to 30% B at 5 min, isocratic till 7 min, linear to 1% B and then re-equilibration at 1% B till 10.9 min. In the nanoESI sprayer, the eluent was mixed with the sheath liquid and measured with positive ESI in full spectrum analysis mode on the q-TOF. Data was recorded in a range from m/z 500 to m/z 2000 with a frequency of 0.5 Hz. The applied LC–MS method is able to distinguish between the IgG subclasses, except for IgG2 and 3 whose tryptic Fc glycopeptides have identical amino acid sequences in Caucasians. IgG4 abundance was often too low to extract meaningful profiles, especially for the early time point samples of the cleavage curves.

Data processing Data were internally calibrated using Data Analysis 4.2 (Bruker) on five glycopeptides and a solvent cluster (see Supporting Information Table S1). Afterwards, the data were transferred to .mzXML format with the MSconvert application of


7


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

Page 8 of 31

ProteoWizard 3.0 suite19 and retention times of the IgG1 and IgG2 glycopeptides manually extracted using MZmine 2.10.20 Glycopeptide compositions and charge states used for targeted feature extraction were determined by manual inspection of selected spectra. The highest intensity of the first three isotopes of every analyte-charge state combination was extracted within a window of ±0.04 Da around the theoretical mass and ±15 s around the manually extracted average retention time. This was done automatically with the 3D Max Extractor script, written in Python 2.7.21 3D Max Extractor takes a list of features and, subsequently, uses a binary search algorithm to isolate the retention time and mass region around each feature. All the spectra within the time window are then examined, retrieving the maximum intensity value within the m/z window, and yielding a single intensity value per spectrum within the time window. The single highest intensity out of the previously acquired list of intensities is then reported as the feature intensity. An output file is generated listing all the specified features and their respective maximum intensity value. Summing of the isotopes and charge states, background subtraction and calculation of the relative intensities was done in Excel 14.0. The absolute intensities were calculated by summing the automatically extracted and background subtracted values of the three isotopes and the two observed charge states, [M+2H]2+ and [M+3H]3+. The relative intensities were calculated from the absolute intensities by total area normalization, meaning that the absolute intensity of each analyte was divided by the sum of the intensities of all analytes.

Statistical Analysis In order to determine the statistical significance of the differences in the IgG glycopeptide profiles, we performed a series of t-tests in R (version 3.0.1).22 The relative abundance of each glycopeptide after either 3 h or 26 h of cleavage under the same denaturing conditions was compared using a paired parametric t-test (two-tailed, under the assumption of unequal variance). An unpaired parametric t-test (again two-tailed and assuming unequal variance) was applied to compare the influence of different denaturing conditions, after either 3 h or 26 h of cleavage. A p-value of less than 0.05 was considered to be significant.


8

Page 9 of 31

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


Results and Discussion We investigated the influence of glycoforms, present on the preserved Fcglycosylation site of IgG, on the efficiency of tryptic cleavage. More specifically, we tested for differences in the production efficiency of the EEQYNSTYR glycopeptide of IgG1 (EEQFNSTFR for IgG2 and IgG3) depending on the glycoform. In order to investigate the influence of the glycoforms on tryptic cleavage, the relative abundances of multiple glycoforms were assessed at different time points during tryptic cleavage of two biopharmaceutical preparations. The relative abundances were calculated by total area normalization. For the IgG1 mAb, 21 glycoforms were registered while in the IVIG preparation Flebogamma 12 and 8 glycoforms were followed for IgG1 and IgG2/3, respectively. The difference in the number of glycoforms is mainly a result of the presence of high mannose and hybrid glycoforms in IgG1 mAb, which are expected in a mAb, but not usually found in plasma IgG of healthy donors.4,18 Incompletely cleaved protein fragments of IgG1 mAb were isolated by gel electrophoresis, fully cleaved by in-gel trypsin treatment, and their glycoprofiles determined for comparison with the glycosylation profile of fully cleaved IgG1 mAb.


9


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

Page 10 of 31

Figure 1: Cleavage time courses for different glycoforms of IgG1 mAb. The graphs show the time courses of the relative abundance of six glycoforms obtained with three different sample pretreatments. Means of triplicates are given, with the error bars denoting standard error of the mean (SEM). For many data points error bars are not visible, as the SEM was smaller than the size of the symbol.

Cleavage time courses of IgG1 mAb including the influence of denaturation Glycoform biases of trypsin The cleavage time courses, which were registered for six glycoforms of IgG1 mAb, are shown in Figure 1. For other glycoforms see Supporting Information Table S2. These data clearly demonstrate preferential cleavage of certain glycoforms, if the samples were not denatured prior to cleavage. The three sialic acid containing glycoforms H4N4F1S1 (G1FS), H5N4F1S1 (G2FS) and H5N4F1S2 (G2FS2) show the highest preference for tryptic cleavage in the non-denatured samples. Their relative


10

Page 11 of 31

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


abundances were initially two to three times higher than in the denatured, reduced and alkylated (DRA) samples, and were constantly decreasing in the cleavage time course of the native sample. This is evident from the example of G2FS in Figure 1A and from Supporting Information Table S2. Statistical comparison of the 3 h and 26 h time points confirmed the decrease of these sialylated glycoforms in the time course (G1FS p=0.028; G2FS p=0.010; G2F2S2 p=0.014; compare also Table S3 in the Supporting Information) indicating preferred cleavage. The second in line with respect to susceptibility to tryptic cleavage are the high mannose and hybrid glycoforms represented by Man5 in Figure 1B. Because Man5 was digested slower than the sialylated glycoforms, but faster than the complex type glycoforms, its relative abundance reached a maximum at around six hours in the nondenatured samples. Man5 showed a statistically significant bias (difference between 3 h and 26 h sample; p=0.011). The other detected compositions which fall into the categories of high mannose, hybrid or mono-antennary glycoforms were H6N2 (Man6), H7N2 (Man7), H8N2 (Man8), H3N3, H4N3, H5N3, H6N3, H3N3F1, H4N3F1, H5N3F1 and H6N3F1. While these glycoforms seemed to show a behavior which either resembled that of Man5 or that of the sialylated glycoforms, they could not always be quantified reliably, especially after long cleavage or after denaturation, and were therefore excluded from statistical analysis. A clear time-dependent change of the relative abundances was observed for the fucosylated complex glycoforms H3N4F1 (G0F), H4N4F1 (G1F) and H5N4F1 (G2F). The curves for G1F are depicted in Figure 1C and the non-denatured curve shows a circa three times lower abundance of G1F in early samples as compared to late samples. G0F and G2F show a similar behavior. Interestingly, the case is much more differentiated for the non-fucosylated complex glycoforms H3N4 (G0), H4N4 (G1) and H5N4 (G2). While G0 is underrepresented (Figure 1D), the relative abundance of G1 is almost constant (Figure 1E) and G2 is overrepresented in the early samples (Figure 1F). Again, all changes from 3 h to 26 h, even in the G1 relative abundance, were found to be statistically significant (G0F p=0.004; G1F p=0.003; G2F p=0.016; G0 p=0.006; G1 p=0.013; G2 p=0.006).


11


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

Page 12 of 31

Absolute glycopeptide intensities Absolute intensities were calculated as described in Materials and methods: Data processing. Though the applied LC–MS method is not quantitative, an analysis of the changes in absolute glycopeptide intensities provided additional confirmation of the findings described above. Due to the number of detected glycoforms and the resulting potential complexity of a kinetic model, we did not attempt to fit the cleavage time curves. Complex-type, non-sialylated glycoforms exhibited the largest increase between the 9 h and 22 h data points while high mannose, hybrid and sialylated glycoforms already showed the strongest increase before the 6 h data point (see Supporting Information Figure S1). These observations are consistent with the changes observed in the relative profiles.

Influence of the denaturation step By comparing the glycopeptide profiles and their time-dependent behavior observed for the non-denatured and the fully denatured peptide chain of IgG1 mAb, we sought to determine the influence of protein higher order structure on the cleavage biases observed. More specifically, we suspected the known influence of Fc Nglycosylation on the antibody’s protein higher order structure to be a key factor in the cleavage of the non-denatured antibodies.13,14 To further study this, we used a denaturation, reduction and alkylation protocol as the gold standard for the destruction of the protein higher order structure. Although the first two time points (1 h and 2 h of incubation) were statistically different from the rest for some glycoforms, the biases were quite small especially when compared to the non-denatured samples (see Figure 1). Notably, there were no more differences detected between the 3 h and the 26 h sample for any glycoform. A comparison between the glycopeptide profiles without denaturation and with DRA showed for all glycoforms significant differences after 3 h of cleavage while after 26 h only G0, G1FS, G2FS, G2FS2 and Man 5 remained different. On the one hand, this again confirms that there is a glycoform bias in trypsin cleavage. On the other hand, it proves that this bias is highly dependent on the intact higher order structure of the antibody or more specifically on the influence the glycan has on it. An influence of the denaturation on reaction speed, for example that the cleavage was not


12

Page 13 of 31

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


complete after 26 h without denaturation, is quite expected. However, we also show experiments proving that the influence of cleavage progression is secondary to that of the denaturation (see below). An example of the extent of the biases is presented in Figure 2. The Man5 species was among the largest peaks in an MS spectrum of the non-denatured sample (Figure 2A) while it had only a low abundance in a spectrum of an acid-denatured sample.


13


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

Page 14 of 31

Figure 2: LC–MS sum spectra of IgG1 mAb glycopeptides. A) A spectrum recorded after 6 h of tryptic cleavage without prior denaturation shows a prominent Man5


14

Page 15 of 31

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


glycopeptide species ([M+2H]2+). B) This spectrum recorded after 22 h of tryptic cleavage with prior acidic denaturation is dominated by the glycopeptides with complex type glycans. The dotted line separates the areas were doubly- and triply-charged versions of the tryptic Fc glycopeptides were observed. The most prominent interferences can be attributed to: * singly-charged interferences; # multiply-charged ions with could not be attributed to a glycopeptide composition.

The glycopeptides profiles after full cleavage As discussed, the glycopeptide profiles did not change any more after 3 h in the DRA sample. Therefore, the most reliable profile at full cleavage was obtained by averaging the DRA profile from 3 h to 26 h. Table 1 depicts these glycopeptides profiles which, for simplicity, we will call final profiles from here on. Only 10 of the 21 IgG1 mAb glycoforms are still found in the final profiles. In the samples in which they are overestimated, these 11 glycoforms can be detected, but in the final profiles they are pushed below the limit of detection.


15


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

Page 16 of 31

Table 1: Final profiles of IgG1 mAb and IVIG (mean and SEM) Glycoform

H3N4F1 (G0F)

m/za of IgG1 [M+3H]3+; [M+2H]2+ 878.683; 1317.523

Structure

Relative abundance IgG1 mAb IVIG DRA Acidic IgG1 IgG2/3 33.74 33.41 17.87 27.20 ±0.25% ±0.19% ±0.15% ±0.29%

H4N4F1 (G1F)

932.700; 1398.551

41.89 ±0.18%

41.48 ±0.28%

30.43 31.89 ±0.16% ±0.10%

H5N4F1 (G2F)

986.717; 1479.578

9.10 ±0.10%

9.10 ±0.13%

16.55 13.75 ±0.11% ±0.16%

H3N4 (G0)

829.996; 1244.493

5.44 ±0.06%

5.52 ±0.07%

1.26 ±0.02%

n.d.

H4N4 (G1)

884.014; 1325.520

3.37 ±0.03%

3.44 ±0.03%

2.47 ±0.02%

n.d.

H5N4 (G2)

938.038; 1406.539

0.81 ± complex bisected glycoforms > complex alpha2-6-sialylated glycoforms ≈ other complex fucosylated glycoforms. Hybrid or monoantennary glycoforms are also preferentially cleaved, while for complex afucosylated glycoforms the situation is more complex and preferential cleavage seems to correlate positively with galactosylation. Effective denaturation of the antibody greatly reduced the bias, but did not fully abolish it. Therefore, proper denaturation and (near to) complete cleavage are essential for biasfree glycopeptide profiling. Beyond glycoproteomics, any proteomics experiment quantitatively comparing two antibody samples or a sample to a(n internal) standard can


26

Page 27 of 31

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


potentially be influenced by this phenomenon, if the glycosylation is not identical. This is also true, if tryptic peptides far away from the glycosylation site are selected for analysis. When unexpected glycosylation profiles are observed in the analysis of recombinantly produced antibodies, it might be prudent to reassess the protease cleavage step in addition to the purification and biotechnological production or sampling. We have provided in depth information on the biases which might be telltale signs of issues with the tryptic cleavage step. In the future, it would be interesting to investigate whether this phenomenon is limited to IgG or whether these biases are the rule rather than the exception. A logical next step would be the analysis of this phenomenon on therapeutic proteins related to antibodies, such as Fc-fusion proteins.4 It would also be interesting to investigate the effect of Fab glycosylation on proteolytic cleavage which should be comparably simpler on Fab glycosylated mAb, such as cetuximab, than on polyclonal IVIG.4 This could provide further crucial information for the troubleshooting of biopharmaceutical analysis protocols. As glycosylation is considered a critical quality attribute, our findings as well as our acidic denaturation-based digestion protocol, are expected to have significant impact on and potential for the analysis of antibody glycosylation for process comparability and biosimilarity assessment.30 We did (also) use original human antibodies, but cannot make statements about possible physiological relevance, because we focused on analytically relevant modified porcine trypsin instead of using human trypsin. However, it is conceivable that the observed bias may also have a physiological function, for example in protein turnover or quality control.

Acknowledgement This research and David Falck, Dietmar Reusch and Markus Haberger were supported by Hoffmann-La Roche. David Falck and Bas C. Jansen acknowledge funding by a Horizon Programme Zenith project funded by The Netherlands Genomic Initiative (Project Number: 93511033), and David Falck and Rosina Plomp by the European Union (Seventh Framework Programme HighGlycan project, grant number 278535). We are very grateful to Paul J. Hensbergen and A. Bondt for helpful discussions.


27


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

Page 28 of 31

Supporting Information • • • • • • • • •

Table S1: Calibration list Table S2: Time course of the tryptic cleavage of IgG1 mAb without denaturation step. Figure S1: Absolute abundances of glycopeptides in the tryptic cleavage of IgG1 mAb without denaturation. Table S3: Results of the statistic comparison of the relative abundances of IgG1 mAb glycoforms obtained with different sample preparation methods. Table S4: Results of the statistic comparison of the relative abundances of IVIG (Flebogamma) glycoforms obtained with different sample preparation methods. Figure S2: Comparison of cleavage time curves of IgG1 mAb including detergent-assisted denaturation. Figure S3: SDS-PAGE analysis of the IgG mAb1 cleavage mixture after 6 h cleavage with prior SDS denaturation. Table S5: Comparison of glycan profiles between fully digested glycopeptides and partially digested fragments at 6 h from the SDS denaturation time curves. Table S6: Time course of the tryptic cleavage of IVIG without denaturation step.

References 1. Lisitsa, A.; Moshkovskii, S.; Chernobrovkin, A.; Ponomarenko, E.; Archakov, A. Profiling proteoforms: promising follow-up of proteomics for biomarker discovery. Expert Rev Proteomics 2014, 11 121-129. 2. Hernandez, C.; Waridel, P.; Quadroni, M. Database construction and peptide identification strategies for proteogenomic studies on sequenced genomes. Curr Top Med Chem 2014, 14 425434. 3. Scherl, A., Clinical protein mass spectrometry. Methods, DOI: 10.1016/j.ymeth.2015.02.015 4. Beck, A.; Wagner-Rousset, E.; Ayoub, D.; Van Dorsselaer, A.; Sanglier-Cianferani, S. Characterization of therapeutic antibodies and related products. Anal Chem 2013, 85 715-736. 5. Switzar, L.; Giera, M.; Niessen, W. M. Protein digestion: an overview of the available techniques and recent developments. J Proteome Res 2013, 12 1067-1077. 6. Hustoft, H. K.; Malerod, H.; Wilson, S. R.; Reubsaet, L.; Lundanes, E.; Greibrokk, T. A Critical Review of Trypsin Digestion for LC-MS Based Proteomics. In Integrative Proteomics; Leung, H., Eds.; InTech: Rijeka, 2012; Vol. 1, 73-82. 7. Coombs, K. M. Quantitative proteomics of complex mixtures. Expert Rev Proteomics 2011, 8 659-677. 8. An, H. J.; Peavy, T. R.; Hedrick, J. L.; Lebrilla, C. B. Determination of N-glycosylation sites and site heterogeneity in glycoproteins. Anal Chem 2003, 75 5628-5637. 9. Lee, J. Y.; Kim, J. Y.; Park, G. W.; Cheon, M. H.; Kwon, K. H.; Ahn, Y. H.; Moon, M. H.; Lee, H. J.; Paik, Y. K.; Yoo, J. S. Targeted mass spectrometric approach for biomarker discovery and validation with nonglycosylated tryptic peptides from N-linked glycoproteins in human plasma. Mol Cell Proteomics 2011, 10 1-12.


28

Page 29 of 31

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


10. Juhasz, P.; Martin, S. A. The utility of nonspecific proteases in the characterization of glycoproteins by high-resolution time-of-flight mass spectrometry. International Journal of Mass Spectrometry 1997, 169 217-230. 11. Du, Y.; Wang, F.; May, K.; Xu, W.; Liu, H. LC-MS analysis of glycopeptides of recombinant monoclonal antibodies by a rapid digestion procedure. J Chromatogr B Analyt Technol Biomed Life Sci 2012, 907 87-93. 12. Deshpande, N.; Jensen, P. H.; Packer, N. H.; Kolarich, D. GlycoSpectrumScan: fishing glycopeptides from MS spectra of protease digests of human colostrum sIgA. J Proteome Res 2010, 9 1063-1075. 13. Krapp, S.; Mimura, Y.; Jefferis, R.; Huber, R.; Sondermann, P. Structural analysis of human IgG-Fc glycoforms reveals a correlation between glycosylation and structural integrity. J Mol Biol 2003, 325 979-989. 14. Buck, P. M.; Kumar, S.; Singh, S. K. Consequences of glycan truncation on Fc structural integrity. MAbs 2013, 5 904-916. 15. Harris, M. Monoclonal antibodies as therapeutic agents for cancer. Lancet Oncol 2004, 5 292-302. 16. Looney, R. J.; Huggins, J. Use of intravenous immunoglobulin G (IVIG). Best Pract Res Clin Haematol 2006, 19 3-25. 17. Damen, C. W.; Chen, W.; Chakraborty, A. B.; van Oosterhout, M.; Mazzeo, J. R.; Gebler, J. C.; Schellens, J. H.; Rosing, H.; Beijnen, J. H. Electrospray ionization quadrupole ion-mobility time-of-flight mass spectrometry as a tool to distinguish the lot-to-lot heterogeneity in Nglycosylation profile of the therapeutic monoclonal antibody trastuzumab. J Am Soc Mass Spectrom 2009, 20 2021-2033. 18. Selman, M. H.; Derks, R. J.; Bondt, A.; Palmblad, M.; Schoenmaker, B.; Koeleman, C. A.; van de Geijn, F. E.; Dolhain, R. J.; Deelder, A. M.; Wuhrer, M. Fc specific IgG glycosylation profiling by robust nano-reverse phase HPLC-MS using a sheath-flow ESI sprayer interface. J Proteomics 2012, 75 1318-1329. 19. Chambers, M. C.; Maclean, B.; Burke, R.; Amodei, D.; Ruderman, D. L.; Neumann, S.; Gatto, L.; Fischer, B.; Pratt, B.; Egertson, J.; Hoff, K.; Kessner, D.; Tasman, N.; Shulman, N.; Frewen, B.; Baker, T. A.; Brusniak, M. Y.; Paulse, C.; Creasy, D.; Flashner, L.; Kani, K.; Moulding, C.; Seymour, S. L.; Nuwaysir, L. M.; Lefebvre, B.; Kuhlmann, F.; Roark, J.; Rainer, P.; Detlev, S.; Hemenway, T.; Huhmer, A.; Langridge, J.; Connolly, B.; Chadick, T.; Holly, K.; Eckels, J.; Deutsch, E. W.; Moritz, R. L.; Katz, J. E.; Agus, D. B.; MacCoss, M.; Tabb, D. L.; Mallick, P. A cross-platform toolkit for mass spectrometry and proteomics. Nat Biotechnol 2012, 30 918-920. 20. Pluskal, T.; Castillo, S.; Villar-Briones, A.; Oresic, M. MZmine 2: modular framework for processing, visualizing, and analyzing mass spectrometry-based molecular profile data. BMC Bioinformatics 2010, 11 395. 21. Van Rossum, G.; Drake Jr, F. L. Python reference manual; Centrum voor Wiskunde en Informatica Amsterdam: 1995; Technical Report CS-R9526 22. R Core Team. R: A language and environment for statistical computing; R Foundation for Statistical Computing: Vienna, Austria, 2012; http://www.R-project.org/ 23. Bondt, A.; Rombouts, Y.; Selman, M. H.; Hensbergen, P. J.; Reiding, K. R.; Hazes, J. M.; Dolhain, R. J.; Wuhrer, M. Immunoglobulin G (IgG) Fab glycosylation analysis using a new mass spectrometric high-throughput profiling method reveals pregnancy-associated changes. Mol Cell Proteomics 2014, 13 3029-3039.


29


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

Page 30 of 31

24. Reusch, D.; Haberger, M.; Falck, D.; Peter, B.; Maier, B.; Gassner, J.; Hook, M.; Wagner, K.; Bonnington, L.; Bulau, P.; Wuhrer, M. Comparison of methods for the analysis of therapeutic immunoglobulin G Fc-glycosylation profiles-Part 2: mass spectrometric methods. MAbs 2015, 0. 25. Mimura, Y.; Sondermann, P.; Ghirlando, R.; Lund, J.; Young, S. P.; Goodall, M.; Jefferis, R. Role of oligosaccharide residues of IgG1-Fc in Fc gamma RIIb binding. J Biol Chem 2001, 276 45539-45547. 26. Umana, P.; Jean-Mairet, J.; Moudry, R.; Amstutz, H.; Bailey, J. E. Engineered glycoforms of an antineuroblastoma IgG1 with optimized antibody-dependent cellular cytotoxic activity. Nat Biotechnol 1999, 17 176-180. 27. Mimura, Y.; Church, S.; Ghirlando, R.; Ashton, P. R.; Dong, S.; Goodall, M.; Lund, J.; Jefferis, R. The influence of glycosylation on the thermal stability and effector function expression of human IgG1-Fc: properties of a series of truncated glycoforms. Mol Immunol 2000, 37 697-706. 28. Vidarsson, G.; Dekkers, G.; Rispens, T. IgG subclasses and allotypes: from structure to effector functions. Front Immunol 2014, 5 520. 29. Raju, T. S.; Lang, S. E. Diversity in structure and functions of antibody sialylation in the Fc. Curr Opin Biotechnol 2014, 30 147-152. 30. Artemenko, N. V.; McDonald, A. G.; Davey, G. P.; Rudd, P. M. Databases and tools in glycobiology. Methods Mol Biol 2012, 899 325-350.


30

Page 31 of 31

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


for TOC only 90x49mm (300 x 300 DPI)


Glycoforms of Immunoglobulin G Based Biopharmaceuticals Are

Recommend Documents