Rate of Asparagine Deamidation in a Monoclonal Antibody

Jan 18, 2017 - Antibodies are an important class of drugs, comprising more than half of all new FDA approvals. Therapeutic antibodies must be chemical...
0 downloads 2 Views 1MB Size
Subscriber access provided by HACETTEPE UNIVERSITESI KUTUPHANESI

Article

The rate of asparagine deamidation in a monoclonal antibody correlates with hydrogen exchange rate at adjacent downstream residues Jonathan James Phillips, Andrew Buchanan, John Andrews, Matthieu Chodorge, Sudharsan Sridharan, Laura Mitchell, Nicole Burmeister, Alistair D. Kippen, Tristan J. Vaughan, Daniel R. Higazi, and David Christopher Lowe Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b04158 • Publication Date (Web): 18 Jan 2017 Downloaded from http://pubs.acs.org on January 20, 2017

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

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

Page 1 of 10

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

Analytical Chemistry

The rate of asparagine deamidation in a monoclonal antibody correlates with hydrogen exchange rate at adjacent downstream residues Jonathan J. Phillips†‡, Andrew Buchanan‡, John Andrews‡, Matthieu Chodorge‡, Sudharsan Sridharan‡, Laura Mitchell‡, Nicole Burmeister‡, Alistair D. Kippen‡, Tristan J. Vaughan‡, Daniel R. Higazi‡, David Lowe‡* †

Department of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge, CB2 3RA, UK. ‡ MedImmune Ltd., Aaron Klug Building, Granta Park, Cambridge, CB21 6GH, UK. *Correspondence to [email protected]

ABSTRACT: Antibodies are an important class of drug molecule, comprising more than half of all new FDA approvals. Therapeutic antibodies must be chemically stable both in storage and in vivo, following administration to patients. Deamidation is a major degradation pathway for all natural and therapeutic proteins circulating in blood. Here, the linkage between deamidation propensity and structural dynamics is investigated by examining two antibodies with differing specificities. While both antibodies share a canonical aspartate-glycine (NG) motif in a structural loop, this is prone to deamidation in one of the antibodies, but not the other. We found that the hydrogen-exchange rate at the adjacent two amides, often the autocatalytic nucleophiles in deamidation, correlated with the rate of degradation. This previously unreported observation was confirmed upon mutation to stabilize the deamidation lability via a generally applicable orthogonal engineering strategy presented here. We anticipate that the structural insight into chemical degradation in full-length monoclonal antibodies and the high-resolution hydrogen-exchange methodology used will have broad application across biology, drug discovery and development.

ment (scFv).6-10 The study of full-length IgG structure is challenging, notably due to their large size (~145 kDa) and dynamic conformation. These dynamic changes include rapid perturbations in CDR loops to slow, solvent-damped, inter-domain motions. Recently, significant efforts have been directed to better understand the conformational dynamics of CDRs and the role of structural mobility in defining target specificity11 and affinity.12 For example, the monoclonal antibody 2F5 recognizes a linear epitope in the HIV-1 coat protein GP41 at the membrane proximal external region (MPER).13,14 Hydrogen/deuterium-exchange mass spectrometry (HDX-MS) data indicated that flexibility of a hinge within CDR3 of the heavy chain (CDRH3) is required for MPER binding in its native lipid environment and subsequent high-affinity binding to a secondary epitope.15 These exemplars illustrate potential insights obtainable from investigating the dynamics of antibodies, despite the considerable analytical challenges. Asparagine deamidation is the non-enzymatic posttranslational modification of asparagine (Asn), typically via a cyclic imide (succinimide) intermediate, to aspartic (Asp) or iso-aspartic (iso-Asp) acids (Scheme 1). 16,17 This results in the appearance of a negative charge and potential distortion of the local structure. It is a common chemical degradation pathway of mAbs that can result in loss of antigen binding and bioactivity.18-22 For several antibody heavy chain variable genes, a particular germline-encoded asparagine (often Asn54) can be highly susceptible to degradation, particularly when followed by a glycine (Gly55) in a so-called ‘NG’ motif. mAb deamidation can occur during production and postadministration in vivo; therefore, novel approaches to design

Since the initial development of mouse hybridomas in 1975, monoclonal antibodies (mAbs) have proven themselves to be a diverse source of highly specific reagents for use as both therapeutics and diagnostics.1 mAbs have subsequently become the biggest growth sector within pharmaceuticals, accounting for five of the ten world-wide best selling prescription drugs in 2012.2 Immunoglobulin G (IgG) antibodies are the dominant therapeutic antibody isotype for therapeutics.3 IgG circulate with notably long half-lives (cf. 29 days for some IgG1,4) related to the particular pH-dependent affinity for the neonatal Fc receptor (FcRn). Natural and therapeutic antibodies are subject to many physical and chemical stresses, such as endosomal recycling, alternating between alkaline and acidic pH. Therapeutic IgG are also exposed to considerable chemical and physical stresses during production, including long bioreactor culture times (cf. 14 days), high pressure transverse flow filtration, acid pH elution from protein A affinity purification, alkaline pH wash steps, high concentration (molecular crowding) and exposure to air-liquid and solid-liquid interfaces.5 Throughout the synthesis and in vivo lifetime of all proteins, both naturally occurring and therapeutic, maintenance of the chemical integrity of the molecule is often vital for appropriate activity. Within the IgG structure, the hypervariable complementarity determining region (CDR) loops in the variable light (VL) and heavy (VH) domains confer antigen binding specificity and affinity. IgG CDR loops have been extensively studied structurally and, to a more limited extent, dynamically; however, the majority of these studies have been restricted to fragments of the IgG – typically the Fab or single chain variable frag-

1

ACS Paragon Plus Environment

Analytical Chemistry antibodies with high resistance to deamidation are important to

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

ensure

manufacturing

Page 2 of 10 consistency,

in

vivo

stability,

Scheme 1. Canonical ‘NG’ motif in antibody CDRH2 deamidation (i) and hydrogen/deuterium-exchange (ii)

(i) Deamidation at the canonical ‘NG’ site of an antibody often proceeds via a cyclic succinimide intermediate. Under basic conditions, the loss of amine can be autocatalysed with the backbone amide nitrogen of the adjacent (i+1) residue as nucleophile. Succinimide hydrolysis at site a is favoured 3:1 over site b. (ii) Base-catalysed hydrogen/deuterium-exchange at the i+1 backbone amide in excess D2O.

and therapeutic efficacy and safety.23 Current strategies to manage deamidation are focused on the control and removal of deamidated proteins during production and storage.19,24 To date, the removal of Asn residues from proteins that are potentially prone to deamidation via mutation has been an empirical process with limited insight into the conformational dynamics or chemical context of the Asn residues.25 Asparagine deamidation in polypeptides, via a succinimide intermediate, requires precise conformational alignment to permit nucleophilic attack of the asparagine gamma carbon by the backbone amide nitrogen of the adjacent “i+1” residue.26 In proteins, it occurs with a rate constant of ~10-5 s-1.27 However, with antibody in vivo half-lives of up to 29 days, a deamidation rate on this order of magnitude can be significant. As such, it is an important process and represents a non-trivial system in which to relate chemistry to structural dynamics, particularly in large protein systems. Given Scheme 1, we hypothesized a potential correlation between rate of hydrogen/deuterium-exchange of the highlighted H atom and rate of deamidation. Recently, hydrogen/deuterium-exchange mass spectrometry (HDX-MS) has proved itself a sensitive probe of protein conformation and dynamics.28,29 The approach is particularly powerful in the study of large proteins that are relatively intractable by other techniques, such as NMR. However, the established HDX-MS workflow uses enzymatic digestion, typically with pepsin, yielding lower resolution data – at the peptide, rather than amino-acid level. Recent work has advanced the coupling of hydrogen-exchange and electron transfer dissociation (ETD).30,31 Soft-fragmentation techniques, such as ETD, have the potential to increase resolution of hy-

drogen-exchange data by fragmenting selectively labeled peptides without inducing deuterium scrambling.32,33 In this study, we present an approach to obtaining highresolution hydrogen-exchange mass spectrometry data in monoclonal antibodies. We have applied this to investigate the hypothesis that asparagine deamidation rate correlates with hydrogen-exchange rate at the adjacent, often autocatalytic, amino acid. This yields a greater understanding of IgG degradation in the context of the full-length protein.

EXPERIMENTAL SECTION The recombinant human monoclonal IgG1 antibodies used in this study were manufactured at MedImmune Ltd and purified using protein A and mixed mode chromatography steps. Antibodies were formulated in a buffer containing 50 mM sodium acetate pH 5.50 and 100 mM sodium chloride. Reagents were of the highest grade commercially available. Accelerated deamidation of IgG. Antibody samples were buffer exchanged by overnight dialysis at 4 °C into 0.1 M Tris pH 8.5, and diluted to 3 mg/ml after recovery. Stressed material was generated by incubation at 40 °C for 0, 0.5, 1 and 2week timepoints. Samples were stored at -80 °C and thawed immediately prior to analysis. Quantification of deamidation in stressed samples. Samples were first digested with trypsin on a centrifugal filter membrane (Ultracel 10 k MWCO, Amicon). Antibody sample (100 µg) was diluted into denaturing buffer (8 M GuHCl, 130 mM Tris, 1 mM EDTA pH 7.6), reduced by addition of 500 mM DTT and incubated for 30 minutes at 37°C. Reduced thiol was capped by addition of 500 mM IAM. Centrifugal filters were

2

ACS Paragon Plus Environment

Page 3 of 10

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

Analytical Chemistry

equilibrated with 100 mM Tris, pH 7.6 prior to sample loading. Samples were desalted by buffer exchange into digestion buffer (2 M Urea, 100 mM Tris, pH 7.6). The digestion was initiated by addition of Trypsin (diluted to 1 µg/µL with suspension buffer provided by Promega) to give a final IgG1:enzyme ratio of 20:1 (w/w). Samples were incubated for 4 h at 37 °C and quenched with 4% TFA (v/v) solution. Peptides were recovered by inversion and centrifugation of the filters. Digested samples were loaded onto a Waters BEH300 C18 UPLC column (2.1 x 150 mm) and eluted with an 82 minute linear gradient of 0-35 % (v/v) acetonitrile, supplemented with 0.02 % (v/v) trifluoroacetic acid, at a flow rate of 200 µL min1 . The extent of peptide deamidation was determined by the shift in centroid for the isotopic envelopes over the timecourse as defined by equation 1. The rate of deamidation at a single site was determined by fitting to equation 2 in Prism (Graphpad). %  

 =

  

× 100

K), such that the final mixture pH was 2.55. The quenched protein was injected into a Waters HDX Manager with an immobilized pepsin column (2.0x30 mm; Poroszyme, Life Technologies), C18 trapping column (VanGuard ACQUITY BEH 2.1x5 mm; Waters) and analytical C18 column (1.0x100 mm ACUITY BEH; Waters). Mobile phases were 0.1 % formic acid in H2O (A) and 0.1 % formic acid in ACN (B), such that their pH was 2.55. Protein was applied to the pepsin and trapping columns in 100 µL/min buffer A and eluted from the analytical column in a linear gradient of 3-40 % B at 40 µL/min. Peptide sequences were assigned from MS^E fragment data with Protein Lynx Global Server (Waters) and DynamX (Waters). Labeling data was acquired as for sequencing, except the mass spectrometer acquired MS scans only. To enable comparison between mutant IgGs, back-exchange was corrected per peptide, per protein following incubation in deuterated quench buffer at 333 K for 24 h. Peptide-level deuteration was evaluated in DynamX (Waters) and ETD data were analyzed in MatLab (Mathworks) and Prism (GraphPad). Electron transfer dissociation (ETD). ETD reagent (1,3dicyanobenzene or 4-nitrotoluene) was infused in a glow discharge source to a total ion count (TIC) of 100-fold more than the TIC at 80% peak height of the analyte ion by varying the discharge current and make-up gas flow in negative ion mode. The mass spectrometer was tuned to maximize c/z product ion efficiency, but minimize H/D-scrambling as follows: capillary 3.0 kV; Source temperature 353 K; sample cone 20 V; source travelling wave velocity 300 m/s and height 0.2 V; trap travelling wave velocity 200 m/s; scan 1 s. Trap pressure and helium cell pressure were optimized per peptide precursor (typical values 6.9 E^-2 mbar and 1.1 E^-3 mbar, respectively), as were ion guide travelling wave height and DC bias in the trap and the pressure and collision energy applied to the transfer ion guide region. Extraction cone voltage was set to 4.0 V and did not induce scrambling in our case.34 Combined spectra of equivalent elution windows were smoothed twice with 5 channels. Deuterium incorporation at specific sites was calculated by subtraction of centroid values of overlapping fragment ions, discounting the N-terminal amino acid of the peptide. Nonlinear regression of deuterium incorporation data. Time-resolved deuterium incorporation per amino acid was fitted to a one-phase exponential association function (equation 2) in Matlab (Mathworks) or Prism (Graphpad). Where Y0 is the average natural isotope abundance of the peptide, Ymax is the maximum possible average mass following deuterium labeling, kobs is the observed first order rate of exchange (s-1) and x is the labeling time (s). Under EX2 conditions, the observed first order rate constant, kobs, relates to a protection factor (Pf) by equations 3-4.

(equation 1)

Where Mt is the peptide centroid at time t, M0 is the initial peptide centroid prior to incubation and Mmax is the maximum calculated centroid based on deamidation of all deamidation sites.  =  +  ! "1 −  $%&'(!) *+ (equation 2) Where Y0 is the average initial deamidation observed in the peptide, Ymax is the maximum observed deamidation, kobs is the observed first order rate of deamidation (s-1) and x is time under accelerated stability conditions (s). Y0 and Ymax were unconstrained in fitting. Site-localized quantification of deamidation per asparagine residue. Localised rates of deamidation for each of four Asn sites in the mAb1 CDRH2 tryptic peptide were found by MS/MS of the precursor triply-charged ion. Fragments that cleaved the peptide bond either side of the Asn residue were subtracted to determine deamidation at that site. Precursor ion corresponding to the H6 tryptic peptide was isolated in the quadrupole and subjected to collision induced dissociation (CID) with a trap collision energy ramp from 15 V – 45 V. H6 retention time was 37.4 min with 0.12 min full width at half maximum (FWHM), yielding 17 scans of 1 s. Deamidation for sequential b or y fragment ions, that differ only by the presence or absence of a terminal Asn residue, was calculated from equation 1. The deamidation value at the Asn residue is given by the subtraction of one sequential fragment from the other. Hydrogen/deuterium-exchange mass spectrometry (HDXMS). Proteins were diluted to 60 µM in potassium phosphate buffer, pH 6.0. This stock was used to initiate labeling experiments by diluting 20-fold with the buffer that was used for accelerated deamidation studies either in protonated (50 mM Tris, pH 8.5) or deuterated (50 mM Tris, pD 8.5) aqueous solvent. Initial mapping experiments were done to assign species from the mass spectra to peptic peptide sequences from the mAbs. This was done largely as described.29 Briefly, protonated diluted protein was mixed 1:1 with a quench solution (100 mM potassium phosphate, pH 2.45, 8 M urea, 0.5 M TCEP, 274

,-./ =

%&0 %12

,345 = 6-7 ,345 (equation 3)

Where kop and kcl are the opening and closing rate constants for the local structure, which has an equilibrium constant Kop. kint is the intrinsic rate of exchange for an amino acid backbone amide, calculated here using SPHERE.35 %&'( %89

= 6-7 =

:

;