Qualitative and Quantitative Characterization of Therapeutic Antibodies by Native Mass Spectrometry Sara Rosati dedicated to mamma e papà ISBN 978-90-8891-928-2 The research in this thesis was performed in the Biomolecular Mass Spectrometry & Proteomics Group, Utrecht University, Utrecht, The Netherlands. This research is supported by the Netherlands Proteomics Centre, embedded in the Netherlands Genomics Initiative and by the Dutch Technology Foundation STW, which is part of the Netherlands Organisation for Scientific Research (NWO) and partly funded by the Ministry of Economic Affairs (project number 10805). Printed by proefschriftmaken.nl || Uitgeverij BOXPress Qualitative and Quantitative Characterization of Therapeutic Antibodies by Native Mass Spectrometry Karakterisering van therapeutische antilichamen met behulp van massaspectrometrie onder natieve condities (met een samenvatting in het Nederlands) Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof. dr. G.J. van der Zwaan, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op woensdag 3 september 2014 des ochtends te 10.30 uur door Sara Rosati geboren op 15 maart 1985 te Termoli, Italië Promotor Prof. dr. A.J.R. Heck Copromotor Dr. E. van Duijn Table of contents Chapter I A Short Introduction into Native Mass Spectrometry and Antibodies 7 Chapter II Analysis of Monoclonal Antibodies by Mass Spectrometry 31 Chapter III Qualitative and Semiquantitative Analysis of Composite Mixtures of Antibodies by Native Mass Spectrometry 51 Chapter IV Exploring an Orbitrap Analyzer for the Characterization of Intact Antibodies by Native Mass Spectrometry 67 Chapter V In-depth Qualitative and Quantitative Analysis of Composite Glycosylation Profiles and Other Micro-Heterogeneity on Intact Monoclonal Antibodies by High-Resolution Native Mass Spectrometry Using a Modified Orbitrap 79 Chapter VI Detailed Mass Analysis of Structural Heterogeneity in Monoclonal Antibodies Using Native Mass Spectrometry 99 Chapter VII Summary and Future Outlook 127 Nederlandse Samenvatting 131 Supplementary Information 134 Appendix Curriculum Vitae 149 List of publications 150 Acknowledgements 152 I r Chapter I e t p a h c A Short Introduction into Native Mass Spectrometry and Antibodies Partially based on: Performing Native Mass Spectrometry Analysis on Therapeutic Antibodies N. J. Thompson, S. Rosati, A.J.R. Heck Methods 65, 11-17 (2014) 7 My thesis collects the work of four years, wherein I enjoyed looking for what is I r e invisible to the naked eye. For the work described in my thesis, mass spectrome- t p ters somehow became my eyes, and, if you hold on and keep on reading, you will a h see what good eyes we managed to develop in these years to see the things we c wanted to study! If you are really willing and have time to read the whole thesis, (and I would be glad if someone really does) be warned: you will read the term “native mass spec- trometry” or “Native MS” hundreds of times! Since I am particularly kind (after all I am a native MS person, gentleness is my specialty!) I’m going to introduce my PhD work by spending a few words on this peculiar terminology, extremely com- mon for my colleagues and me, but maybe a little arduous for most people, even experts in chemistry, pharmacy and biology. The word “native” was coined for the first time as an adjective to describe “mass spectrometry” by my promotor Albert Heck in 20041,2 and has since then become widely accepted in the field, with even workshops and conferences organized around this technique. The unique feature that led to the choice of such adjective is the capability of native mass spectrometry to mass analyze protein and protein complexes in the gas-phase, whereby non-covalent interactions can be retained. Therefore, proteins can be analyzed retaining properties of the tertiary and qua- ternary structures of the samples as it occurs in more conventional and broader known techniques of size-exclusion chromatography and native (non-reducing) gels. Are intrinsic properties really conserved during native MS analysis? Is a question my colleagues and myself get questioned many times. Most people may argue that proteins transferred into the gas-phase, completely desolvated, carrying a number of charges, are very unlikely to retain their “real” physiological structure, conformation and function. Still, cumulating from many reports from our group and others it has now been established that proteins during this transition can keep a certain level of conformational memory3. Notwithstanding, we are aware that a certain level of disruption will occur. However, if this disruption does occur this is evidently of such low extent that most non-covalent interactions remain intact. That is very fortunate for me as it allowed me to investigate interesting properties of proteins, and in particular of therapeutic antibodies, described throughout this thesis. 8 1. Native Mass Spectrometry I r e t Mass spectrometry (MS) is a technique with a broad range of applications, in part p a due to the universal nature of its capability to separate and identify multiple com- h c ponents in mixtures by differences in mass. Although MS has been around for more than a century, it was initially limited to the analysis of small molecules due to the inability to transfer large biomolecules into the gas phase efficiently, without breaking them apart. However, with the introduction of so-called “soft” or “gentle” ionization techniques, i.e., matrix-assisted laser desorption ionization (MALDI)4 and electrospray ionization (ESI)5, MS became capable of studying much larger biomolecules. This has been regarded as such a breakthrough for science that it has been awarded the Nobel Prize in 2002. The application area of ESI was subsequently even further extended via the incorporation of volatile buffers at neutral pH as solvents, e.g., aqueous ammonium acetate. This combination al- lowed the retention of non-covalent interactions and thus protein complexes in the gas phase, a technique as explained above termed native MS1,2,6. Native MS has further evolved to utilize nanoflow ESI for even more gentle ionization condi- tions at much higher sensitivity and coupled to quadrupole-time-of-flight (Q-TOF) analyzers, with a theoretically infinite mass range. The dream of Nobel Laureate John Fenn that electrospray could make “elephants fly” has not been achieved yet, although molecular elephants such as whole intact ribosomes and viruses7,8 can now be ionized and thus studied by native mass spectrometry. Therefore, it is worthwhile to describe the technology used in a bit more detail (Figure 1). Nano-electrospray ionization (nano-ESI) Nowadays, ESI is one of the most widespread methods of ionization used in MS, used for amongst other small molecules, lipids, sugars, polymers and peptides and proteins. In conventional ESI a voltage difference ranging from 3 to 6 kV is applied across a capillary containing the sample and the counter-electrode placed at atmospheric pressure. The flow through the capillary is typically in the µl/min range. The resulting electric field induces the formation of a Taylor-cone consist- ing of small charged liquid droplets at the tip of the capillary. The evaporation of the solvent causes the shrinking of the droplets around the analyte. For large molecules, such as intact proteins, it is believed that ions enter the gas-phase via the so-called charge residue model, in which Coulombic repulsion between charges induces a number of fission events, drastically reducing the droplet size, until the bare, completely desolvated, charged analyte is transferred into the gas- phase10-12. 9 I r e t p a h c Figure 1. Experimental workflow of native MS analysis of mAbs pictured in three main steps: sample preparation, native MS analysis, and data analysis. Starting from the puri- fied sample, a deglycosylation step can be performed followed by buffer exchange. Instru- mental analysis is performed on TOF or Orbitrap instruments, directly infusing the sample using a nano-ESI source. Finally, several software programs can be used to analyze the raw data thus obtaining both qualitative and quantitative information. Reprinted from9. Two unique features have made of ESI “the” ionization method for native MS: i) the mildness of the ionization process, and ii) the formation of multiply charged ions. While the first feature is necessary to preserve non-covalent interactions, the second allows the detection of large biomolecules at (relatively) low m/z. Yet, these two very important features are not enough to make a native protein or 10