Chemoselective Modification Strategies for Peptides and Proteins in Aqueous Media Dissertation zur Erlangung des akademischen Grades des Doktors der Naturwissenschaften (Dr. rer. Nat.) Eingereicht im Fachbereich Biologie, Chemie, Pharmazie der Freien Universität Berlin Vorgelegt von Michaela Mühlberg aus Berlin 2014 Die Arbeit wurde zwischen dem 01. Juli 2009 und dem 31. März 2014 unter der Leitung von Prof. Dr. Christian P. R. Hackenberger am Institut für Chemie und Biochemie der Freien Universität Berlin sowie am Leibniz-Institut für Molekulare Pharmakologie angefertigt. 1. Gutachter: Prof. Dr. Christian P. R. Hackenberger 2. Gutachter: Prof. Dr. Beate Koksch Disputation am: 10. Juli 2014 Declaration I herewith confirm that I have prepared this dissertation without the help of any impermissible resources. All citations are marked as such. The present thesis has neither been accepted in any previous doctorate degree procedure nor has it been evaluated as insufficient. Berlin, 31 th March 2014 Michaela Mühlberg The work on this dissertation resulted so far in the following publications: 1. M. Mühlberg, D. M. M. Jaradat, R. Kleineweischede, I. Papp, D. Dechtrirat, S. Muth, M. Broncel, C. P. R. Hackenberger, Bioorganic & Medicinal Chemistry 2010 , 18 , 3679-3686. Acidic and basic deprotection strategies of borane-protected phosphinothioesters for the traceless Staudinger ligation 2. S. Sowa, M. Mühlberg, K. M. Pietrusiewicz, C. P. R. Hackenberger, Bioorganic & Medicinal Chemistry 2013 , 21 , 3465-3472. Traceless Staudinger Acetylation of Azides in Aqueous Buffers 3. M. Mühlberg, K. D. Siebertz, B. Schlegel, P. Schmieder, C. P. R. Hackenberger, Chemical Communications 2014 , DOI: 10.1039/c4cc00774c. Controlled Thioamide vs. Amide Formation in the Thioacid–Azide Reaction under Acidic Aqueous Conditions 4. M. Mühlberg, M. G. Hoesl, C. Kuehne, J. Dernedde, N. Budisa, C. P. R. Hackenberger, 2014 , submitted. Orthogonal Dual-Modification of Proteins by Oxime Ligation and Cu-Catalyzed Azide– Alkyne Cycloaddition Acknowledgements First, I would like to thank Prof. Dr. Christian Hackenberger for offering me the opportunity to work in his research group on such diverse and interesting projects and for the freedom he gave me to pursue different ideas. I would like to especially thank him also for the support that I got throughout my whole career. I am grateful to Prof. Dr. Beate Koksch for being my second supervisor throughout my Ph.D., for helpful discussions and for being the second referee of my thesis. Special thanks go to Ina Wilkening, Robert Vallée, Paul Majkut, Verena Böhrsch and Lukas Artner for fruitful discussions and their support in the lab, and many pleasant hours outside of the lab throughout all those years. I also would like to thank Maria Glanz, Olaia Nieto-Garcia, Dominik Schumacher and Kristina Siebertz for all their support during the last year of my Ph.D. and their patience with me during my last months. All current and former group members are thanked for the pleasant and friendly atmosphere, especially my lab mates Da’san Jaradat, Jordi Bertran, Simon Reiske and Divya Agrawal. A thank you goes also to Nediljko Budisa, Michael Hösl and Nina Bohlke for their constant supply of proteins, fruitful discussions and the very pleasant collaboration on an interesting and challenging project. I would also like to thank Prof. Pietrusiewicz and Sylwia Sowa for the productive collaboration and discussions. I would like to thank Chris Weise for his time and support during my first MALDI-MS measurements and all the scientific troubles that came with it. I would like to acknowledge Andreas Springer and the whole mass spectrometry team from the FU Berlin as well as Eberhard Krause and the whole mass spectrometry team from the FMP in Buch for their help with all my MS measurements over time and for helpful discussion on more complex problems. I also thank Andreas Schäfer and the team of the NMR service at the FU Berlin as well as Peter Schmieder and the team of the NMR department at the FMP in Buch for their help and support with my NMR measurements with special thanks to Brigitte Schlegel. Many thanks also go to Katrin Wittig, Marianne Dreißigacker and Katharina Tebel for their help and support with all administrative problems, especially Katta for many helpful and motivating conversations throughout my Ph.D. For financial support, I would like to thank the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 765), the Fonds der chemischen Industrie, the Studienstiftung des deutschen Volkes, the Freie Universität Berlin, the Leibniz-Institute for Molecular Pharmacology and the Humboldt Universität Berlin. Many thanks go to my friends, especially Silke Zobel, for all their support and patience with me during the last years. Last but not least, I would like to thank Mark Archibald and my family, especially my mother, for their belief in me during all those years and their motivation and support, which helped me to get to the end. for my mother Abstract In nature, a protein’s structure and function depends amongst other things on very precise posttranslational modifications such as phosphorylations, glycosylations and many others. Our ability to naturally and unnaturally modify a peptide or a protein in a chemoselective and thereby very distinct fashion enables us to examine biological phenomena in a very sophisticated approach. During this thesis, we have focused on a variety of different ways of decorating peptides and proteins in a chemoselective manner. This work involves four different approaches that deal with different chemoselective labelling strategies to introduce either natural or unnatural modifications to peptides or proteins. In the first project, the traceless Staudinger ligation and its application for selective acetylation of azido lysine peptides in aqueous media was probed (Scheme 1A). Different alkyl azides were reacted with two different phosphines: 1) commercially available hydrophobic (diphenylphosphino-)methanethiol acetate and 2) water-soluble bis( p-dimethylaminoethyl)- phosphinomethanethiol acetate. Initial studies showed no chemoselectivity of the reaction in the presence of basic amino acid side chains such as lysine. Consequently, the traceless Staudinger ligation was probed with an azido lysine without free amines present in the peptide. Thereby, only the water-soluble variant showed a good reactivity in aqueous buffered systems. The highest conversion of 61% was achieved with the water-soluble phosphine (10 eq.) in a phosphate buffer (0.4 M, pH 8)/DMF mixture (2:8). Scheme 1: Selective acetylation of alkyl azido peptides by A) traceless Staudinger ligation or B) thioacid–azide reaction. In the second project, the thioacid–azide reaction was probed for selective acetylation of alkyl azido peptides (Scheme 1B). As alkyl azides are more electron-rich and therefore less reactive than the commonly used sulfonyl azides, a side reaction of the thioacid with basic side chains such as lysines was observed. Therefore, the reaction was probed at different pH values with an electron-rich azido butanoyl and a modestly electron-poor azido glycine peptide. The results showed a maximum conversion towards the desired acetylated peptides at pH 5 and conversion rates could be easily increased to > 99% at higher concentrations during the reaction. At lower pH, an unexpected and different main product was observed, which was not the desired amide but a thioamide (Scheme 1B). Thioamide formation could be increased at pH 2 to more than 90% for the modestly electron-poor azido peptide. It was shown that the ratio of thioamide and amide depends on the pH of the reaction mixture as well as on the nature of the azide, which allowed selective control over the desired product. Within the third project, a new multiple-labelling approach for proteins was probed. Selective dual-modification of a thermophilic lipase (TTL) was achieved successfully by combination of two orthogonal functionalisation strategies: oxime ligation and CuAAC (Figure 1). To do so, the TTL was bearing several azidohomoalanine residues, which were incorporated by selective pressure induction as methionine analogues, and an N-terminal serine, which could be post- translationally reacted with periodate to yield an aldehyde as the second bioorthogonal tag. This strategy allowed modification of the protein with two different functional moieties: galactose for carbohydrate–protein binding studies and biotin for immobilisation of the protein. With these modifications, the protein could be applied in surface plasmon resonance measurements to study their different binding to a lectin and to determine lectin binding constants. Thereby, the biotin functionalisation allowed immobilisation of our constructs on a streptavidin-coated chip, which required significantly less amounts of our proteins for the measurements. OH OH O biotin H2N O N 3 N HO O CuAAC 3 oxime OH N3 ligation O R N3 N N3 3 N3 Figure 1: Dual-modification approach on TTL by oxime ligation and CuAAC. In addition, dual-functionalisation a GFP variant was performed successfully by selective combination of thiazolidine formation on the chemically created N-terminal aldehyde and CuAAC on previously incorporated azidohomoalanine residues. It was shown that thiazolidine formation is more efficient than oxime ligation and yields a very stable product. In the fourth project, a new way for selective post-translational aldehyde formation on a distinct position anywhere throughout a protein was planned. To do