«Biochemical Methods for NMR Investigations of Large Proteins Sylvain Tourel Vollständiger Abdruck der von der Fakultät für Chemie der Technischen ...»
Lehrstuhl für Organische Chemie II der
Technischen Universität München
Biochemical Methods for NMR Investigations
of Large Proteins
Vollständiger Abdruck der von der Fakultät für Chemie der Technischen Universität
München zur Erlangung des akademischen Grades eines
Doktors der Naturwissenschaften (Dr. rer. nat.)
Vorsitzender: Univ.-Prof. Dr. Christian F. W. Becker
Prüfer der Dissertation: 1. Priv.- Doz. Dr. Gerd Gemmecker
2. Univ.-Prof. Dr. Sevil Weinkauf Die Dissertation wurde am 19.06.2008 bei der Technischen Universität München eingereicht und durch die Fakultät für Chemie am 22.09.2008 angenommen.
To my parents and Matthias « Ce n'est pas parce que les choses sont difficiles que nous n'osons pas, c'est parce que nous n'osons pas qu'elles sont difficiles. »
« Nicht weil es schwer ist, wagen wir es nicht, sondern weil wir es nicht wagen, ist es schwer. »
« It is not because things are difficult that we do not dare, it is because we do not dare that they are difficult. »
(Lucius Annaeus Seneca) Sénèque Zusammenfassung Die NMR-Spektroskopie stellt neben der Strukturbestimmung eine der wichtigsten Methoden zur Untersuchung der Dynamik von Proteinen dar. In dieser Arbeit werden verschiedene Methoden beschrieben, mit deren Hilfe auch Proteine oberhalb der momentan limitierenden Größe von ungefähr 100kDa untersucht werden können. Neben der selektiven Markierung von Proteinen, welche unter anderem zu einer geringeren Überlappung der Signale im Spektrum führt, wird ein neuer Ansatz für die Präparation von reversen Mizellen vorgestellt. Durch den Transfer von Proteinen in diese Art von Mizellen werden deren Relaxationseigenschaften drastisch verbessert. Eine Kombination dieser Techniken stellt eine vielversprechende Methode zur Untersuchung von großen Proteinen und Protein-Komplexen dar.
Abstract The basic motivation of this work is to make measurements of large proteins accessible to nuclear magnetic resonance, allowing studies of large protein complexes at atomic resolution. Because nuclear magnetic resonance also allows measurements of protein dynamics, it is a powerful biophysical method and complementary to X-ray crystallography.
Our particular approach aims to extend the size limit of proteins suitable for NMR studies, which is currently estimated at 100kDa.
The results of this work describe an improved strategy for isotopic segmental labeling of proteins as signal filter for reducing signal overlap. It introduces a new approach of reverse micelles preparation via phase transfer and thus provides a powerful and widely applicable method for successful encapsulation of proteins in reverse micelles.
Practical examples of this approach demonstrate the feasibility of combining several biophysical and biochemical methods for successfully studying large proteins.
Acknowledgements The work presented in this thesis was prepared from March 2005 until June 2008 under the supervision of PD Dr. Gerd Gemmecker and Prof. Dr. Horst Kessler at the Chemistry department of the Technical University of Munich.
I would like to thank Prof. Dr. Horst Kessler personally for, his trust, his interest and his financial support.
In addition I sincerely thank Dr. Gerd Gemmecker for his technical support and his motivation for the challenging projects.
I would like to thank Prof. Dr. Gerhard Wagner and Dr. Philipp Selenko from the Harvard Medical School for their full support and for the great working atmosphere.
I thank very particularly my colleague and friend Sandra Groscurth for the great technical discussions, team work, and for her full support during the good as well as the bad periods. This work would have not been possible without our close collaboration.
I am indebted to numerous people who have influenced this work in one way or the other (in alphabetical order): Dr. Hari Arthanari, Johannes Beck, Dr. Murray Coles, Lucas Doedens, Janine Eckardt, Katie Edmonds, Andreas Enthart, Alexander Frenzel, Dr.
Dominique Frueh, Dr. Rainer Haessner, Timo Huber, Dr. Dmitri Ivanov, Peter Kaden, Maura Kilcommons, Jochen Klages, Dr. Tom Malia, Dr. Assen Marintchev, Dr. Monica Lopez, Dr. Monika Oberer, Florian Opperer, Dr. Ricard Rodriguez, Dr. Chikako Suzuki, Dr. Koh Takeuchi, Dr. Vincent Truffault and Mona Wolff.
Chapter 1: In-vivo intein segmental labeling
3. Materials and methods
3.1. Plasmid construction
3.2. Expression and purification of segmentally labeled HSP90
3.3. Expression and purification of segmentally labeled eIF4A
4. Results and discussions
Chapter 2: Thermal shift assay
10. Material and methods
10.1. Method implementation
11. Results: unfolding/refolding study of GB1
12. Application to natively unfolded protein A-synuclein
13. Conclusion and future developments
Chapter 3: Reverses micelles
17. Material and methods
Contents 217.1. Proteins
17.4. CD spectroscopy
17.5. Recovery of partially unfolded transferred cCytochrome C
17.6. Efficient extraction of hemoglobin
18. Conclusion and future developments
Abbreviations 1 Abbreviations
General Introduction Modern methods in nuclear magnetic resonance (NMR) spectroscopy provide a wealth of information on the structural and kinetic properties of different biological molecules. Unfortunately, multidomain proteins have long been excluded from modern NMR studies due to their large sizes and because of the resulting spectral complexity with high degrees of resonance overlap. For these reasons NMR spectroscopy has largely focused on isolated protein domains in the past. While those analyses have provided strong evidence that many protein domains exhibit functional and structural characteristics that are indiscernible from their full-length protein propensities, information about inter-domain orientations or about the interactions between individual domains could not be obtained experimentally. Here we present in the first chapter a protocol that allows to selective isotope-labeling of individual protein domains in multi-domain proteins, for the purpose of analyzing them with modern NMR methods. The described approach enables single protein domains to be studied in full-length protein contexts. While advancements in NMR methodology, most notably the introduction of transverse relaxation-optimized spectroscopy (TROSY) methods1, have greatly extended the size limit of proteins suitable for NMR analyses 2,3, the presented method of domain-selective isotope labeling reduces the spectral complexity of large multi-domain proteins to single domain polypeptides.
Nevertheless high resolution NMR of large proteins suffers from adverse relaxation properties due to their slow tumbling in solution.
Reverse micelles (RMs) are aggregates in which nanoscale droplets of a polar liquid, usually water, are surrounded by a surfactant layer in a nonpolar continuous phase. Nonpolar continuous phases such as alkanes provide low viscosity to
General Introduction 2reverse micelles, directly influencing their overall tumbling and therefore the protein ‘s NMR properties. They are widely used as media for reactions or protein encapsulation, in which the extent of confinement or the presence of a surfactant interface play a central role to the protein structure. Our studies on reverse micelle preparation have focused on the effects of protein extraction via phase transfer, which are determined by charged interactions. Specifically, we have examined the effects of water on the secondary structure and a strategy to influence the water amount within the reverse micelles.
Biophysical characterization of proteins is a key factor to understanding biology at a chemical level. Understanding protein folding/unfolding improves our knowledge of protein stability, protein aggregation or fibril formation. Therefore factors influencing protein folding are very important, and high throughput screening assays can facilitate corresponding studies. The thermal shift assay takes advantage of an environmentally sensitive fluorescence dye, such as Sypro Orange®, and follows its signal changes while the protein undergoes thermal unfolding. The assay is widely used for screening compound libraries for ligands of targets proteins. The ligand-binding affinity of any potential inhibitor can be assessed from the shift of the unfolding temperature (Tm) obtained in the presence vs. absence of the potential inhibitor.
The assay can also be used to detect und study essential conditions for protein folding unstability. We have applied this strategy to determine buffer conditions which destabilize GB1 allowing unfolding/refolding NMR experiments via thermal denaturation in a temperature range compatible with NMR experiments.
. Wider, G. & Wüthrich, K. NMR spectroscopy of large molecules and multimolecular assemblies in solution. Curr. Opin. Struct. Biol., 9, 594Yu, H. Extending the size limit of protein nuclear magnetic resonance.
Proc. Natl. Acad. Sci. USA 96, 332-334.
. Iwai H, Zuger S, Jin J, Tam PH. Highly efficient protein trans-splicing by a naturally split DnaE intein from Nostoc punctiforme. FEBS Lett.
2006 Mar 20;580(7),1853-8.
Chapter 1: In-vivo intein segmental labeling 1Chapter 1: In-vivo intein segmental labeling
Nuclear magnetic resonance (NMR) provides, contrary to X-ray crystallography, more information on the kinetics, interactions and conformational state of polypeptide chains. The interdomain structural reorganisation caused by proteinprotein or protein-ligand binding events of large, multidomain proteins is essential for the understanding of biological signal transduction pathways, as well as for drug design strategies intervening with those.
Since isolated domains of proteins tend to be structurally identical to the original multidomain protein, NMR structural studies are usually focused on single domains with a reasonable size for NMR. Nevertheless studies of full length proteins would have the benefit of giving insight into the structural interactions between domains, which are often the molecular basis for regulation and signal transduction.
In order to make such systems accessible for NMR, we have adapted and optimized an intein ligation method based on the recent in vivo ligation protocols developed by Hideo Iwai.
Chapter 1: In-vivo intein segmental labeling 2
Molecular biology has focused on understanding protein functions and interactions on a structural level. For this purpose scientists have developed and improved biophysical methods, allowing the observation of events and rearrangements. Such techniques as NMR, infrared spectroscopy, fluorescence spectroscopy, electrophysiology or crystallography have pushed molecular biology to a field of excellence.
NMR is strongly established in the field of life science. This success is certainly the result of major and constant developments in sampling strategies, acquisition and processing methods in the last 20 years  . Still NMR has not reached yet its full capacity and method developments are a key of the NMR success.
With the introduction of isotopic labeling in the mid 90s, the sensitivity of NMR has increased drastically, making nitrogen and carbon spins available for protein NMR studies. But NMR has maintained one of its main difficulties: the protein size. Large proteins have disadvantageous physical properties, which have a strong impact on the measurement results. Therefore NMR research groups have mainly focused on single domain proteins and structures published and stored in the PDB database rarely excess the 30 kDa (or even 20 kDa) limit. Characterizations of biological pathways were rarely studied using NMR, except for reasonably small full length proteins.
The recent development of TROSY (transverse relaxation-optimized spectroscopy) , RDC (residual dipolar couplings), and new labeling strategies have presumably shifted this limit of protein structure determination up to 100 kDa. This is naturally a theoretical value, while there are practically no examples in PDB.
Polypeptide ligation is one of this recent labeling strategies with very interesting potential for the understanding of protein-protein interactions on a structural level in a full length protein. Such methods are based on reducing the number of signals,
Chapter 1: In-vivo intein segmental labeling 3and hence signal overlap in NMR spectra, which greatly simplifies the structural analysis of large proteins.
In vitro ligation for segmental isotopic labeling has been tested and established, but it has limited application since it requires laborious preparation and optimization of the ligation conditions.
In contrast, the in vivo approach has shown promising results and has been used successfully for different applications. The research group of Hideo Iwai has recently presented a novel in vivo ligation protocol with very promising results .
This method is based on sequential expression of the two precursor fragments, coded in two different expression systems. The ligation relies on the trans-protein splicing activity of split inteins. The Iwai group was able to apply this technique using the split intein DnaE from Synechocystis sp. strain PCC6803 (Ssp Dna E) and from Nostoc punctiforme (Npu DnaE) .
The work presented here has focused on application and optimization of this promising approach to achieve protein overexpression yields suitable for NMR.
We have adapted and combined several methods from cloning to expression protocols to develop a fast and simple strategy accessible to a large number of biochemistry groups.