Membrane proteins are involved in several fundamental biological processes such as transport or signal transduction. Most of them are enzymes, receptors or other important biological macromolecules representing up to 70% of therapeutic targets. Despite the interest in understanding their structures and behaviour the scientific knowledge is still very limited due to several practical difficulties. In 2009 a new platform for membrane protein studies called SMALP (Styrene-Maleic Acid Lipid Particles) nanodiscs was introduced. SMALPs are self-assembled structures formed by a bilayer of phospholipids controlled in diameter by a polystyrene maleic acid (SMA) copolymer belt. The purpose of this research project herein presented was to structurally characterise SMALPs, with analyses aimed to understand the role of both the polymeric and lipid parts in the self-assembly process. A series of investigations were carried out to elucidate the specific copolymer characteristics that allow the assembly into such well-defined, stable and reproducible structures. Experiments performed via small angle X-ray (SAXS) and neutron (SANS) scattering together with nuclear magnetic resonance (NMR), gel-filtration chromatography (GPC), dynamic light scattering (DLS), allowed identification of the specific polymeric characteristics of the copolymer architecture which were revealed to be crucial for the SMALPs assembly process. Investigations performed also addressed the question whether it was possible to assemble nanodiscs with the use of different phospholipids (with different chain length and charged or non-charged heads) and what the impact of the different lipids had on the structures. Finally, further analyses were made to test the physical chemical behaviour of the SMALPs when important environmental parameters such as temperature, pH and salt concentration of the buffer were changed.

Rhomboids are intramembrane serine proteases that cleave transmembrane (TM) protein substrates within the phospholipid bilayer. Since the discovery of the first rhomboid protease, many homologous rhomboids have been identified in all kingdoms illustrating their biological significance. Rhomboids are key players in a variety of biological processes such as, cell signalling, protein degradation, mitochondria health, apoptosis, and pathogenicity. While the mechanism of substrate entry into the rhomboid active site is still not clear, it is thought to involve dynamics around the putative substrate gate, of which appears to be comprised of the fifth transmembrane a-helix. A powerful tool that can be used to investigate conformational dynamics around the substrate gate is solution-state nuclear magnetic resonance (NMR). However, due to the size restriction of solution-state NMR, only detergent micelles have been able to produce well- resolved 1H-15N HSQC spectra of rhomboids. However, the lipid membrane environment has a significant impact on rhomboid structure and function. The use of membrane-scaffolding proteins (MSPs) in the formation of nanodiscs has the potential to allow the study of rhomboid dynamics in lipid bilayers by solution-state NMR. Therefore, this thesis investigates the plausibility of incorporating rhomboid into nanodiscs that would be compatible with solution NMR with a focus on the E. coli rhomboid, ecGlpG. The formation of empty (no ecGlpG) and ecGlpG-encapsulated nanodiscs was attempted using two MSP variants. While some successful nanodisc formation was possible, MSP degradation and low yields were seen for all nanodisc samples. Further optimization or alternate nanodisc systems will be required to incorporate ecGlpG into more membrane-like environments in a state that is compatible with solution-state NMR.

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第22464号 / 工博第4725号 / 新制||工||1738(附属図書館) / 京都大学大学院工学研究科分子工学専攻 / (主査)教授 梅田 眞郷, 教授 水落 憲和, 准教授 菅瀬 謙治 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DGAM

Membrane proteins are important targets that represent more than 50% of current drug targets. However, characterisation of membrane proteins falls behind compared to their soluble counterparts. The most challenging part of membrane protein research is finding a suitable membrane mimetic that stabilises them in solution and maintains their native structure and function. The recently developed saposin-A (SapA) based lipid nanoparticle system seems to be advantageous over existing membrane mimetic system. It provides a native-like lipid bilayer, high incorporation yield and more importantly size adaptability. SapA lipid nanoparticles have been applied to structural studies and two high-resolution structures of membrane proteins were previously obtained using cryo-electron microscopy. This thesis aimed to study small-to-medium sized membrane proteins in SapA lipid nanoparticles using NMR spectroscopy. We first explore the mechanism of SapA lipid nanoparticle formation for the purpose of establishing an incorporation protocol that can be applied to most membrane proteins. The effect of pH and the presence of detergents on the opening of SapA was investigated in Chapter 2. A proposed energy diagram describing the mechanism of SapA opening is reported with which we were able to develop a protocol that can generate different sizes of SapA-1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) nanoparticles. In addition, we also showed that SapA can form lipid nanoparticles with various lipid compositions, showing the versatility of the system. In Chapter 3, we validated the ability of SapA lipid nanoparticles to be used as a membrane mimetic. A -barrel model protein, bacterial outer membrane protein X (OmpX), was incorporated into SapA-DMPC nanoparticles and a 2D 15N-1H correlation NMR spectrum was recorded. Our result was compared to the NMR parameters of the same protein in MSP nanodiscs from the literature, and it was concluded that SapA lipid nanoparticles indeed provide a lipid bilayer environment similar to MSP nanodiscs. Because of high incorporation yield, we were able to incorporate OmpX into different lipid compositions to investigate the effect of lipid head groups and aliphatic chains on the membrane protein's chemical environment. Next, the applicability of SapA lipid nanoparticles was expanded to -helical transmembrane proteins in Chapter 4. Two microbial rhodopsins, Anabaena sensory rhodopsin (ASR) and Natronomonas pharaonis sensory rhodopsin II (pSRII), were tested. The parameters for expression and purification of ASR were first screened for the optimal yield. Although incorporation of ASR resulted in inhomogeneous particles due to imperfect experimental procedure, pSRII in SapA-DMPC nanoparticles showed high sample quality. The 2D NMR spectrum of pSRII in SapA-DMPC nanoparticles shows distinct differences to pSRII in detergent micelles, suggesting substantial effects from the membrane mimetic on the conformation of the membrane protein. Despite the good NMR spectral quality considering the large particle size, perdeuteration of pSRII and the lipids will be necessary for further investigation. With the SapA lipid nanoparticles established, we aimed to use it for the study of a biologically important G protein-coupled receptor, 1-adrenergic receptor (1AR), discussed in Chapter 5. The possibility of expressing 1AR using a cell-free expression system was explored first. Although a good amount of the protein was obtained, only a fraction of it was functional. Therefore, a conventional baculovirus-insect cell expression system was used to produce selective isotope labelled 1AR for NMR studies. NMR spectra of 1AR in SapA-DMPC nanoparticles with activating ligands and an intracellular binding partner were recorded and compared to the spectra of the same protein in detergents. This revealed a more active-like conformation of ligand-bound 1AR in the lipid bilayer, suggesting that certain parts of the protein are sensitive to the membrane mimetic used. This emphasises the importance of using a native-like membrane mimetic to capture the full properties of membrane proteins. In conclusion, I demonstrate in this thesis that SapA lipid nanoparticles are a versatile membrane mimetic system that can accommodate membrane proteins with different sizes and folds. This system is also compatible with solution NMR spectroscopy enabling structure and dynamics studies of biologically important membrane proteins. We believe SapA lipid nanoparticles will have a significant impact on membrane protein research in the future.

Membrane proteins are vital for cellular homeostasis. They maintain the electrochemical gradients that are essential for signaling and control the fine balance of trace elements. In order to fulfill these tasks, they need to undergo controlled conformational transitions within the lipid bilayer of a cell membrane. It is well-recognized that membrane protein structure and function depends on the lipid membrane. However, much less is known about the role of water re-partitioning at the protein–lipid interface and particularly within a membrane protein during functional transitions. Intra-membrane protein hydration is expected to be particularly important for ion transport processes, where the hydration shell of a solvated ion needs to be rearranged and partially removed in order to bind the ion within the transporter before it is re-solvated upon exiting the membrane protein. These processes are spatially and temporally organized in metal-transporting ATPases of the PIB-subtype of P-type ATPases. Here, the functional role of water entry into the transmembrane region of the copper-transporting PIB-type ATPase CopA from Legionella pneumophila (LpCopA) has been investigated. The recombinant protein was affinity-purified and functionally reconstituted into nanodiscs prepared with the extended scaffolding protein MSP1E3D1. Nanodiscs provide a planar native-like lipid bilayer in a water-soluble nanoparticle with advantageous optical properties for spectroscopy. The small polarity-sensitive fluorophore 6-bromoacetyl-2-dimethylaminonaphthalene (BADAN) was used as a probe for the molecular environment of the conserved copper-binding cysteine-proline-cysteine (CPC) motif which is located close to a wide “entry platform” for Cu+ to the transmembrane (TM) channel. The systematic study of proteins with mutated metal-binding motifs using steady-state and time-resolved fluorescence spectroscopy indicates that strong gradients of hydration and protein flexibility can exist across the narrow range of the CPC motif. The data suggest that Cu+ passes a “hydrophobic gate” at the more cytoplasmic C384 provided by rather stable TM helix packing before entering a more flexible and readily hydratable site in the interior of LpCopA around C382 where the polarity is strongly regulated by protein–lipid interactions. This flexibility could also be partly mediated by rearrangements of an adjacent amphipathic protein stretch that runs parallel to the membrane surface as a part of the cytoplasmic entry site. Using tryptophan fluorescence, circular dichroism, and Fourier-transform infrared absorption spectroscopy of a synthetic peptide derived from this segment, its lipid-dependent structural variability could be revealed. Depending on lipid-mediated helix packing interactions, the CPC motif has the potential to support a strong dielectric gradient with about ten units difference in permittivity across the CPC distance. This property may be crucial in establishing the directionality of ion transport by a non-symmetric re-solvation potential in the ion release channel of LpCopA. The experimental elucidation of these molecular details emphasizes not only the importance of intra-membrane protein water which has been hypothesized particularly for PIB-type ATPases. Moreover it is shown here, that the lateral pressure of a cell membrane may provide a force that restores a low hydration state from a transiently formed state of high internal water content at the distal side of the CPC motif. ATP-driven conformational changes that induce intra-membrane protein hydration of a conformational intermediate of the Post-Albers cycle could thus be set back efficiently by lateral pressure of the cell membrane at a later step of the cycle.

Membrane proteins are vital for cellular homeostasis. They maintain the electrochemical gradients that are essential for signaling and control the fine balance of trace elements. In order to fulfill these tasks, they need to undergo controlled conformational transitions within the lipid bilayer of a cell membrane. It is well-recognized that membrane protein structure and function depends on the lipid membrane. However, much less is known about the role of water re-partitioning at the protein–lipid interface and particularly within a membrane protein during functional transitions. Intra-membrane protein hydration is expected to be particularly important for ion transport processes, where the hydration shell of a solvated ion needs to be rearranged and partially removed in order to bind the ion within the transporter before it is re-solvated upon exiting the membrane protein. These processes are spatially and temporally organized in metal-transporting ATPases of the PIB-subtype of P-type ATPases. Here, the functional role of water entry into the transmembrane region of the copper-transporting PIB-type ATPase CopA from Legionella pneumophila (LpCopA) has been investigated. The recombinant protein was affinity-purified and functionally reconstituted into nanodiscs prepared with the extended scaffolding protein MSP1E3D1. Nanodiscs provide a planar native-like lipid bilayer in a water-soluble nanoparticle with advantageous optical properties for spectroscopy. The small polarity-sensitive fluorophore 6-bromoacetyl-2-dimethylaminonaphthalene (BADAN) was used as a probe for the molecular environment of the conserved copper-binding cysteine-proline-cysteine (CPC) motif which is located close to a wide “entry platform” for Cu+ to the transmembrane (TM) channel. The systematic study of proteins with mutated metal-binding motifs using steady-state and time-resolved fluorescence spectroscopy indicates that strong gradients of hydration and protein flexibility can exist across the narrow range of the CPC motif. The data suggest that Cu+ passes a “hydrophobic gate” at the more cytoplasmic C384 provided by rather stable TM helix packing before entering a more flexible and readily hydratable site in the interior of LpCopA around C382 where the polarity is strongly regulated by protein–lipid interactions. This flexibility could also be partly mediated by rearrangements of an adjacent amphipathic protein stretch that runs parallel to the membrane surface as a part of the cytoplasmic entry site. Using tryptophan fluorescence, circular dichroism, and Fourier-transform infrared absorption spectroscopy of a synthetic peptide derived from this segment, its lipid-dependent structural variability could be revealed. Depending on lipid-mediated helix packing interactions, the CPC motif has the potential to support a strong dielectric gradient with about ten units difference in permittivity across the CPC distance. This property may be crucial in establishing the directionality of ion transport by a non-symmetric re-solvation potential in the ion release channel of LpCopA. The experimental elucidation of these molecular details emphasizes not only the importance of intra-membrane protein water which has been hypothesized particularly for PIB-type ATPases. Moreover it is shown here, that the lateral pressure of a cell membrane may provide a force that restores a low hydration state from a transiently formed state of high internal water content at the distal side of the CPC motif. ATP-driven conformational changes that induce intra-membrane protein hydration of a conformational intermediate of the Post-Albers cycle could thus be set back efficiently by lateral pressure of the cell membrane at a later step of the cycle.

Les protéines membranaires représentent environ 2/3 des cibles thérapeutiques. Le développement de nouveaux médicaments est toutefois limité par l'absence de données structurales pour de nombreuses protéines. Les protéines membranaires s'avèrent en effet difficiles à manipuler et à maintenir en solution ce qui complique leur étude structurale. Les protéines sont en général solubilisées grâce à des surfactants comme les détergents, les amphipols, les hémifluorés et les peptergents. Il est aussi possible de les étudier dans des conditions plus physiologiques en les insérant dans des membranes lipidiques telles que des liposomes, des bicelles, ou des nanodisques.Les nanodisques sont des particules protéolipidiques autoassemblées, composées de protéines d'assemblages et de lipides, qui constituent un système de membranes modèles très prometteur permettant de solubiliser des protéines membranaires dans un milieu dépourvu de détergent. D'autres avantages sont aussi la variabilité de la constitution en lipides et l'accessibilité des deux côtés de la membrane.Dans le cadre de ma thèse, j'ai mis au point l'insertion de plusieurs protéines membranaires en nanodisques afin de permettre leur caractérisation fonctionnelle, biophysique et structurale. Nous avons en particulier étudié le transporteur ABC BmrA impliqué dans la résistance aux antibiotiques et cherché à identifier les changements conformationnels de la protéine en nanodisques par microscopie électronique. Les interactions de la protéine YedZ, un homologue de NADPH oxydases, avec ses partenaires solubles potentiels ont été étudiés par différentes méthodes telles que le pontage chimique, la résonance plasmonique de surface et la spectrométrie de masse native. En parallèle, le mécanisme d'assemblage des nanodisques a été investigué. Une interaction entre les protéines d'assemblages et des cations divalents a été mise en évidence. Cette interaction a un effet sur l'oligomérisation de la protéine d'assemblage mais également sur l'homogénéité des nanodisques. Ces observations nous ont permis d'améliorer les conditions de préparation des nanodisques, condition déterminante pour le succès de nombreuses approches structurales. Nous avons pu en particulier explorer la possibilité de cristalliser ces particules en vue d'études cristallographiques. / Membrane proteins represent around 2/3 of therapeutic targets. However, the development of new drugs is hampered by the lack of structural data for many proteins. Membrane proteins are indeed difficult to handle and to maintain stable in solution, which complicates their study by structural methods. Proteins are usually stabilized by surfactants like detergents, amphipols, hemifluorinated compounds and peptergents. It is also possible to study those proteins in an environment mimicking their native conditions by incorporating them in lipid membranes such as liposomes, bicelles or nanodiscs.Nanodiscs are self-assembled proteolipidic particles, composed of a scaffold protein and lipids. This technology is a top-notch model membrane system, which provides a detergent free environment to study membrane proteins in solution. Further advantages are the possibility to vary the lipid composition and the accessibility of the incorporated protein from both sides of the membrane.During my PhD project, I have achieved the insertion of several membrane proteins into nanodiscs for functional, biophysical and structural characterizations. In particular, we have studied Bmra, an ABC transporter involved in multidrug resistance and tried to identify the conformational changes of the protein in nanodiscs by electron microscopy. The interaction of YedZ, a NADPH oxidase homologue, with potential soluble partners has been studied by various methods such as cross-linking, surface plasmon resonance and native mass spectrometry. In parallel, the mechanism of nanodiscs assembly has been investigated. An interaction between the scaffold protein and divalent cations has been revealed. This interaction influences the oligomerization of the scaffold protein but also the nanodiscs homogeneity. Those observations allowed us to improve the preparation of the nanodiscs, which was an essential step torward the success of many structural approaches. In particular, we were able to explore their accessibility to protein crystallography.

ECF (Energy-Coupling Factor)-Transporter sind prokaryotische Aufnahmesysteme für Mikronährstoffe, die eine spezielle Gruppe von Transportern mit ATP-Bindekassette (ABC) darstellen. Sie beinhalten zwei asymmetrische Membranproteine, von denen eins (S) für die spezifische Bindung und Translokation des Substrates und das andere (T) für die Kopplung mit den ATPasen (A1,A2) zuständig ist. Bei ECF-Transportern der Subklasse I bilden diese Komponenten eine Einheit, während bei Vertretern der Subklasse II ein AAT-Modul mit wechselnden S-Einheiten interagiert. In der vorliegenden Arbeit wurde der Transportmechanismus, der eine Drehung der kompletten S-Einheit in der Membran beinhaltet, anhand des Biotintransporters BioMNY erstmals experimentell validiert. Durch Rekonstitution in Lipid-Nanodiscs, chemische Quervernetzung, fluoreszenz- und ESR-spektroskopische Techniken sowie einen Bindungstest mit radioaktivem Biotin wurde gezeigt, dass (i) die ATP-Bindung an die ATPasen zu einer Aufrichtung der S-Einheit (BioY) führt, (ii) diese Bewegung die Substratbeladung ermöglicht und (iii) BioY dabei ununterbrochen mit der T-Einheit (BioN) interagiert. Dies stellt einen Gegensatz zu Systemen der Subklasse II dar, für die ein ATP-abhängiger Austausch von S-Einheiten im Transportzyklus gezeigt worden war. Darüber hinaus wurde ein Escherichia coli-Stamm konstruiert, der durch Blockierung seines hochaffinen Biotintransporters und des -synthesewegs auf Spuren von Biotin nicht wachsen kann. Dieser Stamm ermöglichte einen eindeutigen Nachweis der Transportaktivität einiger solitärer BioY-Proteine. Aufgrund der einheitlichen Topologie von S-Einheiten ist ein Kippen auch für solitäre BioY-Varianten wahrscheinlich. Auch die metallspezifischen S-Einheiten CbiM und NikM besitzen ohne AAT-Modul eine basale Co2+- bzw. Ni2+-Transportaktivität. Ein ESR-spektroskopischer Kobaltnachweises zeigte, dass die aus nur zwei Membranhelices bestehende CbiN-Einheit für die Metallbeladung von CbiM essentiell ist. / ECF (Energy-Coupling Factor) transporters are a subgroup of ABC transporters that mediate uptake of micronutrients into prokaryotic cells. In contrast to canonical ABC importers, ECF transporters comprise two unrelated membrane proteins, one of which is responsible for specific and high affinity substrate binding (S) and the other one constitutes the coupling component (T) between S and the cytosolic ABC-ATPases (A1,A2). Subclass I transporters consist of four dedicated components whereas in subclass II transporters, a central AAT-module may interact with various S units. The biotin specific subclass I ECF transporter BioMNY was used to experimentally verify the hitherto hypothetic transport mechanism, which involves a rotation of the S unit within the membrane. With a series of experiments including reconstitution of BioMNY into lipid nanodiscs, site-specific cross-linking, a substrate binding assay with radioactive biotin and both fluorescence and EPR spectroscopic techniques, the ATP-dependent rotation of BioY (S) as a prerequisite for substrate binding and release was shown for the first time for an ECF transporter. Unlike subclass II transporters, for which an ATP-dependent release of the S unit was proposed, BioY interacts continuously with BioN (T) during the transport cycle. In a second focus of the work, an Escherichia coli reporter strain for biotin transporters was constructed. Due to inactivation of both biotin synthesis and the intrinsic high affinity biotin transporter, this strain was not capable of growing on trace amounts of biotin. With the use of this strain, transport activity of recombinantly produced solitary BioY proteins that naturally lack other ECF components was evidenced. Transport activity in the absence of AAT modules is also a feature of the Co2+ and Ni2+ specific S components CbiM and NikM. An EPR spectroscopic Co2+ detection assay helped underscoring the essential role of the small membrane protein CbiN for Co2+ loading of CbiM.

No description available.

Inwardly rectifying potassium (Kir) channels are essential for controlling the excitability of eukaryotic cells, forming a key part of the inter-cellular signalling system in multi-cellular organisms. However, as prokaryotic (KirBac) channels are less technically challenging to study in vitro and have been shown to be directly homologous to eukaryotic channels, they are often studied in lieu of their mammalian counterparts. A vital feature of Kir and KirBac channels is their mechanism for opening and closing, or their gating: this study predominantly features observations of open and/or closed channel populations. A well-characterised member of the KirBac family, KirBac1.1, has been successfully expressed, purified into detergent micelles, and doubly labelled with fluorescent maleimide dyes in order to enable observation of confocal-in-solution Förster Resonance Energy Transfer (FRET) at the single molecule level. Results demonstrate single-molecule FRET signals from KirBac1.1 and therefore represent the first single-molecule FRET observations from a KirBac channel. Perturbation of the open-closed dynamic equilibrium was performed via activatory point mutations, changes in pH, and ligand binding. A protocol for reconstitution into nanodiscs was optimised in order to more closely approximate native conditions, and the single-molecule FRET observations repeated. This thesis presents a comparison between measurements made using the detergent solubilisation system and those made using nanodiscs.