Refine
Document Type
- Doctoral Thesis (20)
Has Fulltext
- yes (20)
Is part of the Bibliography
- no (20)
Keywords
- rhodopsin (2)
- solid-state NMR (2)
- 2-Photonen (1)
- C. elegans (1)
- Chrimson (1)
- DNP (1)
- KR2 (1)
- Sodium ion pump (1)
- Spinach (1)
- adenylyl cyclases (1)
Institute
- Biochemie, Chemie und Pharmazie (20) (remove)
The focus of this thesis is the integral membrane protein Escherichia coli diacylglycerol kinase (DGK). It is located within the inner membrane, where it catalyzes the ATP-dependent phosphorylation of diacylglycerol (DAG) to phosphatic acid (PA). DGK is a unique enzyme, which does not share any sequence homology with typical kinases. In spite of its small size, it exhibits a notable complexity in structure and function. The aim of this thesis is the investigation of DGK’s structure and function at an atomic level directly within the native-like lipid bilayer using MAS NMR. This way, a deeper understanding of DGK’s catalytic mechanism should be obtained.
First, the preparation of DGK was optimized, leading to a sample, which provides well-resolved MAS NMR spectra. The high quality MAS NMR spectra formed the foundation for the second step, the resonance assignment of DGK’s backbone and side chains. The assignment was performed at high magnetic field (1H frequency 850 MHz). The sequential assignment of immobile domains was carried out using dipolar coupling based 3D experiments, NCACX, NCOCX and CONCA. The measurement time could be reduced by paramagnetic doping with Gd3+-DOTA in combination with an E-free probehead. The sequential assignment was mainly performed using a uniformly labelled sample (U-13C,15N-DGK). Residual ambiguities could be resolved by reverse labelling (U-13C,15N-DGK-I,L,V). Resonances could be assigned for 82% of the residues, from which 74% were completely assigned. For validation, ssFLYA was applied, which is a generally applicable algorithm for the automatic assignment of protein solid state NMR spectra. Its principal applicability for demanding systems as membrane proteins could be proven for the first time. Overall, ~90% of the manually obtained assignments could be confirmed by ssFLYA. For the completion of DGK’s assignment, J-coupling based 2D experiments, 1H-13C/15N HETCOR and 13C-13C TOBSY, were carried out to detect highly mobile residues. This way, residues of the two termini and the cytosolic loop, which were not detectable by dipolar coupling based experiments, could be assigned tentatively. Whereupon, peaks for arginine and lysine were assigned unambiguously to Arg9 and Lys12. Overall, ~84% of the residues could be assigned by the applied NMR strategy. Furthermore, a secondary structure analysis was carried out. It showed substantial similarities between wild-type DGK, its thermostable mutant determined both by MAS NMR and the crystal structure of wtDGK. However, there are few differences around the flexible regions most likely caused by the high mobility of these regions. During the assignment procedure, no systematic peak doublets or triplets were detected, indicating that the DGK trimer adopts a symmetric conformation. This is in contrast to the X-ray structure, which shows asymmetries between the three subunits. Especially, crystal packing may be a potential source for these structural asymmetries.
On the basis of the nearly complete assignment of DGK, the apo state was compared with the substrate bound states. Perturbations in peak position and intensity of the substrate bound states were analysed for all assigned residues in 3D and 2D spectra. The nucleotide-bound state was emulated by adenylylmethylenediphosphonate (AMP-PCP), a non-hydrolysable ATP analogue, whereas the DAG-bound state was mimicked by 1,2-dioctanoyl-sn-glycerol (DOG, chain length n = 8). Upon nucleotide binding, extensive chemical shift perturbations could be observed. These data provide evidence for a symmetric DGK trimer with all of its three active sites concurrently occupied. Additionally, it could be demonstrated that the nucleotide substrate induces a substantial conformational change. This most likely supports the enzyme in binding of the lipid substrate, indicating positive heteroallostery. In contrast, the overall alterations caused by DOG are very minor. They involve mainly changes in peak intensities. For DGK bound with either AMP-PCP+DOG or only AMP-PCP, a similar spectral fingerprint was observed. This implies that binding of the nucleotide seems to set the enzyme into a catalytic active state, triggering the actual phosphoryl transfer reaction.
The investigation of DGK’s remarkable stability and the cross-talk between its subunits forms the last part of this thesis. This demands for the identification of key intra- and interprotomer contacts, which are of structural or functional importance. For this purpose, 13C-13C DARR and 2D NCOCX spectra with long mixing times were recorded using high field MAS NMR. Additionally, DNP-enhanced 13C−15N TEDOR experiments were conducted on mixed labelled DGK trimers to enable the visualization of interprotomer contacts. With the applied NMR strategy, intra- (Arg32 - Trp25/ Glu28/ Ala29 and Trp112 - Ser61) and interprotomer (ArgNn,e - AspCg/ GluCd/ AsnCg) long-range interactions could be identified.
In the past decade, the optogenetic toolbox for the manipulation of ion currents and cNMP levels in Caenorhabditis elegans (C. elegans) expanded. However, the implemented tools for cAMP generation were soluble enzymes (euPAC, bPAC, IlaC22 k27 and PaaC) and thus they do not precisely mimic physiological cAMP signalling occurring in microdomains in close proximity to the plasma membrane. Here, cAMP is predominantly generated by membrane-bound adenylyl cyclases, that are located in microdomains together with G protein-coupled receptors (GPCRs), protein kinase A (PKA) and their targets, enabling spatially and temporal regulation of cAMP signalling. For this reason, one aim of this study was to develop and implement membrane bound photoactivatable adenylyl cyclases for the manipulation of cAMP mediated signalling in close proximity to the plasma membrane. For this purpose, the guanylyl cyclase domains of the Blastocladiella and Catenaria Cyclase Opsins (CyclOps) were mutated to adenylyl cyclases either by introducing the mutations E497K and C566D (abbreviated as (A-2x)) or by the mutations E497K, H564D, and C566T (abbreviated as (A-3x)).
To determine the nucleotide specificity switch from GTP to ATP and the extent of light-dependent cAMP generation, the engineered enzymes were expressed in body wall muscle cells of C. elegans and in vitro cNMP measurements using C. elegans extracts were performed. Here, the highest levels of light induced cAMP generation during sustained stimulation (0.5 mW/mm2; 470 nm, 15 min) were detected for the variants BeCyclOp(A-2x), YFP-BeCyclOp(A-2x), and YFP-CaCyclOp(A-2x) (39, 57, 40 nM, respectively), though they did not reach the extent produced by the soluble bPAC (142 nM). In contrast, low magnitudes of generated cAMP were measured for the versions BeCyclOp(A-3x) and CaCyclOp(A-2x) (8 and 7 nM, respectively). Importantly, no obvious residual cGMP and basal activity was ascertained for any of the engineered enzymes.
To assess their potential to trigger and modulate cAMP mediated cholinergic neurotransmission, and to evaluate the influence of cytosolic and membrane proximal optogenetic cAMP generation, the enzymes were expressed in cholinergic motor neurons and compared to the implemented soluble bPAC via locomotion behaviour analysis on solid and in liquid media. Photoactivation of BeCyclOp(A-2x), YFP-BeCyclOp(A-2x), and YFP-CaCyclOp(A-2x) caused similarly enhanced or even more potent behavioural changes (swimming and crawling) as bPAC, whereas a more rapidly decaying response was observed for the bPAC evoked effects. Moreover, an increased diversity of the behavioural output was detected for cytosolic cAMP production by bPAC, i.e. increased bending angles and a decreased body length.
Confocal fluorescence microscopy was performed to examine the expression levels of YFP-tagged enzymes in cholinergic neurons, whereas both YFP-CyclOp(A-2x)s were expressed at similar levels, but 1.4-fold lower relative to the soluble bPAC-YFP. To compare the amount of light-dependent cAMP generation bPAC and BeCyclOp(A-2x) at light conditions that match the conditions of the behavioural experiments (30 s), cAMP measurements using C. elegans extracts were performed, whereas BeCyclOp(A-2x) depicted a 4-fold lower amount of optogenetic cAMP production than the soluble bPAC.
In sum, local (membrane proximal) cAMP generation by the membrane-bound photoactivatable adenylyl cyclases may more specifically activate cAMP dependent neurotransmission of cholinergic motor neurons than cytosolic cAMP generation, i.e. an increased mobilization and priming/docking of synaptic vesicles and an increased filling of the synaptic vesicles with the neurotransmitter acetylcholine and thus an increase in locomotion behaviour.
The optogenetic toolbox for the manipulation of cGMP mediated signalling in C. elegans consisted of the natural membrane-bound BeCyclOp and the artificial soluble bPGC. The latter generates cGMP with low efficiency and slow kinetics (~0.2 cGMP s-1), whereas BeCyclOp enables the production of much larger amounts of cGMP (L/D = 5000) at a high turnover rate (~17 cGMP s-1). Thus, one aim of this thesis was to implement a tool with features in between those of BeCyclOp and bPGC. Several orthologous CyclOps were assessed by Gao et al., 2015 for light-regulated cGMP production by in vitro assays based on the measurement of the cNMP content from CyclOp containing oocyte membranes. Here, CaCyclOp showed the highest ratio of light versus dark activity (L/D = 230) after BeCyclOp, and thus was selected for characterization in C. elegans...
Verschiedene physikalische Effekte erlauben es Licht so zu führen und zu verändern, dass es Einblicke in für Menschen sonst unzugängliche Bereiche gewährt. Eines von insgesamt drei Elementen dieser Dissertationsschrift ist der Aufbau eines Multiphotonen-Mikroskops. Dieses fortschrittliche Werkzeug erweitert das zur Verfügung stehende Instrumentarium um verschiedene Analysemethoden, allen voran die 2-Photonen-Fluoreszenz-Mikroskopie. Durch geringfügige Modifikationen können auch weitere Methoden, wie beispielsweise stimulierte Raman-Streuung realisiert werden.
Insbesondere die 2-Photonen-Fluoreszenz-Mikroskopie war für das zweite Element dieser Dissertationsschrift von großer Bedeutung. In dieser Studie wurde das Bleichverhalten von Spinach bei 2-Photonen-Absorption untersucht, sowohl an frei in Lösung befindlichen als auch auf einem Träger immobilisierten Spinach-Komplexen. Die Ergebnisse zu den frei in Lösung befindlichen Spinach-Komplexen zeigen, dass die Verstärkung der Fluoreszenz von DFHBI grundsätzlich auch im Fall der 2-Photonen-Absorption eintritt. Dabei wurde ein Ausbleichen der 2-Photonen-induzierten Fluoreszenz für frei in Lösung befindliche Spinach-Komplexe erst bei außerordentlich hohen Intensitäten der Anregungsstrahlung beobachtet. Dieser Befund kann zumindest teilweise auf das Eindiffundieren fluoreszenter Spinach-Komplexe in das sehr kleine Fokalvolumen innerhalb der 2-Photonen-Anregung stattfindet zurückgeführt werden. Für immobilisierte Spinach-Komplexe konnte gezeigt werden, dass eine kontinuierliche Bildaufnahme gegenüber einer Bildaufnahme in Intervallen mit jeweils zusätzlichen Dunkelphasen zur Erholung des reversiblen Bleichens der 2-Photonen-induzierten Fluoreszenz, sowie der generelle Verzicht auf spezielle Belichtungsschemata und Methoden der Datenakquise mit keinen besonderen Nachteilen verbunden ist. Abschließend betrachtet erweist sich Spinach bei 2-Photonen-Anregung als ausgesprochen resistent gegenüber einem irreversiblem Ausbleichen des Fluoreszenzsignals.
Als drittes Element dieser Dissertationsschrift wurde die Dynamik von Chrimson, einem Kanalrhodopsin mit rot-verschobener Absorption mittels zeitaufgelöster Spektroskopie im sichtbaren Spektralbereich untersucht. Sowohl die Anregungswellenlänge als auch der pH-Wert bzw. der Protonierungszustand des Gegenions haben einen messbaren Einfluss auf die Primärreaktion. Diese verlangsamt sich, sobald der pH-Wert abgesenkt oder die Anregungswellenlänge rot-verschoben wird. Darüber hinaus führt eine Rot-Verschiebung der Anregungswellenlänge zu einer geringeren Effizienz der Isomerisation des Retinal-Chromophors. Die Primärreaktion von Chrimson entspricht dabei einem Reaktionsmodell mit einer Verzweigung des Reaktionspfades auf der Energiehyperfläche des angeregten Zustandes. Ein Reaktionspfad führt dabei durch ein lokales Minimum, welches in seiner Ausprägung stark von der elektrostatischen Umgebung des Retinal-Chromophors abhängt. Je nach ursprünglichem Protonierungszustand des Gegenions der Retinal-Schiff-Base wurden große Unterschiede hinsichtlich der beobachteten transienten Absorptionsmuster für den im Anschluss von Chrimson durchlaufenen Photozyklus gefunden. Bei pH 6,0 weist der Photozyklus von Chrimson eine insgesamt deutlich schnellere Kinetik auf, als es für den Photozyklus bei pH 9,5 beobachtet wurde. Es ist bemerkenswert, dass in elektrophysiologischen Messungen für beide Photozyklen eine Öffnung des Ionenkanals gefunden wurde. Die Kanalfunktion von Chrimson ist somit grundsätzlich nicht vom Protonierungszustand des Gegenions abhängig, wenngleich die Kinetik des Ionenkanals durchaus davon beeinflusst wird. Dies deutet auf Unterschiede in den Wechselwirkungen zwischen dem Ionenkanal und dem Gegenion der Retinal-Schiff-Base hin.
In der vorliegenden Arbeit wurde die Dynamik zweier grundlegend verschiedener, deaktivierender Mechanismen von Retinalproteinen untersucht. In einem dritten Projekt wurde die Photodynamik einer Dreifachmutante von visuellem Rhodopsin erforscht, von der eine Mutation zu kongenitaler (angeborener) Nachtblindheit führt und zwei andere Mutationen das Protein über eine Disulfidbrücke stabilisieren. Die Ergebnisse dieser drei Projekte sind im Folgenden zusammengefasst.
Die Aktivität des mikrobiellen Proteorhodopsins als lichtgetriebene Protonenpumpe kann photoinduziert unterbunden werden. Dies erfolgt durch die Absorption von blauem Licht durch das Retinal bei deprotonierter Schiff‘schen Base. Vor dieser Arbeit war allerdings nur wenig über den Mechanismus und die Kinetik dieses Effekts bekannt. Das einzige Retinalprotein, an dem diese Deaktivierungsdynamik auf molekularer Ebene zeitaufgelöst untersucht wurde, ist Bakteriorhodopsin. Doch auch an diesem System wurde die ultraschnelle Primärreaktion in der photoinduzierten Deaktivierungsdynamik - die Photoisomerisierung des 13-cis-Retinals - bisher nicht zeitaufgelöst gemessen.
In dieser Arbeit wurde ein Weg gefunden, diesen Prozess auf einer Sub-Pikosekundenzeitskala zu detektieren. Dazu wurde eine Proteorhodopsinmutante genutzt, in der der primäre Protonendonor E108 durch Glutamin ersetzt ist. Diese Mutante weist eine signifikante Erhöhung der Lebensdauer des M-Intermediats auf. Im photostationären Gleichgewicht führt diese veränderte Kinetik zu einer erheblich erhöhten Akkumulation des Proteins im M-Zustand, die ausreicht, um photoinduzierte Absorptionsänderungen der Deaktivierungsdynamik sowohl im sichtbaren als auch im mittleren Infrarotbereich auf ultrakurzer Zeitskala zu detektieren. Dieses Projekt erfolgte in Kooperation mit dem Arbeitskreis Glaubitz (Goethe-Universität Frankfurt am Main).
Es zeigte sich, dass die Anregung des Retinals von Proteorhodopsin im M-Zustand zur Isomerisierung von 13-cis zu all-trans führt, die nach wenigen Pikosekunden abgeschlossen ist. Der zweite und abschließende Schritt ist die Reprotonierung der Schiff'schen Base. Es stellte sich heraus, dass dieser Prozess auf einer Nanosekundenzeitskala abläuft und über einen Protonentransfer vom primären Protonenakzeptor D97 zur Schiff'schen Base ermöglicht ist.
Die in dieser Arbeit vorgestellte Methodik zur Untersuchung der deaktivierenden Photodynamik von Proteorhodopsin auf ultraschneller Zeitskala, könnte in Zukunft auf weitere mikrobielle Rhodopsine angewandt werden. So ist die Studie der Deaktivierungsdynamik von Channelrhodopsinen von großem Interesse für optogenetische Anwendungen. Eine lichtgesteuerte Kontrolle der Ionenkanalöffnung und -schließung sollte die Präzision in der Regulierung ionischer Permeation erheblich verbessern.
Die Proteorhodopsinmutante E108Q wurde außerdem in ihrer primären Photodynamik sowohl bei grünem als auch blauem Anregungslicht untersucht. Es zeigte sich in beiden Fällen eine Dynamik, die der des Wildtyps sehr ähnlich ist. Eine Beobachtung unterscheidet sich jedoch wesentlich vom Wildtyp. Das K-Intermediat der E108Q-Mutante scheint nach einigen hundert Pikosekunden zumindest partiell zu zerfallen, woraufhin sich eine Signatur im blauen Spektralbereich bildet. Blitzlichtphotolysemessungen lassen vermuten, dass diese blau absorbierende Species im zwei- bis dreistelligen Nanosekundenbereich wieder zerfallen sein muss.
Der zweite Teil dieser Arbeit beschäftigt sich mit dem Photozerfall von visuellem Rhodopsin. Es ist bekannt, dass die Signaltransduktion durch Wechselwirkung zwischen aktiviertem Rhodopsin und Arrestin unterbunden wird. Im ersten Abschnitt wurde der Einfluss der Arrestin-1-Variante p44 auf die Photodynamik visuellen, bovinen Rhodopsins untersucht. In einer Kooperation mit dem Arbeitskreis Schwalbe (Goethe-Universität Frankfurt am Main) konnte gezeigt werden, dass Arrestin erheblichen Einfluss auf die Zerfallsdynamik von Meta II und Meta III hat. Es wurde festgestellt, dass die Wechselwirkung von p44 mit photoaktiviertem Rhodopsin eine erhöhte Population des Intermediats Meta III bewirkt, mit der Folge einer zweifach langsameren Freisetzungskinetik des all-trans-Retinals. Diese Beobachtung weist auf eine physiologische Rolle des Zustands Meta III in der Retinalhomöostase hin.
Gegenstand einer zweiten Studie mit dem Arbeitskreis Schwalbe ist zum einen die Rhodopsinmutation G90D, die mit kongenitaler (angeborener) stationärer Nachtblindheit zusammenhängt, und zum anderen die Doppelmutation N2C und D282C, die zur Ausbildung einer stabilisierenden Disulfidbrücke zwischen den im extrazellulären Bereich eingeführten Cysteinen führt. Im Rahmen dieser Arbeit wurde die Photodynamik des Wildtyps, der Doppelmutante und der stabilisierten G90D-Mutante (Mutationen G90D, N2C und D282C) sowohl auf einer ultrakurzen Zeitskala als auch auf einer Minutenskala untersucht.
The focus of this research was to understand the molecular mechanism that lies behind the insertion of tail-anchored membrane proteins into the ER membrane of yeast cells. State-of-art instruments such as LILBID, and Cryo-EM, combined with the introduction of direct electron detectors, were used to analyze the proteins that capture tail-anchored proteins near the ER membrane and help their releases from a chaperone, an ATPase named Get3. Get3 escorts TA proteins to the ER membrane, where both Get3 and the TA proteins interact sequentially to Get3 membrane bound receptors Get1 and Get2. Get1 and Get2 are homologs of mammalian WRB and CAML.
The native host was used to separately produce Get1, Get2, and the Get2/Get1 single chain constructs. The studies showed that when Get1 is expressed alone, Get1 does not seems to be located in the ER membrane but rather in microbodies like shape organelles (or peroxisome). Interestingly, Get1 seems to be located in the ER membrane when it is linked to Get2 as single chain construct.
The localization study of Get2/Get1 fused to GFP shows from the fluorescence intensity that Get2/Get1.GFP has a tube-like morphology or membrane-enclosed sacs (cisterna), implying that Get2/Get1 is actually targeted to the ER membrane and is likely functional. In other words, Get1 and Get2 stabilize each other in the ER membrane.
The expression of Get2/Get1 was found to be already optimum when expressed as single chain construct because the fluorescence counts did not improve when additives such as DMSO or histidine were added. However, when Get1 and Get2 are expressed separately, additives improve their protein production yield. In 1 liter culture, Get1 yield is increased by about 3 mg and Get2 by 1.8 mg. This can be explained by the space that Get1 and Get2 should occupy within the ER membrane as they must coexist with other membrane components to maintain the homeostasis of the cell. Hence, if there were no gain for single chain construct expression, it meant that Get2/Get1 was already well expressed on its own in ER membrane and has reached its optimum expression without the help of additives. The Get2/Get1 overexpression is more stable, tolerated and less toxic for the cells to express it at a high level.
DDM has proved to be the best detergent from the detergents tested to solubilize Get1, Get2, and Get2/Get1.
Thereafter, Get1, Get2 (data not shown), and Get2/Get1 were successfully purified in DDM micelles.
Furthermore, for the first time using LILBID, the actual study has shown that Get1 and Get2 are predominantly a heterotetramer (2xGet1 and 2xGet2) but higher oligomerization may exist as well.
Get3 binds to Get1 in a biphasic way with a specific strong binding of an affinity of 57 nM and the second of 740 nM nonspecific indicative of heterogeneity within the interaction between Get1 and Get3. This heterogeneity is caused by the presence of different conformation of either protein. However, in order to characterize a high-resolution structure model of a specific target one needs highly homogenous and identical molecules of the target protein or complex in solution. The homogeneity increases the chances of growing crystals during crystallography as the good homogeneity will likely generate a perfect packing of unit cells stack (also known as crystal lattice) in the three-dimensional spaces. The same truth goes for the single particles analysis Cryo-EM, especially for smaller complexes where having less or no conformation alterations of specific targets will enable the researcher to classify the particles in 2D and 3D, therefore improving the signal-to-noise-ratio that will ultimately lead to high-resolution structure determination.
Get1, Get2/Get1 and chimeric variants (tGet2/Get1, T4l.Get2/Get1, T4l.Get2.apocyte.Get1) were crystallized but none of the crystals could diffract due to heterogeneity.
This heterogeneity was not only occurring upon the binding of Get3 to its membrane receptors, but seems to be already present within the receptors themselves through possibly different conformation.
In this Ph.D. thesis, the heterogeneity of purified Get2 and Get1 as complex or individually in detergent is then, so far, the limiting factor for obtaining a high-resolution structure model of Get1 and Get2. As mentioned above, the heterogeneity observed was not due to the quality of the sample preparation but rather to the effect of different conformations that could have been native, or just because of the micelle used, as it was proven by the 3-D heterogeneity classification by Cryo-EM.
In general, crosslinking is one way to keep the integrity of protein complexes, however it appeared not to improve the sample quality when it was analyzed in micelles. Often the integrity of some membrane proteins is affected when they are solubilized and purified in detergents.
Finally, in this study, the structural map of Get2 and Get1 complex linked with chimeric protein T4 lysozyme and apocytochrome C b562RIL gene was obtained at 10 Å. However, this single chain construct has a density map corresponding to heterodimer species (one Get1 and Get2). Therefore, based on those data the tertiary structure of Get2/Get1 in micelle is poorly defined. It could be that the membrane extraction in DDM and the purification destabilizes the structure of the complex.
Gegenstand der vorliegenden Arbeit sind die Untersuchungen lichtgesteuerter Reaktionen der zwei Retinalproteine Channelrhodopsin-2 (ChR-2) und Proteorhodopsin (PR) mit Hilfe zeitaufgelöster Laserspektroskopie.
Da der Mechanismus der Kanalöffnung des ChR-2 bis heute nicht vollständig aufgeklärt werden konnte, beschäftigt sich diese Arbeit insbesondere mit den Prozessen, die direkt nach der Photoanregung des Retinals stattfinden und die Kanalöffnung vorbereiten. Es wurde dabei gezielt auf für die Funktion des Proteins wichtige Faktoren wie strukturelle Besonderheiten des Chromophors und seiner Umgebung eingegangen und deren Auswirkung auf die Dynamik der Photoreaktionen sowie die Veränderungen im Protein nach der Anregung untersucht.
Zunächst wurden die Ergebnisse der vis-pump-IR-probe-Experimente an ChR-2 im Bereich der Carbonylschwingungsbanden protonierter Glutamat- und Aspartat-Reste dargestellt. Dabei wurde insbesondere die Bildungsdynamik der Differenzbanden in diesem Spektralbereich untersucht und in Anlehnung an die vorhandene Literatur eine Bandenzuordnung der für die Funktion des Proteins wichtigen Aminosäurereste vorgenommen. Aus den Messergebnissen konnte geschlossen werden, dass die mit der Kanalöffnung einhergehenden Konformationsänderungen in ChR-2 durch eine effektive Aufnahme der Überschussenergie durch das Protein auf einer sub-Pikosekunden-Zeitskala vorbereitet werden.
Des Weiteren wurden spektroskopische Untersuchungen an der R120H-Mutante des ChR-2 vorgestellt. Da diese Mutante bei elektrophysiologischen Messungen keine Kanalaktivität zeigte, sollte zunächst geklärt werden, ob die Mutation einen Einfluss auf die Retinalisomerisierung und den nachfolgenden Photozyklus hat. Dabei stellte sich heraus, dass die Retinalisomerisierung bei der R120H-Mutante zwar im Vergleich zum Wildtyp etwas verzögert stattfindet, der Einfluss der Punktmutation auf den weiteren Photozyklus jedoch insgesamt gering ist. Mit Hilfe der Kurzzeit-IR-Spektroskopie im Bereich der Amid I-Schwingung des Proteinrückgrats konnten für die Mutante allerdings signifikante Veränderungen der Bildungsdynamik sowie eine deutliche Abnahme der Amplitude des Amid I-Signals detektiert werden. Anhand weiterer Experimente an den Mutanten E123T und D253N in diesem Spektralbereich konnte anschließend ein Zusammenhang zwischen der Intensität der Amid I-Bande und der Kanalaktivität von ChR-2 festgestellt werden. Diese Ergebnisse ließen somit die Schlussfolgerung zu, dass die Aminosäurereste R120 und D253 eine entscheidende Rolle beim schnellen Transfer der Überschussenergie an das Protein nach der Retinalanregung und der so initiierten Kanalöffnung spielen.
Zusätzlich wurde der Frage nachgegangen, inwieweit Veränderungen am Chromophor die Isomerisierungsreaktion, den nachfolgenden Photozyklus sowie die Funktion des ChR-2 als Ionenkanal beeinflussen können. Zu diesem Zweck wurden spektroskopische Untersuchungen an einem mit 9-12-Phenylretinal (PheRet) rekonstituierten ChR-2 vorgestellt. Es konnte gezeigt werden, dass die Isomerisierung des PheRet zu seiner 13-cis-Form in ChR-2 stark verlangsamt ist und verglichen mit dem nicht modifizierten Chromophor deutlich ineffizienter abläuft. Es wurde außerdem festgestellt, dass die Veränderungen am Retinal zu deutlichen Beeinträchtigungen des Photozyklus führen. Zum einen wurde ein sehr schneller Zerfall des ersten Photoprodukts sowie die Bildung eines zusätzlichen, blauverschobenen Px-Zustands detektiert. Außerdem wurde festgestellt, dass nach der Deprotonierung des isomerisierten PheRet der Großteil der modifizierten Retinale in den Ausgangszustand zurückkehrt und der P3-Zustand nur in geringen Mengen gebildet wird. Die Messergebnisse führten somit zu der Schlussfolgerung, dass die all-trans-Konformation des PheRet in ChR-2 deutlich bevorzugt wird. Da elektrophysiologische Untersuchungen des Retinal-Analogons jodach keine signifikanten Verminderungen der Photoströme im Vergleich zum ATR in ChR-2 zeigten, ließ sich schließlich festhalten, dass die vorgenommenen Veränderungen am Chromophor, die zu einer deutlichen Hemmung der Isomerisierungsreaktion führen und einen starken Einfluss auf den nachfolgenden Photozyklus haben, nicht ausreichend sind, um die Kanalaktivität von ChR-2 komplett zu blockieren, solange noch ein kleiner Anteil der Retinale isomerisieren kann.
Der abschließende Teil der Arbeit beschäftigt sich mit der Absorption des UV-Lichts durch das Retinal mit deprotonierter Schiff-Base im grünabsorbierenden Proteorhodopsin, welches in einem alkalischen Medium im Dunkelzustand akkumuliert werden kann. Die Untersuchungen der Primärreaktion zeigten einen langsamen biexponentiellen Zerfall des angeregten Zustands der UV-absorbierenden Spezies mit anschließender Bildung des 13-cis-Photoprodukts. Aufgrund dieser Ergebnisse konnte ein Reaktionsmodell für die ersten Prozesse nach der UV-Anregung des Retinals im GPR aufgestellt werden, welches möglicherweise für weitere UV-Rezeptoren genutzt werden kann.
Probing the photointermediates of light-driven sodium ion pump KR2 by DNP-enhanced solid-state NMR
(2021)
KR2 is a light-driven sodium ion pump found in marine flavobacterium Krokinobacter Eikastus. The protein belongs to the microbial rhodopsin family, which is characterized by seven transmembrane helices and a retinal cofactor covalently bound to a conserved lysine residue through a Schiff base linkage. Specific features of KR2 and other sodium pumping rhodopsins are the NDQ motif, the N-terminal helix capping the protein at the extracellular side, and the sodium ion bound at the protomer interface in the pentameric structure. The ability to pump sodium ions was a surprising discovery since the positive charge at the Schiff base was long thought to hinder the transport of non-proton cations and the Grotthuss mechanism could not be applied to explain the Na+ transport. The photocycle of KR2 revealed by flashed photolysis and ultrafast femtosecond absorption spectroscopy consists of consecutive intermediates, named K, L, M, and O.
Here, DNP-enhanced ssNMR was used to analyze various aspects of these intermediate states. The K/L-state can be generated and trapped by in-situ illumination inside the magnet at 110 K. The trapping of L-state together with the K-state at this temperature is unexpected as this usually leads to the trapping of only K-state in bacteriorhodopsin (BR), proteorhodopsin (PR), and channelrhodopsin 2 (ChR2). This observation suggests a lower energy barrier between K- and L-state in KR2. For the O-state, the intermediate was generated by illuminating outside the magnet, followed by rapid freezing in liquid nitrogen and transfer to the magnet. Based on these procedures, the retinal conformation, and the electrostatic environment at the Schiff base in KR2 dark, K-, L- and O-intermediates were probed using 13C-labeled retinals bound to 15N-labeled KR2 by both 1D and 2D magic angle spinning (MAS) NMR experiments.
The obtained data show an all-trans retinal conformation with the distortion of 150° at H-C14-C15-H in the dark state whereas the retinal has a 13-cis, 15-anti conformation in the K- and L-state after light activation. Differences between K- and L-intermediates were observed. The retinal chemical shifts of the K-state show a large deviation from the model compound behavior between the middle and end part of the polyene chain. In the L-state, these differences are much less pronounced. These observations indicate that the light energy stored in the K-state dissipates into the protein in the subsequent photointermediate states. Furthermore, an additional shielding observed for C14 in L-state indicates the slight rotation toward a more compact 13-cis, 15-syn conformation. The distortion of the H-C14-C15-H angle in the L-state (136°) is larger than in the dark state. This twist of the retinal in the L-state would play an important role in lowering the pKa of the Schiff base, which is a prerequisite for the proton transfer from the Schiff base to the proton acceptor (D116). The electrostatic environments at the Schiff base in K- and L-states cause a de-shielding of the 15N nitrogen compared to the dark state. This indicates a stepwise stronger interaction with the counterion as the Schiff base proton moves away from the Schiff base and comes closer to the D116 in the transition from K- to L-state and approaches the proton transfer step during the M-state formation. In the O-state, the retinal was found to be in the all-trans conformation but differed to the dark state in the C13, C20, and Schiff base nitrogen chemical shifts. The largest effect (9 ppm) was observed for the Schiff base nitrogen, which could be explained by the effect of the positive charge of bound Na+ near the Schiff base in the O-state, coordinated by N112 and D116 as observed in the O-state crystal structure in the pentameric form.
The structural change at the opsin followed the retinal isomerization and the energy transfer from the chromophore to the surrounding were also investigated in this thesis using various amino acids labeling schemes. Moreover, 1H-13C hNOE in combination with CE-DNP was applied to probe the dynamics of retinylidene methyl groups and 23Na MAS NMR was employed to detect the bound sodium ion at the protomer interface in KR2 dark state.
Resistant microbes are a growing concern. It was estimated that about 33,000 of people die because of the infections caused by multidrug resistant bacteria each year in Europe (ECDC, 2018, https://www.ecdc.europa.eu/). Bacteria can acquire resistance against toxic compounds via different mechanisms and intrinsic active efflux is one of the first mechanisms deployed by bacterial cells. The membrane-localized efflux pumps catalysing this reaction, extract toxic compounds from the interior of the cell and transport these to the outside, thereby maintaining sub-lethal toxin levels in the cytoplasm, periplasm and membranes. Gram-negative three-component efflux pumps, analysed in this study, are composed of an inner membrane protein, a member of the Resistance-Nodulation cell Division (RND) superfamily, an Outer Membrane Factor (OMF) protein and a Membrane Fusion Protein (MFP) that connects the two afore mentioned components into an active efflux pump. The pumps described in this work, AcrAB-TolC and EmrAB-TolC, are drug efflux pumps belonging to the RND and MFS superfamilies, respectively, while CusCBA is an efflux pump that belongs to the RND heavy metal efflux family. Another efflux pump that was used as a model for the design of an in vitro assay for the silver ion transport studies, CopA, belongs to the P-type ATPase superfamily. All pumps analysed in this study are part of the resistance system of Escherichia coli, which is a highly clinically relevant pathogen.
In order to examine the AcrAB-TolC, CopA and CusA efflux pumps, the individual components were separately produced in E. coli, purified to monodispersity and reconstituted in large unilamellar vesicles, LUVs. Means for the optimized production and adequate conditions for efficient reconstitution were presented in this study. The activity of AcrB in LUVs was detected using fluorescence quenching of the dye 8-hydroxy-1,3,6 pyrenetrisulfonate (pyranine), which is incorporated inside the proteoliposomes and is sensitive to the pH changes in its surrounding. The inactive AcrB variant with a substitution in the proton relay network, D407N, showed no activity in proteoliposomes, which correlates with the measurements done in empty liposomes. When AcrA was co-reconstituted with AcrB D407N proteoliposomes it did not restore protein activity. To test the assembly of the AcrAB-TolC pump out of its single components, an in vitro assay was established where the complex assembly was tested with AcrAB- and TolC-containing liposomes. These experiments showed putative AcrAB-TolC formation in the presence or absence of a pump substrate, taurocholate, as well as in the presence of the pump inhibitor, MBX3132. The assembly appeared stable over time and results were invariant in the presence or absence of a pH gradient across the AcrAB-containing membrane.
After determination of the ATPase activity of the P-type ATPase, CopA, in detergent micelles, the protein was reconstituted in LUVs. Quenching of the Ag+-sensitive dye Phen Green SK (PGSK), present on the inside of the CopA-containing proteoliposomes, was observed in presence of ATP and Ag+. Under the same conditions, but in absence of Ag+-ions, quenching was reduced by 80 % after 300 seconds. No PGSK-quenching was observed in control liposomes in the presence of ATP and Ag+. The additional presence of sodium azide led to minimal reduction of the PGSK-quenching as expected since sodium azide is not an inhibitor of P-type ATPases, but the quenching rate was similar to that of the same experimental condition with control liposomes.
The RND superfamily member CusA, as part of the tripartite CusCBA efflux pump, has been proposed to sequester Ag+ or Cu+ from either the cytoplasmic or periplasmic side of the inner membrane. The periplasmic transport of silver ions was implied from an in vitro assay where the quenching of a pH sensitive dye, 9-amino-6-chloro-2-methoxyacridine (ACMA), indicates acidification of the lumen of the proteoliposomes containing CusA when an inwardly directed pH was imposed. The same experiment with the CusA D405N variant, which was previously reported to be an inactive variant, also led to ACMA quenching, although at a slightly lower rate. Under application of an inwardly directed pH and a (negative inside), CusA-containing proteoliposomes showed a strong quenching of the incorporated PGSK dye, suggesting strong Ag+ influx.
The Major Facilitator Superfamily-(MFS-) type EmrAB-TolC pump has an analogous structural setup as the RND-type AcrAB-TolC pump. To examine the efflux of one of its substrates, carbonyl - cyanide m-chlorophenylhydrazone (CCCP), a plate-based susceptibility assay was used. The presence of the EmrAB-TolC pump confers lower susceptibility levels towards CCCP in E. coli, compared to cells not expressing the pump or cells expressing only the MFS component, indicating that EmrAB-TolC extrudes CCCP.
The work done in this study opens up a path towards investigation of drug and metal resistance in vitro. The methodologies to obtain proteoliposomal samples of multicomponent efflux pumps and subsequent measurements of drug/metal ion and H+ fluxes, as well as the determination of pump assembly are crucial for the future research on pump catalysis and transport kinetics. The in vivo drug-plate assays done in this work provide initial insights for future investigations of the drug susceptibility of E. coli expressing the MFS-type tripartite efflux pumps.
The phospholipid bilayers are the primary constituents of the membrane in living cells in which lipids are hold together in bilayer leaflets through a combination of different forces into the liquid crystalline (Lα) phase. Despite their thin fragile formations, the phospholipid bilayers are responsible for performing a variety of important tasks in the cells, some of which are carried out directly by the lipid bilayers and some by various integral proteins embedded within the bilayers. There have been continues efforts over the past decades to replicate the compound biophysical properties of living cell membranes in model lipid bilayers.
An important question remains unanswered: is it possible to replicate physical properties under “non-equilibrium” conditions as found in cell membranes in model lipid bilayers? In almost all previous studies, the model lipid bilayers were under static conditions – for instance, at zero lateral pressure. However, in living organisms, the cell membranes are involved in continuous (nonequilibrium) exchange and (or) transport of lipid species with the surrounding environment which consequently leads them to experience continuous lateral pressure variations. One suitable in vitro approach is to spatiotemporally control the model lipid bilayers over a time period during which they can be spatially stimulated at a level compatible to that found under in vivo conditions. This can be achieved with high spatiotemporal resolution by making lipids light-dependent through implementation of azobenzene photoswitch in their structures.
In this study, a specific azobenzene containing photolipid (AzoPC) is integrated into POPE:POPG bilayers (POPE: 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine, POPG: 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-(1'-rac-glycerol)) at ~14 mol% to construct a photo responsive model bilayers entitled as photoliposomes. Magic angle spinning solid-state NMR spectroscopy (MAS-NMR) at high field (850 MHz) is the measurement technique of choice by which it is possible to pursue the dynamics (fluidity) of the bulk lipids within the photoliposomes at atomistic resolution. It is shown that the AzoPCs undergo an efficient trans-to-cis isomerization (~85%) within the photoliposomes as the result of UV light absorption, and thermally relax back to the trans state during a period of ~65 h under the MAS measurement conditions. The order parameter measurements based on the C−H dipolar couplings reveal that the non-equilibrium cis-to-trans thermal isomerization impact of AzoPC on the fluidity of the bulk lipid is highly localized – the fluidity perturbations originate from specific order parameter changes in the middle section of the bulk lipid acyl chains. Further 1H NOESY measurements confirm the hypothesis that the azoswitch topologies in either cis and trans conformer of the photolipid is the key parameter in localized alteration of the C−H order parameters along the bulk lipid acyl chains.
Diacylglycerol kinase (DgkA) from E. coli is an enzyme responsible for the phosphorylation of diacylglycerol to phosphatidic acid, at the expense of adenosine triphosphate. Structurally, DgkA is a homo oligomer composed of three symmetric 14 kDa protomers, each of which has three transmembrane helices and one surface helix. Upon embedding within the photoliposomes, it is shown that DgkA enhances the AzoPC localization impact on the fluidity of the bulk lipids. In this regard, the results of a series of statistical simulations of lipid lateral diffusions along the bilayer leaflets in presence and absence of embedded proteins are accompanied with those of experimentally measured based upon which it is justified that membrane proteins markedly limit lipid lateral diffusions in the bilayers. In case of the DgkA proteo-liposomes with lipid-to-protein ratio of 50, it is estimated that the diffusion coefficient of lipids is above 2-fold lower compared to that of the protein free liposomes.
The cis-to-trans AzoPC isomerization and its following consequence in localized alteration of the bulk lipid fluidity is further investigated on the structural dynamics and enzymatic functionality of the embedded DgkA within the proteo-photoliposomes. It is revealed that DgkA structural dynamics are perturbated in a multi-scale, complex manner. The dynamics of residues located in different regions of DgkA changes with the light-induced AzoPC isomerization, but their time courses differ from residue to residue. For example, 29Ala, a residue on the hinge between the surface helix and membrane helix-1, exhibits the steepest time-dependent cross peak intensity changes in time-resolved NCA spectra. The impact of the lasting membrane fluidity perturbation on the enzymatic functionality of the embedded DgkA is subsequently measured which demonstrates a significant variation under cis- and trans-AzoPC conformations within the proteo-photoliposomes.
Transport mechanism of a multidrug resistance protein investigated by pulsed EPR spectroscopy
(2019)
In human several diseases result from malfunctions of ATP-binding cassette (ABC) systems, which form one of the largest transport system superfamily. Many ABC exporters contain asymmetric nucleotide-binding sites (NBSs) and some of them are inhibited by the transported substrate.1 For the active transport of diverse chemically substrates across biological membranes, ABC transport complexes use the energy of ATP binding and subsequent hydrolysis. In this thesis, the heterodimeric ABC exporter TmrAB2,3 from Thermus thermophilus, a functional homolog of the human antigen translocation complex TAP, was investigated by using pulsed electron-electron double resonance (PELDOR/DEER) spectroscopy. In the presence of ATP, TmrAB exists in an equilibrium between inward- and outward-facing conformations. This equilibrium can be modulated by changing the ATP concentration, showing asymmetric behaviour in the open-to-close equilibrium between the consensus and the degenerate NBSs. At the degenerate NBS the closed conformation is more preferred and closure of one of the NBSs is sufficient to open the periplasmic gate at the transmembrane domain (TMD).3 By determining the temperature dependence of this conformational equilibrium, the thermodynamics of the energy coupling during ATP-induced conformational changes in TmrAB were investigated. The results demonstrate that ATP-binding alone drives the global conformational switching to the outward-facing state and allows the determination of the entropy and enthalpy changes for this step. With this knowledge, the Gibbs free energy of this ATP induced transition was calculated. Furthermore, an excess of substrate, meaning trans-inhibition of the transporter is resulting mechanistically in a reverse transition from the outward-facing state to an occluded conformation predominantly.3 This work unravels the central role of the reversible conformational equilibrium in the function and regulation of an ABC exporter. For the first time it is shown that the conformational thermodynamics of a large membrane protein complex can be investigated. The presented experiments give new possibilities to investigate other related medically important transporters with asymmetric NBSs or other similar protein complexes.