Combining site-directed spin labeling EPR spectroscopy and biomolecular simulations to study conformation and dynamics of membrane proteins

Bitte benutzen Sie diese Kennung, um auf die Ressource zu verweisen:
Titel: Combining site-directed spin labeling EPR spectroscopy and biomolecular simulations to study conformation and dynamics of membrane proteins
Autor(en): Klose, Daniel
Erstgutachter: Prof. Dr. Heinz-Jürgen Steinhoff
Zweitgutachter: Prof. Dr. Christopher W. M. Kay
Zusammenfassung: Understanding the conformational and dynamic changes of biomacromolecular complexes in different states, such as the membrane protein photoreceptor-transducer complex NpSRII/NpHtrII, is a key step to gaining insight into the functional mechanism of these important classes of protein complexes, since ~30 % of the human proteome are membrane proteins, yet they are largely underrepresented in terms of structural information with <1 % of all structures in the protein data bank. Hence for the development of methods suitable to study the conformation and dynamics of such complexes there is a strong demand and a vast potential field of applications. Here we combined method development at the interface between biomolecular simulations and model-based analysis of EPR- and fluorescence spectroscopic data with application studies using state-of-the-art spectroscopic techniques in conjunction with site-directed spin- or fluorescence labeling. In an initial benchmark study on the rigid globular protein complex Rpo4/7, we compared experimental inter fluorescence label distances or spin label distance distributions to a variety of predicted inter label distances based on molecular dynamics simulations, Monte Carlo sampling and a discrete rotamer library analysis. We found that while for the molecular dynamics simulations with explicit solvent considerable sampling challenges have to be overcome to reproduce the experimentally observed inter label distance distributions, the Monte Carlo sampling performed well when compared to the experimental data and was computationally less demanding. Significantly more efficient and equally accurate for our examples was the so-called rotamer library analysis available for the spin labels since it relies on a pre-calculated set of rotational isomers. In general, predictions for the mean distances were in agreement within the error margins while distribution shapes were more challenging to reproduce. Overall this study shows a positive evaluation for the assessed tools and the developed simulation protocols as well as their potential applications. Using the combination of EPR and fluorescence spectroscopy for distance determination we studied the structural influence of RNA binding on Rpo4/7, and showed that the protein complex stays conformationally rigid and thereby serves as a guiding rail for the nascent RNA chain that leaves the RNA polymerase along the Rpo4/7 RNA binding interface. To enhance the interpretation of experimentally determined changes of conformation and dynamics in protein complexes and to discuss the observed changes in terms of structural information, we built models of the two transcription factors TFE and the Spt4/5 complex, as well as of Argonaute, a 713 amino acid four-domain protein nuclease from Methanocaldococcus jannaschii. These structural models not only allowed a more accurate planning of fluorescence or EPR labeling experiments, but also the models enabled the discussion of the experimental data in structural terms. Based on such an initial structure further computational analysis techniques may be applied to identify putative structural changes or dynamic modes. This was shown for the histidine transporter HisQMP2, where we combined normal mode analysis to model protein flexibility with the rotamer library analysis to screen for possible conformational changes in comparison to experimental inter spin distance data. The most prominent agreement with one mode led to a working hypothesis of a conformational change and provides the basis for validation in future experiments. Due to the inherent synergy effects, we applied a combined experimental and simulation approach for the EPR-based distance determination in the globular DNA-binding protein LexA to probe conformation and dynamics of the N-terminal DNA-binding domains with respect to the C-terminal domains within the LexA homodimer. While the C-terminal dimerization domains exhibit a well-defined conformation that proved to be independent of DNA-binding, large-scale changes in conformation and dynamics were detected for the N-terminal domains. They were only found in a defined conformation when bound to DNA while in its absence a large rotational freedom of the entire N-terminal domains contributed to the conformational ensemble. Combined with a biochemical characterization of the autocatalytic cleavage of LexA, our data explains how LexA induces the SOS response after DNA damage or under latent antibiotic stress. We further studied the membrane photoreceptor-transducer complex NpSRII/NpHtrII that governs the light-dependent swimming behavior in Natronomonas pharaonis by a two-component signaling system. This system comprises extraordinary features of sensitivity, signal amplification, integration and transducer cooperativity, yet the molecular details of these features are poorly understood, as is signal propagation itself. By combining time-resolved cw EPR spectroscopy of NpSRII/NpHtrII variants spin labeled in the HAMP1 domain with time-resolved optical absorbance spectroscopy to report on the receptor signaling state, we found a tight kinetic coupling of receptor and transducer during the relaxation back to the ground state and hence a prolonged activation period, that with ~500 - ~700 ms is sufficiently long to cause phosphorylation bursts of the cognate kinase CheA. This explains signal amplification already on the level of the NpSRII/NpHtrII dimers. We further determined the transient difference spectra from the time-resolved EPR data that show local differences in dynamics and steric restrictions upon light-activation. Comparing these experimentally observed differences to predictions confirms the assumed two-state structural model and shows this transition between the two states for a single HAMP domain in a light-dependent manner. Additionally, our approach integrates a dynamic view into the model, since the two states are shown to exhibit different local dynamics in a fashion described previously as a competing model for signaling by dynamic differences based on biochemical studies. Here we show unification of the two models into one congruent description encompassing a transition between the two previously suggested states by concerted structural and dynamic changes. In an independent analysis using all-atom and coarse grained molecular dynamics of the NpSRII/NpHtrII complex in the minimal unit that can exert kinase control, the trimer of receptor-transducer dimers, we revealed a distinct dynamical pattern encoded in the primary sequence of the coiled-coil heptad-repeats. Upon receptor activation, these segments alter their dynamics in a concerted fashion with regions such as HAMP1 and the adaptation region becoming more compact, while HAMP2 and the tip become more dynamic, leading to dynamic and to limited structural changes at the CheA-kinase binding sites. Together with an extensive validation against experimental data, these findings suggest the altered dynamics as the mechanism for signal propagation along the extended coiled-coil structure of NpHtrII. This working model, that explains the current body of experimental data, allows for further refinement by all-atom molecular dynamics and provides a basis to devise future experiments for validation. The presented studies outline the versatile methodology of combined experimental and simulation approaches to analyze the conformation and dynamics of biomacromolecules including membrane protein complexes.
Schlagworte: SDSL EPR spectroscopy; molecular dynamics; membrane protein conformation and dynamics; FRET
Erscheinungsdatum: 29-Jan-2015
Enthalten in den Sammlungen:FB04 - E-Dissertationen

Dateien zu dieser Ressource:
Datei Beschreibung GrößeFormat 
thesis_klose.pdfPräsentationsformat12,6 MBAdobe PDFMiniaturbild

Alle Ressourcen im repOSitorium sind urheberrechtlich geschützt, soweit nicht anderweitig angezeigt.