Session P41: Physics of Proteins: Structure and Dynamics I

Infrared Structural Biology: Detect Functionally Important Structural Motions of Proteins

Proteins are dynamic. Lack of dynamic structures of proteins hampers our understanding of protein functions. Infrared structural biology (IRSB) is an emerging technology. There are several advantages of IRSB for mechanistic studies of proteins: (1) its excellent dynamic range (detecting structural motions from picoseconds to $\ge $ seconds); (2) its high structural sensitivity (detect tiny but functionally important structural motions such as proton transfer and changes in hydrogen bonding interaction); (3) its ability to detect different structural motions simultaneously. Successful development of infrared structural biology demands not only new experimental techniques (from infrared technologies to chemical synthesis and cell biology), but also new data processing (how to translate infrared signals into quantitative structural information of proteins). These topics will be discussed as well as examples of how to use IRSB to study structure-function relationship of proteins.   

Quantifying the Energy Landscape Statistics in Proteins - a Relaxation Mode Analysis

Energy landscape, the hypersurface in the configurational space, has been a useful concept in describing complex processes that occur over a very long time scale, such as the multistep slow relaxations of supercooled liquids and folding of polypeptide chains into structured proteins. Despite extensive simulation studies, its experimental characterization still remains a challenge. To address this challenge, we developed a relaxation mode analysis (RMA) for liquids under a framework analogous to the normal mode analysis for solids. Using RMA, important statistics of the activation barriers of the energy landscape becomes accessible from experimentally measurable two-point correlation functions, e.g. using quasi-elastic and inelastic scattering experiments. We observed a prominent coarsening effect of the energy landscape. The results were further confirmed by direct sampling of the energy landscape using a metadynamics-like adaptive autonomous basin climbing computation. We first demonstrate RMA in a supercooled liquid when dynamical cooperativity emerges in the landscape-influenced regime. Then we show this framework reveals encouraging energy landscape statistics when applied to proteins.   

The Onset of Collective Structural Vibrations at the Protein Dynamical Transition

X-ray, neutron scattering and terahertz measurements [1,2] found a rapid increase in dynamics of critically hydrated proteins above ~220 K, termed the protein dynamical transition. Protein function ceases below the DT. It has been suggested that protein dynamics is slaved to the solvent and the DT originates from thermally activated solvent motions. Since previous measurements did not distinguish local diffusive and librational motions from long-range collective vibrations of proteins, it has not been determined how long-range motions are affected by the DT. Using a recently developed technique, crystal anisotropy terahertz microscopy [3] we directly measured the long-range motions for lysozyme and examined the temperature dependence in the 180-290 K range. We find that the distinct intramolecular vibrations do not follow the expected phonon-like behavior of solid state systems where the vibrational bands sharpen and blue shift with decreasing temperature, rather decrease in intensity as the DT is approached and disappear below the DT. This suggests the surrounding solvent below the DT acts as a frozen cage preventing long-range motions. 1.Doster,W.,et al. Phys.Rev.Lett., 2010.104(9):098101. 2.Niessen,K.,et al. Biophys.Rev., 2015.7,201. 3.Acbas,G.,et al. Nat.Commun., 2014.5,3076.   

Structural and Dynamic Analysis on IDPs by Modified AWSEM-MD

Unlike globular proteins, intrinsically disordered proteins (IDPs) lack both secondary and tertiary structures and can play key roles in various biological processes, including transcriptional regulation, molecular recognition and cellular signaling. These functions can be potentially elucidated by structural heterogeneity of IDPs. Because of their flexibility and disordered nature, it has been difficult to investigate IDPs both computationally and experimentally. In particular, it is desirable to develop coarse-grained, yet accurate models of IDPs, such that simulations exploring sufficient conformational ensembles could be carried out within feasible times. To achieve this goal, we modified the associative memory, water mediated, structure and energy model (AWSEM-MD), which is typically used for folding of globular proteins or binding studies. We tested modified AWSEM-MD on several well-studied IDPs and found the transient secondary structure propensity is consistent with NMR experimental results. The rugged free energy landscapes obtained also show structural heterogeneity of these IDPs. Our proposed extension of AWSEM-MD may allow simulating a wider range of IDPs with high accuracy and computational efficiency.   

Capturing high temperature protein conformations for low-temperature study using ultra-fast cooling

protocols for cooling biomolecular crystals for x-ray cryocrystallography are poorly controlled, leading to crystal-to-crystal and within-crystal non-isomorphism. Furthermore, cooling times below the protein-solvent glass transition of .1 s provide ample time for biological temperature conformations to depopulate and shift. To address these issues, methods and apparatus for cooling biomolecular crystals at rates approaching 100,000 K/s have been developed. These cooling rates are sufficient to eliminate ice formation on cooling without use of cryoprotectants, and to quench additional high-temperature conformations for low-temperature study. Time scales for conformational relaxation can be characterized using variable cooling rates. Possible extension of these methods to maximize conformational quenching will be discussed.   

Protein Conformational Entropy is Independent of Solvent .

Proteins exhibit most of their conformational entropy in individual bond vector motions on the ps-ns timescale. These motions can be examined through determination of the Lipari-Szabo generalized squared order parameter (O$^{\mathrm{2}})$ using NMR spin relaxation measurements. It is often argued that most protein motions are intimately dependent on the nature of the solvating environment. Here the solvent dependence of the fast protein dynamics is directly assessed. Using the model protein ubiquitin, the order parameters of the backbone and methyl groups are shown to be generally unaffected by up to a six-fold increase in bulk viscosity or by encapsulation in the nanoscale interior of a reverse micelle. In addition, the reverse micelle condition permits direct comparison of protein dynamics to the mobility of the hydration layer; no correlation is observed. The dynamics of aromatic side chains are also assessed and provide an estimate of the length- and timescale of protein motions where solvent dependence is seen. These data demonstrate the solvent independence of conformational entropy, clarifying a long-held misconception in the fundamental behavior of biological macromolecules. Supported by the National Science Foundation.   

Moving in the Right Direction: Evolution of Protein Structural Vibrations with Functional State and Mutation

Long-range intramolecular vibrations may enable efficient access to functionally important conformations. We examine how these motions change with inhibitor binding and mutation using terahertz anisotropic absorption and molecular modeling [1,2]. The measured anisotropic absorption dramatically changes with 3NAG inhibitor binding for wild type (WT) free chicken egg white lysozyme (CEWL). We examine the evolution of internal motions with binding using normal mode analysis to calculate an ensemble averaged vibrational density of states (VDOS) and isotropic and anisotropic absorptions for both WT and a two residue (R14 and H15) deletion mutant which has a 1.4 higher activity rate [3]. While the VDOS and isotropic response are largely unchanged with inhibitor binding, the anisotropic response changes dramatically with binding. However, for the mutant the calculated unbound anisotropic absorption more closely resembles its bound spectrum, and it has increased calculated mean squared fluctuations in regions overlapping those in its bound state. These results indicate that the mutant's enhanced activity may be due to a shift in the \textit{direction} of vibrations toward those of the bound state, increasing the sampling rate of the bound conformation. [1] DOI: 10.1007/s12551-015-0168-4 [2] DOI: 10.1038/ncomms4076 [3] DOI: 10.1006/jmbi.1999.2572   

The Molecular Dynamics Study of the Structural Conversions in the Transformer Protein RfaH

Recently, a class of multi-domain proteins such as RfaH transcription factor are labelled as the transformer proteins as they undergo major conformational transformation for performing multiple functions. In the absence of the inter-domain contacts, the C-terminal domain of RfaH transforms from its alpha-helix conformation to a beta-barrel structure. Each of these states have their own functional role: in its alpha-helx state, RfaH-CTD inhibits the transcription by masking the binding site of RNAP, but in its beta state it facilitates the translation. We used various molecular dynamics simulations to study its transformer-like behavior of full-RfaH and identified key amino acid residues that are important in modulating such behavior. Our results show that the inter domain interactions constitute the major barrier in the alpha-helix to beta-barrel conversion. Once the interfacial interactions are broken, structural conversion is easier. The structural conversion from beta-barrel to alpha-helix proceeds with the rearrangement of the hydrophobic residues followed by the inter domain contacts formation via non-native, transient salt-bridge formation, leading to the formation of the native inter domain salt-bridge and hydrophobic contacts to give the final alpha-helix structure.   

Motional displacements in proteins incorporating dynamical diversity

The average mean square displacement (MSD), $\langle r^2 \rangle$, of hydrogen \textit{H} in proteins is measured using incoherent neutron scattering methods. The observed MSD shows a marked increase in magnitude at a temperature $T_D$ $\simeq$ 240 K. This is widely interpreted as a dynamical transition to large MSDs which make function possible in proteins. However, when the data is interpreted in terms of a single averaged MSD, the extracted $\langle r^2 \rangle$ depends on the neutron momentum transfer, $\hbar Q$, used in the measurement. We have shown recently that this apparent dependence on $Q$ arises because the dynamical diversity of the \textit{H} in the protein is neglected[2]. We present models of the dynamical diversity of \textit{H} in Lysosyme that when used in the analysis of simulated neutron data lead to consistent, $Q$ independent values for the average MSD and for the diversity model.\\ 2. D. Vural and L. Hong, J. C. Smith and H. R. Glyde. {\it Phys. Rev. E} {\bf 91}, 052705 (2015).   

Microsecond dynamics of mismatch repair proteins

We will present the results of long-time simulations (250ns-1microsecond) of the mismatch repair protein complexes Mutsalpha bound to various substrates, both normal and damaged. We do so to demonstrate the importance of long-range fluctuations and generalized allostery in such systems and how long-scale GPU-enabled simulations can enabled such analysis.   

Intermediate State Dependence of Intramolecular Vibrations in Photoactive Yellow Protein

Photoactive proteins provide a testbed for the role of long-range collective motions in protein function. Long-range intramolecular vibrations of the protein scaffold may provide efficient energy relaxation[1], enhancement of chromophore vibrations that promote structural transitions[2] and assistance in electron energy transfer[3]. Photoactive yellow protein (PYP) is a cytoplasmic photocycling protein associated with the negative phototactic response to blue light in halohodospira halophile. We measure the intramolecular vibrations of PYP using crystal anisotropy terahertz microscopy (CATM)[4] as a function of photoexcitation. Room temperature CATM measurements are performed in the dark and with continuous illumination at 488 nm, which is found to result in an approximately 20{\%} steady photoexcited state (pB). We find a decrease in anisotropic absorption in frequency range 20-60 cm$^{\mathrm{-1}}$ with photoexcitation. This result may be due to an increase in sample disorder associated with the structural change in pB state. We compare the measured and calculated spectra using molecular dynamics and normal mode/quasiharmonic analysis to identify the nature of the motions giving rise to the resonant absorption bands. .1. Levantino, M., et al. Nat Commun, 2015. \textbf{6}. 2. Mataga, N., et al. Chem. Phys. Lett., 2002. \textbf{352}(3-4): p. 220-225. 3. Fokas, A.S., et al. Photosynth. Res., 2014. \textbf{122}(3): p. 275-292. 4. Acbas, G., et al. Nat Commun, 2014. \textbf{5}.   

Structure and dynamics of Ebola virus matrix protein VP40 by a coarse-grained Monte Carlo simulation

Ebola virus matrix protein VP40 (consisting of 326 residues) plays a critical role in viral assembly and its functions such as regulation of viral transcription, packaging, and budding of mature virions into the plasma membrane of infected cells. How does the protein VP40 go through structural evolution during the viral life cycle remains an open question? Using a coarse-grained Monte Carlo simulation we investigate the structural evolution of VP40 as a function of temperature with the input of a knowledge-based residue-residue interaction. A number local and global physical quantities (e.g. mobility profile, contact map, radius of gyration, structure factor) are analyzed with our large-scale simulations. Our preliminary data show that the structure of the protein evolves through different state with well-defined morphologies which can be identified and quantified via a detailed analysis of structure factor.   

\textbf{Effects of pressure on the dynamics of a hyperthermophilic protein revealed by quasielastic neutron scattering}

Inorganic pyrophosphatase (IPPase) from \textit{Thermococcus thioreducens} is a large oligomeric protein derived from hyperthermophilic microorganism that is found near hydrothermal vents deep under the sea, where the pressure is nearly 100 MPa. Here we study the effects of pressure on the conformational flexibility and relaxation dynamics of IPPase over a wide temperature range using quasielastic neutron scattering (QENS) technique. Two spectrometers were used to investigate the $\beta $-relaxation dynamics of proteins in time ranges from 2 to 25 ps, and from 100 ps to 2 ns. Our results reveal that, under the pressure of 100 MPa, IPPase displays much faster relaxation dynamics than a mesophilic model protein, hen egg white lysozyme (HEWL) [1], opposite to what we observed previously under the ambient pressure [2]. These contradictory observations imply that high pressure affects the dynamical properties of proteins by distorting their energy landscapes. Accordingly, we derived a general schematic denaturation phase diagram that can be used as a general picture to understand the effects of pressure on protein dynamics and activities. [1] Shrestha et al. (2015), \textit{Proc Natl Acad Sci USA}, doi: 10.1073/pnas.1514478112. [2] Chu et al. (2012), \textit{J Phys Chem B} 116(33): 9917-9921.