Ribosome
Direct applications to the largest known molecular structure, the ribosome are producing important information about how the ribosome structure performs its complex functions as a complex machine. The protein-synthesizing ribosome undergoes large motions to effect the translocation of tRNAs and mRNA; the domain motions of this system were explored with a coarse-grained elastic network model using normal mode analysis. Crystal structures were used to construct various model systems of the 70S complex with/without tRNA, elongation factor Tu and the ribosomal proteins. Computed motions reveal the well-known ratchet-like rotational motion of the large subunits, as well as the head rotation of the small subunit and the high flexibility of the L1 and L7/L12 stalks, even in the absence of ribosomal proteins. This result indicates that these experimentally observed motions during translocation are inherently controlled by the ribosomal shape and only partially dependent upon GTP hydrolysis. The ribosomal proteins S7, S11 and S18 may also be involved in assuring translation fidelity by constraining the mRNA at the exit site of the channel. The mRNA also interacts with the 16S 3' end forming the Shine-Dalgarno complex at the initiation step; the 3' end may act as a 'hook' to reel in the mRNA to facilitate its exit.
See: Kurkcuoglu O, Doruker P, Sen TZ, Kloczkowski A, and Jernigan RL The Ribosome Structure Controls and Directs mRNA Entry, Translocation and Exit Dynamics. Phys Biol. 2008;5:46005.; Kurkcuoglu O, Kurkcuoglu Z, Doruker P, Jernigan RL. Collective dynamics of the ribosomal tunnel revealed by elastic network modeling. Proteins. 2009; 75:837-45; Kurkcuoglu O, Turgut OT, Cansu S, Jernigan RL, Doruker P Focused functional dyanmics of supramolecules by use of a mixed-resolution elastic network model. Biophys J 2009;97:1178-87.
Collective Dynamics of the Ribosomal Tunnel. The collective dynamics of the nascent polypeptide exit tunnel was investigated. The calculated normal modes are considered individually and in linear combinations with different coefficients mimicking the phase angles between modes, in order to follow the mechanistic motions of the tunnel wall residues. The low frequency fluctuations indicate three distinct regions in the tunnel - the entrance, the neck and the exit – the linings of which each having distinctly different motions. Generally the lining of the entrance region moves in the exit direction to assist in the removal of the peptide, with the exit region having significantly larger motions, but in a perpendicular direction, whereas the confined neck region generally has rotational motions. Especially the universally conserved extensions of ribosomal proteins L4 and L22 located at the narrowest and mechanistically strategic region of tunnel undergo generally anti- or non-correlated motions, which may have an important role in the nascent polypeptide gating mechanism, and corresponds closely to the peristaltic motions that Joachim Frank has reported observing at this site.
mRNA Dynamics within the Ribosome.. The ribosome structure appears to exert strong control and directs mRNA entry, translocation and exit dynamics. The ribosome undergoes large motions to effect the translocation of the tRNAs and mRNA. This result indicates that these experimentally observed motions during translocation are inherently controlled by the ribosomal shape and only partially dependent upon GTP hydrolysis. Normal mode analysis further reveals the mobility of A- and P-tRNAs increasing in the absence of the E-tRNA. In addition, the dynamics of the E-tRNA is affected by the absence of the ribosomal protein L1. The mRNA in the entrance tunnel of the mRNA interacts directly with helicase proteins S3 and S4, which constrain the mRNA in a clamp-like fashion, as well as with protein S5, which likely orients the mRNA to ensure high fidelity of translation. The ribosomal proteins S7, S11 and S18 also are involved in translation fidelity by constraining the mRNA at the exit site of the channel. The mRNA also interacts with the 3’ end of the 16S RNA forming the Shine-Dalgarno complex at the initiation step; the 3’ end may act as a ‘hook’ to reel in the mRNA to facilitate its exit.
Despite capabilities of quantum computations to inform us about reaction mechanisms there is still a need for understanding the mechanics of enzymes. And many questions remain to be answered. How important are the domain motions of enzymes for the reaction? For binding? Do these motions conspire to assist the reaction? Our investigations will focus on these mechanistic aspects of enzyme reactions.
Enzyme mechanism of triosephosphate isomerase (TIM)
TIM is an important enzyme in the glycolytic pathway, catalyzing reversible isomerization of dihydroxyacetone phosphate (DHAP) to D-glyceraldehyde 3-phosphate (GAP). TIM is functional as a homodimer with each monomer comprising 245 residues and an active site. Figure 3 (a-b) shows alternative conformations in the slowest internal mode for apo-TIM with uniform coarse-graining of the dimer for a single-node-per-residue. Fig. 3c show the magnitudes and directions of motion of atoms near catalytic site with respect to DHAP computed using mixed coarse graining. Our computations indicated that using the whole protein structure in the computational model, regardless of its level of detail, is crucial for obtaining global dynamics that relate closely to the enzyme mechanism.
Processing mechanisms - Reverse transcriptase
This is one of the first proteins we studied with the ENM gave evidence of motions that relate to its processing steps. The slowest internal mode of motion was a hinge at a site centrally located between the two enzyme sites, and because it can push the two enzyme sites further apart when it opens, this motion can be utilized to pull the nucleic acid strand ahead to its next position. Another slow mode corresponds to a hinge that opens/closes the polymerase site. A coordinated motion alternating between these two states open/pulling and closed/pushing could directly account for this enzyme’s processive motions, where one base is copied onto a second nucleic acid strand at the polymerase site and one base is cut off of it at the RNase H site.
Single molecule pulling experiments
Until now, elastic network models have always been used to study fluctuational dynamics of proteins either around a single native state or between multiple states of proteins exhibiting conformational transitions. In an important new application to single molecule experiments, we have found that elastic network models can also be successfully applied to predict the transient conformations formed during the mechanically-enforced unfolding of proteins. We have combined two different theoretical methods to enable the prediction of the residual structures of proteins under external stretching forces in single molecule experiments: the Gō-like models for energy and the elastic network models for entropy. We have studied in detail the proteins titin and green fluorescent protein and have tested ten additional proteins. Our hypothesis for identifying the breaks is that these will occur at the most strained parts of the structure, which is where motions are most restricted.
Predicting the order in which contacts are broken during single molecule protein stretching experiments
We investigated several different conformations of Ig-I27, the 27th immunoglobulin domain of the I band of titin (PDB :1tit), during stretching. Ig-I27 contains 89 residues with eight beta strands, indicated by letters A(4-7), A’(11-15), B(18-25), C(32-36), D(47-52), E(5-61), F(69-75) and G(78-88). The results show the force-displacement F/du curve, and the main figure shows displacement du vs the sequential distance | i - j | between amino acids i and j, which are in native contact. By comparing these results with Elastic Network Models we found that the smallest fluctuations <(Ri)2> for some particular residues are correlated with the highest forces confirms our hypothesis that these particular contacts are broken in the next step during protein stretching. Our results show that the simple GNM is able to predict contacts that break in the next time-stage of stretching. See: Sułkowska JI, Kloczkowski A, Sen TZ, Cieplak M, Jernigan RL Predicting the order in which contacts are broken during single molecule protein stretching experiments. Proteins 2007, 71:45-60.
Computation of the time dependent dynamics of protein residues for elastic network models
We have studied the time dependent fluctuations of loops in the native structure of tubulin (PDB: 1tub). Tubulin consists of 6907 atoms with 867 C atoms. Figure (left) shows the time evolution of fluctuations in time units of _____for several arbitrary residues: (residue 1 – grey line, residue 167 – green line, residue 334 – brown line, residue 667 – blue line, and residue 834 – pink line. We want to know which normal modes have the largest influence on the motions of the loops, and can directly see these in this type of computation.
The importance of slow motions for protein functional loops
Loops in proteins that connect secondary structures such as alpha-helix and beta-sheet, are often on the surface and may play a critical role in the details of how proteins interact with one another. The mobility of loops is central for the motional freedom and flexibility requirements of active-site loops and may play a critical role for some functions. The structures and behaviours of loops have not been studied much in the context of the whole structure and its overall motions, especially how these might be coupled. We investigated loop motions by using coarse-grained structures (Cα atoms only) to solve the motions of the system by applying Lagrange equations with elastic network models to learn about which loops move in an independent fashion and which move in coordination with domain motions, faster and slower, respectively. The normal modes of the system are calculated using the Eigen-decomposition of the stiffness matrix. The contribution of individual modes and groups of modes is investigated for their effects on all residues in each loop by using Fourier analyses. Our results indicate overall that the motions of functional sets of loops behave in similar ways as the whole structure. But overall only a relatively few loops move in coordination with the dominant slow modes of motion, and these are often closely related to function.
Triose Phosphate Isomerase and Aldolase – steps on glycolysis pathway
Two important steps in glycolysis pathway are – the cleaving of sugar by Fructose Bisphosphate Aldolase (FBA) from a six carbon sugar into two smaller three carbon components components –Dihydroxy Acetone Phosphate (DHAP) and Glyceraldehyde 3-Phosphate (GAP). One of these products, GAP, is the right substrate for continuing along the pathway. But DHAP is not, and in a kind of shunt step, Triosephosphate Isomerase (TIM) converts DHAP into GAP, following which GAP from both reactions rejoin to continue along the pathway to the following enzyme in the cycle.
Aldolases Utilize Different Oligomeric States to Preserve Their Functional Dynamics
We employed coarse-grained elastic network normal-mode analyses to investigate the dynamics of Escherichia coli fructose 1,6-bisphosphate aldolase, E. coli tagatose 1,6-bisphosphate aldolase, and Thermus aquaticus fructose 1,6-bisphosphate aldolase and compared their motions in different oligomeric states. The first one is a dimer, and the second and third are tetramers. Our analyses suggested that oligomerization not only stabilizes the aldolase structures, showing fewer fluctuations at the subunit interfaces, but also allows the enzyme to achieve the required dynamics for its functional loops. The essential mobility of these loops in the functional oligomeric states can facilitate the enzymatic mechanism, substrate recruitment in the open state, bringing the catalytic residues into their required configuration in the closed bound state, and moving back to the open state to release the catalytic products and repositioning the enzyme for its next catalytic cycle. These findings show that global motions are conserved among aldolases within their different oligomeric states to preserve its catalytic mechanism. This suggests that assembled proteins will have modified dynamics that will be important for their functioning within assembled states.
Investigations of allosteric effects on dynamics upon assembly
Predicting dynamics and conformational changes to proteins is needed in order to fully comprehend the effects of assembly. The most striking observation about the distributions of the structures is that bound proteins have modified dynamics – most fluctuations at the binding interface are reduced upon binding, but not all. Perhaps more importantly there are certain locations remote from the binding site that show increases in their fluctuations. These allosteric sites are important outcomes of allosteric activation originating in protein assembly, and represent a type of entropic conservation – as some binding site residue become more confined with lower entropies, these remote sites heat up with increased entropies.
FBA and TIM Architectures
Active form of these enzyme are oligomers. In S.cerevisiae the functional state of FBA is a dimer. In each subunit, this enzyme has an active site cavity where the action takes place. The opening/closing of this cavity and the catalytic activity of this enzyme is controlled by three functional loops (loop 6, loop 7, and loop 8). And, these are similar between FBA and TIM. The active form of TIM is also an oligomer and in S.cerevisiae, it is a dimer. Each subunit contains one catalytic cavity where isomerization between DHAP and GAP takes place. Very similar to the FBA structure, the cavity opening/closing and catalysis are controlled by three functional loops (loop 6, loop 7, and loop 8).
Structural Similarity of FBA and TIM
FBA and TIM have different sizes: each subunit of FBA has 358 residues whereas each subunit of TIM has 247 residues. The extra residues in FBA are mainly used in creating its dimerization interface. However, superposition of the cores of these two enzyme structures gives an RMSD value of 4.88Å which is quite small considering their very low sequence similarity. It also aligns the functional loops of FBA with the corresponding functional loops of TIM. FBA and TIM are two closely interrelated enzymes. Substrate for FBA is a six-carbon sugar and one of the FBA products, DHAP, then goes into the TIM catalytic cavity for isomerization to GAP. So it is expected that there should be some synchronization in the production rates between FBA and TIM. For this synchronization to take occur, there should be a correlation between the motions of the functionally important components of these two proteins. We have used elastic network coarse-grained models to investigate the correlations between FBA and TIM motions. Elastic Network Model generated fluctuations have shown to be in excellent agreement with the fluctuations about the native conformations of proteins for a wide variety of proteins. (Katebi AR, Jernigan RL. The critical role of the loops of triosephosphate isomerase for its oligomerization, dynamics, and functionality . Prot Sci 2014;23:231-228.)
ACP movement pathway within the Fatty Acid Synthesis Machine
Fatty Acids are synthesized on the Fatty Acid Synthesis Pathway from acetyl-CoA and malonyl-CoA precursors. Each cycle adds two carbon atoms to the growing chain. There are two types of Fatty Acid Synthases: FAS I (Type I) and FAS II (Type II). FAS II is generally found in prokaryotes, plants, fungi, and parasites, as well as in mitochondria. FAS II is a set of separate enzymes. However, in FAS I, all catalytic domains are integrated into one large structure, where all synthesis steps occur in the separate domains. Mammalian FAS I is a large dimeric structure, whereas fungal and yeast FAS I is a much larger dodecameric protein.
Fungal FAS I is a dodecameric structure with over 20,000 residues. The structure consists of a very rigid wheel and two domes with the wheel sitting between the two domes. The activation, priming, elongation, and termination of fungal FAS is carried out by seven domains: (1) Phosphopantetheine Transferase (PPT), (2) Acyl Transferase (AT), (3) Malonyl/palmitoyl Transferase (MPT), (4) Ketoacyl Reductase (KR), (5) Dehydratase (DH), and (6) Enoyl Reductase (ER). Acyl Carrier Protein is an important flexible part of both structures, but it is missing within the reported structures, however, there are some available related structures (Fig. 6). The challenge here will be to insert this structure into the interior of the FAS and to model its motions to show how it shuttles substrate between the sequential active sites on the different enzymes. It is activated by binding with Phosphopantethein in the Phosphopantethein Transferase (PPT) domain of the synthase. The elongating fatty acid is attached at one end of the phosphophantetheine (PP) while the other end is bound shown in the Figure.
ACP is believed to exhibit large scale dynamics in moving successively from one active site to the next active site. From initial simulations of FAS dynamics, we obtain results that show how the ACP moves sequentially through the cavity inside the FAS cavity. In this way we are able to simulate the pathway for ACP moving from one active site to the next active site, which will provide a detailed mechanism of the functional steps in fungal fatty acid synthesis.
Computing the transport pathways through membrane transport proteins
Substrate transport through membrane transporters is critical for many biological processes. One of the most interesting questions is how to understand the substrate specificity of transporters. Due to the limitations of experimental methods, computational approaches can be applied advantageously to screen a large number of possible transported molecules. The experimental determination of the mechanistic details of transport is difficult. We have employed steered molecular dynamics simulations to determine the critical factors responsible for the transport and how they interact with protein components along the pathway. Systems that we have investigated include the transport of 1) sulfonamide drugs by the AbgT transporter YdaH protein and 2) long-chain cyclic lipids (Hopanoid and Steroid) by the RND-like HpnN protein. Protein-embedded lipid systems were minimized and equilibrated, followed by Steered Molecular dynamics simulations at constant velocity (cv-SMD) were carried out on the equilibrated systems with NAMD. Based on the simulation trajectories, we have determined the transport pathway through the protein for the ligand transport in these two cases, as well as the roles of the important residues along the pathway, which importantly are consistent with the experimentally identified residues.
Lipid transport via HpnN (RND-Family protein)
HpnN is RND-family transporter beneficial to Rhodopseudomonas palustris. TIE-1. Rhodopseudomonas palustris is a Gram-negative bacterium that produces structurally diverse hopanoid lipids that are similar to eukaryotic steroids. Previous studies suggested HpnN is essential for the movement of hopanoids from the cytoplasmic to the outer membrane. In our computational study, we explore the transport mechanism of diploptene, a 30-Carbon chain hopanoid and a eukaryotic homologue, taurodeoxycholic acid. The computational approach applied is constant velocity-steered molecular dynamics simulation, where the lipid is pulled at a constant velocity in a specific direction through the channel. The simulations for the two ligands were carried out with similar parameters, where the long-chain lipids followed the same channel to pass through, but the two showed distinctively different exit paths.
Molecular Docking of lipids
In preparation for the simulation studies, Diploptene and Taurodeoxycholic acid were docked separately with HpnN protein via Autodock4.2.The grid, in the case of both molecules, was devised using Autogrid4 and was limited/constrained to the region near the channel start. Grid determines the region where the ligand is subjected to docking. Further, a Genetic algorithm has been implemented to carry out the molecular docking. Based on the results of binding energies for the docking process, the best conformation/pose of the ligand at the channel opening was considered for subsequent simulations.
Steered Molecular Dynamics Simulations
To prepare the system for simulations, the protein with the ligand enclosed, was partly immersed in 120 X 120 Å2 palmitoyl-oleyl-phosphatidylethanolamine (POPE) membrane bilayer normal to the z-axis. The membrane was built using membrane builder plugin in NAMD. Transmembrane helices were placed within the membrane, and the other helices outside. The topology and parameter files for Diploptene and Taurodeoxycholic acid were generated with the SwissParam server. The system was solvated considering the entire molecule dimensions, including the extra-membrane part of the protein. Further, a padding of 5 Å of water was added and the system was ionized with the addition of appropriate number of Na+ and Cl- ions to maintain charge balance. The system was subjected to minimization for 250,000 steps, followed by equilibration with the initial temperature of 11̊ K that was gradually increased by 1̊ K at each time step, until 310̊ K. During the equilibration process, a harmonic restraint of 1 kcal/mol Å2 was applied to the protein-enclosed with the ligand. The system was set up with periodic boundary conditions using the particle mesh Ewald method and grid spacing of 1 Å. Default parameters for NAMD with the CHARMM27 force-field was used.
We performed steered MD simulations, by applying a constant force to Diploptene and Taurodeoxycholic acid to study their permeation pathways through the membrane transporter. The force-position profile for both the lipid transport were deter,omed. The 3-dimensional trajectory for the ligands suggests they followed the same pathway during transport, but have different exit points. The change in their permeation pathways occurred after they came into the vicinity of F541, F117 and W661.These residues have been experimentally determined to be functionally important residues for the lipid transport. Diploptene passed out of the channel by going past F541 and W661, whereas Taurodeoxycholic acid interacted with F117 and F541, on a somewhat longer exit pathway. Diploptene was transported in 1 ns across the channel, whereas Taurodeoxycholic acid took 50% longer. The longer time-duration for Taurodeoxycholic acid can be attributed to its bulkier size as well as the longer path travelled by the steroid after the interaction with the trio of amino acid residues. In addition, Taurodeoxycholic acid underwent a change in orientation as it entered the channel protruding outside the lipid membrane.
Cadherin dynamics for cell adhesion
Manibog K, Sankar K, Kim S-A, Zhang Y, Jernigan RL, Sivasankar S. Molecular determinants of cadherin ideal bond formation: Conformation-dependent unbinding on a multidimensional landscape. Proc Natl Acad Sci USA. 2016; 113 : E5711-E5720
The goal of this project has been to provide an understanding of how cells withstand mechanical stress and maintain homeostasis. We believe there are two key mechanisms involved:
- force-induced switching between alternate Ecad conformations and
- force-induced clustering of Ecad at cell-cell contacts.
We are testing these hypotheses on multiple length scales, ranging from single molecules to live cells, by using an integrated approach that combines predictive computer simulations and quantitative single molecule experiments.
We generated improved Ecad elastic network models based on information from multiple experimental structures (see figure below). We calculated structure-based Ecad elastic network models from an X-ray crystallographic dataset of 11 Ecad structures (available from the PDB) by directly using the inverse of the variance of internal distance changes between pairs of residues as the spring stiffnesses. This variable spring elastic network approach allowed us to straightforwardly investigate the dynamics of Ecad and the results were entirely based on experimental structures, and not on simulations. One of the most significant gains from this approach is that it yielded some very weak springs that actually break, whereas with a uniform spring model, all springs are infinitely extensible. There were also significantly improved agreements in the cumulative overlaps between the Principal Components from the dataset and these new ENM normal modes.
Discussions of how proteins assemble lead naturally to considerations of their initial interactions. Classical cadherins are an essential family of cell-cell adhesion proteins that play critical roles in maintaining the integrity of tissues and proteins interacting between cells that are exposed to mechanical force. Cadherins mediate adhesion by interacting in two distinct trans conformations: strand-swap dimers and X-dimers. As cadherins convert between these conformations, they form ideal bonds – i.e. their adhesion is insensitive to force. However, the structural basis for ideal bond formation is unknown. Here, we combine findings from single molecule force measurements with coarse-grained and atomistic simulations, to resolve the mechanistic basis for cadherin ideal bond formation. Using simulations, we predict the energy landscape for cadherin adhesion, the transition pathways for interconversion between X-dimers and strand-swap dimers and the cadherin structures that form ideal bonds. Based on these predictions, we engineered cadherin mutants that promote or inhibit ideal bond formation and measure their bond mechanics using single molecule force clamp measurements with an atomic force microscope. Our data establishes that cadherins adopt a metastable, intermediate conformation as they shuttle between X-dimers and strand-swap dimers; pulling this conformation induces a torsional motion perpendicular to the pulling direction that unbinds the proteins and forms force independent ideal bonds. This torsional motion is blocked when cadherins associate laterally on the cell surface in a cis orientation suggesting that ideal bonds play a role in promoting cis interactions on the cell surface.