Substrate Specificities of Variants of Barley (1,3)- and (1,3;1,4)-ß-d-Glucanases Resulting from Mutagenesis and Segment Hybridization.

Barley (1,3;1,4)-ß-d-glucanase is believed to have evolved from an ancestral monocotyledon (1,3)-ß-d-glucanase, enabling the hydrolysis of (1,3;1,4)-ß-d-glucans in the cell walls of leaves and germinating grains. In the present study, we investigated the substrate specificities of variants of the barley enzymes (1,3;1,4)-ß-d-glucan endohydrolase [(1,3;1,4)-ß-d-glucanase] isoenzyme EII (HvEII) and (1,3)-ß-d-glucan endohydrolase [(1,3)-ß-d-glucanase] isoenzyme GII (HvGII) obtained by protein segment hybridization and site-directed mutagenesis. Using protein segment hybridization, we obtained three variants of HvEII in which the substrate specificity was that of a (1,3)-ß-d-glucanase and one variant that hydrolyzed both (1,3)-ß-d-glucans and (1,3;1,4)-ß-d-glucans; the wild-type enzyme hydrolyzed only (1,3;1,4)-ß-d-glucans. Using substitutions of specific amino acid residues, we obtained one variant of HvEII that hydrolyzed both substrates. However, neither protein segment hybridization nor substitutions of specific amino acid residues gave variants of HvGII that could hydrolyze (1,3;1,4)-ß-d-glucans; the wild-type enzyme hydrolyzed only (1,3)-ß-d-glucans. Other HvEII and HvGII variants showed changes in specific activity and their ability to degrade the (1,3;1,4)-ß-d-glucans or (1,3)-ß-d-glucans to larger oligosaccharides. We also used molecular dynamics simulations to identify amino-acid residues or structural regions of wild-type HvEII and HvGII that interact with (1,3;1,4)-ß-d-glucans and (1,3)-ß-d-glucans, respectively, and may be responsible for the substrate specificities of the two enzymes.

Target-template relationships in protein structure prediction and their effect on the accuracy of thermostability calculations.

Improving protein thermostability has been a labor- and time-consuming process in industrial applications of protein engineering. Advances in computational approaches have facilitated the development of more efficient strategies to allow the prioritization of stabilizing mutants. Among these is FEP+, a free energy perturbation implementation that uses a thoroughly tested physics-based method to achieve unparalleled accuracy in predicting changes in protein thermostability. To gauge the applicability of FEP+ to situations where crystal structures are unavailable, here we have applied the FEP+ approach to homology models of 12 different proteins covering 316 mutations. By comparing predictions obtained with homology models to those obtained using crystal structures, we have identified that local rather than global sequence conservation between target and template sequence is a determining factor in the accuracy of predictions. By excluding mutation sites with low local sequence identity (<40%) to a template structure, we have obtained predictions with comparable performance to crystal structures (R2 of 0.67 and 0.63 and an RMSE of 1.20 and 1.16 kcal/mol for crystal structure and homology model predictions, respectively) for identifying stabilizing mutations when incorporating residue scanning into a cascade screening strategy. Additionally, we identify and discuss inherent limitations in sequence alignments and homology modeling protocols that translate into the poor FEP+ performance of a few select examples. Overall, our retrospective study provides detailed guidelines for the application of the FEP+ approach using homology models for protein thermostability predictions, which will greatly extend this approach to studies that were previously limited by structure availability.

Differences in protein structural regions that impact functional specificity in GT2 family ß-glucan synthases.

Most cell wall and secreted ß-glucans are synthesised by the CAZy Glycosyltransferase 2 family (www.cazy.org), with different members catalysing the formation of (1,4)-ß-, (1,3)-ß-, or both (1,4)- and (1,3)-ß-glucosidic linkages. Given the distinct physicochemical properties of each of the resultant ß-glucans (cellulose, curdlan, and mixed linkage glucan, respectively) are crucial to their biological and biotechnological functions, there is a desire to understand the molecular evolution of synthesis and how linkage specificity is determined. With structural studies hamstrung by the instability of these proteins to solubilisation, we have utilised in silico techniques and the crystal structure for a bacterial cellulose synthase to further understand how these enzymes have evolved distinct functions. Sequence and phylogenetic analyses were performed to determine amino acid conservation, both family-wide and within each sub-family. Further structural analysis centred on comparison of a bacterial curdlan synthase homology model with the bacterial cellulose synthase crystal structure, with molecular dynamics simulations performed with their respective ß-glucan products bound in the trans-membrane channel. Key residues that differentially interact with the different ß-glucan chains and have sub-family-specific conservation were found to reside at the entrance of the trans-membrane channel. The linkage-specific catalytic activity of these enzymes and hence the type of ß-glucan chain built is thus likely determined by the different interactions between the proteins and the first few glucose residues in the channel, which in turn dictates the position of the acceptor glucose. The sequence-function relationships for the bacterial ß-glucan synthases pave the way for extending this understanding to other kingdoms, such as plants.

Simulations of Cellulose Synthesis Initiation and Termination in Bacteria.

The processivity of cellulose synthesis in bacterial cellulose synthase (CESA) was investigated using molecular dynamics simulations and the hybrid quantum mechanics and molecular mechanics approach. Our results suggested that cellulose synthesis in bacterial CESA can be initiated with H2O molecules. The chain length or degree of polymerization (DOP) of the product cellulose is related to the affinity of the cellulose chain to the transmembrane tunnel of the enzyme. This opens up the possibility of generating mutants that would produce cellulose chains with desired chain lengths that could have applications in the biofuel and textile fields that depend on the DOP of cellulose chains.

The Shape of Native Plant Cellulose Microfibrils.

Determining the shape of plant cellulose microfibrils is critical for understanding plant cell wall molecular architecture and conversion of cellulose into biofuels. Only recently has it been determined that these cellulose microfibrils are composed of 18 cellulose chains rather than 36 polymers arranged in a diamond-shaped pattern. This study uses density functional theory calculations to model three possible habits for the 18-chain microfibril and compares the calculated energies, structures, 13C NMR chemical shifts and WAXS diffractograms of each to evaluate which shape is most probable. Each model is capable of reproducing experimentally-observed data to some extent, but based on relative theoretical energies and reasonable reproduction of all variables considered, a microfibril based on 5 layers in a 34443 arrangement is predicted to be the most probable. A habit based on a 234432 arrangement is slightly less favored, and a 6 × 3 arrangement is considered improbable.

Hydrogen bonds and twist in cellulose microfibrils.

There is increasing experimental and computational evidence that cellulose microfibrils can exist in a stable twisted form. In this study, atomistic molecular dynamics (MD) simulations are performed to investigate the importance of intrachain hydrogen bonds on the twist in cellulose microfibrils. We systematically enforce or block the formation of these intrachain hydrogen bonds by either constraining dihedral angles or manipulating charges. For the majority of simulations a consistent right handed twist is observed. The exceptions are two sets of simulations that block the O2-O6' intrachain hydrogen bond, where no consistent twist is observed in multiple independent simulations suggesting that the O2-O6' hydrogen bond can drive twist. However, in a further simulation where exocyclic group rotation is also blocked, right-handed twist still develops suggesting that intrachain hydrogen bonds are not necessary to drive twist in cellulose microfibrils.

Evolutionary Dynamics of the Cellulose Synthase Gene Superfamily in Grasses.

Phylogenetic analyses of cellulose synthase (CesA) and cellulose synthase-like (Csl) families from the cellulose synthase gene superfamily were used to reconstruct their evolutionary origins and selection histories. Counterintuitively, genes encoding primary cell wall CesAs have undergone extensive expansion and diversification following an ancestral duplication from a secondary cell wall-associated CesA. Selection pressure across entire CesA and Csl clades appears to be low, but this conceals considerable variation within individual clades. Genes in the CslF clade are of particular interest because some mediate the synthesis of (1,3;1,4)-ß-glucan, a polysaccharide characteristic of the evolutionarily successful grasses that is not widely distributed elsewhere in the plant kingdom. The phylogeny suggests that duplication of either CslF6 and/or CslF7 produced the ancestor of a highly conserved cluster of CslF genes that remain located in syntenic regions of all the grass genomes examined. A CslF6-specific insert encoding approximately 55 amino acid residues has subsequently been incorporated into the gene, or possibly lost from other CslFs, and the CslF7 clade has undergone a significant long-term shift in selection pressure. Homology modeling and molecular dynamics of the CslF6 protein were used to define the three-dimensional dispositions of individual amino acids that are subject to strong ongoing selection, together with the position of the conserved 55-amino acid insert that is known to influence the amounts and fine structures of (1,3;1,4)-ß-glucans synthesized. These wall polysaccharides are attracting renewed interest because of their central roles as sources of dietary fiber in human health and for the generation of renewable liquid biofuels.

Unique aspects of the structure and dynamics of elementary Iß cellulose microfibrils revealed by computational simulations.

The question of how many chains an elementary cellulose microfibril contains is critical to understanding the molecular mechanism(s) of cellulose biosynthesis and regulation. Given the hexagonal nature of the cellulose synthase rosette, it is assumed that the number of chains must be a multiple of six. We present molecular dynamics simulations on three different models of Iß cellulose microfibrils, 18, 24, and 36 chains, to investigate their structure and dynamics in a hydrated environment. The 36-chain model stays in a conformational space that is very similar to the initial crystalline phase, while the 18- and 24-chain models sample a conformational space different from the crystalline structure yet similar to conformations observed in recent high-temperature molecular dynamics simulations. Major differences in the conformations sampled between the different models result from changes to the tilt of chains in different layers, specifically a second stage of tilt, increased rotation about the O2-C2 dihedral, and a greater sampling of non-TG exocyclic conformations, particularly the GG conformation in center layers and GT conformation in solvent-exposed exocyclic groups. With a reinterpretation of nuclear magnetic resonance data, specifically for contributions made to the C6 peak, data from the simulations suggest that the 18- and 24-chain structures are more viable models for an elementary cellulose microfibril, which also correlates with recent scattering and diffraction experimental data. These data inform biochemical and molecular studies that must explain how a six-particle cellulose synthase complex rosette synthesizes microfibrils likely comprised of either 18 or 24 chains.

Can cyclic HIV protease inhibitors bind in a non-preferred form? An ab initio, DFT and MM-PB(GB)SA study.

X-ray crystallography studies have identified that most cyclic inhibitors of HIV protease (including cyclic ureas) bind in a symmetric manner, however some cyclic inhibitors, such as cyclic sulfamides, bind in a non-symmetric manner. This raises the question as to whether it is possible for cyclic sulfamides to bind symmetrically and conversely for cyclic ureas to bind non-symmetrically. Herein we report an analysis of the conformational preference of cyclic ureas and sulfamides both free in solution and bound to HIV protease, including an investigation of the effect of branching. Quantum chemical calculations (B3LYP, M06-2X, MP2, CCSD(T)) predict the cyclic urea to prefer a symmetric conformation in solution, with a large activation barrier towards inter-conversion to the non-symmetric conformation. This differs from the cyclic sulfamides, which marginally prefer a non-symmetric conformation with a much smaller barrier to inter-conversion making it more likely for a non-preferred conformation to be observed. It is predicted that the cyclic scaffold itself favours a symmetric form, while branching induces a preference for a non-symmetric form. MD simulations on the free inhibitors identified inter-conversion with the cyclic sulfamides but not the cyclic ureas, in support of the quantum chemical results. MM-PB(GB)SA calculations on the cyclic inhibitors bound to HIV protease corroborate the X-ray crystallography studies, identifying the cyclic ureas to bind symmetrically and the cyclic sulfamides in a non-symmetrical manner. While the non-preferred form of the sulfamide may well be present as a free molecule in solution, our results suggest that it is unlikely to bind to HIV protease in a symmetric manner.

Effect of atomic charge, solvation, entropy, and ligand protonation state on MM-PB(GB)SA binding energies of HIV protease.

The molecular mechanics-Poisson-Boltzmann surface area (MM-PBSA) and MM-generalized-Born surface area (MM-GBSA) approaches are commonly used in molecular modeling and drug design. Four critical aspects of these approaches have been investigated for their effect on calculated binding energies: (1) the atomic partial charge method used to parameterize the ligand force field, (2) the method used to calculate the solvation free energy, (3) inclusion of entropy estimates, and (4) the protonation state of the ligand. HIV protease has been used as a test case with six structurally different inhibitors covering a broad range of binding strength to assess the effect of these four parameters. Atomic charge methods are demonstrated to effect both the molecular dynamics (MD) simulation and MM-PB(GB)SA binding energy calculation, with a greater effect on the MD simulation. Coefficients of determination and Spearman rank coefficients were used to quantify the performance of the MM-PB(GB)SA methods relative to the experimental data. In general, better performance was achieved using (i) atomic charge models that produced smaller mean absolute atomic charges (Gasteiger, HF/STO-3G and B3LYP/cc-pVTZ), (ii) the MM-GBSA approach over MM-PBSA, while (iii) inclusion of entropy had a slightly positive effect on correlations with experiment. Accurate representation of the ligand protonation state was found to be important. It is demonstrated that these approaches can distinguish ligands according to binding strength, underlining the usefulness of these approaches in computer-aided drug design.

Effect of structural stress on the flexibility and adaptability of HIV-1 protease.

Resistance remains a major issue with regards to HIV-1 protease, despite the availability of numerous HIV-1 protease inhibitors and copious amounts of structural and binding data. In an effort to improve our understanding of how HIV-1 protease is able to "outsmart" new drugs, we have investigated the flexibility of HIV-1 protease and in particular how it adapts to different structural stresses. Our analysis has highlighted the effects of space group on the variability between structures of HIV-1 protease and suggests that consideration of multiple structures and appropriate consideration of different conformations of the Ile50 residue is necessary in any structural analysis. Calculation of the root-mean-square deviation on a per-residue basis has been used to identify 'natural variation', while mutational and ligand analyses have been carried out to identify the effect on structure as a result of specific stresses. It was observed that mutations readily cause changes to occur at sites both close to and distant from a mutation site, with changes more likely to occur at residues that are sites of other major mutations. It is also revealed that HIV-1 protease adaption is dependent on the type and the structure of any bound ligand. Identification of the specific changes that occur due to these stresses will aid in the understanding of resistance and also aid in the design of new drugs.

Binding of an RNA pol II Ligand to the WW Domain of Pin1 Using Molecular Dynamics Docking Simulations.

A novel docking protocol using a long, all atom molecular dynamics (MD) simulation, in an explicit solvent medium, without using any distance constraints is presented. This MD docking protocol is able to dock ligands, based on the C-terminal domain (CTD) of RNA polymerase II, into the tryptophan-tryptophan (WW) domain of Pin1. In this docking process, a significant loop-bending event occurs in order to encircle the ligand into its solvent exposed binding site, which cannot be simulated using current protocols. The simulations were validated structurally and energetically against an X-ray structure to confirm correct sampling of conformational space. Based on these simulations, and justification of the starting structure as a valid intermediate structure, a potential molecular basis for binding was predicted as well as confirming the key residues involved in the formation of the final strong and stable Pin1 WW domain-ligand complex.