Protein Engineering Design and Selection - Advance Access

Discovering rules for protein-ligand specificity using support vector inductive logic programming

Kelley, L. A., Shrimpton, P. J., Muggleton, S. H., Sternberg, M. J.E. — 2009-07-02

Structural genomics initiatives are rapidly generating vast numbers of protein structures. Comparative modelling is also capable of producing accurate structural models for many protein sequences. However, for many of the known structures, functions are not yet determined, and in many modelling tasks, an accurate structural model does not necessarily tell us about function. Thus, there is a pressing need for high-throughput methods for determining function from structure. The spatial arrangement of key amino acids in a folded protein, on the surface or buried in clefts, is often the determinants of its biological function. A central aim of molecular biology is to understand the relationship between such substructures or surfaces and biological function, leading both to function prediction and to function design. We present a new general method for discovering the features of binding pockets that confer specificity for particular ligands. Using a recently developed machine-learning technique which couples the rule-discovery approach of inductive logic programming with the statistical learning power of support vector machines, we are able to discriminate, with high precision (90%) and recall (86%) between pockets that bind FAD and those that bind NAD on a large benchmark set given only the geometry and composition of the backbone of the binding pocket without the use of docking. In addition, we learn rules governing this specificity which can feed into protein functional design protocols. An analysis of the rules found suggests that key features of the binding pocket may be tied to conformational freedom in the ligand. The representation is sufficiently general to be applicable to any discriminatory binding problem. All programs and data sets are freely available to non-commercial users at http://www.sbg.bio.ic.ac.uk/svilp_ligand/.

Protein design in biological networks: from manipulating the input to modifying the output

Van der Sloot, A. M., Kiel, C., Serrano, L., Stricher, F. — 2009-07-02

Protein engineering has been an invaluable tool for the deciphering of protein folding and function and in the understanding of biological signaling networks. From an applied point of view it has been of paramount importance in biotechnological and biopharmaceutical products and applications. Traditionally, the protein engineering tools of choice were ‘classical’ rational design, or directed evolution-based methods. In recent years, a third tool has matured: computational protein design (CPD). In this review, we summarize the underlying principles of CPD and discuss its application for understanding and modifying biological systems. Three main applications of the use of protein design will be highlighted and reviewed: artificially rewiring of signal transduction networks, prediction and generation of large-scale in silico interaction networks and using protein design to manipulate gene expression.

Human variant Creutzfeldt-Jakob disease and sheep scrapie PrPres detection using seeded conversion of recombinant prion protein

2009-07-01

The pathological isoform of the prion protein (PrP^res) can serve as a marker for prion diseases, but more practical tests are needed for preclinical diagnosis and sensitive detection of many prion infections. Previously we showed that the quaking-induced conversion (QuIC) assay can detect sub-femtogram levels of PrP^res in scrapie-infected hamster brain tissue and distinguish cerebral spinal fluid (CSF) samples from normal and scrapie-infected hamsters. We now report the adaptation of the QuIC reaction to prion diseases of medical and agricultural interest: human variant Creutzfeldt-Jakob disease (vCJD) and sheep scrapie. PrP^res-positive and -negative brain homogenates from humans and sheep were discriminated within 1–2 days with a sensitivity of 10–100 fg PrP^res. More importantly, in as little as 22 h we were able to distinguish CSF samples from scrapie-infected and uninfected sheep. These results suggest the presence of prions in CSF from scrapie-infected sheep. This new method enables the relatively rapid and sensitive detection of human CJD and sheep scrapie PrP^res and may facilitate the development of practical preclinical diagnostic and high-throughput interference tests.

Assessing computational methods for predicting protein stability upon mutation: good on average but not in the details

Potapov, V., Cohen, M., Schreiber, G. — 2009-06-26

Methods for protein modeling and design advanced rapidly in recent years. At the heart of these computational methods is an energy function that calculates the free energy of the system. Many of these functions were also developed to estimate the consequence of mutation on protein stability or binding affinity. In the current study, we chose six different methods that were previously reported as being able to predict the change in protein stability (G) upon mutation: CC/PBSA, EGAD, FoldX, I-Mutant2.0, Rosetta and Hunter. We evaluated their performance on a large set of 2156 single mutations, avoiding for each program the mutations used for training. The correlation coefficients between experimental and predicted G values were in the range of 0.59 for the best and 0.26 for the worst performing method. All the tested computational methods showed a correct trend in their predictions, but failed in providing the precise values. This is not due to lack in precision of the experimental data, which showed a correlation coefficient of 0.86 between different measurements. Combining the methods did not significantly improve prediction accuracy compared to a single method. These results suggest that there is still room for improvement, which is crucial if we want forcefields to perform better in their various tasks.

Towards temperature-dependent coarse-grained potentials of side-chain interactions for protein folding simulations. I: Molecular dynamics study of a pair of methane molecules in water at various temperatures

2009-06-25

By means of molecular dynamics simulations of a pair of methane molecules in a TIP3P periodic water box with the NVT scheme at six temperatures and, additionally, the NPT scheme at three temperatures ranging from T = 283 to 373 K, we determined the potential of mean force (PMF) of pairs of interacting methane molecules in water as functions of distance between the methane molecules. The PMFs converge to a single baseline only for r> 11 Å at all temperatures. The curves of the dimensionless PMF obtained at different temperatures with the NVT scheme overlap almost perfectly in the region of the contact minimum and still very well in the regions of the desolvation maximum and the solvent-separated minimum, which suggests that the temperature-dependent hydrophobic interaction potentials at constant volume in united-residue force fields can be obtained by scaling the respective dimensionless potentials by RT, R being the universal gas constant. For the dimensionless potentials of mean force obtained with the NPT scheme, the depth of the contact minimum increases, whereas the height of the desolvation maximum and the depth of the solvent-separated minimum decrease with temperature, in agreement with results reported in the literature.

Differential modification of Cys10 alters transthyretin's effect on beta-amyloid aggregation and toxicity

2009-06-23

Tg2576 mice produce high levels of beta-amyloid (Aβ) and develop amyloid deposits, but lack neurofibrillary tangles and do not suffer the extensive neuronal cell loss characteristic of Alzheimer's disease. Protection from Aβ toxicity has been attributed to up-regulation of transthyretin (TTR), a normal component of plasma and cerebrospinal fluid. We compared the effect of TTR purified from human plasma (pTTR) with that produced recombinantly (rTTR) on Aβ aggregation and toxicity. pTTR slowed Aβ aggregation but failed to protect primary cortical neurons from Aβ toxicity. In contrast, rTTR accelerated aggregation, while effectively protecting neurons. This inverse correlation between Aβ aggregation kinetics and toxicity is consistent with the hypothesis that soluble intermediates rather than insoluble fibrils are the most toxic Aβ species. We carried out a detailed comparison of pTTR with rTTR to ascertain the probable cause of these different effects. No differences in secondary, tertiary or quaternary structure were detected. However, pTTR differed from rTTR in the extent and nature of modification at Cys10. We hypothesize that differential modification at Cys10 regulates TTR's effect on Aβ aggregation and toxicity.

The impact of ataxin-1-like histidine insertions on polyglutamine aggregation

Jayaraman, M., Kodali, R., Wetzel, R. — 2009-06-18

Spinocerebellar ataxia type 1 (SCA1) is one of a group of nine expanded CAG repeat diseases, in which polyglutamine (polyQ) expansion above a threshold is associated with increased disease risk and aggregation. SCA1 is unique in which the polyQ in the disease protein, ataxin1, often contains a few His residues that appear to block toxicity. Here, we ask how His insertions affect aggregation by comparing a Q₃₀ peptide with and without a centrally inserted His-Gln-His sequence. We found that at pH 7.5–8.5, His interruptions decrease polyQ aggregation rates but do not change the spontaneous growth mechanism: nucleated growth polymerization with a critical nucleus of one without non-fibrillar intermediates. The decreased aggregation rates are because of reductions in nucleation equilibrium constants. At pH 6, however, the His-interrupted peptide aggregates by a different mechanism that involves a low ThT-binding intermediate and produces a polymorphic amyloid product. In aggregates grown at pH 7.5, the His residues are solvent-accessible. Aggregates of His-inserted polyQ are good seeds for Q₃₀ elongation, suggesting the potential to recruit polyQ proteins in the cell. Our data are therefore most consistent with His insertions blocking toxicity by suppressing rates and/or altering pathways of spontaneous aggregation.

Computational design-based molecular engineering of the glycosyl hydrolase family 11 B. subtilis XynA endoxylanase improves its acid stability

Belien, T., Joye, I. J., Delcour, J. A., Courtin, C. M. — 2009-06-16

Rational protein engineering was applied to improve the limited stability of the glycosyl hydrolase family 11 (GH11) endo-β-1,4-xylanase from Bacillus subtilis under acidic conditions. Since the pH dependence of protein stability is governed by the ionisation states of the side chains of its titrable amino acid residues, we explored the strategy of changing pH-stability profiles by altering pK_a values of key residues through in silico designed mutations. To this end, computational predictions and molecular modelling were carried out using the recently developed pKD software package. Four endoxylanase variants, in which the pK_a values of either Asp4 and Asp11 or His149 were targeted to shift downwards through incorporation of three to five point mutations, were generated and recombinantly expressed in the cytoplasm of Escherichia coli. All four mutants showed considerably increased functional stability at acid pH levels. They retained ~30–70% and ~75–95% of their activity after incubation at pH 3 and 4, respectively, in comparison with only ~23% and ~57%, respectively, for the wild-type enzyme under the experimental conditions. No acidophilic adaptation of the catalytic activity had occurred. In addition, their functional stability and catalytic activity profiles under different temperature and ionic strength conditions were significantly altered. These findings contribute to general understanding of the molecular mechanisms governing the pH-dependent stability of GH11 proteins, and hence they can be applied to enhance the stability and effectiveness of many GH11 endoxylanases used in industry today.