Computational design of ligand binding is not a solved problem

De novo design of drug-binding proteins with predictable binding energy and specificity

The de novo design of small molecule-binding proteins has seen exciting recent progress; however, high-affinity binding and tunable specificity typically require laborious screening and optimization after computational design. We developed a computational procedure to design a protein that recognizes a common pharmacophore in a series of poly(ADP-ribose) polymerase-1 inhibitors. One of three designed proteins bound different inhibitors with affinities ranging from <5 nM to low micromolar. X-ray crystal structures confirmed the accuracy of the designed protein-drug interactions. Molecular dynamics simulations informed the role of water in binding. Binding free energy calculations performed directly on the designed models were in excellent agreement with the experimentally measured affinities. We conclude that de novo design of high-affinity small molecule-binding proteins with tuned interaction energies is feasible entirely from computation.

Protein sequence design on given backbones with deep learning

Deep learning methods for protein sequence design focus on modeling and sampling the many- dimensional distribution of amino acid sequences conditioned on the backbone structure. To produce physically foldable sequences, inter-residue couplings need to be considered properly. These couplings are treated explicitly in iterative methods or autoregressive methods. Non-autoregressive models treating these couplings implicitly are computationally more efficient, but still await tests by wet experiment. Currently, sequence design methods are evaluated mainly using native sequence recovery rate and native sequence perplexity. These metrics can be complemented by sequence-structure compatibility metrics obtained from energy calculation or structure prediction. However, existing computational metrics have important limitations that may render the generalization of computational test results to performance in real applications unwarranted. Validation of design methods by wet experiments should be encouraged.

De novo design of drug-binding proteins with predictable binding energy and specificity

The de novo design of small-molecule-binding proteins has seen exciting recent progress; however, the ability to achieve exquisite affinity for binding small molecules while tuning specificity has not yet been demonstrated directly from computation. Here, we develop a computational procedure that results in the highest affinity binders to date with predetermined relative affinities, targeting a series of PARP1 inhibitors. Two of four designed proteins bound with affinities ranging from < 5 nM to low μM, in a predictable manner. X-ray crystal structures confirmed the accuracy of the designed protein-drug interactions. Molecular dynamics simulations informed the role of water in binding. Binding free-energy calculations performed directly on the designed models are in excellent agreement with the experimentally measured affinities, suggesting that the de novo design of small-molecule-binding proteins with tuned interaction energies is now feasible entirely from computation. We expect these methods to open many opportunities in biomedicine, including rapid sensor development, antidote design, and drug delivery vehicles.

Retracing the evolution of a modern periplasmic binding protein

Investigating the evolution of structural features in modern multidomain proteins helps to understand their immense diversity and functional versatility. The class of periplasmic binding proteins (PBPs) offers an opportunity to interrogate one of the main processes driving diversification: the duplication and fusion of protein sequences to generate new architectures. The symmetry of their two-lobed topology, their mechanism of binding, and the organization of their operon structure led to the hypothesis that PBPs arose through a duplication and fusion event of a single common ancestor. To investigate this claim, we set out to reverse the evolutionary process and recreate the structural equivalent of a single-lobed progenitor using ribose-binding protein (RBP) as our model. We found that this modern PBP can be deconstructed into its lobes, producing two proteins that represent possible progenitor halves. The isolated halves of RBP are well folded and monomeric proteins, albeit with a lower thermostability, and do not retain the original binding function. However, the two entities readily form a heterodimer in vitro and in-cell. The x-ray structure of the heterodimer closely resembles the parental protein. Moreover, the binding function is fully regained upon formation of the heterodimer with a ligand affinity similar to that observed in the modern RBP. This highlights how a duplication event could have given rise to a stable and functional PBP-like fold and provides insights into how more complex functional structures can evolve from simpler molecular components.

A L-glutamine binding protein modified MNM structured optical fiber biosensor based on surface plasmon resonance sensing for detection of L-glutamine metabolism in vitro embryo culture

A surface plasmon resonance (SPR) optical fiber sensor with multimode-coreless-multimode (MNM) structure was developed, which modified by L-glutamine-binding protein (QBP) for detection of L-glutamine (Gln). The QBP was immobilized on the surface of gold films by chemical cross-linking and exhibited a binding affinity for L-glutamine. The conformation of QBP can be changed from the "open" to the "closed", which led to a red-shift of the SPR peak when QBP bounded to L-glutamine. There was a good linear correlation between is a dependence of the SPR peak on and the concentration of L-glutamine concentration in the range 10-100 μM, with a sensitivity of 10.797nm/log10[Gln] for L-glutamine in the in vitro embryo culture (IVC) medium environment, and the limit of detection (LOD) is 1.187 μM. This QBP-modified MNM structure optical fiber SPR sensor provides a new idea for the developmental potential assessment of embryos in the process of in vitro embryo culture.

Subcellular Localization Defects Characterize Ribose-Binding Mutant Proteins with New Ligand Properties in Escherichia coli

Periplasmic binding proteins have been previously proclaimed as a general scaffold to design sensor proteins with new recognition specificities for nonnatural compounds. Such proteins can be integrated in bacterial bioreporter chassis with hybrid chemoreceptors to produce a concentration-dependent signal after ligand binding to the sensor cell. However, computationally designed new ligand-binding properties ignore the more general properties of periplasmic binding proteins, such as their periplasmic translocation, dynamic transition of open and closed forms, and interactions with membrane receptors. In order to better understand the roles of such general properties in periplasmic signaling behavior, we studied the subcellular localization of ribose-binding protein (RbsB) in Escherichia coli in comparison to a recently evolved set of mutants designed to bind 1,3-cyclohexanediol. As proxies for localization, we calibrated and deployed C-terminal end mCherry fluorescent protein fusions. Whereas RbsB-mCherry coherently localized to the periplasmic space and accumulated in (periplasmic) polar regions depending on chemoreceptor availability, mutant RbsB-mCherry expression resulted in high fluorescence cell-to-cell variability. This resulted in higher proportions of cells devoid of clear polar foci and of cells with multiple fluorescent foci elsewhere, suggesting poorer translocation, periplasmic autoaggregation, and mislocalization. Analysis of RbsB mutants and mutant libraries at different stages of directed evolution suggested overall improvement to more RbsB-wild-type-like characteristics, which was corroborated by structure predictions. Our results show that defects in periplasmic localization of mutant RbsB proteins partly explain their poor sensing performance. Future efforts should be directed to predicting or selecting secondary mutations outside computationally designed binding pockets, taking folding, translocation, and receptor interactions into account. IMPORTANCE Biosensor engineering relies on transcription factors or signaling proteins to provide the actual sensory functions for the target chemicals. Since for many compounds there are no natural sensory proteins, there is a general interest in methods that could unlock routes to obtaining new ligand-binding properties. Bacterial periplasmic binding proteins (PBPs) form an interesting family of proteins to explore for this purpose, because there is a large natural variety suggesting evolutionary trajectories to bind new ligands. PBPs are conserved and amenable to accurate computational binding pocket predictions. However, studying ribose-binding protein in Escherichia coli, we discovered that designed variants have defects in their proper localization in the cell, which can impair appropriate sensor signaling. This indicates that functional sensing capacity of PBPs cannot be obtained solely through computational design of the ligand-binding pocket but must take other properties of the protein into account, which are currently very difficult to predict.

Tailoring Escherichia coli Chemotactic Sensing towards Cadmium by Computational Redesign of Ribose-Binding Protein

Periplasmic binding proteins such as ribose-binding proteins (RBPs) are involved in the bacterial chemotaxis two-component system. RBP selectively identifies and interacts with ribose to induce a conformational change that leads to chemotaxis. Here, we report the development of an engineered Escherichia coli (E. coli) strain expressing a redesigned RBP that can effectively sense cadmium ions and regulate chemotactic movement of cells toward a cadmium ion gradient. RBP was computationally redesigned to bind cadmium ions and produce the conformational change required for chemoreceptor binding. The successful design, CdRBP1, binds to cadmium ions with a dissociation constant of 268 nM. When CdRBP1 was expressed in the periplasmic space of E. coli, the bacteria became live cadmium ion hunters with high selectivity over other divalent metal ions. This work presents an example of making cadmium ions, which are toxic for most organisms, as an attractant to regulate cells movement. Our approach also demonstrates that RBP can be precisely designed to develop metal-detecting living systems for potential applications in synthetic biology and environmental studies.

IMPORTANCE Cadmium pollution is one of the major environmental problems due to excessive release and accumulation. New technologies that can auto-detect cadmium ions with good biocompatibility are in urgent need. In this study, we engineered the bacterial chemotaxis system to positively sense cadmium ions by redesigning ribose-binding protein (RBP) to tightly bind cadmium ion and produce the right conformational change for receptor binding and signaling. Our engineered E. coli cells can auto-detect and chase cadmium ions with divalent metal ion selectivity. Many attempts have been carried out to redesign RBP at the ribose binding site with little success. Instead of the ribose binding site, we introduced the cadmium binding site in the opening of the ribose binding pocket by a specially developed computational algorithm. Our design strategy can be applied to engineer live bacteria with autonomous detection and remediation abilities for metal ions or other chemicals in the future.

De novo protein design by an energy function based on series expansion in distance and orientation dependence

MOTIVATION

Despite many successes, de novo protein design is not yet a solved problem as its success rate remains low. The low success rate is largely because we do not yet have an accurate energy function for describing the solvent-mediated interaction between amino acid residues in a protein chain. Previous studies showed that an energy function based on series expansions with its parameters optimized for side-chain and loop conformations can lead to one of the most accurate methods for side chain (OSCAR) and loop prediction (LEAP). Following the same strategy, we developed an energy function based on series expansions with the parameters optimized in four separate stages (recovering single-residue types without and with orientation dependence, selecting loop decoys and maintaining the composition of amino acids). We tested the energy function for de novo design by using Monte Carlo simulated annealing.

RESULTS

The method for protein design (OSCAR-Design) is found to be as accurate as OSCAR and LEAP for side-chain and loop prediction, respectively. In de novo design, it can recover native residue types ranging from 38% to 43% depending on test sets, conserve hydrophobic/hydrophilic residues at ∼75%, and yield the overall similarity in amino acid compositions at more than 90%. These performance measures are all statistically significantly better than several protein design programs compared. Moreover, the largest hydrophobic patch areas in designed proteins are near or smaller than those in native proteins. Thus, an energy function based on series expansion can be made useful for protein design.

AVAILABILITY AND IMPLEMENTATION

The Linux executable version is freely available for academic users at http://zhouyq-lab.szbl.ac.cn/resources/.

Ribose-Binding Protein Mutants With Improved Interaction Towards the Non-natural Ligand 1,3-Cyclohexanediol

Bioreporters consist of genetically modified living organisms that respond to the presence of target chemical compounds by production of an easily measurable signal. The central element in a bioreporter is a sensory protein or aptamer, which, upon ligand binding, modifies expression of the reporter signal protein. A variety of naturally occurring or modified versions of sensory elements has been exploited, but it has proven to be challenging to generate elements that recognize non-natural ligands. Bacterial periplasmic binding proteins have been proposed as a general scaffold to design receptor proteins for non-natural ligands, but despite various efforts, with only limited success. Here, we show how combinations of randomized mutagenesis and reporter screening improved the performance of a set of mutants in the ribose binding protein (RbsB) of Escherichia coli, which had been designed based on computational simulations to bind the non-natural ligand 1,3-cyclohexanediol (13CHD). Randomized mutant libraries were constructed that used the initially designed mutants as scaffolds, which were cloned in an appropriate E. coli bioreporter system and screened for improved induction of the GFPmut2 reporter fluorescence in presence of 1,3-cyclohexanediol. Multiple rounds of library screening, sorting, renewed mutagenesis and screening resulted in 4.5-fold improvement of the response to 1,3-cyclohexanediol and a lower detection limit of 0.25 mM. All observed mutations except one were located outside the direct ligand-binding pocket, suggesting they were compensatory and helping protein folding or functional behavior other than interaction with the ligand. Our results thus demonstrate that combinations of ligand-binding-pocket redesign and randomized mutagenesis can indeed lead to the selection and recovery of periplasmic-binding protein mutants with non-natural compound recognition. However, current lack of understanding of the intermolecular movement and ligand-binding in periplasmic binding proteins such as RbsB are limiting the rational production of further and better sensory mutants.

New computational protein design methods for de novo small molecule binding sites

Protein binding to small molecules is fundamental to many biological processes, yet it remains challenging to predictively design this functionality de novo. Current state-of-the-art computational design methods typically rely on existing small molecule binding sites or protein scaffolds with existing shape complementarity for a target ligand. Here we introduce new methods that utilize pools of discrete contacts between protein side chains and defined small molecule ligand substructures (ligand fragments) observed in the Protein Data Bank. We use the Rosetta Molecular Modeling Suite to recombine protein side chains in these contact pools to generate hundreds of thousands of energetically favorable binding sites for a target ligand. These composite binding sites are built into existing scaffold proteins matching the intended binding site geometry with high accuracy. In addition, we apply pools of side chain rotamers interacting with the target ligand to augment Rosetta's conventional design machinery and improve key metrics known to be predictive of design success. We demonstrate that our method reliably builds diverse binding sites into different scaffold proteins for a variety of target molecules. Our generalizable de novo ligand binding site design method provides a foundation for versatile design of protein to interface previously unattainable molecules for applications in medical diagnostics and synthetic biology.

Identification and Analysis of Natural Building Blocks for Evolution-Guided Fragment-Based Protein Design

Natural evolution has generated an impressively diverse protein universe via duplication and recombination from a set of protein fragments that served as building blocks. The application of these concepts to the design of new proteins using subdomain-sized fragments from different folds has proven to be experimentally successful. To better understand how evolution has shaped our protein universe, we performed an all-against-all comparison of protein domains representing all naturally existing folds and identified conserved homologous protein fragments. Overall, we found more than 1000 protein fragments of various lengths among different folds through similarity network analysis. These fragments are present in very different protein environments and represent versatile building blocks for protein design. These data are available in our web server called F(old P)uzzle (fuzzle.uni-bayreuth.de), which allows to individually filter the dataset and create customized networks for folds of interest. We believe that our results serve as an invaluable resource for structural and evolutionary biologists and as raw material for the design of custom-made proteins.

Computational redesign of the Escherichia coli ribose-binding protein ligand binding pocket for 1,3-cyclohexanediol and cyclohexanol

Bacterial periplasmic-binding proteins have been acclaimed as general biosensing platform, but their range of natural ligands is too limited for optimal development of chemical compound detection. Computational redesign of the ligand-binding pocket of periplasmic-binding proteins may yield variants with new properties, but, despite earlier claims, genuine changes of specificity to non-natural ligands have so far not been achieved. In order to better understand the reasons of such limited success, we revisited here the Escherichia coli RbsB ribose-binding protein, aiming to achieve perceptible transition from ribose to structurally related chemical ligands 1,3-cyclohexanediol and cyclohexanol. Combinations of mutations were computationally predicted for nine residues in the RbsB binding pocket, then synthesized and tested in an E. coli reporter chassis. Two million variants were screened in a microcolony-in-bead fluorescence-assisted sorting procedure, which yielded six mutants no longer responsive to ribose but with 1.2-1.5 times induction in presence of 1 mM 1,3-cyclohexanediol, one of which responded to cyclohexanol as well. Isothermal microcalorimetry confirmed 1,3-cyclohexanediol binding, although only two mutant proteins were sufficiently stable upon purification. Circular dichroism spectroscopy indicated discernable structural differences between these two mutant proteins and wild-type RbsB. This and further quantification of periplasmic-space abundance suggested most mutants to be prone to misfolding and/or with defects in translocation compared to wild-type. Our results thus affirm that computational design and library screening can yield RbsB mutants with recognition of non-natural but structurally similar ligands. The inherent arisal of protein instability or misfolding concomitant with designed altered ligand-binding pockets should be overcome by new experimental strategies or by improved future protein design algorithms.

NMR Analysis of Apo Glutamine-Binding Protein Exposes Challenges in the Study of Interdomain Dynamics

Glutamine-binding protein (GlnBP) displays an apo, "open" and a holo, "closed" crystal form, mutually related by a rigid-body reorientation of its domains. A fundamental question about such large-scale conformational transitions, whether the closed state exists in the absence of ligand, is controversial in the case of GlnBP. NMR observations have indicated no evidence of the closed form, whereas experimentally validated computations have suggested a remarkable ca. 40 % population. Herein, a paramagnetic NMR strategy designed to detect the putative apo-closed species shows that a major population of the latter is highly improbable. Further, NMR residual dipolar couplings collected under three anisotropic conditions do not reveal differential domain alignment and establish that the average solution conformation is satisfied by the apo-open crystal structure. Our results indicate that the computational prediction of large-scale interdomain motions is not trivial and may lead to erroneous conclusions without proper experimental validation.

Accurate Estimation of Ligand Binding Affinity Changes upon Protein Mutation

The design of proteins with novel ligand-binding functions holds great potential for application in biomedicine and biotechnology. However, our ability to engineer ligand-binding proteins is still limited, and current approaches rely primarily on experimentation. Computation could reduce the cost of the development process and would allow rigorous testing of our understanding of the principles governing molecular recognition. While computational methods have proven successful in the early stages of the discovery process, optimization approaches that can quantitatively predict ligand affinity changes upon protein mutation are still lacking. Here, we assess the ability of free energy calculations based on first-principles statistical mechanics, as well as the latest Rosetta protocols, to quantitatively predict such affinity changes on a challenging set of 134 mutations. After evaluating different protocols with computational efficiency in mind, we investigate the performance of different force fields. We show that both the free energy calculations and Rosetta are able to quantitatively predict changes in ligand binding affinity upon protein mutations, yet the best predictions are the result of combining the estimates of both methods. These closely match the experimentally determined ΔΔG values, with a root-mean-square error of 1.2 kcal/mol for the full benchmark set and of 0.8 kcal/mol for a subset of protein systems providing the most reproducible results. The currently achievable accuracy offers the prospect of being able to employ computation for the optimization of ligand-binding proteins as well as the prediction of drug resistance.

Structure-Based Design of Versatile Biosensors for Small Molecules Based on the PAS Domain of a Thermophilic Histidine Kinase

The development of biosensors for in vitro quantification of small molecules such as metabolites or man-made chemicals is still a major challenge. Here we show that engineered variants of the sensory PAS domain of the histidine kinase CitA of the thermophilic bacterium Geobacillus thermoleovorans represent promising alternatives to established biorecognition elements. By combining binding site grafting and rational design we constructed protein variants binding l-malate, ethylmalonate, or the aromatic compound phthalate instead of the native ligand citrate. Due to more favorable entropy contributions, the wild-type protein and its engineered variants exhibited increased (nano- to micromolar) affinities and improved enantioselectivity compared to CitA homologues of mesophilic organisms. Ligand binding was directly converted into an optical signal that was preserved after immobilization of the protein. A fluorescently labeled variant was used to quantify ethylmalonate, an urinary biomarker for ethylmalonic encephalopathy, in synthetic urine, thereby demonstrating the applicability of the sensor in complex samples.

Redesign of LAOBP to bind novel l-amino acid ligands

Computational protein design is still a challenge for advancing structure-function relationships. While recent advances in this field are promising, more information for genuine predictions is needed. Here, we discuss different approaches applied to install novel glutamine (Gln) binding into the Lysine/Arginine/Ornithine binding protein (LAOBP) from Salmonella typhimurium. We studied the ligand binding behavior of two mutants: a binding pocket grafting design based on a structural superposition of LAOBP to the Gln binding protein QBP from Escherichia coli and a design based on statistical coupled positions. The latter showed the ability to bind Gln even though the protein was not very stable. Comparison of both approaches highlighted a nonconservative shared point mutation between LAOBP_graft and LAOBP_sca. This context dependent L117K mutation in LAOBP turned out to be sufficient for introducing Gln binding, as confirmed by different experimental techniques. Moreover, the crystal structure of LAOBP_L117K in complex with its ligand is reported.

Osmolyte-Like Stabilizing Effects of Low GdnHCl Concentrations on d-Glucose/d-Galactose-Binding Protein

The ability of d-glucose/d-galactose-binding protein (GGBP) to reversibly interact with its ligands, glucose and galactose, makes this protein an attractive candidate for sensing elements of glucose biosensors. This potential is largely responsible for attracting researchers to study the conformational properties of this protein. Previously, we showed that an increase in the fluorescence intensity of the fluorescent dye 6-bromoacetyl-2-dimetylaminonaphtalene (BADAN) is linked to the holo-form of the GGBP/H152C mutant in solutions containing sub-denaturing concentrations of guanidine hydrochloride (GdnHCl). It was hypothesized that low GdnHCl concentrations might lead to compaction of the protein, thereby facilitating ligand binding. In this work, we utilize BADAN fluorescence spectroscopy, intrinsic protein UV fluorescence spectroscopy, and isothermal titration calorimetry (ITC) to show that the sub-denaturing GdnHCl concentrations possess osmolyte-like stabilizing effects on the structural dynamics, conformational stability, and functional activity of GGBP/H152C and the wild type of this protein (wtGGBP). Our data are consistent with the model where low GdnHCl concentrations promote a shift in the dynamic distribution of the protein molecules toward a conformational ensemble enriched in molecules with a tighter structure and a more closed conformation. This promotes the increase in the configurational complementarity between the protein and glucose molecules that leads to the increase in glucose affinity in both GGBP/H152C and wtGGBP.

Complete alanine scanning of the Escherichia coli RbsB ribose binding protein reveals residues important for chemoreceptor signaling and periplasmic abundance

The Escherichia coli RbsB ribose binding protein has been used as a scaffold for predicting new ligand binding functions through in silico modeling, yet with limited success and reproducibility. In order to possibly improve the success of predictive modeling on RbsB, we study here the influence of individual residues on RbsB-mediated signaling in a near complete library of alanine-substituted RbsB mutants. Among a total of 232 tested mutants, we found 10 which no longer activated GFPmut2 reporter expression in E. coli from a ribose-RbsB hybrid receptor signaling chain, and 13 with significantly lower GFPmut2 induction than wild-type. Quantitative mass spectrometry abundance measurements of 25 mutants and wild-type RbsB in periplasmic space showed four categories of effects. Some (such as D89A) seem correctly produced and translocated but fail to be induced with ribose. Others (such as N190A) show lower induction probably as a result of less efficient production, folding and translocation. The third (such as N41A or K29A) have defects in both induction and abundance. The fourth category consists of semi-constitutive mutants with increased periplasmic abundance but maintenance of ribose induction. Our data show how RbsB modeling should include ligand-binding as well as folding, translocation and receptor binding.

A growth- and bioluminescence-based bioreporter for the in vivo detection of novel biocatalysts

The use of bioreporters in high-throughput screening for small molecules is generally laborious and/or expensive. The technology can be simplified by coupling the generation of a desired compound to cell survival, causing only positive cells to stay in the pool of generated variants. Here, a dual selection/screening system was developed for the in vivo detection of novel biocatalysts. The sensor part of the system is based on the transcriptional regulator AraC, which controls expression of both a selection reporter (LeuB or KmR; enabling growth) for rapid reduction of the initially large library size and a screening reporter (LuxCDABE; causing bioluminescence) for further quantification of the positive variants. Of four developed systems, the best system was the medium copy system with KmR as selection reporter. As a proof of principle, the system was tested for the selection of cells expressing an l-arabinose isomerase derived from mesophilic Escherichia coli or thermophilic Geobacillus thermodenitrificans. A more than a millionfold enrichment of cells with l-arabinose isomerase activity was demonstrated by selection and exclusion of false positives by screening. This dual selection/screening system is an important step towards an improved detection method for small molecules, and thereby for finding novel biocatalysts.

Synthetic Protein Switches: Theoretical and Experimental Considerations

Synthetic protein switches with tailored response functions are finding increasing applications as tools in basic research and biotechnology. With a number of successful design strategies emerging, the construction of synthetic protein switches still frequently necessitates an integrated approach that combines detailed biochemical and biophysical characterization in combination with high-throughput screening to construct tailored synthetic protein switches. This is increasingly complemented by computational strategies that aim to reduce the need for costly empirical optimization and thus facilitate the protein design process. Successful computational design approaches range from analyzing phylogenetic data to infer useful structural, biophysical, and biochemical information to modeling the structure and function of proteins ab initio. The following chapter provides an overview over the theoretical considerations and experimental approaches that have been successful applied in the construction of synthetic protein switches.

Computational design of ligand-binding proteins

Custom-designed ligand-binding proteins with novel functions hold the potential for numerous applications. In recent years, the developments of computational methods together with high-throughput experimental screening techniques have led to the generation of novel, high-affinity ligand-binding proteins for given ligands. In addition, naturally occurring ligand-binding proteins have been computationally designed to recognize new ligands while keeping their original biological functions at the same time. Furthermore, metalloproteins have been successfully designed for novel functions and applications. Though much has been learned in these successful design cases, advances in our understanding of protein dynamics and functions related to ligand binding and development of novel computational strategies are necessary to further increase the success rate of computational protein-ligand binding design.

An Atomistic Statistically Effective Energy Function for Computational Protein Design

Shortcomings in the definition of effective free-energy surfaces of proteins are recognized to be a major contributory factor responsible for the low success rates of existing automated methods for computational protein design (CPD). The formulation of an atomistic statistically effective energy function (SEEF) suitable for a wide range of CPD applications and its derivation from structural data extracted from protein domains and protein-ligand complexes are described here. The proposed energy function comprises nonlocal atom-based and local residue-based SEEFs, which are coupled using a novel atom connectivity number factor to scale short-range, pairwise, nonbonded atomic interaction energies and a surface-area-dependent cavity energy term. This energy function was used to derive additional SEEFs describing the unfolded-state ensemble of any given residue sequence based on computed average energies for partially or fully solvent-exposed fragments in regions of irregular structure in native proteins. Relative thermal stabilities of 97 T4 bacteriophage lysozyme mutants were predicted from calculated energy differences for folded and unfolded states with an average unsigned error (AUE) of 0.84 kcal mol(-1) when compared to experiment. To demonstrate the utility of the energy function for CPD, further validation was carried out in tests of its capacity to recover cognate protein sequences and to discriminate native and near-native protein folds, loop conformers, and small-molecule ligand binding poses from non-native benchmark decoys. Experimental ligand binding free energies for a diverse set of 80 protein complexes could be predicted with an AUE of 2.4 kcal mol(-1) using an additional energy term to account for the loss in ligand configurational entropy upon binding. The atomistic SEEF is expected to improve the accuracy of residue-based coarse-grained SEEFs currently used in CPD and to extend the range of applications of extant atom-based protein statistical potentials.

Protein:Ligand binding free energies: A stringent test for computational protein design

A computational protein design method is extended to allow Monte Carlo simulations where two ligands are titrated into a protein binding pocket, yielding binding free energy differences. These provide a stringent test of the physical model, including the energy surface and sidechain rotamer definition. As a test, we consider tyrosyl-tRNA synthetase (TyrRS), which has been extensively redesigned experimentally. We consider its specificity for its substrate l-tyrosine (l-Tyr), compared to the analogs d-Tyr, p-acetyl-, and p-azido-phenylalanine (ac-Phe, az-Phe). We simulate l- and d-Tyr binding to TyrRS and six mutants, and compare the structures and binding free energies to a more rigorous "MD/GBSA" procedure: molecular dynamics with explicit solvent for structures and a Generalized Born + Surface Area model for binding free energies. Next, we consider l-Tyr, ac- and az-Phe binding to six other TyrRS variants. The titration results are sensitive to the precise rotamer definition, which involves a short energy minimization for each sidechain pair to help relax bad contacts induced by the discrete rotamer set. However, when designed mutant structures are rescored with a standard GBSA energy model, results agree well with the more rigorous MD/GBSA. As a third test, we redesign three amino acid positions in the substrate coordination sphere, with either l-Tyr or d-Tyr as the ligand. For two, we obtain good agreement with experiment, recovering the wildtype residue when l-Tyr is the ligand and a d-Tyr specific mutant when d-Tyr is the ligand. For the third, we recover His with either ligand, instead of wildtype Gln.

CSAR Benchmark Exercise 2013: Evaluation of Results from a Combined Computational Protein Design, Docking, and Scoring/Ranking Challenge

Community Structure-Activity Resource (CSAR) conducted a benchmark exercise to evaluate the current computational methods for protein design, ligand docking, and scoring/ranking. The exercise consisted of three phases. The first phase required the participants to identify and rank order which designed sequences were able to bind the small molecule digoxigenin. The second phase challenged the community to select a near-native pose of digoxigenin from a set of decoy poses for two of the designed proteins. The third phase investigated the ability of current methods to rank/score the binding affinity of 10 related steroids to one of the designed proteins (pKd = 4.1 to 6.7). We found that 11 of 13 groups were able to correctly select the sequence that bound digoxigenin, with most groups providing the correct three-dimensional structure for the backbone of the protein as well as all atoms of the active-site residues. Eleven of the 14 groups were able to select the appropriate pose from a set of plausible decoy poses. The ability to predict absolute binding affinities is still a difficult task, as 8 of 14 groups were able to correlate scores to affinity (Pearson-r > 0.7) of the designed protein for congeneric steroids and only 5 of 14 groups were able to correlate the ranks of the 10 related ligands (Spearman-ρ > 0.7).

Insertion of a xylanase in xylose binding protein results in a xylose-stimulated xylanase

BACKGROUND

Product inhibition can reduce catalytic performance of enzymes used for biofuel production. Different mechanisms can cause this inhibition and, in most cases, the use of classical enzymology approach is not sufficient to overcome this problem. Here we have used a semi-rational protein fusion strategy to create a product-stimulated enzyme.

RESULTS

A semi-rational protein fusion strategy was used to create a protein fusion library where the Bacillus subtilis GH11 xylanase A (XynA) was inserted at 144 surface positions of the Escherichia coli xylose binding protein (XBP). Two XynA insertions at XBP positions 209 ([209]XBP-Xyn-XBP) and 262 ([262]XBP-Xyn-XBP) showed a 20% increased xylanolytic activity in the presence of xylose, conditions where native XynA is inhibited. Random linkers of 1-4 Gly/Ala residues were inserted at the XynA N- and C-termini in the [209]XBP and [262]XBP, and the chimeras 2091A and 2621B were isolated, showing a twofold increased xylanolytic activity in the presence of xylose and k cat values of 200 and 240 s(-1) in the 2091A and 2621B, respectively, as compared to 70 s(-1) in the native XynA. The xylose affinity of the XBP was unchanged in the chimeras, showing that the ~3- to 3.5-fold stimulation of catalytic efficiency by xylose was the result of allosteric coupling between the XBP and XynA domains. Molecular dynamics simulations of the chimeras suggested conformation alterations in the XynA on xylose binding to the XBP resulted in exposure of the catalytic cavity and increased mobility of catalytic site residues as compared to the native XynA.

CONCLUSIONS

These results are the first report of engineered glycosyl hydrolase showing allosteric product stimulation and suggest that the strategy may be more widely employed to overcome enzyme product inhibition and to improve catalytic performance. Graphical abstractProtein fusion of a GH11 xylanase (in red) and a xylose binding protein (XBP, in blue) results in a xylanase-XBP chimera that presents allosteric activation of the xylanase activity by xylose (shown as a space-filled molecule bound to the xylanase-XBP chimera).

Computational design of enzyme-ligand binding using a combined energy function and deterministic sequence optimization algorithm

Enzyme amino-acid sequences at ligand-binding interfaces are evolutionarily optimized for reactions, and the natural conformation of an enzyme-ligand complex must have a low free energy relative to alternative conformations in native-like or non-native sequences. Based on this assumption, a combined energy function was developed for enzyme design and then evaluated by recapitulating native enzyme sequences at ligand-binding interfaces for 10 enzyme-ligand complexes. In this energy function, the electrostatic interaction between polar or charged atoms at buried interfaces is described by an explicitly orientation-dependent hydrogen-bonding potential and a pairwise-decomposable generalized Born model based on the general side chain in the protein design framework. The energy function is augmented with a pairwise surface-area based hydrophobic contribution for nonpolar atom burial. Using this function, on average, 78% of the amino acids at ligand-binding sites were predicted correctly in the minimum-energy sequences, whereas 84% were predicted correctly in the most-similar sequences, which were selected from the top 20 sequences for each enzyme-ligand complex. Hydrogen bonds at the enzyme-ligand binding interfaces in the 10 complexes were usually recovered with the correct geometries. The binding energies calculated using the combined energy function helped to discriminate the active sequences from a pool of alternative sequences that were generated by repeatedly solving a series of mixed-integer linear programming problems for sequence selection with increasing integer cuts.

Solution NMR studies of periplasmic binding proteins and their interaction partners

Periplasmic binding proteins (PBPs) are a crucial part of ATP-binding cassette import systems in Gram-negative bacteria. Central to their function is the ability to undergo a large-scale conformational rearrangement from open-unliganded to closed-liganded, which signals the presence of substrate and starts its translocation. Over the years, PBPs have been extensively studied not only owing to their essential role in nutrient uptake but also because they serve as excellent models for both practical applications (e.g., biosensor technology) and basic research (e.g., allosteric mechanisms). Although much of our knowledge at atomic level has been inferred from the detailed, static pictures afforded by crystallographic studies, nuclear magnetic resonance (NMR) has been able to fill certain gaps in such body of work, particularly with regard to dynamic processes. Here, we review NMR studies on PBPs, and their unique insights on conformation, dynamics, energetics, substrate binding, and interactions with related transport proteins. Based on the analysis of recent paramagnetic NMR results, as well as crystallographic and functional observations, we propose a mechanism that could explain the ability of certain PBPs to achieve a closed conformation in absence of ligand while others seem to remain open until ligand-mediated closure.

X-ray crystallographic validation of structure predictions used in computational design for protein stabilization

Protein engineering aimed at enhancing enzyme stability is increasingly supported by computational methods for calculation of mutant folding energies and for the design of disulfide bonds. To examine the accuracy of mutant structure predictions underlying these computational methods, crystal structures of thermostable limonene epoxide hydrolase variants obtained by computational library design were determined. Four different predicted effects indeed contributed to the obtained stabilization: (i) enhanced interactions between a flexible loop close to the N-terminus and the rest of the protein; (ii) improved interactions at the dimer interface; (iii) removal of unsatisfied hydrogen bonding groups; and (iv) introduction of additional positively charged groups at the surface. The structures of an eightfold and an elevenfold mutant showed that most mutations introduced the intended stabilizing interactions, and side-chain conformations were correctly predicted for 72-88% of the point mutations. However, mutations that introduced a disulfide bond in a flexible region had a larger influence on the backbone conformation than predicted. The enzyme active sites were unaltered, in agreement with the observed preservation of catalytic activities. The structures also revealed how a c-Myc tag, which was introduced for facile detection and purification, can reduce access to the active site and thereby lower the catalytic activity. Finally, sequence analysis showed that comprehensive mutant energy calculations discovered stabilizing mutations that are not proposed by the consensus or B-FIT methods.

Design of allosterically regulated protein catalysts

Activity of allosteric protein catalysts is regulated by an external stimulus, such as protein or small molecule binding, light activation, pH change, etc., at a location away from the active site of the enzyme. Since its original introduction in 1961, the concept of allosteric regulation has undergone substantial expansion, and many, if not most, enzymes have been shown to possess some degree of allosteric regulation. The ability to create new catalysts that can be turned on and off using allosteric interactions would greatly expand the chemical biology toolbox and will allow for detection of environmental pollutants and disease biomarkers and facilitate studies of cellular processes and metal homeostasis. Thus, design of allosterically regulated protein catalysts represents an actively growing area of research. In this paper, we describe various approaches to achieving regulation of catalysis.

Escherichia [corrected] coli ribose binding protein based bioreporters revisited

Bioreporter bacteria, i.e., strains engineered to respond to chemical exposure by production of reporter proteins, have attracted wide interest because of their potential to offer cheap and simple alternative analytics for specified compounds or conditions. Bioreporter construction has mostly exploited the natural variation of sensory proteins, but it has been proposed that computational design of new substrate binding properties could lead to completely novel detection specificities at very low affinities. Here we reconstruct a bioreporter system based on the native Escherichia coli ribose binding protein RbsB and one of its computationally designed variants, reported to be capable of binding 2,4,6-trinitrotoluene (TNT). Our results show in vivo reporter induction at 50 nM ribose, and a 125 nM affinity constant for in vitro ribose binding to RbsB. In contrast, the purified published TNT-binding variant did not bind TNT nor did TNT cause induction of the E. coli reporter system.

Design of catalytically amplified sensors for small molecules

Catalytically amplified sensors link an allosteric analyte binding site with a reactive site to catalytically convert substrate into colored or fluorescent product that can be easily measured. Such an arrangement greatly improves a sensor’s detection limit as illustrated by successful application of ELISA-based approaches. The ability to engineer synthetic catalytic sites into non-enzymatic proteins expands the repertoire of analytes as well as readout reactions. Here we review recent examples of small molecule sensors based on allosterically controlled enzymes and organometallic catalysts. The focus of this paper is on biocompatible, switchable enzymes regulated by small molecules to track analytes both in vivo and in the environment.

A new role for Escherichia coli DsbC protein in protection against oxidative stress

We report a new function for Escherichia coli DsbC, a protein best known for disulfide bond isomerization in the periplasm. We found that DsbC regulates the redox state of the single cysteine of the L-arabinose-binding protein AraF. This cysteine, which can be oxidized to a sulfenic acid, mediates the formation of a disulfide-linked homodimer under oxidative stress conditions, preventing L-arabinose binding. DsbC, unlike the homologous protein DsbG, reduces the intermolecular disulfide, restoring AraF binding properties. Thus, our results reveal a new link between oxidative protein folding and the defense mechanisms against oxidative stress.

Computational design of ligand-binding proteins with high affinity and selectivity

The ability to design proteins with high affinity and selectivity for any given small molecule is a rigorous test of our understanding of the physiochemical principles that govern molecular recognition. Attempts to rationally design ligand-binding proteins have met with little success, however, and the computational design of protein-small-molecule interfaces remains an unsolved problem. Current approaches for designing ligand-binding proteins for medical and biotechnological uses rely on raising antibodies against a target antigen in immunized animals and/or performing laboratory-directed evolution of proteins with an existing low affinity for the desired ligand, neither of which allows complete control over the interactions involved in binding. Here we describe a general computational method for designing pre-organized and shape complementary small-molecule-binding sites, and use it to generate protein binders to the steroid digoxigenin (DIG). Of seventeen experimentally characterized designs, two bind DIG; the model of the higher affinity binder has the most energetically favourable and pre-organized interface in the design set. A comprehensive binding-fitness landscape of this design, generated by library selections and deep sequencing, was used to optimize its binding affinity to a picomolar level, and X-ray co-crystal structures of two variants show atomic-level agreement with the corresponding computational models. The optimized binder is selective for DIG over the related steroids digitoxigenin, progesterone and β-oestradiol, and this steroid binding preference can be reprogrammed by manipulation of explicitly designed hydrogen-bonding interactions. The computational design method presented here should enable the development of a new generation of biosensors, therapeutics and diagnostics.

Computational design of protein-small molecule interfaces

The computational design of proteins that bind small molecule ligands is one of the unsolved challenges in protein engineering. It is complicated by the relatively small size of the ligand which limits the number of intermolecular interactions. Furthermore, near-perfect geometries between interacting partners are required to achieve high binding affinities. For apolar, rigid small molecules the interactions are dominated by short-range van der Waals forces. As the number of polar groups in the ligand increases, hydrogen bonds, salt bridges, cation-π, and π-π interactions gain importance. These partial covalent interactions are longer ranged, and additionally, their strength depends on the environment (e.g. solvent exposure). To assess the current state of protein-small molecule interface design, we benchmark the popular computer algorithm Rosetta on a diverse set of 43 protein-ligand complexes. On average, we achieve sequence recoveries in the binding site of 59% when the ligand is allowed limited reorientation, and 48% when the ligand is allowed full reorientation. When simulating the redesign of a protein binding site, sequence recovery among residues that contribute most to binding was 52% when slight ligand reorientation was allowed, and 27% when full ligand reorientation was allowed. As expected, sequence recovery correlates with ligand displacement.

Trastuzumab-Peptide interactions: mechanism and application in structure-based ligand design

Understanding of protein-ligand interactions and its influences on protein stability is necessary in the research on all biological processes and correlative applications, for instance, the appropriate affinity ligand design for the purification of bio-drugs. In this study, computational methods were applied to identify binding site interaction details between trastuzumab and its natural receptor. Trastuzumab is an approved antibody used in the treatment of human breast cancer for patients whose tumors overexpress the HER2 (human epidermal growth factor receptor 2) protein. However, rational design of affinity ligands to keep the stability of protein during the binding process is still a challenge. Herein, molecular simulations and quantum mechanics were used on protein-ligand interaction analysis and protein ligand design. We analyzed the structure of the HER2-trastuzumab complex by molecular dynamics (MD) simulations. The interaction energies of the mutated peptides indicate that trastuzumab binds to ligand through electrostatic and hydrophobic interactions. Quantitative investigation of interactions shows that electrostatic interactions play the most important role in the binding of the peptide ligand. Prime/MM-GBSA calculations were carried out to predict the binding affinity of the designed peptide ligands. A high binding affinity and specificity peptide ligand is designed rationally with equivalent interaction energy to the wild-type octadecapeptide. The results offer new insights into affinity ligand design.

Rational design of a ligand-controlled protein conformational switch

Design of a regulatable multistate protein is a challenge for protein engineering. Here we design a protein with a unique topology, called uniRapR, whose conformation is controlled by the binding of a small molecule. We confirm switching and control ability of uniRapR in silico, in vitro, and in vivo. As a proof of concept, uniRapR is used as an artificial regulatory domain to control activity of kinases. By activating Src kinase using uniRapR in single cells and whole organism, we observe two unique phenotypes consistent with its role in metastasis. Activation of Src kinase leads to rapid induction of protrusion with polarized spreading in HeLa cells, and morphological changes with loss of cell-cell contacts in the epidermal tissue of zebrafish. The rational creation of uniRapR exemplifies the strength of computational protein design, and offers a powerful means for targeted activation of many pathways to study signaling in living organisms.

Computational enzyme design

Recent developments in computational chemistry and biology have come together in the "inside-out" approach to enzyme engineering. Proteins have been designed to catalyze reactions not previously accelerated in nature. Some of these proteins fold and act as catalysts, but the success rate is still low. The achievements and limitations of the current technology are highlighted and contrasted to other protein engineering techniques. On its own, computational "inside-out" design can lead to the production of catalytically active and selective proteins, but their kinetic performances fall short of natural enzymes. When combined with directed evolution, molecular dynamics simulations, and crowd-sourced structure-prediction approaches, however, computational designs can be significantly improved in terms of binding, turnover, and thermal stability.