Strategies for Selection from Protein Libraries Composed of de Novo Designed Secondary Structure Modules

As more and more protein structures are determined, it has become clear that there is only a limited number of protein folds in nature. To explore whether the protein folds found in nature are the only solutions to the protein folding problem, or that a lack of evolutionary pressure causes the paucity of different protein folds found, we set out to construct protein libraries without any restriction on topology. We generated different libraries (all alpha-helix, all beta-strand and alpha-helix plus beta-strand) with an average length of 100 amino acid residues, composed of designed secondary structure modules (alpha-helix, beta-strand and beta-turn) in various proportions, based primarily on the patterning of polar and non-polar residues. From the analysis of proteins chosen randomly from the libraries, we found that a substantial portion of pure alpha-helical proteins show properties similar to native proteins. Using these libraries as a starting point, we aim to establish a selection system which allows us to enrich proteins with favorable folding properties (non-aggregating, compactly folded) from the libraries. We have developed such a method based on ribosome display. This selection is based on two concepts: (1) misfolded proteins are more sensitive to proteolysis, (2) misfolded and/or aggregated proteins are more hydrophobic. We show that by applying each of these selection criteria proteins that are compactly folded and soluble can be enriched over insoluble and random coil proteins.     

DNA display of biologically active proteins for in vitro protein selection

In vitro display technologies are powerful tools for screening peptides with desired functions. We previously proposed a DNA display system in which streptavidin-fused peptides are linked with their encoding DNAs via biotin labels in emulsion compartments and successfully applied it to the screening of random peptide libraries. Here we describe its application to functional and folded proteins. By introducing peptide linkers between streptavidin and fused proteins, we achieved highly efficient (>95%) formation of DNA-protein conjugates. Furthermore, we successfully enriched a glutathione-S-transferase gene by a factor of 20-30-fold per round on glutathione-coupled beads. Thus, DNA display should be useful for rapidly screening or evolving proteins based on affinity selection.     

Selection of stably folded proteins by phage-display with proteolysis.

To facilitate the process of protein design and learn the basic rules that control the structure and stability of proteins, combinatorial methods have been developed to select or screen proteins with desired properties from libraries of mutants. One such method uses phage-display and proteolysis to select stably folded proteins. This method does not rely on specific properties of proteins for selection. Therefore, in principle it can be applied to any protein. Since its first demonstration in 1998, the method has been used to create hyperthermophilic proteins, to evolve novel folded domains from a library generated by combinatorial shuffling of polypeptide segments and to convert a partially unfolded structure to a fully folded protein.     

Binding cavities and druggability of intrinsically disordered proteins

To assess the potential of intrinsically disordered proteins (IDPs) as drug design targets, we have analyzed the ligand-binding cavities of two datasets of IDPs (containing 37 and 16 entries, respectively) and compared their properties with those of conventional ordered (folded) proteins. IDPs were predicted to possess more binding cavity than ordered proteins at similar length, supporting the proposed advantage of IDPs economizing genome and protein resources. The cavity number has a wide distribution within each conformation ensemble for IDPs. The geometries of the cavities of IDPs differ from the cavities of ordered proteins, for example, the cavities of IDPs have larger surface areas and volumes, and are more likely to be composed of a single segment. The druggability of the cavities was examined, and the average druggable probability is estimated to be 9% for IDPs, which is almost twice that for ordered proteins (5%). Some IDPs with druggable cavities that are associated with diseases are listed. The optimism versus obstacles for drug design for IDPs is also briefly discussed.     

Conformation-specific affinity purification of proteins using engineered binding proteins: application to the estrogen receptor

Affinity chromatography coupled with an "affinity tag" has become a powerful and routine technology for the purification of recombinant proteins. However, such tag-based affinity chromatography usually cannot separate different conformational states (e.g., folded and misfolded) of a protein to be purified. Here, we describe a strategy to separate different conformations of a protein by using "tailor-made" affinity chromatography based on engineered binding proteins. Our method involves: (i) engineering of a binding protein specific to a particular conformation of the protein of interest, and (ii) production and immobilization of the binding protein to prepare conformation-specific affinity chromatography media. Using "monobodies," small antibody mimics based on the fibronectin type III domain, as the target-binding proteins, we demonstrated the effectiveness of our method by separating the active form of the estrogen receptor alpha ligand-binding domain (ERalpha-LBD) from a mixture of active and misfolded species and by discriminating two different conformations of ERalpha-LBD bound to different ligands. Our strategy should be generally applicable to the preparation of conformationally homogeneous protein samples.     

Improved prediction of protein ligand-binding sites using random forests

This article describes a novel method for predicting ligand-binding sites of proteins. This method uses only 8 structural properties as input vector to train 9 random forest classifiers which are combined to predict binding residues. These predicted binding residues are then clustered into some predicted ligand-binding sites. According to our measurement criterion, this method achieved a success rate of 0.914 in the bound state dataset and 0.800 in the unbound state dataset, which are better than three other methods: Q-SiteFinder, SCREEN and Morita's method. It indicates that the proposed method here is successful for predicting ligand-binding sites.     

Investigation of de novo totally random biosequences, Part II: On the folding frequency in a totally random library of de novo proteins obtained by phage display

We present an investigation on theoretically possible protein structures which have not been selected by evolution and are, therefore, not present on our Earth ('Never Born Proteins' (NBP)). In particular, we attempt to assess whether and to what extent such polypeptides might be folded, thus acquiring a globular protein status. A library (ca. 10(9) clones) of totally random polypeptides, with a length of 50 amino acids, has been produced by phage display. The only structural bias in these sequences is a tripeptide substrate for thrombin: PRG, chosen according to the criteria described in the preceding Part I of this series. The presence of this substrate in an otherwise totally random sequence forms the basis for a qualitative experimental criterion which distinguishes unfolded from folded proteins, as folded proteins are more protected from protease digestion than unfolded ones. The investigation of 79 sequences, randomly selected from the initially large library, shows that over 20% of this population is thrombin-resistant, likely due to folding. Analysis of the amino acid sequences of these clones shows no significant homology to extant proteins, which indicates that they are indeed totally de novo. A few of these sequences have been expressed, and here we describe the structural properties of two thrombin-resistant randomly selected ones. These two de novo proteins have been characterized by spectroscopic methods and, in particular, by circular dichroism. The data show a stable three-dimensional folding, which is temperature-resistant and can be reversibly denatured by urea. The consequences of this finding within a library of 'Never Born Proteins' are discussed in terms of molecular evolution.     

Bacterial expression,purification and biophysical characterization of wheat germ agglutinin and its four hevein-like domains

Wheat germ agglutinin (WGA), a chitin binding lectin, has attracted increasing interest because of its unique characteristics such as conformational stability, binding specificity and transcytosis capacity. To pave the way for the study of the molecular basis of WGA's structural stability and binding capacity, as well as to facilitate its use in biomedical and biotechnological developments, we produced recombinant WGA and its 4 isolated hevein-like domains in a bacterial system. All the proteins were expressed as fusion constructs linked to a thioredoxin domain, which was enzymatically or chemically released. The structural and ligand-binding properties of recombinant WGA were similar to the wild lectin. The 4 isolated domains folded and were ligand-binding competent, indicating that each domain constitutes an independent folding unity. The biophysical characterization of the recombinant domains sheds new light on the intricate folding and binding behavior of this emblematic lectin.     

On optima: the case of myoglobin-facilitated oxygen diffusion

The process of myoglobin/leghemoglobin-facilitated oxygen diffusion is adapted to function in different environments in diverse organisms. We enquire how the functional parameters of the process are optimized in particular organisms. The ligand-binding properties of the proteins, myoglobin and plant symbiotic hemoglobins, we discover, suggest that they have been adapted under genetic selection pressure for optimal performance. Since carrier-mediated oxygen transport has probably evolved independantly many times, adaptation of diverse proteins for a common functionality exemplifies the process of convergent evolution. The progenitor proteins may be built on the myoglobin scaffold or may be very different.     

Positive selection for loss-of-function tat mutations identifies critical residues required for TatA activity

The Tat system, found in the cytoplasmic membrane of many bacteria, is a general export pathway for folded proteins. Here we describe the development of a method, based on the transport of chloramphenicol acetyltransferase, that allows positive selection of mutants defective in Tat function. We have demonstrated the utility of this method by selecting novel loss-of-function alleles of tatA from a pool of random tatA mutations. Most of the mutations that were isolated fall in the amphipathic region of TatA, emphasizing the pivotal role that this part of the protein plays in TatA function.     

Ribosome display: selecting and evolving proteins in vitro that specifically bind to a target

Ribosome display is an in vitro selection and evolution technology for proteins and peptides from large libraries. As it is performed entirely in vitro, there are two main advantages over other selection technologies. First, the diversity of the library is not limited by the transformation efficiency of bacterial cells, but only by the number of ribosomes and different mRNA molecules present in the test tube. Second, random mutations can be introduced easily after each selection round, as no library must be transformed after any diversification step. This allows facile directed evolution of binding proteins over several generations. A prerequisite for the selection of proteins from libraries is the coupling of genotype (RNA, DNA) and phenotype (protein). In ribosome display, this link is accomplished during in vitro translation by stabilizing the complex consisting of the ribosome, the mRNA and the nascent, correctly folded polypeptide. The DNA library coding for a particular library of binding proteins is genetically fused to a spacer sequence lacking a stop codon. This spacer sequence, when translated, is still attached to the peptidyl tRNA and occupies the ribosomal tunnel, and thus allows the protein of interest to protrude out of the ribosome and fold. The ribosomal complexes are allowed to bind to surface-immobilized target. Whereas non-bound complexes are washed away, mRNA of the complexes displaying a binding polypeptide can be recovered, and thus, the genetic information of the binding polypeptides is available for analysis. Here we describe a step-by-step procedure to perform ribosome display selection using an Escherichia coli S30 extract for in vitro translation, based on the work originally described and further refined in our laboratory. A protocol that makes use of eukaryotic in vitro translation systems for ribosome display is also included in this issue.     

Redesign of a four-helix bundle protein by phage display coupled with proteolysis and structural characterization by NMR and X-ray crystallography

To test whether it is practical to use phage display coupled with proteolysis for protein design, we used this approach to convert a partially unfolded four-helix bundle protein, apocytochrome b(562), to a stably folded four-helix bundle protein. Four residues expected to form a hydrophobic core were mutated. One residue was changed to Trp to provide a fluorescence probe for studying the protein's physical properties and to partially fill the void left by the heme. The other three positions were randomly mutated. In addition, another residue in the region to be redesigned was substituted with Arg to provide a specific cutting site for protease Arg-c. This library of mutants was displayed on the surface of phage and challenged with protease Arg-c to select stably folded proteins. The consensus sequence that emerged from the selection included hydrophobic residues at only one of the three positions and non-hydrophobic residues at the other two. Nevertheless, the selected proteins were thermodynamically very stable. The structure of a selected protein was characterized using multi-dimensional NMR. All four helices were formed in the structure. Further, site-directed mutagenesis was used to change one of the two non-hydrophobic residues to a hydrophobic residue, which increased the stability of the protein, indicating that the selection result was not based solely on the protein's global stability and that local structural characteristics may also govern the selection. This conclusion is supported by the crystal structure of another mutant that has two hydrophobic residues substituted for the two non-hydrophobic residues. These results suggest that the hydrophobic interactions in the core are not sufficient to dictate the selection and that the location of the cutting site of the protease also influences the selection of structures.     

1H nuclear magnetic relaxation dispersion of Cu,Zn superoxide dismutase in the native and guanidinium-induced unfolded forms

Potentialities and limitations of the use of (1)H NMRD technique for the characterization of the hydration properties of unfolded or partially folded states of proteins are discussed. The copper(I) form of monomeric Cu,Zn superoxide dismutase in its folded state and in the presence of 4M guanidinium chloride is taken as case system. The dispersion profile, analyzed with an extended relaxation matrix analysis, indicates the presence of long-lived water molecules in the folded state. The observed increase in relaxation at high field upon addition of guanidinium chloride indicates an increase in the number of solvation protons interacting with the protein and exchanging with a time shorter than the protein reorientational time. The observed effect is consistent with an exposed protein surface of SOD in the presence of 4M guanidinium chloride smaller than what could be expected for a random coil.     

The use of isothermal titration calorimetry to unravel chemotactic signalling mechanisms

Chemotaxis is based on the action of chemosensory pathways and is typically initiated by the recognition of chemoeffectors at chemoreceptor ligand-binding domains (LBD). Chemosensory signalling is highly complex; aspect that is not only reflected in the intricate interaction between many signalling proteins but also in the fact that bacteria frequently possess multiple chemosensory pathways and often a large number of chemoreceptors, which are mostly of unknown function. We review here the usefulness of isothermal titration calorimetry (ITC) to study this complexity. ITC is the gold standard for studying binding processes due to its precision and sensitivity, as well as its capability to determine simultaneously the association equilibrium constant, enthalpy change and stoichiometry of binding. There is now evidence that members of all major LBD families can be produced as individual recombinant proteins that maintain their ligand-binding properties. High-throughput screening of these proteins using thermal shift assays offer interesting initial information on chemoreceptor ligands, providing the basis for microcalorimetric analyses and microbiological experimentation. ITC has permitted the identification and characterization of many chemoreceptors with novel specificities. This ITC-based approach can also be used to identify signal molecules that stimulate members of other families of sensor proteins.     

Mining mammalian genomes for folding competent proteins using Tat‐dependent genetic selection in Escherichia coli

Recombinant expression of eukaryotic proteins in Escherichia coli is often limited by poor folding and solubility. To address this problem, we employed a recently developed genetic selection for protein folding and solubility based on the bacterial twin‐arginine translocation (Tat) pathway to rapidly identify properly folded recombinant proteins or soluble protein domains of mammalian origin. The coding sequences for 29 different mammalian polypeptides were cloned as sandwich fusions between an N‐terminal Tat export signal and a C‐terminal selectable marker, namely β‐lactamase. Hence, expression of the selectable marker and survival on selective media was linked to Tat export of the target mammalian protein. Since the folding quality control feature of the Tat pathway prevents export of misfolded proteins, only correctly folded fusion proteins reached the periplasm and conferred cell survival. In general, the ability to confer growth was found to relate closely to the solubility profile and molecular weight of the protein, although other features such as number of contiguous hydrophobic amino acids and cysteine content may also be important. These results highlight the capacity of Tat selection to reveal the folding potential of mammalian proteins and protein domains without the need for structural or functional information about the target protein.     

Prediction of ligand-binding sites of proteins by molecular docking calculation for a random ligand library

A new approach to predicting the ligand-binding sites of proteins was developed, using protein-ligand docking computation. In this method, many compounds in a random library are docked onto the whole protein surface. We assumed that the true ligand-binding site would exhibit stronger affinity to the compounds in the random library than the other sites, even if the random library did not include the ligand corresponding to the true binding site. We also assumed that the affinity of the true ligand-binding site would be correlated to the docking scores of the compounds in the random library, if the ligand-binding site was correctly predicted. We call this method the molecular-docking binding-site finding (MolSite) method. The MolSite method was applied to 89 known protein-ligand complex structures extracted from the Protein Data Bank, and it predicted the correct binding sites with about 80-99% accuracy, when only the single top-ranked site was adopted. In addition, the average docking score was weakly correlated to the experimental protein-ligand binding free energy, with a correlation coefficient of 0.44.     

Design of protein conformational switches

Protein conformational switches are ubiquitous in nature and often regulate key biological processes. To design new proteins that can switch conformation, protein designers have focused on the two key components of protein switches: the amino acid sequence must be compatible with the multiple target states and there must be a mechanism for perturbing the relative stability of these states. Proteins have been designed that can switch between folded and disordered states, between distinct folded states and between different aggregation states. A variety of trigger mechanisms have been used, including pH shifts, post-translational modification and ligand binding. Recently, computational protein design methods have been applied to switch design. These include algorithms for designing novel ligand-binding sites and simultaneously optimizing a sequence for multiple target structures.     

Investigation of de novo totally random biosequences, Part IV: Folding Properties of de novo, totally random RNAs

This work lies within the framework of a broader project aimed at exploring the realm of all possible folded polypeptides; the main question addressed here is whether the corresponding RNAs also assume a folded conformation. We present an investigation on the structural properties of de novo, totally random RNAs by means of the 'RNA Foster' assay. Experimental results show that all RNAs studied are folded at 37 degrees , so that fold seems to be a common feature of RNAs. Random RNAs' fold shows a surprising thermal stability with an average T(m) value at ca. 50 degrees which prompts the idea that thermo-stable structures might not be as rare as they are commonly thought to be. The results are discussed within the general framework of random RNA properties such as those that might have been produced in a prebiotic scenario.     

Covariation of backbone motion throughout a small protein domain

The synchronization (correlation) of conformational fluctuations in folded proteins may influence the rates of enzyme catalysis and ligand binding as well as the stabilities of native proteins and their complexes. However, experimental detection of correlated motions remains difficult. Herein, we present an analysis of the covariation of NMR-derived backbone dynamical parameters among a family of ten mutants of a small protein. Both the spatial restriction and the time scales of backbone motions exhibit a higher degree of covariation than would be expected if the internal motions of each group were independent, providing experimental support for correlated dynamics. Application of this approach to other proteins may reveal dynamical correlations that influence catalysis, ligand-binding and/or protein stability.     

On predicting foldability of a protein from its sequence

Several properties of amino acid sequences corresponding to proteins that are known to fold are compared to those of randomly generated sequences and to sequences of intrinsically disordered proteins in order to find properties that distinguish folding sequences from the rest. The properties studied included helix and sheet propensities from secondary structure prediction, adjacency correlations, directionality correlations, as well as propensities of all possible triplets and quadruplets. Small differences between known folded and random sequences were observed for the adjacency and directional correlations, and significant differences were seen on the triplet and especially on the quadruplet propensities. Based on the differences in the adjacency, triplet or quadruplet propensities folding scores were defined and used to test the accuracy of foldability prediction based on these statistics. The best predictions were obtained from the quadruplet propensities.