Computational protein design.

The field of computational protein design is reaching its adolescence. Protein design algorithms have been applied to design or engineer proteins that fold, fold faster, catalyze, catalyze faster, signal, and adopt preferred conformational states. Further developments of scoring functions, sampling strategies, and optimization methods will expand the range of applicability of computational protein design to larger and more varied systems, with greater incidence of success. Developments in this field are beginning to have significant impact on biotechnology and chemical biology.

Long-range effects on calcium binding and conformational change in the N-domain of calmodulin.

Proteins within the EF-hand protein family exhibit different conformational responses to Ca(2+) binding. Calmodulin and other members of the EF-hand protein family undergo major changes in conformation upon binding Ca(2+). However, some EF-hand proteins, such as calbindin D9k (Clb), bind Ca(2+) without a significant change in conformation. Here, we investigate the effects of replacement of a leucine at position 39 of the N-terminal domain of calmodulin (N-Cam) with a phenylalanine derived from Clb. This variant is studied alone and in the context of other mutations that affect the conformational properties of N-Cam. Strikingly, the introduction of Phe39, which is distant from the calcium binding sites, leads to a significant enhancement of Ca(2+) binding affinity, even in the context of other mutations which trap the protein in the closed form. The results yield novel insights into the evolution of EF-hand proteins as calcium sensors versus calcium buffers.

Solvation energetics and conformational change in EF-hand proteins.

Calmodulin and other members of the EF-hand protein family are known to undergo major changes in conformation upon binding Ca(2+). However, some EF-hand proteins, such as calbindin D9k, bind Ca(2+) without a significant change in conformation. Here, we show the importance of a precise balance of solvation energetics to conformational change, using mutational analysis of partially buried polar groups in the N-terminal domain of calmodulin (N-cam). Several variants were characterized using fluorescence, circular dichroism, and NMR spectroscopy. Strikingly, the replacement of polar side chains glutamine and lysine at positions 41 and 75 with nonpolar side chains leads to dramatic enhancement of the stability of the Ca(2+)-free state, a corresponding decrease in Ca(2+)-binding affinity, and an apparent loss of ability to change conformation to the open form. The results suggest a paradigm for conformational change in which energetic strain is accumulated in one state in order to modulate the energetics of change to the alternative state.

Virtual interaction profiles of proteins.

We have developed a new method for the prediction of peptide sequences that bind to a protein, given a three-dimensional structure of the protein in complex with a peptide. By applying a recently developed sequence prediction algorithm and a novel ensemble averaging calculation, we generate a diverse collection of peptide sequences that are predicted to have significant affinity for the protein. Using output from the simulations, we create position-specific scoring matrices, or virtual interaction profiles (VIPs). Comparison of VIPs for a collection of binding motifs to sequences determined experimentally indicates that the prediction algorithm is accurate and applicable to a diverse range of structures. With these VIPs, one can scan protein sequence databases rapidly to seek binding partners of potential biological significance. Overall, this method can significantly enhance the information contained within a protein- peptide crystal structure, and enrich the data obtained by experimental selection methods such as phage display.

Prediction of amino acid sequence from structure.

We have developed a method for the prediction of an amino acid sequence that is compatible with a three-dimensional backbone structure. Using only a backbone structure of a protein as input, the algorithm is capable of designing sequences that closely resemble natural members of the protein family to which the template structure belongs. In general, the predicted sequences are shown to have multiple sequence profile scores that are dramatically higher than those of random sequences, and sometimes better than some of the natural sequences that make up the superfamily. As anticipated, highly conserved but poorly predicted residues are often those that contribute to the functional rather than structural properties of the protein. Overall, our analysis suggests that statistical profile scores of designed sequences are a novel and valuable figure of merit for assessing and improving protein design algorithms.

Solution structure and dynamics of a designed hydrophobic core variant of ubiquitin.

BACKGROUND: The recent merger of computation and protein design has resulted in a burst of success in the generation of novel proteins with native-like properties. A critical component of this coupling between theory and experiment is a detailed analysis of the structures and stabilities of designed proteins to assess and improve the accuracy of design algorithms. RESULTS: Here we report the solution structure of a hydrophobic core variant of ubiquitin, referred to as 1D7, which was designed with the core-repacking algorithm ROC. As a measure of conformational specificity, we also present amide exchange protection factors and backbone and sidechain dynamics. The results indicate that 1D7 is similar to wild-type (WT) ubiquitin in backbone structure and degree of conformational specificity. We also observe a good correlation between experimentally determined sidechain structures and those predicted by ROC. However, evaluation of the core sidechain conformations indicates that, in general, 1D7 has more sidechains in less statistically favorable conformations than WT. CONCLUSIONS: Our results provide an explanation for the lower stability of 1D7 compared to WT, and suggest modifications to design algorithms that may improve the accuracy with which structure and stability are predicted. The results also demonstrate that core packing can affect conformational flexibility in subtle ways that are likely to be important for the design of function and protein-ligand interactions.

Side-chain and backbone flexibility in protein core design.

We have developed a computational approach for the design and prediction of hydrophobic cores that includes explicit backbone flexibility. The program consists of a two-stage combination of a genetic algorithm and monte carlo sampling using a torsional model of the protein. Backbone structures are evaluated either by a canonical force-field or a constraining potential that emphasizes the preservation of local geometry. The utility of the method for protein design and engineering is explored by designing three novel hydrophobic core variants of the protein 434 cro. We use the new method to evaluate these and previously designed 434 cro variants, as well as a series of phage T4 lysozyme variants. In order to properly evaluate the influence of backbone flexibility, we have also analyzed the effects of varying amounts of side-chain flexibility on the performance of fixed backbone methods. Comparison of results using a fixed versus flexible backbone reveals that, surprisingly, the two methods are almost equivalent in their abilities to predict relative experimental stabilities, but only when full side-chain flexibility is allowed. The prediction of core side-chain structure can vary dramatically between methods. In some, but not all, cases the flexible backbone method is a better predictor of structure. The development of a flexible backbone approach to core design is particularly important for attempts at de novo protein design, where there is no prior knowledge of a precise backbone structure.

Rotamer strain as a determinant of protein structural specificity.

We present direct evidence for a change in protein structural specificity due to hydrophobic core packing. High resolution structural analysis of a designed core variant of ubiquitin reveals that the protein is in slow exchange between two conformations. Examination of side-chain rotamers indicates that this dynamic response and the lower stability of the protein are coupled to greater strain and mobility in the core. The results suggest that manipulating the level of side-chain strain may be one way of fine tuning the stability and specificity of proteins.

Computer search algorithms in protein modification and design.

The computer-aided design of protein sequences requires efficient search algorithms to handle the enormous combinatorial complexity involved. A variety of different algorithms have now been applied with some success. The choice of algorithm can influence the representation of the problem in several important ways--the discreteness of the configuration, the types of energy terms that can be used and the ability to find the global minimum energy configuration. The use of dead end elimination to design the complete sequence for a small protein motif and the use of genetic and mean-field algorithms to design hydrophobic cores for proteins represent the major themes of the past year.

From coiled coils to small globular proteins: design of a native-like three-helix bundle.

A monomolecular native-like three-helix bundle has been designed in an iterative process, beginning with a peptide that noncooperatively assembled into an antiparallel three-helix bundle. Three versions of the protein were designed in which specific interactions were incrementally added. The hydrodynamic and spectroscopic properties of the proteins were examined by size exclusion chromatography, sedimentation equilibrium, fluorescence spectroscopy, and NMR. The thermodynamics of folding were evaluated by monitoring the thermal and guanidine-induced unfolding transitions using far UV circular dichroism spectroscopy. The attainment of a unique, native-like state was achieved through the introduction of: (1) helix capping interactions; (2) electrostatic interactions between partially exposed charged residues; (3) a diverse collection of apolar side chains within the hydrophobic core.

De novo design of the hydrophobic core of ubiquitin.

We have previously reported the development and evaluation of a computational program to assist in the design of hydrophobic cores of proteins. In an effort to investigate the role of core packing in protein structure, we have used this program, referred to as Repacking of Cores (ROC), to design several variants of the protein ubiquitin. Nine ubiquitin variants containing from three to eight hydrophobic core mutations were constructed, purified, and characterized in terms of their stability and their ability to adopt a uniquely folded native-like conformation. In general, designed ubiquitin variants are more stable than control variants in which the hydrophobic core was chosen randomly. However, in contrast to previous results with 434 cro, all designs are destabilized relative to the wild-type (WT) protein. This raises the possibility that beta-sheet structures have more stringent packing requirements than alpha-helical proteins. A more striking observation is that all variants, including random controls, adopt fairly well-defined conformations, regardless of their stability. This result supports conclusions from the cro studies that non-core residues contribute significantly to the conformational uniqueness of these proteins while core packing largely affects protein stability and has less impact on the nature or uniqueness of the fold. Concurrent with the above work, we used stability data on the nine ubiquitin variants to evaluate and improve the predictive ability of our core packing algorithm. Additional versions of the program were generated that differ in potential function parameters and sampling of side chain conformers. Reasonable correlations between experimental and predicted stabilities suggest the program will be useful in future studies to design variants with stabilities closer to that of the native protein. Taken together, the present study provides further clarification of the role of specific packing interactions in protein structure and stability, and demonstrates the benefit of using systematic computational methods to predict core packing arrangements for the design of proteins.

De novo design of the hydrophobic cores of proteins.

We have developed and experimentally tested a novel computational approach for the de novo design of hydrophobic cores. A pair of computer programs has been written, the first of which creates a "custom" rotamer library for potential hydrophobic residues, based on the backbone structure of the protein of interest. The second program uses a genetic algorithm to globally optimize for a low energy core sequence and structure, using the custom rotamer library as input. Success of the programs in predicting the sequences of native proteins indicates that they should be effective tools for protein design. Using these programs, we have designed and engineered several variants of the phage 434 cro protein, containing five, seven, or eight sequence changes in the hydrophobic core. As controls, we have produced a variant consisting of a randomly generated core with six sequence changes but equal volume relative to the native core and a variant with a "minimalist" core containing predominantly leucine residues. Two of the designs, including one with eight core sequence changes, have thermal stabilities comparable to the native protein, whereas the third design and the minimalist protein are significantly destabilized. The randomly designed control is completely unfolded under equivalent conditions. These results suggest that rational de novo design of hydrophobic cores is feasible, and stress the importance of specific packing interactions for the stability of proteins. A surprising aspect of the results is that all of the variants display highly cooperative thermal denaturation curves and reasonably dispersed NMR spectra. This suggests that the non-core residues of a protein play a significant role in determining the uniqueness of the folded structure.

New strategies in protein design.

Initially, it was hoped that very simple rules could be sued to design proteins that embody all the characteristics of natural proteins. Indeed, with single-domain proteins as targets, it has been possible to design proteins that adopt the desired global fold. Yet, designed proteins with well defined structures and properties that mimic those of natural proteins remain elusive. Recent efforts in protein design have been directed toward addressing the basis for non-native characteristics in most protein designs. Although it is clear that specific tertiary interactions between all residues in a protein contribute to the final folded state, much attention has been placed on optimizing the packing of side chains in the hydrophobic core, with substantial success.

Length-encoded multiplex binding site determination: application to zinc finger proteins.

The screening of combinatorial libraries is becoming a powerful method for identifying or refining the structures of ligands for binding proteins, enzymes, and other receptors. We describe an oligonucleotide library search procedure in which the identity of each member is encoded in the length of oligonucleotides. This encoding scheme allows binding-site preferences to be evaluated via DNA length determination by denaturing gel electrophoresis. We have applied this method to determine the binding-site preferences for 18 Cys2His2 zinc finger domains as the central domain within a fixed context of flanking zinc fingers. An advantage of the method is that the relative affinities of all members of the library can be estimated in addition to simply determining the sequence of the optimal or consensus ligand. The zinc finger domain specificities determined will be useful for modular zinc finger protein design.

Structural studies of the engrailed homeodomain.

The structure of the Drosophila engrailed homeodomain has been solved by molecular replacement and refined to an R-factor of 19.7% at a resolution of 2.1 A. This structure offers a high-resolution view of an important family of DNA-binding proteins and allows comparison to the structure of the same protein bound to DNA. The most significant difference between the current structure and that of the 2.8-A engrailed-DNA complex is the close packing of an extended strand against the rest of the protein in the unbound protein. Structural features of the protein not previously noted include a "herringbone" packing of 4 aromatic residues in the core of the protein and an extensive network of salt bridges that covers much of the helix 1-helix 2 surface. Other features that may play a role in stabilizing the native state include the interaction of buried carbonyl oxygen atoms with the edge of Phe 49 and a bias toward statistically preferred side-chain dihedral angles. There is substantial disorder at both ends of the 61 amino acid protein. A 51-amino acid variant of engrailed (residues 6-56) was synthesized and shown by CD and thermal denaturation studies to be structurally and thermodynamically similar to the full-length domain.

Use of a zinc-finger consensus sequence framework and specificity rules to design specific DNA binding proteins.

We have designed three zinc-finger proteins with different DNA binding specificities. The design strategy combines a consensus zinc-finger framework sequence with previously characterized recognition regions such that the specificity of each protein is predictable. The first protein consists of three identical zinc fingers, each of which was expected to recognize the subsite GCG. This protein binds specifically to the sequence 5'-GCG-GCG-GCG-3' with a dissociation constant of approximately 11 microM. The second protein has three zinc fingers with different predicted preferred subsites. This protein binds to the predicted recognition site 5'-GGG-GCG-GCT-3' with a dissociation constant of 2 nM. Furthermore, selection experiments indicate that this is the optimal binding site. A permuted version of the second protein was also constructed and shown to preferentially recognize the corresponding permuted site 5'-GGG-GCT-GCG-3' over the non-permuted site. These results indicate that earlier observations on the specificity of zinc fingers can be extended to generalized zinc-finger structures and realize the use of zinc fingers for the design of site-specific DNA binding proteins. This consensus-based design system provides a useful model system with which to study details of zinc-finger-DNA specificity.

Toward rules relating zinc finger protein sequences and DNA binding site preferences.

Zinc finger proteins of the Cys2-His2 type consist of tandem arrays of domains, where each domain appears to contact three adjacent base pairs of DNA through three key residues. We have designed and prepared a series of variants of the central zinc finger within the DNA binding domain of Sp1 by using information from an analysis of a large data base of zinc finger protein sequences. Through systematic variations at two of the three contact positions (underlined), relatively specific recognition of sequences of the form 5'-GGGGN(G or T)GGG-3' has been achieved. These results provide the basis for rules that may develop into a code that will allow the design of zinc finger proteins with preselected DNA site specificity.

Redesigning the DNA-binding specificity of a zinc finger protein: a data base-guided approach.

A peptide corresponding to the three zinc finger domains of the human transcription factor Sp1 has been expressed and found to bind a consensus Sp1 binding site with the sequence 5'-GGGGCGGGG-3'. Examination of the amino acid distributions within a large zinc finger sequence data base and chemical arguments suggested that a particular Arg to Gln sequence change might convert binding specificity to 5'-GGGGCAGGG-3'. Experimental tests of this hypothesis revealed that such a change could be induced only when two other sequence changes, deduced from examination of sequence correlations, were made as well. These results provide the most direct information to date about how zinc finger proteins might recognize adenine-containing binding sites and bear on the existence and nature of any code between zinc finger protein and binding site sequences.