Potential of genetic algorithms in protein folding and protein engineering simulations

Genetic algorithms are very efficient search mechanisms whichmutate, recombine and select amongst tentative solutions toa problem until a near optimal one is achieved. We introducethem as a new tool to study proteins. The identification andmotivation for different fitness functions is discussed. Theevolution of the zinc finger sequence motif from a random startis modelled. User specified changes of the repressor structurewere simulated and critical sites and exchanges for mutagenesisidentified. Vast conformational spaces are efficiently searchedas illustrated by the ab initio folding of a model protein ofa four ß strand bundle. The genetic algorithm simulationwhich mimicked important folding constraints as overall hydrophobicpackaging and a propensity of the betaphilic residues for transpositions achieved a unique fold. Cooperativity in the ßstrand regions and a length of 3–5 for the interconnectingloops was critical. Specific interaction sites were considerablyless effective in driving the fold.     

Directed evolution converts subtilisin E into a functional equivalent of thermitase

We used directed evolution to convert Bacillus subtilis subtilisinE into an enzyme functionally equivalent to its thermophilichomolog thermitase from Thermoactinomyces vulgaris. Five generationsof random mutagenesis, recombination and screening created subtilisinE 5-3H5, whose half-life at 83°C (3.5 min) and temperatureoptimum for activity (T_opt, 76°C) are identical with thoseof thermitase. The T_opt of the evolved enzyme is 17°C higherand its half-life at 65°C is >200 times that of wild-typesubtilisin E. In addition, 5-3H5 is more active towards thehydrolysis of succinyl-Ala-Ala-Pro-Phe-p-nitroanilide than wild-typeat all temperatures from 10 to 90°C. Thermitase differsfrom subtilisin E at 157 amino acid positions. However, onlyeight amino acid substitutions were sufficient to convert subtilisinE into an enzyme equally thermostable. The eight substitutions,which include known stabilizing mutations (N218S, N76D) andalso several not previously reported, are distributed over thesurface of the enzyme. Only two (N218S, N181D) are found inthermitase. Directed evolution provides a powerful tool to unveilmechanisms of thermal adaptation and is an effective and efficientapproach to increasing thermostability without compromisingenzyme activity.     

Biased mutation-assembling: an efficient method for rapid directed evolution through simultaneous mutation accumulation

We have developed an efficient optimization technique, 'biased mutation-assembling', for improving protein properties such as thermostability. In this strategy, a mutant library is constructed using the overlap extension polymerase chain reaction technique with DNA fragments from wild-type and phenotypically advantageous mutant genes, in which the number of mutations assembled in the wild-type gene is stochastically controlled by the mixing ratio of the mutant DNA fragments to wild-type fragments. A high mixing ratio results in a mutant composition biased to favor multiple-point mutants. We applied this strategy to improve the thermostability of prolyl endopeptidase from Flavobacterium meningosepticum as a case study and found that the proportion of thermostable mutants in a library increased as the mixing ratio was increased. If the proportion of thermostable mutants increases, the screening effort needed to find them should be reduced. Indeed, we isolated a mutant with a 1200-fold longer activity half-life at 60 degrees C than that of wild-type prolyl endopeptidase after screening only 2000 mutants from a library prepared with a high mixing ratio. Our results indicate that an aggressive accumulation of advantageous mutations leads to an increase in the quality of the mutant library and a reduction in the screening effort required to find superior mutants.     

Alteration of product specificity of Aeropyrum pernix farnesylgeranyl diphosphate synthase (Fgs) by directed evolution

Directed evolution of the C25 farnesylgeranyl diphosphate synthase of Aeropyrum pernix (Fgs) was carried out by error-prone PCR with an in vivo color complementation screen utilizing carotenoid biosynthetic pathway enzymes. Screening yielded 12 evolved clones with C20 geranylgeranyl diphosphate synthase activity which were isolated and characterized in order to understand better the chain elongation mechanism of this enzyme. Analysis of these mutants revealed three different mechanisms of product chain length specificity. Two mutants (A64T and A64V) have a single mutation at the 8th amino acid upstream of a conserved first aspartate-rich motif (FARM), which is involved in the mechanism for chain elongation reaction of all prenyl diphosphate synthases. One mutant (A135T) carries a single mutation at the 7th amino acid upstream of another conserved region (141GQ142), which was recently found to be another important region controlling chain elongation of a type III C20 geranylgeranyl diphosphate synthase and Escherichia coli C15 farnesyl diphosphate synthase. Finally, one mutant carrying four mutations (V84I, H88R, I177 M and M191V) is of interest. Molecular modeling, site-directed mutagenesis and in vitro assays of this mutant suggest that product chain-length distribution can be also controlled by a structural change provoked by a cooperative interaction of amino acids.     

Affinity maturation of a Taq DNA polymerase specific affibody by helix shuffling

The possibility of increasing the affinity of a Taq DNA polymerasespecific binding protein (affibody) was investigated by an -helixshuffling strategy. The primary affibody was from a naive combinatoriallibrary of the three-helix bundle Z domain derived from staphylococcalprotein A. A hierarchical library was constructed through selectivere-randomization of six amino acid positions in one of the two-helices of the domain, making up the Taq DNA polymerase bindingsurface. After selections using monovalent phage display technology,second generation variants were identified having affinities(K_D) for Taq DNA polymerase in the range of 30–50 nM asdetermined by biosensor technology. Analysis of binding dataindicated that the increases in affinity were predominantlydue to decreased dissociation rate kinetics. Interestingly,the affinities observed for the second generation Taq DNA polymerasespecific affibodies are of similar strength as the affinitybetween the original protein A domain and the Fc domain of humanimmunoglobulin G. Further, the possibilities of increasing theapparent affinity through multimerization of affibodies wasdemonstrated for a dimeric version of one of the second generationaffibodies, constructed by head-to-tail gene fusion. As comparedwith its monomeric counterpart, the binding to sensor chip immobilizedTaq DNA polymerase was characterized by a threefold higher apparentaffinity, due to slower off-rate kinetics. The results showthat the binding specificity of the protein A domain can bere-directed to an entirely different target, without loss ofbinding strength.     

Directed evolution of RuBisCO hypermorphs through genetic selection in engineered E.coli

The Calvin Cycle is the primary conduit for the fixation of carbon dioxide into the biosphere; ribulose 1,5-bisphosphate carboxylase/oxygenase (RuBisCO) catalyzes the rate-limiting fixation step. Our goal is to direct the evolution of RuBisCO variants with improved kinetic and biophysical properties. The Calvin Cycle was partially reconstructed in Escherichia coli; the engineered strain requires the Synechococcus PCC6301 RuBisCO for growth in minimal media supplemented with a pentose. We randomly mutated the gene encoding the large subunit of RuBisCO (rbcL), co-expressed the resulting library with the small subunit (rbcS) and the Synechococcus PCC7492 phosphoribulokinase (prkA), and selected hypermorphic variants. The RuBisCO variants that evolved during three rounds of random mutagenesis and selection were over-expressed, and exhibited 5-fold improvement in specific activity relative to the wild-type enzyme. These results demonstrate a new strategy for the artificial selection of RuBisCO and other non-native metabolic enzymes.     

Improved mutants from directed evolution are biased to orthologous substitutions

We have engineered human epidermal growth factor (EGF) by directed evolution through yeast surface display for significantly enhanced affinity for the EGF receptor (EGFR). Statistical analysis of improved EGF mutants isolated from randomly mutated yeast-displayed libraries indicates that mutations are biased towards substitutions at positions exhibiting significant phylogenetic variation. In particular, mutations in high-affinity EGF mutants are statistically biased towards residues found in orthologous EGF species. This same trend was also observed with other proteins engineered through directed evolution in our laboratory (EGFR, interleukin-2) and in a meta-analysis of reported results for engineered subtilisin. By contrast, reported loss-of-function mutations in EGF were biased towards highly conserved positions. Based on these findings, orthologous mutations were introduced into a yeast-displayed EGF library by a process we term shotgun ortholog scanning mutagenesis (SOSM). EGF mutants with a high frequency of the introduced ortholog mutations were isolated through screening the library for enhanced binding affinity to soluble EGFR ectodomain. These mutants possess a 30-fold increase in binding affinity over wild-type EGF to EGFR-transfected fibroblasts and are among the highest affinity EGF proteins to be engineered to date. Collectively, our findings highlight a general approach for harnessing information present in phylogenetic variability to create useful genetic diversity for directed evolution. Our SOSM method exploits the benefits of library diversity obtained through complementary methods of error-prone PCR and DNA shuffling, while circumventing the need for acquisition of multiple genes for family or synthetic shuffling.     

A system based on specific protein-RNA interactions for analysis of target protein-protein interactions in vitro: successful selection of membrane-bound Bak-Bcl-xL proteins in vitro

Ribosome display systems are very effective and powerful tools for in vitro screening of transcribed mRNAs that encode proteins (or peptides) with specific (known or unknown) functions. We have modified such a system by exploiting the interaction between a tandemly fused MS2 coat-protein (MSp) dimer and the RNA sequence of the corresponding specific binding motif, C-variant (or Cv). We placed the MSp dimer at the N-terminus of a nascent protein and the Cv binding motif was attached to the 5' end of the protein's mRNA. This configuration enhanced the stability of the ribosome-mRNA complex. We demonstrate here that this improved ribosome display system provides an effective method for identifying the gene for a protein that binds to a protein of interest. We visualized the formation of polysome complexes in this advanced polysome display by atomic force microscopy (AFM) and found that the AFM images of polysomes in our system were different from those observed in the case of conventional ribosome display systems. Our results suggest that our technology might usefully complement yeast two-hybrid assays.     

Directed evolution of Tk-subtilisin from a hyperthermophilic archaeon: identification of a single amino acid substitution responsible for low-temperature adaptation

Tk-subtilisin from the hyperthermophilic archaeon Thermococcus kodakaraensis is synthesized in a prepro-form (prepro-Tk-subtilisin), secreted in a pro-form (pro-Tk-subtilisin), and matured to an active form (mat-Tk-subtilisin*; a Ca(2+)-bound active form of matured domain of Tk-subtilisin) upon autoprocessing and degradation of the propeptide [Tk-propeptide; propeptide of Tk-subtilisin (Gly1-Leu69)]. Pro-Tk-subtilisin exhibited halo-forming activity only at 80 degrees C, but not at 70 and 60 degrees C, because Tk-propeptide is not effectively degraded by mat-Tk-subtilisin* and forms an inactive complex with mat-Tk-subtilisin* at <80 degrees C. Random mutagenesis in the entire prepro-Tk-subtilisin gene, followed by screening for mutant proteins with halo-forming activity at 70 and 60 degrees C, allowed us to identify single Gly56 --> Ser mutation in the propeptide region responsible for low-temperature adaptation of pro-Tk-subtilisin. SDS-PAGE analyses and mat-Tk-subtilisin* activity assay of pro-G56S-subtilisin indicated more rapid maturation than pro-Tk-subtilisin. The resultant active form was indistinguishable from mat-Tk-subtilisin* in activity and stability, indicating that Gly56 --> Ser mutation does not seriously affect the folding of the mature domain. However, this mutation greatly destabilized the propeptide, making it unstructured in an isolated form. As a result, Tk-propeptide with Gly56 --> Ser mutation (G56S-propeptide) was more susceptible to proteolytic degradation and less effectively inhibited mat-Tk-subtilisin* activity than Tk-propeptide. These results suggest that pro-G56S-subtilisin is more effectively matured than pro-Tk-subtilisin at lower temperatures, because autoprocessed G56S-propeptide is unstructured upon dissociation from mat-Tk-subtilisin* and is therefore effectively degraded by mat-Tk-subtilisin*.     

In vitro DNA recombination by L-Shuffling during ribosome display affinity maturation of an anti-Fas antibody increases the population of improved variants

The use of random mutagenesis in concert with protein display technologies to rapidly select high affinity antibody variants is an established methodology. In some cases, DNA recombination has been included in the strategy to enable selection of mutations which act cooperatively to improve antibody function. In this study, the impact of L-Shuffling DNA recombination on the eventual outcome of an in vitro affinity maturation has been experimentally determined. Parallel evolution strategies, with and without a recombination step, were carried out and both methods improved the affinity of an anti-Fas single chain variable fragment (scFv). The recombination step resulted in an increased population of affinity-improved variants. Moreover, the most improved variant, with a 22-fold affinity gain, emerged only from the recombination-based approach. An analysis of mutations preferentially selected in the recombined population demonstrated strong cooperative effects when tested in combination with other mutations but small, or even negative, effects on affinity when tested in isolation. These results underline the ability of combinatorial library approaches to explore very large regions of sequence space to find optimal solutions in antibody evolution studies.     

Directed evolution of O6-alkylguanine-DNA alkyltransferase for applications in protein labeling

The specific reaction of O6-alkylguanine-DNA alkyltransferase (AGT) with O6-benzylguanine (BG) derivatives allows for a specific labeling of AGT fusion proteins with chemically diverse compounds in living cells and in vitro. The efficiency of the labeling depends on a number of factors, most importantly on the reactivity, selectivity and stability of AGT. Here, we report the use of directed evolution and two different selection systems to further increase the activity of AGT towards BG derivatives by a factor of 17 and demonstrate the advantages of this mutant for the specific labeling of AGT fusion proteins displayed on the surface of mammalian cells. The results furthermore identify two regions of the protein outside the active site that influence the activity of the protein towards BG derivatives.     

Development of a screening platform for directed evolution using the reef coral fluorescent protein ZsGreen as a solubility reporter

Soluble proteins, with high expression levels, are preferred candidates for structural and functional studies. In cases of low expression, aggregation or inclusion body formation, time-consuming searches for optimal expression or refolding conditions are required. We have developed a high-throughput solubility engineering and screening platform for proteins that are expressed in an insoluble form in Escherichia coli with the aim of obtaining a broad spectrum of best hits with increased solubility in difficult to express target proteins. This process has been developed using error-prone PCR to introduce random base changes in genes of interest. Expression of mutated proteins in fusion with the reef coral fluorescent protein ZsGreen as a solubility marker has enabled the selection of more soluble variants. We have used a colony picker to achieve high-throughput selection of E.coli expressing more soluble target protein-ZsGreen fusions, with increased fluorescence. The whole process enables us to complete one round of mutation, screening and analysis of 20,000 potential soluble clones within approximately 8 weeks. We describe the development of the methods using different model proteins and show one example, the kinase domain from the human EphB2 receptor, as a successful application of the whole platform.     

Directed evolution of a type I antifreeze protein expressed in Escherichia coli with sodium chloride as selective pressure and its effect on antifreeze tolerance

Both freezing tolerance and NaCI tolerance are improved whenantifreeze proteins are expressed as fusion proteins with twodomains of staphylococcal protein A (SPA) in Escherichia coli.To characterize these properties further we created a randomlymutated expression library in E.coli, based on the winter flounderantifreeze protein HPLC-8 component gene. Low-fidelity PCR productsof this gene were fused to the spa gene encoding two domainsof the SPA. The library was screened for enhanced NaCl toleranceand four clones were selected. The freezing tolerance of eachof the selected clones was enhanced to varying extents. DNAsequencing of the isolated mutants revealed that the amphiphilicproperties of the native antifreeze protein were essentiallyconserved. Furthermore, by studying the primary sequence ofthe randomly mutated clones, in comparison with the degree offreezing tolerance, we have identified clues which help in understandingthe relationship between salt and freezing tolerance.     

Construction and optimization of a CC49-based scFv-beta-lactamase fusion protein for ADEPT

CC49 is a clinically validated antibody with specificity for TAG-72, a carbohydrate epitope that is over-expressed and exposed on a large fraction of solid malignancies. We constructed a single chain fragment (scFv) based on CC49 and fused it to beta-lactamase. The first generation fusion protein, TAB2.4, was expressed at low levels in Escherichia coli and significant degradation was observed during production. We optimized the scFv domain of TAB2.4 by Combinatorial Consensus Mutagenesis (CCM). An improved variant TAB2.5 was identified that resulted in an almost 4-fold improved expression and 2.5 degrees higher thermostability relative to its parent molecule. Soluble TAB2.5 can be manufactured in low-density E.coli cultures at 120 mg/l. Our studies suggest that CCM is a rapid and efficient method to generate antibody fragments with improved stability and expression. The fusion protein TAB2.5 can be used for antibody directed enzyme prodrug therapy (ADEPT).     

Comparison of solvent-inaccessible cores of homologous proteins: definitions useful for protein modelling

The three-dimensional structures of 41 homologous proteins (belongingto eight families) were compared by pairwise superposition.A subset of "core" residues was defined as those whose sidechains have <7% of their surface exposed to solvent. Thissubset has significantly higher sequence identity and lowerroot mean square (RMS) carbon separation than for all topologicallyequivalent residues in the structure, when members of a proteinfamily are superposed. For such superpositions the relationshipbetween RMS distance and percentage sequence identity of thissubset of residues is similar to that for all equivalent residues,although some variation is observed between families of proteinswhich are predominantly ß sheet and those which aremainly helix. The definition of a structurally more conservedcore may be ueful in model building proteins from an homologousfamily. The RMS differences of coordinates of structures ofproteins with identical sequences are found to be related tothe resolutions of the structures.     

Multivariate-activity mining for molecular quasi-species in a glutathione transferase mutant library

A library of recombinant glutathione transferases (GSTs) generated by shuffling of DNA encoding human GST M1-1 and GST M2-2 was screened with eight alternative substrates, and the activities were subjected to multivariate analysis. Assays were made in lysates of bacteria in which the GST variants had been expressed. The primary data showed clustering of the activities in eight-dimensional substrate-activity space. For an incisive analysis, the rows of the data matrix, corresponding to the different enzyme variants, were individually scaled to unit length, thus accounting for different expression levels of the enzymes. The columns representing the activities with alternative substrates were subsequently individually normalized to unit variance and a zero mean. By this standardization, the data were adjusted to comparable orders of magnitude. Three molecular quasi-species were recognized by multivariate K-means and principal component analyses. Two of them encompassed the parental GST M1-1 and GST M2-2. A third one diverged functionally by displaying enhanced activities with some substrates and suppressed activities with signature substrates for GST M1-1 and GST M2-2. A fourth cluster contained mutants with impaired functions and was not regarded as a quasi-species. Sequence analysis of representatives of the mutant clusters demonstrated that the majority of the variants in the diverging novel quasi-species were structurally similar to the M1-like GSTs, but distinguished themselves from GST M1-1 by a Ser to Thr substitution in the active site. The data show that multivariate analysis of functional profiles can identify small structural changes influencing the evolution of enzymes with novel substrate-activity profiles.     

Understanding protein lids: structural analysis of active hinge mutants in triosephosphate isomerase

The conformational switch from open to closed of the flexible loop 6 of triosephosphate isomerase (TIM) is essential for the catalytic properties of TIM. Using a directed evolution approach, active variants of chicken TIM with a mutated C-terminal hinge tripeptide of loop 6 have been generated (Sun,J. and Sampson,N.S., Biochemistry, 1999, 38, 11474-11481). In chicken TIM, the wild-type C-terminal hinge tripeptide is KTA. Detailed enzymological characterization of six variants showed that some of these (LWA, NPN, YSL, KTK) have decreased catalytic efficiency, whereas others (KVA, NSS) are essentially identical with wild-type. The structural characterization of these six variants is reported. No significant structural differences compared with the wild-type are found for KVA, NSS and LWA, but substantial structural adaptations are seen for NPN, YSL and KTK. These structural differences can be understood from the buried position of the alanine side chain in the C-hinge position 3 in the open conformation of wild-type loop 6. Replacement of this alanine with a bulky side chain causes the closed conformation to be favored, which correlates with the decreased catalytic efficiency of these variants. The structural context of loop 6 and loop 7 and their sequence conservation in 133 wild-type sequences is also discussed.     

Physicochemical factors for discriminating between soluble and membrane proteins: hydrophobicity of helical segments and protein length

The average hydrophobicity of a polypeptide segment is consideredto be the most important factor in the formation of transmembranehelices, and the partitioning of the most hydrophobic (MH) segmentinto the alternative nonpolar environment, a membrane or hydrophobiccore of a globular protein may determine the type of proteinproduced. In order to elucidate the importance of the MH segmentin determining which of the two types of protein results froma given amino acid sequence, we statistically studied the characteristicsof MH helices, longer than 19 residues in length, in 97 membraneproteins whose three-dimensional structure or topology is known,as well as 397 soluble proteins selected from the Protein DataBank. The average hydrophobicity of MH helices in membrane proteinshad a characteristic relationship with the length of the protein.All MH helices in membrane proteins that were longer than 500residues had a hydrophobicity greater than 1.75 (Kyte and Doolittlescale), while the MH helices in membrane proteins smaller than100 residues could be as hydrophilic as 0.1. The possibilityof developing a method to discriminate membrane proteins fromsoluble ones, based on the effect of size on the type of proteinproduced, is discussed.     

Direct expression in E.coli of a functionally active protein A-snake toxin fusion protein

We constructed a recombinant expression plasmid encoding a proteinA–neurotoxin fusion protein. The fused toxin is directlyexpressed in the periplasmic space of Escherichia coli and canbe purified in the milligram range by a single immuno-affinitystep. The LD₅₀ values of the fused toxin and native toxin are130 and 20 nmol/kg mouse respectively. The K_d values characterizingtheir binding to the nicotinic acetylcholine receptor (AcChoR)are respectively 4.8 ± 0.8 and 0.07 ± 0.03 nM.In contrast, the fused and native toxins are equally well recognizedby a toxin-specific monoclonal antibody which recognizes theAcChoR binding site. The lower toxicity of the fused toxin mightresult, therefore, from a steric hindrance, due to the presenceof the bulky protein A moiety (mol. wt = 31 kd) rather thanto a direct alteration of the ‘toxic’ site. Thefused toxin is more immunogenic than native toxin, since 1 nmolof hybrid toxin and 14 nmol of native toxin give rise to comparabletiters of antitoxin antibodies which, furthermore, are equallypotent at neutralizing neurotoxicity. The work described inthis paper shows that the use of fused toxins may be of paramountimportance for future development of serotherapy against envenomationby snake bites.