Searching sequence space by definably random mutagenesis: improving the catalytic potency of an enzyme

Ab initio models for receptor-ligand interactions in proteins. 4. Model assembly study of the catalytic mechanism of triosephosphate isomerase

The catalytic mechanism of triosephosphate isomerase (TIM) was investigated with ab initio quantum mechanical calculations. Electrostatic interactions between the quantum mechanical active site and the protein and solvent environment were modeled using the finite difference Poission-Boltzman method. The complexes of TIM with the substrate dihydroxyacetone phosphate (DHAP), five possible intermediates and the product glyceraldehyde-3-phosphate (GAP) were optimized in the active-site model at the 3-21G(*) level and energy profile for the proton abstraction from DHAP by the active-site Glu 167 was calculated at the MP2/3-21G(*)13-21G(*) level. Calculated energetics of the enzyme reaction were found to be in reasonable agreement with the experimental findings. Calculations revealed that an enediol of the substrate is a probable intermediate in the enzyme reaction. It was suggested that the proton abstracted from the substrate by the active-site glutamate goes to the carbonyl oxygen of the substrate producing enediol intermediate either directly or after it is exchanged with solvent.

Exploring the active site of chorismate mutase by combinatorial mutagenesis and selection: the importance of electrostatic catalysis

Chorismate mutase (EC 5.4.99.5) catalyzes the intramolecular rearrangement of chorismate to prephenate. Arg-90 in the active site of the enzyme from Bacillus subtilis is in close proximity to the substrate's ether oxygen and may contribute to efficient catalysis by stabilizing the presumed dipolar transition state that would result upon scission of the C--O bond. To test this idea, we have developed a novel complementation system for chorismate mutase activity in Escherichia coli by reengineering parts of the aromatic amino acid biosynthetic pathway. The codon for Arg-90 was randomized, alone and in combination with that for Cys-88, and active clones were selected. The results show that a positively charged residue either at position 88 (Lys) or 90 (Arg or Lys) is essential. Our data provide strong support for the hypothesis that the positive charge is required for stabilization of the transition state of the enzymatic chorismate rearrangement. The new selection system, in conjunction with combinatorial mutagenesis, renders the mechanism of the natural enzyme(s) accessible to further exploration and opens avenues for the improvement of first generation catalytic antibodies with chorismate mutase activity.

Second-site suppression of regulatory phosphorylation in Escherichia coli isocitrate dehydrogenase

Inactivation of Escherichia coli isocitrate dehydrogenase upon phosphorylation at S113 depends upon the direct electrostatic repulsion of the negatively charged gamma-carboxylate of isocitrate by the negatively charged phosphoserine. The effect is mimicked by replacing S113 with aspartate or glutamate, which reduce performance (kcat/K(i).isocitrat/ Km.NADP) by a factor of 10(7). Here, we demonstrate that the inactivating effects of the electrostatic repulsion are completely eliminated by a second-site mutation, and provide the structural basis for this striking example of intragenic suppression. N115 is adjacent to S113 on one face of the D-helix, interacts with isocitrate and NADP+, and has been postulated to serve in both substrate binding and in catalysis. The single N115L substitution reduces affinity for isocitrate by a factor of 50 and performance by a factor of 500. However, the N115L substitution completely suppresses the inactivating electrostatic effects of S113D or S113E: the performance of the double mutants is 10(5) higher than the S113D and S113E single mutants. These mutations have little effect on the kinetics of alternative substrates, which lack the charged gamma-carboxylate of isocitrate. Both glutamate and aspartate at site 113 remain fully ionized in the presence of leucine. In the crystal structure of the N115L mutant, the leucine adopts a different conformer from the wild-type asparagine. Repacking around the leucine forces the amino-terminus of the D-helix away from the rest of the active site. The hydrogen bond between E113 and N115 in the S113E single mutant is broken in the S113E/N115L mutant, allowing the glutamate side chain to move away from the gamma-carboxylate of isocitrate. These movements increase the distance between the carboxylates, diminish the electrostatic repulsion, and lead to the remarkably high activity of the S113E/N115L mutant.

Interactions between residues in staphylococcal alpha-hemolysin revealed by reversion mutagenesis

alpha-Hemolysin (alpha HL), a pore-forming polypeptide of 293 amino acids, is secreted by Staphylococcus aureus as a water-soluble monomer. Residues that play key roles in the formation of functional heptameric pores on rabbit red blood cells (rRBC) have been identified previously by site-directed mutagenesis. alpha HL-H35N, in which the histidine at position 35 of the wild-type sequence is replaced with asparagine, is nonlytic and is arrested in assembly as a heptameric prepore. In this study, second-site revertants of H35N that have the ability to lyse rRBC were generated by error-prone PCR under conditions designed to produce single base changes. The analysis of 22 revertants revealed new codons clustered predominantly in three distinct regions of the H35N gene. One cluster includes amino acids 107-111 (four revertants) and another residues 144-155 (five revertants). These two clusters flank the central glycine-rich loop of alpha HL, which previously has been implicated in formation of the transmembrane channel, and encompass residues Lys-110 and Asp-152 that, like His-35, are crucial for lytic activity. The third cluster lies in the region spanning amino acids 217-228 (eight revertants), a region previously unexplored by mutagenesis. Single revertants were found at amino acid positions 84 and 169. When compared with H35N, the heptameric prepores formed by the revertants underwent more rapid conversion to fully assembled pores, as determined by conformational analysis by limited proteolysis. The rate of conversion to the fully assembled pore was strongly correlated with hemolytic activity. Previous work has suggested that the N terminus of alpha HL and the central loop cooperate in the final step of assembly. The present study suggests that the key N-terminal residue His-35 operates in conjunction with residues flanking the loop and C-terminal residues in the region 217-228. Hence, reversion mutagenesis extends the linear analysis that has been provided by direct point mutagenesis.

Cytochrome P450 proteins and potential utilization in biodegradation

The cytochrome P450 enzymes are major catalysts involved in the oxidations of xenobiotic chemicals in microorganisms as well as higher animals and plants. Because of their functional roles, they offer potential in biodegradation technology. A number of microbial P450s have already been characterized and offer advantages in terms of their high catalytic rates and facile expression in microorganisms. One approach to extending the catalytic selectivity to more compounds in the environment is rational design. In three cases, the three-dimensional structures of bacterial cytochrome P450 enzymes are available and can be further understood through studies with molecular dynamics. Many mammalian cytochrome P450 enzymes have been studied extensively and have potential for biodegradation because of their broad catalytic selectivities (e.g., P450 2E1). Several advances have been made in the heterologous expression of these proteins in microorganisms. Improvements under development include electron transfer from flavodoxin and the use of cytochrome P450:NADPH-cytochrome P450 reductase fusion proteins. Random mutagenesis offers the potential of improving the catalytic activities of some of these proteins. Future challenges include the use of cytochrome P450 expression vectors in microorganisms capable of thriving in the environment; recent success in expression of vectors in Salmonella genotoxicity tester strains may be encouraging in this regard.

Termination-altering mutations in the second-largest subunit of yeast RNA polymerase III

In order to identify catalytically important amino acid changes within the second-largest subunit of yeast RNA polymerase III, we mutagenized selected regions of its gene (RET1) and devised in vivo assays for both increased and decreased transcription termination by this enzyme. Using as the reporter gene a mutant SUP4-o tRNA gene that in one case terminates prematurely and in the other case fails to terminate, we screened mutagenized RET1 libraries for reduced and increased transcription termination, respectively. The gain in suppression phenotype was in both cases scored as a reduction in the accumulation of red pigment in yeast strains harboring the ade2-1 ochre mutation. Termination-altering mutations were obtained in regions of the RET1 gene encoding amino acids 300 to 325, 455 to 486, 487 to 521, and 1061 to 1082 of the protein. In degree of amino acid sequence conservation, these range from highly variable in the first to highly conserved in the last two regions. Residues 300 to 325 yielded mainly reduced-termination mutants, while in region 1061 to 1082, increased-termination mutants were obtained exclusively. All mutants recovered, while causing gain of suppression with one SUP4 allele, brought about a reduction in suppression with the other allele, thus confirming that the phenotype is due to altered termination rather than an elevated level of transcription initiation. In vitro transcription reactions performed with extracts from several strong mutants demonstrated that the mutant polymerases respond to RNA terminator sequences in a manner that matches their in vivo termination phenotypes.

DNA shuffling by random fragmentation and reassembly: in vitro recombination for molecular evolution

Computer simulations of the evolution of linear sequences have demonstrated the importance of recombination of blocks of sequence rather than point mutagenesis alone. Repeated cycles of point mutagenesis, recombination, and selection should allow in vitro molecular evolution of complex sequences, such as proteins. A method for the reassembly of genes from their random DNA fragments, resulting in in vitro recombination is reported. A 1-kb gene, after DNase I digestion and purification of 10- to 50-bp random fragments, was reassembled to its original size and function. Similarly, a 2.7-kb plasmid could be efficiently reassembled. Complete recombination was obtained between two markers separated by 75 bp; each marker was located on a separate gene. Oligonucleotides with 3' and 5' ends that are homologous to the gene can be added to the fragment mixture and incorporated into the reassembled gene. Thus, mixtures of synthetic oligonucleotides and PCR fragments can be mixed into a gene at defined positions based on homology. As an example, a library of chimeras of the human and murine genes for interleukin 1 beta has been prepared. Shuffling can also be used for the in vitro equivalent of some standard genetic manipulations, such as a backcross with parental DNA. The advantages of recombination over existing mutagenesis methods are likely to increase with the numbers of cycles of molecular evolution.

Rapid evolution of a protein in vitro by DNA shuffling

DNA shuffling is a method for in vitro homologous recombination of pools of selected mutant genes by random fragmentation and polymerase chain reaction (PCR) reassembly. Computer simulations called genetic algorithms have demonstrated the importance of iterative homologous recombination for sequence evolution. Oligonucleotide cassette mutagenesis and error-prone PCR are not combinatorial and thus are limited in searching sequence space. We have tested mutagenic DNA shuffling for molecular evolution in a beta-lactamase model system. Three cycles of shuffling and two cycles of backcrossing with wild-type DNA, to eliminate non-essential mutations, were each followed by selection on increasing concentrations of the antibiotic cefotaxime. We report here that selected mutants had a minimum inhibitory concentration of 640 micrograms ml-1, a 32,000-fold increase and 64-fold greater than any published TEM-1 derived enzyme. Cassette mutagenesis and error-prone PCR resulted in only a 16-fold increase.

Phage display of enzymes and in vitro selection for catalytic activity

Despite recent progress, our understanding of enzymes remains limited: the prediction of the changes that should be introduced to alter their properties or catalytic activities in an expected direction remains difficult. An alternative to rational design is selection of mutants endowed with the anticipated properties from a large collection of possible solutions generated by random mutagenesis. We describe here a new technique of in vitro selection of genes on the basis of the catalytic activity of the encoded enzymes. The gene coding for the enzyme to be engineered is cloned into the genome of a filamentous phage, whereas the enzyme itself is displayed on its surface, creating a phage enzyme. A bifunctional organic label containing a suicide inhibitor of the enzyme and a ligand with high affinity for an immobilized receptor are constructed. On incubation of a mixture of phage enzymes, those phages showing an activity on the inhibitor under the conditions of the experiment are labeled. These phages can be recovered by affinity chromatography. The design of the label and the factors controlling the selectivity of the selection are analyzed. The advantages of the technique and its scope in terms of the enzymes that can be engineered are discussed.

Microbial systems and directed evolution of protein activities

Recent advances in recombinant DNA methodology have had an important impact on the capacity to manipulate protein-coding sequences. The appearance of new, powerful screening systems completes a scenario for conducting directed evolution experiments. We review here some of the latest developments in experimental approaches to directed evolution, utilizing microbial systems. These include phage display, surface display, operator-repressor systems, and novel mutagenesis approaches. We also highlight the achievements and limitations of current methodologies. We present strategies used by our own group that permitted isolation of specificity mutants of beta-lactamase. Possible improvements for the future of the variation-selection approach to the study and manipulation of proteins are presented.

Tuning the activity of an enzyme for unusual environments: sequential random mutagenesis of subtilisin E for catalysis in dimethylformamide

Random mutagenesis has been used to engineer the protease subtilisin E to function in a highly nonnatural environment--high concentrations of a polar organic solvent. Sequential rounds of mutagenesis and screening have yielded a variant (PC3) that hydrolyzes a peptide substrate 256 times more efficiently than wild-type subtilisin in 60% dimethylformamide. PC3 subtilisin E and other variants containing different combinations of amino acid substitutions are effective catalysts for transesterification and peptide synthesis in dimethylformamide and other organic media. Starting with a variant containing four effective amino acid substitutions (D60N, D97G, Q103R, and N218S; where, for example, D60N represents Asp-60-->Asn), six additional mutations (G131D, E156G, N181S, S182G, S188P, and T255A) were generated during three sequential rounds of mutagenesis and screening. The 10 substitutions are clustered on one face of the enzyme, near the active site and substrate binding pocket, and all are located in loops that connect core secondary structure elements and exhibit considerable sequence variability in subtilisins from different sources. These variable surface loops are effective handles for "tuning" the activity of subtilisin. Seven of the 10 amino acid substitutions in PC3 are found in other natural subtilisins. Great variability is exhibited among naturally occurring sequences that code for similar three-dimensional structures--it is possible to make use of this sequence flexibility to engineer enzymes to exhibit features not previously developed (or required) for function in vivo.

Structure/function analysis of the Ala116-->Lys121 region of endonuclease V by random targeted mutagenesis

Endonuclease V is the product of the denV gene of bacteriophage T4 and is responsible for the recognition and repair of pyrimidine dimers due to UV irradiation of DNA. This is accomplished by a two-step mechanism involving incision at the site of the lesion followed by cleavage of the phosphate backbone. In order to better understand this molecule, and to validate our new mutagenesis procedure, we have constructed a series of random mutations within the region Ala116-->Lys121 using a random targeted mutagenesis procedure developed for this study. The results presented here suggest an important role for this region in the stabilization of the thymine dimer-containing substrate. These mutants also confirm a direct correlation between survival and both DNA binding and pyrimidine dimer-DNA glycosylase activity. No such correlation exists between survival and AP lyase activity. The results are consistent with the recently published X-ray crystal structure.

An algorithm for protein engineering: simulations of recursive ensemble mutagenesis

An algorithm for protein engineering, termed recursive ensemble mutagenesis, has been developed to produce diverse populations of phenotypically related mutants whose members differ in amino acid sequence. This method uses a feedback mechanism to control successive rounds of combinatorial cassette mutagenesis. Starting from partially randomized "wild-type" DNA sequences, a highly parallel search of sequence space for peptides fitting an experimenter's criteria is performed. Each iteration uses information gained from the previous rounds to search the space more efficiently. Simulations of the technique indicate that, under a variety of conditions, the algorithm can rapidly produce a diverse population of proteins fitting specific criteria. In the experimental analog, genetic selection or screening applied during recursive ensemble mutagenesis should force the evolution of an ensemble of mutants to a targeted cluster of related phenotypes.

Optimizing Nucleotide Mixtures to Encode Specific Subsets of Amino Acids for Semi-Random Mutagenesis

In random mutagenesis, synthesis of an NNN triplet (i.e. equiprobable A, C, G, and T at each of the three positions in the codon) could be considered an optimal nucleotide mixture because all 20 amino acids are encoded. NN(G,C) might be considered a slightly more intelligent "dope" because the entire set of amino acids is still encoded using only half as many codons. Using a general algorithm described herein, it is possible to formulate more complex doping schemes which encode specific subsets of the twenty amino acids, excluding others from the mix. Maximizing the equiprobability of amino acid residues contributing to such a subset is suggested as an optimal basis for performing semi-random mutagenesis. This is important for reducing the nucleotide complexity of combinatorial cassettes so that "sequence space" can be searched more efficiently. Computer programs have been developed to provide tables of optimized dopes compatible with automated DNA synthesizers.

In vivo catalysis of a metabolically essential reaction by an antibody

We have established a growth selection requirement for a catalytic antibody with modest chorismate mutase activity. Conversion of (-)-chorismate into prephenate is the key step in the biosynthesis of the aromatic amino acids tyrosine and phenylalanine. Strains of the yeast Saccharomyces cerevisiae containing an insertion mutation in the structural gene for the enzyme chorismate mutase (EC 5.4.99.5) require exogenous supplements of these two amino acids for efficient growth. Intracellular expression of the heterologous antibody catalyst in one such strain, identified by random mutagenesis and genetic selection, provides a substantial growth advantage under auxotrophic conditions; complementation was not observed with an unrelated esterolytic antibody. In addition to demonstrating that tailored immunoglobulin catalysts can carry out vital biochemical reactions in vivo, these experiments provide a powerful selection assay for identifying genetic changes within the antibody molecule itself that augment chemical efficiency.

Selection of small molecules by the Tetrahymena catalytic center

The catalytic center in group I RNAs contains a selective binding site that accommodates both guanosine and L-arginine. In order to understand the specificity of the RNA for small molecules, we analyzed 6 RNAs that vary in this region. Specificity for nucleotides resides substantially in G264 rather than its paired nucleotide C311, and is expressed substantially in Km, with comparatively little variation in kcat. kcat is not notably perturbed even for RNAs with mispairs in the active-site helix. For 5 of 6 sequences, effects of RNA substitutions on arginine binding and GTP reactivity are proportional, confirming that arginine contacts a subset of the groups occupied by G. As a result of particular mutations, reaction with GTP is decreased, and reaction with the natural nucleotides UTP and ATP is enhanced. Molecular modeling of these effects suggests that exceptionally flexible placement of reactants may be an essential quality of RNA-catalyzed splicing. The specificity of the intron can be rationalized by a type of binding model not previously considered, in which the G/arginine site includes adjacent nucleotides (an 'axial' site), rather than a single nucleotide, G264.

The adaptability of the active site of trypanosomal triosephosphate isomerase as observed in the crystal structures of three different complexes

Crystals of triosephosphate isomerase from Trypanosoma brucei brucei have been used in binding studies with three competitive inhibitors of the enzyme's activity. Highly refined structures have been deduced for the complexes between trypanosomal triosephosphate isomerase and a substrate analogue (glycerol-3-phosphate to 2.2 A), a transition state analogue (3-phosphonopropionic acid to 2.6 A), and a compound structurally related to both (3-phosphoglycerate to 2.2 A). The active site structures of these complexes were compared with each other, and with two previously determined structures of triosephosphate isomerase either free from inhibitor or complexed with sulfate. The comparison reveals three conformations available to the "flexible loop" near the active site of triosephosphate isomerase: open (no ligand), almost closed (sulfate), and fully closed (phosphate/phosphonate complexes). Also seen to be sensitive to the nature of the active site ligand is the catalytic residue Glu-167. The side chain of this residue occupies one of two discrete conformations in each of the structures so far observed. A "swung out" conformation unsuitable for catalysis is observed when sulfate, 3-phosphoglycerate, or no ligand is bound, while a "swung in" conformation ideal for catalysis is observed in the complexes with glycerol-3-phosphate or 3-phosphonopropionate. The water structure of the active site is different in all five structures. The results are discussed with respect to the triosephosphate isomerase structure function relationship, and with respect to an on-going drug design project aimed at the selective inhibition of glycolytic enzymes of T. brucei.

Substrate specificity of trypsin investigated by using a genetic selection

The structural determinants of the primary substrate specificity of rat anionic trypsin were examined by using oligonucleotide-directed mutagenesis coupled to a genetic selection. A library was created that encoded trypsins substituted at amino acid positions 189 and 190 at the base of the substrate binding pocket. A genetic selection, with a dynamic range of 5 orders of proteolytic activity, was used to search 90,000 transformants of the library. Rapid screening for arginyl amidolysis and esterolysis confirmed the activity of the purified isolates. Trypsin and 15 mutant trypsins with partially preserved function were identified and characterized kinetically on arginyl and lysyl peptide substrates. Alternative arrangements of amino acids in the substrate binding pocket sustained efficient catalysis. A negative charge at amino acid position 189 or 190 was shown to be essential for high-level catalysis. With the favored aspartic acid residue at position 189, several amino acids could replace serine at position 190. Modulation of the specificity for arginine and lysine substrates was shown to depend on the amino acid at position 190. The regulatory effect of the amino acid side chain at position 190 on the substrate specificity is also reflected in substrate binding pockets of naturally occurring trypsin homologs.

Peptides on phage: a vast library of peptides for identifying ligands

We have constructed a vast library of peptides for finding compounds that bind to antibodies and other receptors. Millions of different hexapeptides were expressed at the N terminus of the adsorption protein (pIII) of fd phage. The vector fAFF1, derived from the tetracycline resistance-transducing vector fd-tet, allows cloning of oligonucleotides in a variety of locations in the 5' region of gene III. A library of 3 x 10(8) recombinants was generated by cloning randomly synthesized oligonucleotides. The library was screened for high-avidity binding to a monoclonal antibody (3-E7) that is specific for the N terminus of beta-endorphin (Tyr-Gly-Gly-Phe). Fifty-one clones selected by three rounds of the affinity purification technique called panning were sequenced and found to differ from previously known ligands for this antibody. The striking finding is that all 51 contained tyrosine as the N-terminal residue and that 48 contained glycine as the second residue. The binding affinities of six chemically synthesized hexapeptides from this set range from 0.35 microM (Tyr-Gly-Phe-Trp-Gly-Met) to 8.3 microM (Tyr-Ala-Gly-Phe-Ala-Gln), compared with 7.1 nM for a known high-affinity ligand (Tyr-Gly-Gly-Phe-Leu). These results show that ligands can be identified with no prior information concerning antibody specificity. Peptide libraries are also likely to be useful in finding ligands that bind to other classes of receptors and in discovering pharmacologic agents.