Screening combinatorial antibody libraries for catalytic acyl transfer reactions

A bacteriophage lambda vector system for the expression of Fab fragments from the mouse antibody repertoire in Escherichia coli has been described. We have used this system to generate a catalytic antibody from a combinatorial antibody library. Monoclonal antibody 43C9 was raised against a transition state analogue of the hydrolysis of carboxyamide. mRNA from hybridoma cells expressing this antibody was cloned into phage lambda and clones that expressed the mRNA for either the heavy or the light chain of the antibody were isolated. These individual libraries were then crossed to generate a combinatorial library in which clones coexpressed the heavy and light chains. This library was screened for antibodies/Fab fragments that bound to the original antigen with high affinity. DNA sequencing showed that these fragments were the same as those in antibody 43C9. Three different clones were found to catalyse the hydrolysis of carboxyamide. More efficient expression vectors and improved screening techniques should lead to the isolation of many more catalytic antibodies from combinatorial antibody libraries.

Replisome-mediated DNA replication

The elaborate process of genomic replication requires a large collection of proteins properly assembled at a DNA replication fork. Several decades of research on the bacterium Escherichia coli and its bacteriophages T4 and T7 have defined the roles of many proteins central to DNA replication. These three different prokaryotic replication systems use the same fundamental components for synthesis at a moving DNA replication fork even though the number and nature of some individual proteins are different and many lack extensive sequence homology. The components of the replication complex can be grouped into functional categories as follows: DNA polymerase, helix destabilizing protein, polymerase accessory factors, and primosome (DNA helicase and DNA primase activities). The replication of DNA derives from a multistep enzymatic pathway that features the assembly of accessory factors and polymerases into a functional holoenzyme; the separation of the double-stranded template DNA by helicase activity and its coupling to the primase synthesis of RNA primers to initiate Okazaki fragment synthesis; and the continuous and discontinuous synthesis of the leading and lagging daughter strands by the polymerases. This review summarizes and compares and contrasts for these three systems the types, timing, and mechanism of reactions and of protein-protein interactions required to initiate, control, and coordinate the synthesis of the leading and lagging strands at a DNA replication fork and comments on their generality.

A zinc ribbon protein in DNA replication: primer synthesis and macromolecular interactions by the bacteriophage T4 primase

The gene product 61 primase protein from bacteriophage T4 was expressed as an intein fusion and purified to homogeneity. The primase binds one zinc ion, which is coordinated by four cysteine residues to form a zinc ribbon motif. Factors that influence the rate of priming were investigated, and a physiologically relevant priming rate of approximately 1 primer per second per primosome was achieved. Primase binding to the single-stranded binding protein (1 primase:4 gp32 monomers; K(d) approximately 860 nM) and to the helicase protein in the presence of DNA and ATP-gamma-S (1 primase:1 helicase monomer; K(d) approximately 100 nM) was investigated by isothermal titration calorimetry (ITC). Because the helicase is hexameric, the inferred stoichiometry of primase binding as part of the primosome is helicase hexamer:primase in a ratio of 1:6, suggesting that the active primase, like the helicase, might have a ring-like structure. The primase is a monomer in solution but binds to single-stranded DNA (ssDNA) primarily as a trimer (K(d) approximately 50-100 nM) as demonstrated by ITC and chemical cross-linking. Magnesium is required for primase-ssDNA binding. The minimum length of ssDNA required for stable binding is 22-24 bases, although cross-linking reveals transient interactions on oligonucleotides as short as 8 bases. The association is endothermic at physiologically relevant temperatures, which suggests an overall gain in entropy upon binding. Some possible sources of this gain in entropy are discussed.

Unexpected formation of an epoxide-derived multisubstrate adduct inhibitor on the active site of GAR transformylase

Multisubstrate adduct inhibitors (MAI) of glycinamide ribonucleotide transformylase (GAR Tfase), which incorporate key features of the folate cofactor and the beta-GAR substrate, typically exhibit K(i)'s in the picomolar range. However, these compounds have reduced bioavailability due to the incorporation of a negatively charged phosphate moiety that prevents effective cellular uptake. Thus, a folate analogue that is capable of adduct formation with the substrate on the enzyme active site could lead to a potent GAR Tfase inhibitor that takes advantage of the cellular folate transport systems. We synthesized a dibromide folate analogue, 10-bromo-10-bromomethyl-5,8,10-trideazafolic acid, that was an intermediate designed to assemble with the substrate beta-GAR on the enzyme active site. We have now determined the crystal structure of the Escherichia coli GAR Tfase/MAI complex at 1.6 A resolution to ascertain the nature and mechanism of its time-dependent inhibition. The high-resolution crystal structure clearly revealed the existence of a covalent adduct between the substrate beta-GAR and the folate analogue (K(i) = 20 microM). However, the electron density map surprisingly indicated a C10 hydroxyl in the adduct rather than a bromide and suggested that the multisubstrate adduct is not formed directly from the dibromide but proceeds via an epoxide. Subsequently, we demonstrated the in situ conversion of the dibromide to the epoxide. Moreover, synthesis of the authentic epoxide confirmed that its inhibitory, time-dependent, and cytotoxic properties are comparable to those of the dibromide. Further, inhibition was strongest when the dibromide or epoxide is preincubated with both enzyme and substrate, indicating that inhibition occurs via the enzyme-dependent formation of the multisubstrate adduct. Thus, the crystal structure revealed the successful formation of an enzyme-assembled multisubstrate adduct and highlighted a potential application for epoxides, and perhaps aziridines, in the design of efficacious GAR Tfase inhibitors.

Unexpected formation of an epoxide-derived multisubstrate adduct inhibitor on the active site of GAR transformylase

Multisubstrate adduct inhibitors (MAI) of glycinamide ribonucleotide transformylase (GAR Tfase), which incorporate key features of the folate cofactor and the beta-GAR substrate, typically exhibit K(i)'s in the picomolar range. However, these compounds have reduced bioavailability due to the incorporation of a negatively charged phosphate moiety that prevents effective cellular uptake. Thus, a folate analogue that is capable of adduct formation with the substrate on the enzyme active site could lead to a potent GAR Tfase inhibitor that takes advantage of the cellular folate transport systems. We synthesized a dibromide folate analogue, 10-bromo-10-bromomethyl-5,8,10-trideazafolic acid, that was an intermediate designed to assemble with the substrate beta-GAR on the enzyme active site. We have now determined the crystal structure of the Escherichia coli GAR Tfase/MAI complex at 1.6 A resolution to ascertain the nature and mechanism of its time-dependent inhibition. The high-resolution crystal structure clearly revealed the existence of a covalent adduct between the substrate beta-GAR and the folate analogue (K(i) = 20 microM). However, the electron density map surprisingly indicated a C10 hydroxyl in the adduct rather than a bromide and suggested that the multisubstrate adduct is not formed directly from the dibromide but proceeds via an epoxide. Subsequently, we demonstrated the in situ conversion of the dibromide to the epoxide. Moreover, synthesis of the authentic epoxide confirmed that its inhibitory, time-dependent, and cytotoxic properties are comparable to those of the dibromide. Further, inhibition was strongest when the dibromide or epoxide is preincubated with both enzyme and substrate, indicating that inhibition occurs via the enzyme-dependent formation of the multisubstrate adduct. Thus, the crystal structure revealed the successful formation of an enzyme-assembled multisubstrate adduct and highlighted a potential application for epoxides, and perhaps aziridines, in the design of efficacious GAR Tfase inhibitors.

Intricacies in ATP-dependent clamp loading: variations across replication systems

DNA replication requires the coordinated effort of many proteins to create a highly processive biomachine able to replicate entire genomes in a single process. The clamp proteins confer on replisomes this property of processivity but in turn require clamp loaders for their functional assembly onto DNA. A more detailed view of the mechanisms for holoenzyme assembly in replication systems has been obtained from the advent of novel solution experiments and the appearance of low- and high-resolution structures for the clamp loaders.

Building a replisome solution structure by elucidation of protein-protein interactions in the bacteriophage T4 DNA polymerase holoenzyme

Assembly of DNA replication systems requires the coordinated actions of many proteins. The multiprotein complexes formed as intermediates on the pathway to the final DNA polymerase holoenzyme have been shown to have distinct structures relative to the ground-state structures of the individual proteins. By using a variety of solution-phase techniques, we have elucidated additional information about the solution structure of the bacteriophage T4 holoenzyme. Photocross-linking and mass spectrometry were used to demonstrate interactions between I107C of the sliding clamp and the DNA polymerase. Fluorescence resonance energy transfer, analytical ultracentrifugation, and isothermal titration calorimetry measurements were used to demonstrate that the C terminus of the DNA polymerase can interact at two distinct locations on the sliding clamp. Both of these binding modes may be used during holoenzyme assembly, but only one of these binding modes is found in the final holoenzyme. Present and previous solution interaction data were used to build a model of the holoenzyme that is consistent with these data.

Creating multiple-crossover DNA libraries independent of sequence identity

We have developed, experimentally implemented, and modeled in silico a methodology named SCRATCHY that enables the combinatorial engineering of target proteins, independent of sequence identity. The approach combines two methods for recombining genes: incremental truncation for the creation of hybrid enzymes and DNA shuffling. First, incremental truncation for the creation of hybrid enzymes is used to create a comprehensive set of fusions between fragments of genes in a DNA homology-independent fashion. This artificial family is then subjected to a DNA-shuffling step to augment the number of crossovers. SCRATCHY libraries were created from the glycinamide-ribonucleotide formyltransferase (GART) genes from Escherichia coli (purN) and human (hGART). The developed modeling framework eSCRATCHY provides insight into the effect of sequence identity and fragmentation length on crossover statistics and draws contrast with DNA shuffling. Sequence analysis of the naive shuffled library identified members with up to three crossovers, and modeling predictions are in good agreement with the experimental findings. Subsequent in vivo selection in an auxotrophic E. coli host yielded functional hybrid enzymes containing multiple crossovers.

Dynamic protein interactions in the bacteriophage T4 replisome

The bacteriophage T4 DNA replisome is a complex dynamic system employing a variety of proteins to orchestrate the synthesis of DNA on both the leading and lagging strands. Assembly of the protein complexes responsible for DNA synthesis and priming requires the coordination of transient biomolecular interactions. This interplay of proteins has been dissected through the use of small molecules including fluorescent probes and crosslinkers, enabling the development of a complex dynamic structural and kinetic model for DNA polymerase holoenzyme assembly and primosome formation.

Backbone dynamics in dihydrofolate reductase complexes: role of loop flexibility in the catalytic mechanism

To elucidate the influence of local motion of the polypeptide chain on the catalytic mechanism of an enzyme, we have measured (15)N relaxation data for Escherichia coli dihydrofolate reductase in three different complexes, representing different stages in the catalytic cycle of the enzyme. NMR relaxation data were analyzed by the model-free approach, corrected for rotational anisotropy, to provide insights into the backbone dynamics. There are significant differences in the backbone dynamics in the different complexes. Complexes in which the cofactor binding site is occluded by the Met20 loop display large amplitude motions on the picosecond/nanosecond time scale for residues in the Met20 loop, the adjacent betaF-betaG loop and for residues 67-69 in the adenosine binding loop. Formation of the closed Met20 loop conformation in the ternary complex with folate and NADP(+), results in attenuation of the motions in the Met20 loop and the betaF-betaG loop but leads to increased flexibility in the adenosine binding loop. New fluctuations on a microsecond/millisecond time scale are observed in the closed E:folate:NADP(+) complex in regions that form hydrogen bonds between the Met20 and the betaF-betaG loops. The data provide insights into the changes in backbone dynamics during the catalytic cycle and point to an important role of the Met20 and betaF-betaG loops in controlling access to the active site. The high flexibility of these loops in the occluded conformation is expected to promote tetrahydrofolate-assisted product release and facilitate binding of the nicotinamide ring to form the Michaelis complex. The backbone fluctuations in the Met20 loop become attenuated once it closes over the active site, thereby stabilizing the nicotinamide ring in a geometry conducive to hydride transfer. Finally, the relaxation data provide evidence for long-range motional coupling between the adenosine binding loop and distant regions of the protein.

Structural requirements for the biosynthesis of backbone cyclic peptide libraries

BACKGROUND

Combinatorial methods for the production of molecular libraries are an important source of ligand diversity for chemical biology. Synthetic methods focus on the production of small molecules that must traverse the cell membrane to elicit a response. Genetic methods enable intracellular ligand production, but products must typically be large molecules in order to withstand cellular catabolism. Here we describe an intein-based approach to biosynthesis of backbone cyclic peptide libraries that combines the strengths of synthetic and genetic methods.

RESULTS

Through site-directed mutagenesis we show that the DnaE intein from Synechocystis sp. PCC6803 is very promiscuous with respect to peptide substrate composition, and can generate cyclic products ranging from four to nine amino acids. Libraries with five variable amino acids and either one or four fixed residues were prepared, yielding between 10(7) and 10(8) transformants. The majority of randomly selected clones from each library gave cyclic products.

CONCLUSIONS

We have developed a versatile method for producing intracellular libraries of small, stable cyclic peptides. Genetic encoding enables facile manipulation of vast numbers of compounds, while low molecular weight ensures ready pharmacophore identification. The demonstrated flexibility of the method towards both peptide length and composition makes it a valuable addition to existing methods for generating ligand diversity.

Creating a dynamic picture of the sliding clamp during T4 DNA polymerase holoenzyme assembly by using fluorescence resonance energy transfer

The coordinated assembly of the DNA polymerase (gp43), the sliding clamp (gp45), and the clamp loader (gp44/62) to form the bacteriophage T4 DNA polymerase holoenzyme is a multistep process. A partially opened toroid-shaped gp45 is loaded around DNA by gp44/62 in an ATP-dependent manner. Gp43 binds to this complex to generate the holoenzyme in which gp45 acts to topologically link gp43 to DNA, effectively increasing the processivity of DNA replication. Stopped-flow fluorescence resonance energy transfer was used to investigate the opening and closing of the gp45 ring during holoenzyme assembly. By using two site-specific mutants of gp45 along with a previously characterized gp45 mutant, we tracked changes in distances across the gp45 subunit interface through seven conformational changes associated with holoenzyme assembly. Initially, gp45 is partially open within the plane of the ring at one of the three subunit interfaces. On addition of gp44/62 and ATP, this interface of gp45 opens further in-plane through the hydrolysis of ATP. Addition of DNA and hydrolysis of ATP close gp45 in an out-of-plane conformation. The final holoenzyme is formed by the addition of gp43, which causes gp45 to close further in plane, leaving the subunit interface open slightly. This open interface of gp45 in the final holoenzyme state is proposed to interact with the C-terminal tail of gp43, providing a point of contact between gp45 and gp43. This study further defines the dynamic process of bacteriophage T4 polymerase holoenzyme assembly.

Identification and mapping of protein-protein interactions between gp32 and gp59 by cross-linking

The bacteriophage T4 59 protein (gp59) plays a vital role in recombination and replication by promoting the assembly of the gene 41 helicase (gp41) onto DNA, thus enabling replication as well as strand exchange in recombination. Loading of the helicase onto gp32 (the T4 single strand binding protein)-coated single-stranded DNA requires gp59 to remove gp32 and replace it with gp41. Cross-linking studies between gp32 and gp59 reveal an interaction between Cys-166 of gp32 and Cys-42 of gp59. Since Cys-166 lies in the DNA binding core domain of gp32, this interaction may affect the association of gp32 with DNA. In the presence of gp32 or DNA, gp59 is capable of forming a multimer consisting of at least five gp59 subunits. Kinetics studies suggest that gp59 and gp41 exist in a one-to-one ratio, predicting that gp59 is capable of forming a hexamer (Raney, K. D., Carver, T. E., and Benkovic, S. J. (1996) J. Biol. Chem. 271, 14074-14081). The C-terminal A-domain of gp32 is needed for gp59 oligomer formation. Cross-linking has established that gp59 can interact with gp32-A (a truncated form of gp32 lacking the A-domain) but cannot form higher species. The results support a model in which gp59 binds to gp32 on a replication fork, destabilizing the gp32-single-stranded DNA interaction concomitant with the oligomerization of gp59 that results in a switching of gp41 for gp32 at the replication fork.

Multimeric structure of the secreted meprin A metalloproteinase and characterization of the functional protomer

Meprin A secreted from kidney and intestinal epithelial cells is capable of cleaving growth factors, extracellular matrix proteins, and biologically active peptides. The secreted form of meprin A is a homo-oligomer composed of alpha subunits, a multidomain protease of 582 amino acids coded for near the major histocompatibility complex of the mouse and human genome. Analyses of the recombinant homo-oligomeric form of mouse meprin A by gel filtration, nondenaturing gel electrophoresis, and cross-linking (with disuccinimidyl suberate or N-(4-azido-2,3,5,6-tetraflourobenzyl)-3-maleimidylpropionamide) indicate that the secreted enzyme forms high molecular weight multimers, with a predominance of decamers. The multimers are composed of disulfide-linked dimers attached noncovalently by interactions involving the meprin, A5 protein, receptor protein-tyrosine phosphatase mu (MAM) domain. The active protomer is the noncovalently linked dimer. Linkage of active protomers by disulfide-bonds results in an oligomer of approximately 900 kDa, which is unique among proteases and distinguishes meprin A as the largest known secreted protease. Electron microscopy revealed that the protein was present in two states, a crescent-shaped structure and a closed ring. It is concluded from this and other data that the covalent attachment of the protomers enables noncovalent associations of the native enzyme to form higher oligomers that are critical for hydrolysis of protein substrates.

Localization of GAR transformylase in Escherichia coli and mammalian cells

Enzymes of the de novo purine biosynthetic pathway may form a multienzyme complex to facilitate substrate flux through the ten serial steps constituting the pathway. One likely strategy for complex formation is the use of a structural scaffold such as the cytoskeletal network or subcellular membrane of the cell to mediate protein-protein interactions. To ascertain whether this strategy pertains to the de novo purine enzymes, the localization pattern of the third purine enzyme, glycinamide ribonucleotide transformylase (GAR Tfase) was monitored in live Escherichia coli and mammalian cells. Genes encoding human as well as E. coli GAR Tfase fused with green fluorescent protein (GFP) were introduced into their respective cells with regulated expression of proteins and localization patterns monitored by using confocal fluorescence microscopy. In both instances images showed proteins to be diffused throughout the cytoplasm. Thus, GAR Tfase is not localized to an existing cellular architecture, so this device is probably not used to concentrate the members of the pathway. However, discrete clusters of the pathway may still exist throughout the cytoplasm.

Evaluation of the catalytic mechanism of AICAR transformylase by pH-dependent kinetics, mutagenesis, and quantum chemical calculations

The catalytic mechanism of 5-aminoimidazole-4-carboxamide ribonucleotide transformylase (AICAR Tfase) is evaluated with pH dependent kinetics, site-directed mutagenesis, and quantum chemical calculations. The chemistry step, represented by the burst rates, was not pH-dependent, which is consistent with our proposed mechanism that the 4-carboxamide of AICAR assists proton shuttling. Quantum chemical calculations on a model system of 5-amino-4-carboxamide imidazole (AICA) and formamide using the B3LYP/6-31G level of theory confirmed that the 4-carboxamide participated in the proton-shuttling mechanism. The result also indicated that the amide-assisted mechanism is concerted such that the proton transfers from the 5-amino group to the formamide are simultaneous with nucleophilic attack by the 5-amino group. Because the process does not lead to a kinetically stable intermediate, the intramolecular proton transfer from the 5-amino group through the 4-carboxamide to the formamide proceeds in the same transition state. Interestingly, the calculations predicted that protonation of the N3 of the imidazole of AICA would reduce the energy barrier significantly. However, the pK(a) of the imidazole of AICAR was determined to be 3.23 +/- 0.01 by NMR titration, and AICAR is likely to bind to the enzyme with its imidazole in the free base form. An alternative pathway was suggested by modeling Lys266 to have a hydrogen-bonding interaction with the N3 of the imidazole of AICAR. Lys266 has been implicated in catalysis based on mutagenesis studies and the recent X-ray structure of AICAR Tfase. The quantum chemical calculations on a model system that contains AICA complexed with CH3NH3+ as a mimic of the Lys residue confirmed that such an interaction lowered the activation energy of the reaction and likewise implicated the 4-carboxamide. To experimentally verify this hypothesis, we prepared the K266R mutant and found that its kcat is reduced by 150-fold from that of the wild type without changes in substrate and cofactor Km values. The kcat-pH profile indicated virtually no pH-dependence in the pH range 6-10.5. The results suggest that the ammonium moiety of Lys or Arg is important in catalysis, most likely acting as a general acid catalyst with a pK(a) value greater than 10.5. The H267A mutant was also prepared since His267 has been found in the active site and implicated in catalysis. The mutant enzyme showed no detectable activity while retaining its binding affinity for substrate, indicating that it plays a critical role in catalysis. We propose that His267 interacts with Lys266 to aid in the precise positioning of the general acid catalyst to the N3 of the imidazole of AICAR.

The Unexpected Catalytic Properties of a Heterodimer of GAR Transformylase

We have developed an efficient expression and purification protocol for a heterodimer of glycinamide ribonucleotide transformylase that was identified in incremental truncation libraries, a general combinatorial method for protein fragment complementation (M. Ostermeier, A. E. Nixon, J. H. Shim, and S. J. Benkovic, [1999], Proc. Natl. Acad. Sci. USA 96, 3562-3567). This heterodimer (B13) containing both a bisection point and a deletion in conserved residues close to the active site was expressed and purified in high yield using Intein methodology. The N-terminus fragment (1-111) and C-terminus fragment (M114-212) were also expressed separately as stable proteins. When these two fragments were mixed together, they associate at a highly specific 1:1 ratio to give only the active heterodimer, B13. The activity of B13 is comparable to that of the wild type and the pH-dependent kinetics of B13 turned out to be nearly identical to those of the wild type, indicating that B13 operates in the same mechanism as the wild type. This result demonstrated that cutting within conserved regions is a viable domain separation and confirmed the generality of using incremental truncation for protein fragment complementation.

Identification of the active oligomeric state of an essential adenine DNA methyltransferase from Caulobacter crescentus

Caulobacter crescentus contains one of the two known prokaryotic DNA methyltransferases that lacks a cognate endonuclease. This endogenous cell cycle regulated adenine DNA methyltransferase (CcrM) is essential for C. crescentus cellular viability. DNA methylation catalyzed by CcrM provides an obligatory signal for the proper progression through the cell cycle. To further our understanding of the regulatory role played by CcrM, we sought to investigate its biophysical properties. In this paper we employed equilibrium ultracentrifugation, velocity ultracentrifugation, and chemical cross-linking to show that CcrM is dimeric at physiological concentrations. However, surface plasmon resonance experiments in the presence of S-adenosyl-homocysteine evince that CcrM binds as a monomer to a defined hemi-methylated DNA substrate containing the canonical methylation sequence, GANTC. Initial velocity experiments demonstrate that dimerization of CcrM does not affect DNA methylation. Collectively, these findings suggest that CcrM is active as a monomer and provides a possible in vivo role for dimerization as a means to stabilize CcrM from premature catabolism.

Crystal structure of a bifunctional transformylase and cyclohydrolase enzyme in purine biosynthesis

ATIC, the product of the purH gene, is a 64 kDa bifunctional enzyme that possesses the final two activities in de novo purine biosynthesis, AICAR transformylase and IMP cyclohydrolase. The crystal structure of avian ATIC has been determined to 1.75 A resolution by the MAD method using a Se-methionine modified enzyme. ATIC forms an intertwined dimer with an extensive interface of approximately 5,000 A(2) per monomer. Each monomer is composed of two novel, separate functional domains. The N-terminal domain (up to residue 199) is responsible for the IMPCH activity, whereas the AICAR Tfase activity resides in the C-terminal domain (200-593). The active sites of the IMPCH and AICAR Tfase domains are approximately 50 A apart, with no structural evidence of a tunnel connecting the two active sites. The crystal structure of ATIC provides a framework to probe both catalytic mechanisms and to design specific inhibitors for use in cancer chemotherapy and inflammation.

Predicting crossover generation in DNA shuffling

We introduce a quantitative framework for assessing the generation of crossovers in DNA shuffling experiments. The approach uses free energy calculations and complete sequence information to model the annealing process. Statistics obtained for the annealing events then are combined with a reassembly algorithm to infer crossover allocation in the reassembled sequences. The fraction of reassembled sequences containing zero, one, two, or more crossovers and the probability that a given nucleotide position in a reassembled sequence is the site of a crossover event are estimated. Comparisons of the predictions against experimental data for five example systems demonstrate good agreement despite the fact that no adjustable parameters are used. An in silico case study of a set of 12 subtilases examines the effect of fragmentation length, annealing temperature, sequence identity and number of shuffled sequences on the number, type, and distribution of crossovers. A computational verification of crossover aggregation in regions of near-perfect sequence identity and the presence of synergistic reassembly in family DNA shuffling is obtained.

Rapid generation of incremental truncation libraries for protein engineering using alpha-phosphothioate nucleotides

Incremental truncation for the creation of hybrid enzymes (ITCHY) is a novel tool for the generation of combinatorial libraries of hybrid proteins independent of DNA sequence homology. We herein report a fundamentally different methodology for creating incremental truncation libraries using nucleotide triphosphate analogs. Central to the method is the polymerase catalyzed, low frequency, random incorporation of alpha-phosphothioate dNTPs into the region of DNA targeted for truncation. The resulting phosphothioate internucleotide linkages are resistant to 3'-->5' exonuclease hydrolysis, rendering the target DNA resistant to degradation in a subsequent exonuclease III treatment. From an experimental perspective the protocol reported here to create incremental truncation libraries is simpler and less time consuming than previous approaches by combining the two gene fragments in a single vector and eliminating additional purification steps. As proof of principle, an incremental truncation library of fusions between the N-terminal fragment of Escherichia coli glycinamide ribonucleotide formyltransferase (PurN) and the C-terminal fragment of human glycinamide ribonucleotide formyltransferase (hGART) was prepared and successfully tested for functional hybrids in an auxotrophic E.coli host strain. Multiple active hybrid enzymes were identified, including ones fused in regions of low sequence homology.

Familial mutations and zinc stoichiometry determine the rate-limiting step of nitrocefin hydrolysis by metallo-beta-lactamase from Bacteroides fragilis

The diverse members of the metallo-beta-lactamase family are a growing clinical threat evolving under considerable selective pressure. The enzyme from Bacillus cereus differs from the Bacteroides fragilis enzyme in sequence, zinc stoichiometry, and mechanism. To chart the evolution of the more reactive B. fragilis enzyme, we have made changes in an active site cysteine residue as well as in zinc content to mimic that which occurs in the B. cereus enzyme. Specifically, by introducing a C104R mutation into the B. fragilis enzyme, binding of two zinc ions is maintained, but the k(cat) value for nitrocefin hydrolysis is decreased from 226 to 14 s(-)(1). Removal of 1 equiv of zinc from this mutant further decreases k(cat) to 4.4 s(-)(1). In both cases, the observed k(cat) closely approximates that found in the di- and monozinc forms of the B. cereus enzyme (12 and 6 s(-)(1), respectively). Pre-steady-state stopped-flow studies using nitrocefin as a substrate indicate that these enzyme forms share a similar mechanism featuring an anionic intermediate but that the rate-limiting step changes from protonation of that species to the C-N bond cleavage leading to the intermediate. Overall, features that contribute 3.7 kcal/mol toward the acceleration of the C-N bond cleavage step have been uncovered although some of the total acceleration is masked in the steady-state by a change in rate-limiting step. These experiments illustrate one step in the evolution of a catalytic mechanism and, in a larger perspective, one step in the evolution of antibiotic resistance mechanisms.

Interloop contacts modulate ligand cycling during catalysis by Escherichia coli dihydrofolate reductase

As a continuation to our studies on the importance of interloop interactions in the Escherichia coli DHFR catalytic cycle, we have investigated the role of the betaG-betaH loop in modulating the closed and occluded conformations of the Met20 loop during the DHFR catalytic cycle. Specifically, to assess the importance of the hydrogen bond formed between Ser148 in the betaG-betaH loop and the Met20 loop, Ser148 was independently substituted with aspartic acid, alanine, and lysine. Moreover, the betaG-betaH loop was deleted entirely to yield the Delta(146-148) DHFR mutant. Steady-state turnover rates for all mutants were at most 3-fold lower than the wild-type rate. Lack of an isotope effect on this rate indicated the chemistry step does not contribute to the steady-state turnover. Consistent with this finding, hydride transfer rates for the DHFR mutants were at least 10-fold greater than the observed steady-state rates. The values ranged from a 30% decrease (Ser148Ala and Ser148Lys) to a 50% increase (Ser148Asp) in rate relative to that of the wild type. Modifications of the betaG-betaH loop enhanced the affinity for the cofactor and decreased the affinity for pterin, as determined by the K(D) values of the mutant proteins. Further analysis of Ser148Ala and Delta(146-148) DHFRs indicated these effects were manifest mainly in ligand off rates, although in some cases the on rate was affected. The Ser148Asp and Delta(146-148) mutations perturbed the preferred catalytic cycle through the introduction of branching at key intermediates. Rather than following the single WT pathway which involves loss of NADP(+) and rebinding of NADPH to precede loss of the product H4F (negative cooperativity), the mutants can reenter the catalytic cycle through different pathways. These findings suggest that the role of the interloop interaction between the betaG-betaH loop and the Met20 loop is to modulate ligand off rates allowing for proper cycling through the preferred kinetic pathway.

Human AICAR transformylase: role of the 4-carboxamide of AICAR in binding and catalysis

We have prepared 4-substituted analogues of 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR) to investigate the specificity and mechanism of AICAR transformylase (AICAR Tfase). Of the nine analogues of AICAR studied, only one analogue, 5-aminoimidazole-4-thiocarboxamide ribonucleotide, was a substrate, and it was converted to 6-mercaptopurine ribonucleotide. The other analogues either did not bind or were competitive inhibitors, the most potent being 5-amino-4-nitroimidazole ribonucleotide with a K(i) of 0.7 +/- 0.5 microM. The results show that the 4-carboxamide of AICAR is essential for catalysis, and it is proposed to assist in mediating proton transfer, catalyzing the reaction by trapping of the addition compound. AICAR analogues where the nitrogen of the 4-carboxamide was derivatized with a methyl or an allylic group did not bind AICAR Tfase, as determined by pre-steady-state burst kinetics; however, these compounds were potent inhibitors of IMP cyclohydrolase (IMP CHase), a second activity of the bifunctional mammalian enzyme (K(i) = 0.05 +/- 0.02 microM for 4-N-allyl-AlCAR). It is proposed that the conformation of the carboxamide moiety required for binding to AICAR Tfase is different than the conformation required for binding to IMP CHase, which is supported by inhibition studies of purine ribonucleotides. It is shown that 5-formyl-AICAR (FAICAR) is a product inhibitor of AICAR Tfase with K(i) of 0.4 +/- 0.1 microM. We have determined the equilibrium constant of the transformylase reaction to be 0.024 +/- 0.001, showing that the reaction strongly favors AICAR and the 10-formyl-folate cofactor. The coupling of the AICAR Tfase and IMP CHase activities on a single polypeptide allows the overall conversion of AICAR to IMP to be favorable by coupling the unfavorable formation of FAICAR with the highly favorable cyclization reaction. The current kinetic studies have also indicated that the release of FAICAR is the rate-limiting step, under steady-state conditions, in the bifunctional enzyme and channeling is not observed between AICAR Tfase and IMP CHase.

Molecular structure of Escherichia coli PurT-encoded glycinamide ribonucleotide transformylase

In Escherichia coli, the PurT-encoded glycinamide ribonucleotide transformylase, or PurT transformylase, catalyzes an alternative formylation of glycinamide ribonucleotide (GAR) in the de novo pathway for purine biosynthesis. On the basis of amino acid sequence analyses, it is known that the PurT transformylase belongs to the ATP-grasp superfamily of proteins. The common theme among members of this superfamily is a catalytic reaction mechanism that requires ATP and proceeds through an acyl phosphate intermediate. All of the enzymes belonging to the ATP-grasp superfamily are composed of three structural motifs, termed the A-, B-, and C-domains, and in each case, the ATP is wedged between the B- and C-domains. Here we describe two high-resolution X-ray crystallographic structures of PurT transformylase from E. coli: one form complexed with the nonhydrolyzable ATP analogue AMPPNP and the second with bound AMPPNP and GAR. The latter structure is of special significance because it represents the first ternary complex to be determined for a member of the ATP-grasp superfamily involved in purine biosynthesis and as such provides new information about the active site region involved in ribonucleotide binding. Specifically in PurT transformylase, the GAR substrate is anchored to the protein via Glu 82, Asp 286, Lys 355, Arg 362, and Arg 363. Key amino acid side chains involved in binding the AMPPNP to the enzyme include Arg 114, Lys 155, Glu 195, Glu 203, and Glu 267. Strikingly, the amino group of GAR that is formylated during the reaction lies at 2.8 A from one of the gamma-phosphoryl oxygens of the AMPPNP.

Homology-independent protein engineering

Is structure, rather than sequence, the key to the successful generation of truly novel proteins? While protein evolution by homologous recombination has become an established tool to explore confined regions in sequence space, the generation of functional hybrid proteins by homology-independent methods further expands the scope of protein engineering.

Design, synthesis, and biological evaluation of fluoronitrophenyl substituted folate analogues as potential inhibitors of GAR transformylase and AICAR transformylase

The examination results of a novel series of potential inhibitors of glycinamide ribonucleotide transformylase (GAR Tfase) and aminoimidazole carboxamide transformylase (AICAR Tfase) are reported. These agents incorporate an electrophilic fluoronitrophenyl group that can potentially react with an active site nucleophile or the substrate GAR/AICAR amine via nucleophilic aromatic substitution.

Characterization of bacteriophage T4-coordinated leading- and lagging-strand synthesis on a minicircle substrate

The DNA replication complex of bacteriophage T4 has been assembled as a single unit on a minicircle substrate with a replication fork that permits an independent measurement of the amount of DNA synthesis on both the leading and lagging strands. The assembled replisome consists of the T4 polymerase [gene product 43 (gp43)], clamp protein (gp45), clamp loader (gp44/62), helicase (gp41), helicase accessory factor (gp59), primase (gp61), and single-stranded DNA binding protein (gp32). We demonstrate that on the minicircle the synthesis of the leading and lagging strands are coordinated and that the C-terminal domain of the gp32 protein regulates this coordination. We show that the reconstituted replisome encompasses two coupled holoenzyme complexes and present evidence that this coupling might include a gp43 homodimer interaction.

Conformationally restricted analogues designed for selective inhibition of GAR Tfase versus thymidylate synthase or dihydrofolate reductase

The synthesis and evaluation of a series of conformationally restricted analogues of 10-formyl-tetrahydrofolate as potential inhibitors of glycinamide ribonucleotide transformylase (GAR Tfase) or aminoimidazole carboxamide transformylase (AICAR Tfase) are reported.

Using an AraC-based three-hybrid system to detect biocatalysts in vivo

Recent methods to create large libraries of proteins have greatly advanced the discovery of proteins with novel functions. However, one limitation in the discovery of new biocatalysts is the screening or selection methods employed to find enzymes from these libraries. We have developed a potentially general method termed QUEST (QUerying for EnzymeS using the Three-hybrid system), which allows the construction of an easily screened or selected phenotype for, in theory, any type of enzymatic reaction. The method couples the in vivo concentration of an enzyme's substrate to changes in the transcriptional level of a reporter operon. Using the arabinose operon activator AraC, we constructed a system capable of detecting the fungal enzyme scytalone dehydratase (SD) in bacteria, and demonstrated its sensitivity and usefulness in library screening.

Improvement in the efficiency of formyl transfer of a GAR transformylase hybrid enzyme

A hybrid glycinamide ribonucleotide transformylase was assembled from two protein domains that were treated as discrete modules. One module contained the ribonucleotide binding domain from the purN glycinamide ribonucleotide transformylase; the second module contained the catalytic machinery and the formyl tetrahydrofolate binding domain from the enzyme encoded by purU, formyl tetrahydrofolate hydrolase. The resultant enzyme showed 0.1% catalytic activity of the wild-type glycinamide ribonucleotide transformylase enzyme but had a formyl transfer efficiency of 10%. A combinatorial mutagenesis approach was used to improve the solubility and formyl transfer properties of the hybrid enzyme. The mutagenized hybrid glycinamide ribonucleotide transformylase was initially expressed as a fusion to the alpha-peptide of beta-galactosidase. Clones were selected for improvement in solubility by determining which clones were capable of alpha-complementation using a blue/white screen. One clone was further characterized and found to have an improved efficiency of transfer of the ribonucleotide increasing this property to >95%.

A two-phagemid system for the creation of non-phage displayed antibody libraries approaching one trillion members

We have designed a two-phagemid system for the construction of very large non-phage displayed Fab antibody libraries in E. coli approaching 10(12) members. The system can accommodate both periplasmic and cytoplasmic Fab expression and should prove useful for the direct selection of functional antibodies by genetic techniques. We successfully alleviate problems of Fab vector instability and report a set of improved 5' primers for the amplification of mouse Ig V(H)95% of mouse Ig V(H) genes and minimize the amount of N-terminal amino acid changes while maintaining the flexibility of periplasmic or cytoplasmic antibody expression in E. coli.

Tracking sliding clamp opening and closing during bacteriophage T4 DNA polymerase holoenzyme assembly

The bacteriophage T4 DNA polymerase holoenzyme, consisting of the DNA polymerase (gp43), the sliding clamp (gp45), and the clamp loader (gp44/62), is loaded onto DNA in an ATP-dependent, multistep reaction. The trimeric, ring-shaped gp45 is loaded onto DNA such that the DNA passes through the center of the ring. gp43 binds to this complex, thereby forming a topological link with the DNA and increasing its processivity. Using stopped-flow fluorescence-resonance energy transfer, we have investigated opening and closing of the gp45 ring during the holoenzyme assembly process. Two amino acids that lie on opposite sides of the gp45 subunit interface, W91 and V162C labeled with coumarin, were used as the fluorescence donor and acceptor, respectively. Free in solution, gp45 has two closed subunit interfaces with W91 to V162-coumarin distances of 19 A and one open subunit interface with a W91 to V162C-coumarin distance of 40 A. Making the assumption that the distance across the two closed subunit interfaces is unchanged during the holoenzyme assembly process, we have found that the distance across the open subunit interface is first increased to greater than 45 A and is then decreased to 30 A during a 10-step assembly mechanism. The gp45 ring is not completely closed in the holoenzyme complex, consistent with previous evidence suggesting that the C-terminus of gp43 is inserted into the gp45 subunit interface. Unexpectedly, ATP-hydrolysis events are coupled to only a fraction of the total distance change, with conformational changes linked to binding DNA and gp43 coupled to the majority of the total distance change. Using the nonhydrolyzable ATP analogue ATP-gamma-S results in formation of a nonproductive gp45 x gp44/62 complex; however, adding an excess of ATP to this nonproductive complex results in rapid ATP/ATP-gamma-S exchange to yield a productive gp45 x gp44/62 complex within seconds.

Cyclic peptide formation catalyzed by an antibody ligase

Cyclic hexapeptides represent a class of compounds with important, diverse biological activities. We report herein that the antibody 16G3 catalyzes the cyclization of d-Trp-Gly-Pal-Pro-Gly-Phe small middle dotp-nitrophenyl ester (8a) to give c-(d-Trp-Gly-Pal-Pro-Gly-l-Phe) (11a). The antibody does not, however, catalyze either epimerization or hydrolysis. The resulting rate enhancement of the cyclization by 16G3 (22-fold) was sufficient to form the desired product in greater than 90% yield. In absolute rate terms, the turnover of 16G3 is estimated to be 2 min(-1). The background rate of epimerization of 8a was reduced from 10 to 1% and hydrolysis from 50 to 4% in the presence of 16G3. As expected, the catalytic effects of 16G3 were blocked by the addition of an amount of the hapten equal to twice the antibody concentration. We also synthesized three diastereomers of 8a: the d-Trp(1)-d-Phe(6) (8b), l-Trp(1)-l-Phe(6) (8c), and l-Trp(1)-d-Phe(6) (8d) hexapeptides as well as d-Trp'-l-Trp(6) (12) and d-Phe'-l-Phe(6) (13). As expected, the rate enhancement by 16G3 was greatest for 8a, because the stereochemistry of Trp(1) and Phe(6) matches that of the corresponding residues on the hapten used to induce the biosynthesis of 16G3. A model of the variable domain of 16G3 was generated from the primary sequence using the antibody structural database to guide the model construction. The resulting model provided support for some previously proposed interpretations of the kinetic data, while providing valuable new insights for others.