Stereo-specific synthesis of analogs of nerve agents and their utilization for selection and characterization of paraoxonase (PON1) catalytic scavengers

Fluorogenic organophosphate inhibitors of acetylcholinesterase (AChE) homologous in structure to nerve agents provide useful probes for high throughput screening of mammalian paraoxonase (PON1) libraries generated by directed evolution of an engineered PON1 variant with wild-type like specificity (rePON1). Wt PON1 and rePON1 hydrolyze preferentially the less-toxic R(P) enantiomers of nerve agents and of their fluorogenic surrogates containing the fluorescent leaving group, 3-cyano-7-hydroxy-4-methylcoumarin (CHMC). To increase the sensitivity and reliability of the screening protocol so as to directly select rePON1 clones displaying stereo-preference towards the toxic S(P) enantiomer, and to determine accurately K(m) and k(cat) values for the individual isomers, two approaches were used to obtain the corresponding S(P) and R(P) isomers: (a) stereo-specific synthesis of the O-ethyl, O-n-propyl, and O-i-propyl analogs and (b) enzymic resolution of a racemic mixture of O-cyclohexyl methylphosphonylated CHMC. The configurational assignments of the S(P) and R(P) isomers, as well as their optical purity, were established by X-ray diffraction, reaction with sodium fluoride, hydrolysis by selected rePON1 variants, and inhibition of AChE. The S(P) configuration of the tested surrogates was established for the enantiomer with the more potent anti-AChE activity, with S(P)/R(P) inhibition ratios of 10-100, whereas the R(P) isomers of the O-ethyl and O-n-propyl were hydrolyzed by wt rePON1 about 600- and 70-fold faster, respectively, than the S(P) counterpart. Wt rePON1-induced R(P)/S(P) hydrolysis ratios for the O-cyclohexyl and O-i-propyl analogs are estimated to be >>1000. The various S(P) enantiomers of O-alkyl-methylphosphonyl esters of CHMC provide suitable ligands for screening rePON1 libraries, and can expedite identification of variants with enhanced catalytic proficiency towards the toxic nerve agents.

Directed evolution of phosphotriesterase from Pseudomonas diminuta for heterologous expression in Escherichia coli results in stabilization of the metal-free state

Phosphotriesterase from Pseudomonas diminuta (PTE) is an extremely efficient metalloenzyme that hydrolyses a variety of compounds including organophosphorus nerve agents. Study of PTE has been hampered by difficulties with efficient expression of the recombinant form of this highly interesting and potentially useful enzyme. We identified a low-level esterolytic activity of PTE and then screened PTE gene libraries for improvements in 2-naphthyl acetate hydrolysis. However, the attempt to evolve this promiscuous esterase activity led to a variant (S5) containing three point mutations that resulted in a 20-fold increase in functional expression. Interestingly, the zinc holoenzyme form of S5 appears to be more sensitive than wild-type PTE to both thermal denaturation and addition of metal chelators. Higher functional expression of the S5 variant seems to lie in a higher stability of the metal-free apoenzyme. The results obtained in this work point out another-and often overlooked-possible determinant of protein expression and purification yields, i.e. the stability of intermediates during protein folding and processing.

Catalytic and binding poly-reactivities shared by two unrelated proteins: The potential role of promiscuity in enzyme evolution

It is generally accepted that enzymes evolved via gene duplication of existing proteins. But duplicated genes can serve as a starting point for the evolution of a new function only if the protein they encode happens to exhibit some activity towards this new function. Although the importance of such catalytic promiscuity in enzyme evolution has been proposed, little is actually known regarding how common promiscuous catalytic activities are in proteins or their origins, magnitudes, and potential contribution to the survival of an organism. Here we describe a pattern of promiscuous activities in two completely unrelated proteins-serum albumins and a catalytic antibody (aldolase antibody 38C2). Despite considerable structural dissimilarities-in the shape of the cavities and the position of catalytic lysine residues-both active sites are able to catalyze the Kemp elimination, a model reaction for proton transfer from carbon. We also show that these different active sites can bind promiscuously an array of hydrophobic negatively charged ligands. We suggest that the basic active-site features of an apolar pocket and a lysine residue can act as a primitive active site allowing these promiscuous activities to take place. We also describe, by modelling product formation at different substrate concentrations, how promiscuous activities of this kind- inefficient and rudimentary as they are-can provide a considerable selective advantage and a starting point for the evolution of new functions.

On the magnitude and specificity of medium effects in enzyme-like catalysts for proton transfer

Medium effects are normally studied by comparing the rates of reactions in different solvents. However, medium effects at the active site of enzymes differ dramatically from bulk solvents, both in their diversity (the presence of more than one type of "solvent") and in their spatial arrangement. We describe medium effects in a simple catalytic system, obtained by systematic alkylation of a polymeric scaffold bearing amine groups to give synzymes that catalyze the Kemp elimination of benzisoxazoles with remarkable efficiency. Our analysis indicates that catalysis by these synzymes is driven primarily by specific, localized enzyme-like medium effects, and these effects seem to differ dramatically from the nonspecific medium effects (i.e., desolvation activation) exhibited by solvents. Ligand-binding studies indicate that the synzyme active sites provide localized microenvironments affording a combination of hydrophobic and apolar regions on one hand and dipolar, protic, and positively charged on the other. Such localized microenvironments are not available in bulk solvents. A Brønsted (leaving group) analysis indicates that, in comparison to solvent catalysis, the efficiency of synzyme catalysis shows little sensitivity to leaving group pK(a). We show that enzyme-like medium effects alone, in the absence of efficient positioning of the catalytic amine base relative to the substrate, can give rise to rate accelerations as high as 10(5), for both activated and nonactivated substrates. Supported by the accidental identification of active sites on the surfaces of noncatalytic proteins and the promiscuous activities found in many enzymes, our findings suggest that the interfaces of protein surfaces and their hydrophobic cores provide a microenvironment that is intrinsically active and may serve as a basis for further evolutionary improvements to give proficient and selective enzymes.

Man-made enzymes--from design to in vitro compartmentalisation

In the past few years, a variety of methods have been developed to allow the in vitro evolution of a range of biomolecules including novel and improved biocatalysts (enzymes). These methods for directed evolution differ in the size and characteristics of the gene repertoire, in the way of linking genotype and phenotype, and in the selection approach. Selections for enzymes can be performed indirectly (for binding of a transition-state analogue or mechanism-based inhibitor), and directly using either intramolecular single-turnover selections (e.g. with SELEX) or the normal (intermolecular, multiple turnover) mode of enzymatic reactions. Each of these methods has distinct strengths and weaknesses. The best system (or combinations of systems) to use depends on the specific target for evolution and the evolutionary distance that needs to be crossed.

Nonspecific catalysis by protein surfaces

Catalytic antibodies are the best available all-around enzyme mimics. They provide a unique experimental approach and some special insights into general questions about catalysis by enzymes. They offer enantiospecific reactions and levels of substrate binding that compare well with typical enzyme reactions, but not--so far--comparable catalytic efficiency. We and others have used the Kemp elimination as a probe of catalytic efficiency in antibodies. We compare these reactions with nonspecific catalysis by other proteins, and with catalysis by enzymes. Several simple reactions are catalyzed by the serum albumins with Michaelis-Menten kinetics, and can be shown to involve substrate binding and catalysis by local functional groups. Here we report the details of one investigation, which implicate known binding sites on the protein surface, and discuss implications for catalyst design and efficiency.

Conformational changes affect binding and catalysis by ester-hydrolysing antibodies

D2.3, D2.4 and D2.5 are ester-hydrolysing antibodies raised against a phosphonate transition state analogue (TSA). All three antibody-TSA binding kinetics, as monitored by fluorescence quenching, indicate an "induced-fit" mechanism: fast bimolecular association followed by a unimolecular isomerisation (k=1-7 s-1). Isomerisation leads to a 30-170-fold increase in affinity towards the TSA and, consequently, to higher catalytic rates. Antibody D2.3 exhibits a complex three-step binding mechanism, in which the last step is a "very slow" isomerisation (k<0.02 s-1). This very slow isomerisation is limiting the rate of catalysis by D2.3, as indicated by the kinetics of product release which show characteristics of enzyme "conformational memory" or "hysteresis". The results support a mechanism consisting of pre-equilibrium between "nether-active" (low affinity) and "active" (high affinity) antibody conformers (prior to ligand addition) as well as induced-fit, i.e. isomerisation of the nether-active ligand-antibody complex to give the active complex. Crystal structures of these antibodies, free and complexed, have previously indicated that their conformation does not change upon binding. Here, we show that the buffer used to crystallise the antibodies, and in particular its polyethylene glycol component, alters the pre-equilibrium in favour of the active conformer, leading to its crystallisation both in the presence and in the absence of the TSA.

Man-made cell-like compartments for molecular evolution

Cellular compartmentalization is vital for the evolution of all living organisms. Cells keep together the genes, the RNAs and proteins that they encode, and the products of their activities, thus linking genotype to phenotype. We have reproduced this linkage in the test tube by transcribing and translating single genes in the aqueous compartments of water-in-oil emulsions. These compartments, with volumes close to those of bacteria, can be recruited to select genes encoding catalysts. A protein or RNA with a desired catalytic activity converts a substrate attached to the gene that encodes it to product. In other compartments, substrates attached to genes that do not encode catalysts remain unmodified. Subsequently, genes encoding catalysts are selectively enriched by virtue of their linkage to the product. We demonstrate the linkage of genotype to phenotype in man-made compartments using a model system. A selection for target-specific DNA methylation was based on the resistance of the product (methylated DNA) to restriction digestion. Genes encoding HaeIII methyltransferase were selected from a 10(7)-fold excess of genes encoding another enzyme.

Expression and characterization of recombinant single-chain Fv and Fv fragments derived from a set of catalytic antibodies

The generation of catalytic antibodies should enable the catalysis of reactions for which no enzymatic or chemical catalyst is currently available. In previous studies, we established a series of catalytic antibodies capable of hydrolysing p-nitrobenzyl (pNB) and p-nitrophenyl (pNP) esters. A group of these catalytic antibodies exhibited high reactivity and substrate specificity, yet each individual antibody demonstrated different kinetic parameters. In order to study the molecular basis for these differences, we have cloned, sequenced and expressed the variable regions of this group of antibodies as functional scFv and Fv in bacteria. The variable region of the heavy chain is derived from a novel germline gene of the J558 family whereas the light chain comes from a germline gene previously found in our catalytic antibodies catalysing the hydrolysis of only nitrophenyl esters, demonstrating that the heavy chain determines the specificity for the nitrobenzyl esters. Several different expression systems were examined for their ability to produce catalytically active antibodies. When expressed as an scFv, both refolded and secreted scFvs exhibited catalytic activity although yields of expressed protein were low. The secreted scFvs had higher specific activity. On the other hand, Fv fragments were expressed in sufficient quantities to allow kinetic analysis. Levels of expression were dependent on the sequence of VL used. Using this expression system, the relative contributions of the individual light and heavy chains to catalysis and binding could be evaluated. Both original VH and VL regions are required for hapten binding, although the VH is more crucial for catalysis. By replacing the CDR3 of the heavy chain with a random sequence, it was shown to be essential for both binding and catalysis. This expression system together with site-directed mutagenesis should enable a more detailed study of the catalytic mechanism of this set of antibodies.

Efficient and selective p-nitrophenyl-ester-hydrolyzing antibodies elicited by a p-nitrobenzyl phosphonate hapten

A number of monoclonal antibodies elicited against a nitrobenzyl (Nbzl)-phosphonate transition-state analogue (TSA), and which were selected for the hydrolysis of the corresponding Nbzl-ester, were also found to catalyze the hydrolysis of the analogous p-nitrophenyl(Np) ester with notable efficiency and specificity. The activity towards the Np-ester is higher in terms of rates (k(cat); as expected from the higher intrinsic reactivity of Np-esters); however, the rate acceleration (k(cat)/k(uncat)) is close to or lower than that observed with the Nbzl-ester. Unexpectedly, the affinity to the Np-ester substrate (1/K(M)) and therefore k(cat)/K(M) are significantly higher. The best example is antibody D2.4 having a k(cat)/K(M) value of 64 s(-1) x M(-1) with the Nbzl-ester and 9400 s(-1) x M(-1) with the Np-ester. Moreover, due to a lower product inhibition by p-nitrophenol relative to p-nitrobenzyl alcohol, these antibodies exhibit more than 1000 turnovers with the Np-ester. The differential affinity of these antibodies to the Nbzl-phosphonate TSA versus the Nbzl-ester substrate (K(S)/K(TSA) or K(M)/K(i)) correlates well with the observed rate enhancement (k(cat)/k(uncat)). For the Np-ester, however, stabilisation of the transition state (as reflected by K(S)/K(TSA) and by the catalytic proficiencies, k(cat)/K(M)/k(uncat)) does not fully account for the catalytic power (k(cat)/k(uncat)), indicating a more complex catalytic mechanism than simply transition-state stabilization. A comparison of the kinetic parameters of D2.4 with other Np-ester-hydrolyzing antibodies raised against Np-phosphonate haptens emphasizes the marked advantage of this antibody which was elicited against an Nbzl-phosphonate hapten. These results appear to be general: anti-(Nbzl-phosphonate TSA) antibodies obtained from other mouse strains and using different immunization protocols are also efficient Np-esterases. They demonstrate the use of an expanded TSA-hapten, where a spacer (a methylene group) mimics bonds that are partially cleaved in the transition state of the catalyzed reaction.

Structural convergence in the active sites of a family of catalytic antibodies

The x-ray structures of three esterase-like catalytic antibodies identified by screening for catalytic activity the entire hybridoma repertoire, elicited in response to a phosphonate transition state analog (TSA) hapten, were analyzed. The high resolution structures account for catalysis by transition state stabilization, and in all three antibodies a tyrosine residue participates in the oxyanion hole. Despite significant conformational differences in their combining sites, the three antibodies, which are the most efficient among those elicited, achieve catalysis in essentially the same mode, suggesting that evolution for binding to a single TSA followed by screening for catalysis lead to antibodies with structural convergence.

Off-the-shelf proteins that rival tailor-made antibodies as catalysts

Mimicking the efficiency of enzyme catalysis is a daunting challenge. An enzyme selectively binds and stabilizes the transition state (s) for a particular reaction. Artificial host systems can bind ground states just as efficiently, and rate enhancements comparable to those in enzymatic reactions can be achieved by bringing catalytic and substrate groups together in intramolecular reactions. But the combination of selective binding and efficient catalysis remains elusive. The best enzyme mimics currently known are catalytic antibodies. They bind transition-state analogues with high affinity, but their catalytic efficiency generally falls far short of that of enzymes. Thorn et al. recently described an antibody that catalyses the eliminative ring-opening of a benziosoxazole "exceptionally efficiently" using carboxylate as the general base, raising the intriguing possibility that this high efficiency derives from precise positioning of catalytic and substrate groups. Here we show that familiar 'off-the-shelf' proteins--serum albumins--catalyse the same reaction at similar rates, using a lysine side-chain amino group as the catalytic general base. Comparisons suggest that formal general base catalysis is of only modest efficiency in both systems, and that the antibody catalysis is boosted by a non-specific medium effect.

Unexpectedly high occurrence of catalytic antibodies in MRL/lpr and SJL mice immunized with a transition-state analog: is there a linkage to autoimmunity?

Upon testing the ability of several strains of mice to elicit esterolytic antibodies after immunization with a p-nitrobenzyl phosphonate hapten, we have found that the occurrence of catalytic antibodies in SJL and MRL/lpr autoimmune mice is dramatically higher than in normal mouse strains (e.g., the wild-type MRL/++ or BALB/c). Fewer than 10 catalytic clones are usually obtained from a single fusion of lymphocytes taken from normal mice, whereas several hundred catalytic clones are obtained in SJL or MRL/lpr mice. Differences in the numbers of hapten-binding clones do not account for the high occurrences of catalytic clones in these strains. This phenomenon prevailed in the early responses; in both SJL and MRL/lpr mice a significant decline in the appearance of catalytic clones was observed after multiple immunizations. Esterolytic antibodies were not found in MRL/lpr mice immunized with haptens that do not mimic the transition state for the hydrolysis of the ester substrate (e.g., with a substrate analog). The catalytic antibodies manifest high specificity to the antigen and variability in their binding and catalytic properties. The use of autoimmunity-prone mice may greatly expand the repertoire of catalytic clones elicited against a transition-state analog hapten. More intriguing is the possible linkage between autoimmunity and the appearance of catalytic antibodies. These results suggest that there is normally a selection against the expression of certain variable genes encoding antibodies with catalytic activity.

Crystal structure of a catalytic antibody Fab with esterase-like activity

BACKGROUND

Antibodies with catalytic properties can be prepared by eliciting an antibody response against 'transition state analog' haptens. The specificity, rate and number of reaction cycles observed with these antibodies more closely resemble the properties of enzymes than any of the many other known enzyme-mimicking systems.

RESULTS

We have determined to 3 A resolution the first X-ray structure of a catalytic antibody Fab. This antibody catalyzes the hydrolysis of a p-nitrophenyl ester. In conjunction with binding studies in solution, this structure of the uncomplexed site suggests a model for transition state fixation where two tyrosines mimic the oxyanion binding hole of serine proteases. A comparison with the structures of known Fabs specific for low molecular weight haptens reveals that this catalytic antibody has an unusually long groove at its combining site.

CONCLUSION

Since transition state analogs contain elements of the desired product, product inhibition is a severe problem in antibody catalysis. The observation of a long groove at the combining site may relate to the ability of this catalytic antibody to achieve multiple cycles of reaction.

pH on-off switching of antibody-hapten binding by site-specific chemical modification of tyrosine

Tetranitromethane (TNM) chemically mutates the binding sites of antibodies so that the nitrated antibodies exhibit pH-dependent binding near physiological pH. Three monoclonal antibodies were selectively modified, each under different conditions, with the resultant loss of binding activity at pH > 8 which is recovered at pH < 6. Recovery and loss of binding are ascribed to the protonation and deprotonation, respectively, of the hydroxyl group of the resulting 3-nitrotyrosine side chain (pKa approximately 7) at the binding site of these antibodies. pH on-off dependency of binding activity, common to all TNM-modified antibodies studied by us so far, may find use in a variety of applications in which controlled modulation under mild conditions is required.

Differences in the biochemical properties of esterolytic antibodies correlate with structural diversity

A prerequisite to the design and engineering of catalytic antibodies is the knowledge of their structure and in particular which residues are involved in binding and catalysis. We compared the structure and catalytic properties of a series of six monoclonal antibodies which were all raised against a p-nitrophenyl (PNP) phosphonate and which catalyze the hydrolysis of p-nitrophenyl esters. Three of the antibodies (Group I) have similar light and heavy chain variable regions. The other three antibodies have similar VL regions of which two (Group II) have VH regions from the MOPC21 gene family and the remaining one (Group III) a VH from the MC101 gene family making a total of three different groups based on their V region sequences. The structural division into groups is paralleled by the differences in binding constants to hapten analogs, substrate specificity and the susceptibility of the catalytic activity of the antibodies to chemical modification of tryptophan and arginine residues. The relative binding of a transition state analog to the binding of substrate is much higher for the Group I antibodies than for the other groups. Only the Group I antibodies can catalyze the hydrolysis of a carbonate substrate. However all of the antibodies lose catalytic activity upon specific tyrosine modification which highlights the importance of tyrosine in the active site of the antibodies. Thus, antibodies raised against a single hapten can give antibodies with different structures, and correspondingly different specificities and catalytic properties.

Catalytic antibodies: a critical assessment

In the past few years, antibodies that catalyze a variety of reactions with enzyme-like properties have been produced. The present review is of a critical nature, rather than a survey or an introduction to the field of catalytic antibodies. Here, we examine the performance of catalytic antibodies in light of the features that define an enzyme: substrate specificity, rate enhancement, and turnover. We also refer to some limitations of the technologies currently used for their generation. In the future, antibodies may provide a new repertoire of tailor-made, enzyme-like, catalysts with possible applications in biology, medicine, and biotechnology. In the following sections, we emphasize that these applications will require far more efficient catalysts than are presently available, and we point to several trends for future research that may offer more efficient catalytic antibodies.

catELISA: a facile general route to catalytic antibodies

The low abundance and activity of catalytic antibodies are major obstacles to their selection from the virtually unlimited repertoire of antibody binding sites. The requirement for new screening methodologies is further emphasized by the availability of combinatorial libraries, in which a functional polypeptide has to be selected out of millions of possibilities. We present a simple and sensitive screening approach (termed catELISA) based on immobilized substrates and immunodetection of the end product of the catalyzed reaction. The feasibility of catELISA is demonstrated here by the generation of potent ester-hydrolyzing antibodies by direct screening of hybridoma supernatants. We show that this approach is not only facile but general: it is not limited by type of reaction, substrate, or catalyst (enzymes, catalytic antibodies, chemical catalysts). catELISA opens a route to catalytic antibodies that replaces existing lengthy and arduous methods, thus allowing us to expand their number and improve their quality and to address questions that would otherwise be difficult to answer.

Detection of catalytic monoclonal antibodies

Several laboratories have now shown that monoclonal antibodies having enzyme-like properties can be generated. The generation of catalytic antibodies makes use of the same basic procedures that have been used for the generation of binding monoclonal antibodies, yet the process involves an additional crucial step: screening for catalytic activity. In this paper we address the unique problems involved in the detection of inefficient catalytic activity that is accompanied by uncatalyzed background reaction. An analysis that allows optimization of assay conditions and estimation of the minimal antibody concentration required to observe catalysis is presented. The results indicate that the structure of the substrate should be optimized to increase its affinity (i.e., decrease its Km) and reduce its concentration to pseudo-first-order conditions (S(O) much less than Km) so that the signal observed in the presence of a catalytic antibody (delta Pcat) is significantly higher than that of the background (delta P(uncat)). Other factors involved in the screening procedures, e.g., sensitivity of the assay, solubility and reactivity of the substrate, and purity of the antibody preparation, are also discussed. The effect of these assay parameters on the ability to detect catalytic activity is demonstrated with p-nitrophenyl ester-hydrolyzing antibodies.

Simple method for selecting catalytic monoclonal antibodies that exhibit turnover and specificity

Monoclonal antibodies were raised against a mono-p-nitrophenyl phosphonate ester to elicit catalytic antibodies capable of hydrolyzing the analogous p-nitrophenyl ester or carbonate. Potential catalytic antibody producing clones were selected, by use of a competitive inhibition assay, on the basis of their affinity for a "short" transition-state analogue, a truncated hapten which maximizes the relative contribution of the transition-state structural elements to binding. Of 30-40 clones that would have been examined on the basis of hapten binding alone, 7 were selected and 4 of these catalyzed the hydrolysis of the relevant p-nitrophenyl ester. This competitive inhibition technique represents a general approach for selecting potential catalytic antibodies and significantly increases the probability of obtaining efficient catalytic monoclonal antibodies. Further study of the catalytic antibodies revealed significant rate enhancement (kcat/kuncat approximately 10(4)) and substrate specificity for the hydrolysis of the analogous ester and, for three of the antibodies, of the analogous carbonate. The antibodies displayed turnover, an essential feature of enzymes. Evidence that catalysis occurred at the antibody combining sites was provided by the identity of the binding and the catalysis-inhibition specificity patterns.