Asymmetric mutations in the tetrameric R67 dihydrofolate reductase reveal high tolerance to active-site substitutions.

Type II R67 dihydrofolate reductase (DHFR) is a bacterial plasmid-encoded enzyme that is intrinsically resistant to the widely-administered antibiotic trimethoprim. R67 DHFR is genetically and structurally unrelated to E. coli chromosomal DHFR and has an unusual architecture, in that four identical protomers form a single symmetrical active site tunnel that allows only one substrate binding/catalytic event at any given time. As a result, substitution of an active-site residue has as many as four distinct consequences on catalysis, constituting an atypical model of enzyme evolution. Although we previously demonstrated that no single residue of the native active site is indispensable for function, library selection here revealed a strong bias toward maintenance of two native protomers per mutated tetramer. A variety of such "half-native" tetramers were shown to procure native-like catalytic activity, with similar KM values but kcat values 5- to 33-fold lower, illustrating a high tolerance for active-site substitutions. The selected variants showed a reduced thermal stability (Tm ∼12°C lower), which appears to result from looser association of the protomers, but generally showed a marked increase in resilience to heat denaturation, recovering activity to a significantly greater extent than the variant with no active-site substitutions. Our results suggest that the presence of two native protomers in the R67 DHFR tetramer is sufficient to provide native-like catalytic rate and thus ensure cellular proliferation.

Lactone-bound structures of cyclohexanone monooxygenase provide insight into the stereochemistry of catalysis.

The Baeyer-Villiger monooxygenases (BVMOs) are microbial enzymes that catalyze the synthetically useful Baeyer-Villiger oxidation reaction. The available BVMO crystal structures all lack a substrate or product bound in a position that would determine the substrate specificity and stereospecificity of the enzyme. Here, we report two crystal structures of cyclohexanone monooxygenase (CHMO) with its product, ε-caprolactone, bound: the CHMO(Tight) and CHMO(Loose) structures. The CHMO(Tight) structure represents the enzyme state in which substrate acceptance and stereospecificity is determined, providing a foundation for engineering BVMOs with altered substrate spectra and/or stereospecificity. The CHMO(Loose) structure is the first structure where the product is solvent accessible. This structure represents the enzyme state upon binding and release of the substrate and product. In addition, the role of the invariant Arg329 in chaperoning the substrate/product during the catalytic cycle is highlighted. Overall, these data provide a structural framework for the engineering of BVMOs with altered substrate spectra and/or stereospecificity.

Different active-site loop orientation in serine hydrolases versus acyltransferases.

Acyl transfer is a key reaction in biosynthesis, including synthesis of antibiotics and polyesters. Although researchers have long recognized the similar protein fold and catalytic machinery in acyltransferases and hydrolases, the molecular basis for the different reactivity has been a long-standing mystery. By comparison of X-ray structures, we identified a different oxyanion-loop orientation in the active site. In esterases/lipases a carbonyl oxygen points toward the active site, whereas in acyltransferases a NH of the main-chain amide points toward the active site. Amino acid sequence comparisons alone cannot identify such a difference in the main-chain orientation. To identify how this difference might change the reaction mechanism, we solved the X-ray crystal structure of Pseudomonas fluorescens esterase containing a sulfonate transition-state analogue bound to the active-site serine. This structure mimics the transition state for the attack of water on the acyl-enzyme and shows a bridging water molecule between the carbonyl oxygen mentioned above and the sulfonyl oxygen that mimics the attacking water. A possible mechanistic role for this bridging water molecule is to position and activate the attacking water molecule in hydrolases, but to deactivate the attacking water molecule in acyl transferases.

Switching catalysis from hydrolysis to perhydrolysis in Pseudomonas fluorescens esterase.

Many serine hydrolases catalyze perhydrolysis, the reversible formation of peracids from carboxylic acids and hydrogen peroxide. Recently, we showed that a single amino acid substitution in the alcohol binding pocket, L29P, in Pseudomonas fluorescens (SIK WI) aryl esterase (PFE) increased the specificity constant of PFE for peracetic acid formation >100-fold [Bernhardt et al. (2005) Angew. Chem., Int. Ed. 44, 2742]. In this paper, we extend this work to address the three following questions. First, what is the molecular basis of the increase in perhydrolysis activity? We previously proposed that the L29P substitution creates a hydrogen bond between the enzyme and hydrogen peroxide in the transition state. Here we report two X-ray structures of L29P PFE that support this proposal. Both structures show a main chain carbonyl oxygen closer to the active site serine as expected. One structure further shows acetate in the active site in an orientation consistent with reaction by an acyl-enzyme mechanism. We also detected an acyl-enzyme intermediate in the hydrolysis of epsilon-caprolactone by mass spectrometry. Second, can we further increase perhydrolysis activity? We discovered that the reverse reaction, hydrolysis of peracetic acid to acetic acid and hydrogen peroxide, occurs at nearly the diffusion limited rate. Since the reverse reaction cannot increase further, neither can the forward reaction. Consistent with this prediction, two variants with additional amino acid substitutions showed 2-fold higher k(cat), but K(m) also increased so the specificity constant, k(cat)/K(m), remained similar. Third, how does the L29P substitution change the esterase activity? Ester hydrolysis decreased for most esters (75-fold for ethyl acetate) but not for methyl esters. In contrast, L29P PFE catalyzed hydrolysis of epsilon-caprolactone five times more efficiently than wild-type PFE. Molecular modeling suggests that moving the carbonyl group closer to the active site blocks access for larger alcohol moieties but binds epsilon-caprolactone more tightly. These results are consistent with the natural function of perhydrolases being either hydrolysis of peroxycarboxylic acids or hydrolysis of lactones.

Extracellular production of Streptomyces lividans acetyl xylan esterase A in Escherichia coli for rapid detection of activity.

Acetyl xylan esterase A (AxeA) from Streptomyces lividans belongs to a large family of industrially relevant polysaccharide esterases. AxeA and its truncated form containing only the catalytically competent domain, AxeA(tr), catalyze both the deacetylation of xylan and the N-deacetylation of chitosan. This broad substrate specificity lends additional interest to their characterization and production. Here, we report three systems for extracellular production of AxeA(tr): secretion from the native host S. lividans with the native signal peptide, extracellular production in Escherichia coli with the native signal peptide, and in E. coli with the OmpA signal peptide. Over five to seven days of a shake flask culture, the native host S. lividans with the native signal peptide secreted AxeA(tr) into the extracellular medium in high yield (388 mg/L) with specific activity of 19 U/mg corresponding to a total of 7000 U/L. Over one day of shake flask culture, E. coli with the native secretion signal peptide produced 84-fold less in the extracellular medium (4.6 mg/L), but the specific activity was higher (100 U/mg) corresponding to a total of 460 U/L. A similar E. coli culture using the OmpA signal peptide, produced 10mg/L with a specific activity of 68 U/mg, corresponding to a total of 680 U/L. In 96-well microtiter plates, extracellular production with E. coli gave approximately 30 and approximately 86 microg/mL in S. lividans. Expression in S. lividans with the native signal peptide is best for high level production, while expression in E. coli using the OmpA secretion signal peptide is best for high-throughput expression and screening of variants in microtiter plate format.

Improving enzyme properties: when are closer mutations better?

Study of mutations that improve enzyme properties reveals that in many, but not all, cases closer mutations are more effective than distant ones. For enantioselectivity, substrate selectivity and new catalytic activity (catalytic promiscuity) closer mutations improved enzymes more effectively than distant ones. However, both close and distant mutations can improve activity, thermal stability and also probably stability toward organic solvents. Typical random mutagenesis methods, such as error-prone PCR, create greater numbers of distant mutations than close mutations because enzymes contain more amino acids distant from the active site than close to the active site. This suggests that instead of mutating the entire enzyme, focusing mutations near the substrate-binding site might dramatically increase the success rate in many directed evolution experiments.

Focusing mutations into the P. fluorescens esterase binding site increases enantioselectivity more effectively than distant mutations.

Rational design of enzymes with improved properties, such as enantioselectivity, usually focuses mutations within the substrate binding site. On the other hand, directed evolution of enzymes usually targets the entire protein and discovers beneficial mutations far from the substrate binding site. In this paper, we propose an explanation for this discrepancy and show that a combined approach--random mutagenesis within the substrate binding site--is better. To increase the enantioselectivity (E) of a Pseudomonas fluorescens esterase (PFE) toward methyl 3-bromo-2-methylpropionate, we focused mutagenesis into the substrate binding site at Trp28, Val121, Phe198, and Val225. Five of the catalytically active mutants (13%) showed better enantioselectivity than wild-type PFE. The increases in enantioselectivity were higher (up to 5-fold, reaching E = 61) than with mutants identified by random mutagenesis of the entire enzyme.