Computationally designed libraries for rapid enzyme stabilization

Methods for the directed evolution of proteins

Directed evolution has proved to be an effective strategy for improving or altering the activity of biomolecules for industrial, research and therapeutic applications. The evolution of proteins in the laboratory requires methods for generating genetic diversity and for identifying protein variants with desired properties. This Review describes some of the tools used to diversify genes, as well as informative examples of screening and selection methods that identify or isolate evolved proteins. We highlight recent cases in which directed evolution generated enzymatic activities and substrate specificities not known to exist in nature.

Metabolism of β-valine via a CoA-dependent ammonia lyase pathway

Pseudomonas species strain SBV1 can rapidly grow on medium containing β-valine as a sole nitrogen source. The tertiary amine feature of β-valine prevents direct deamination reactions catalyzed by aminotransferases, amino acid dehydrogenases, and amino acid oxidases. However, lyase- or aminomutase-mediated conversions would be possible. To identify enzymes involved in the degradation of β-valine, a PsSBV1 gene library was prepared and used to complement the β-valine growth deficiency of a closely related Pseudomonas strain. This resulted in the identification of a gene encoding β-valinyl-coenzyme A ligase (BvaA) and two genes encoding β-valinyl-CoA ammonia lyases (BvaB1 and BvaB2). The BvaA protein demonstrated high sequence identity to several known phenylacetate CoA ligases. Purified BvaA enzyme did not convert phenyl acetic acid but was able to activate β-valine in an adenosine triphosphate (ATP)- and CoA-dependent manner. The substrate range of the enzyme appears to be narrow, converting only β-valine and to a lesser extent, 3-aminobutyrate and β-alanine. Characterization of BvaB1 and BvaB2 revealed that both enzymes were able to deaminate β-valinyl-CoA to produce 3-methylcrotonyl-CoA, a common intermediate in the leucine degradation pathway. Interestingly, BvaB1 and BvaB2 demonstrated no significant sequence identity to known CoA-dependent ammonia lyases, suggesting they belong to a new family of enzymes. BLAST searches revealed that BvaB1 and BvaB2 show high sequence identity to each other and to several enoyl-CoA hydratases, a class of enzymes that catalyze a similar reaction with water instead of amine as the leaving group.

X-ray crystallographic validation of structure predictions used in computational design for protein stabilization

Protein engineering aimed at enhancing enzyme stability is increasingly supported by computational methods for calculation of mutant folding energies and for the design of disulfide bonds. To examine the accuracy of mutant structure predictions underlying these computational methods, crystal structures of thermostable limonene epoxide hydrolase variants obtained by computational library design were determined. Four different predicted effects indeed contributed to the obtained stabilization: (i) enhanced interactions between a flexible loop close to the N-terminus and the rest of the protein; (ii) improved interactions at the dimer interface; (iii) removal of unsatisfied hydrogen bonding groups; and (iv) introduction of additional positively charged groups at the surface. The structures of an eightfold and an elevenfold mutant showed that most mutations introduced the intended stabilizing interactions, and side-chain conformations were correctly predicted for 72-88% of the point mutations. However, mutations that introduced a disulfide bond in a flexible region had a larger influence on the backbone conformation than predicted. The enzyme active sites were unaltered, in agreement with the observed preservation of catalytic activities. The structures also revealed how a c-Myc tag, which was introduced for facile detection and purification, can reduce access to the active site and thereby lower the catalytic activity. Finally, sequence analysis showed that comprehensive mutant energy calculations discovered stabilizing mutations that are not proposed by the consensus or B-FIT methods.

Rational stabilization of complex proteins: a divide and combine approach

Increasing the thermostability of proteins is often crucial for their successful use as analytic, synthetic or therapeutic tools. Most rational thermostabilization strategies were developed on small two-state proteins and, unsurprisingly, they tend to fail when applied to the much more abundant, larger, non-fully cooperative proteins. We show that the key to stabilize the latter is to know the regions of lower stability. To prove it, we have engineered apoflavodoxin, a non-fully cooperative protein on which previous thermostabilizing attempts had failed. We use a step-wise combination of structure-based, rationally-designed, stabilizing mutations confined to the less stable structural region, and obtain variants that, according to their van't Hoff to calorimetric enthalpy ratios, exhibit fully-cooperative thermal unfolding with a melting temperature of 75°C, 32 degrees above the lower melting temperature of the non-cooperative wild type protein. The ideas introduced here may also be useful for the thermostabilization of complex proteins through formulation or using specific stabilizing ligands (e.g. pharmacological chaperones).

Structure-based design of combinatorial mutagenesis libraries

The development of protein variants with improved properties (thermostability, binding affinity, catalytic activity, etc.) has greatly benefited from the application of high-throughput screens evaluating large, diverse combinatorial libraries. At the same time, since only a very limited portion of sequence space can be experimentally constructed and tested, an attractive possibility is to use computational protein design to focus libraries on a productive portion of the space. We present a general-purpose method, called "Structure-based Optimization of Combinatorial Mutagenesis" (SOCoM), which can optimize arbitrarily large combinatorial mutagenesis libraries directly based on structural energies of their constituents. SOCoM chooses both positions and substitutions, employing a combinatorial optimization framework based on library-averaged energy potentials in order to avoid explicitly modeling every variant in every possible library. In case study applications to green fluorescent protein, β-lactamase, and lipase A, SOCoM optimizes relatively small, focused libraries whose variants achieve energies comparable to or better than previous library design efforts, as well as larger libraries (previously not designable by structure-based methods) whose variants cover greater diversity while still maintaining substantially better energies than would be achieved by representative random library approaches. By allowing the creation of large-scale combinatorial libraries based on structural calculations, SOCoM promises to increase the scope of applicability of computational protein design and improve the hit rate of discovering beneficial variants. While designs presented here focus on variant stability (predicted by total energy), SOCoM can readily incorporate other structure-based assessments, such as the energy gap between alternative conformational or bound states.

Simultaneous use of in silico design and a correlated mutation network as a tool to efficiently guide enzyme engineering

In order to improve the efficiency of directed evolution experiments, in silico multiple-substrate clustering was combined with an analysis of the variability of natural enzymes within a protein superfamily. This was applied to a Pseudomonas fluorescens esterase (PFE I) targeting the enantioselective hydrolysis of 3-phenylbutyric acid esters. Data reported in the literature for nine substrates were used for the clustering meta-analysis of the docking conformations in wild-type PFE I, and this highlighted a tryptophan residue (W28) as an interesting target. Exploration of the most frequently, naturally occurring amino acids at this position suggested that the reduced flexibility observed in the case of the W28F variant leads to enhancement of the enantioselectivity. This mutant was subsequently combined with mutations identified in a library based on analysis of a correlated mutation network. By interrogation of <80 variants a mutant with 15-fold improved enantioselectivity was found.

Enantioselective enzymes by computational design and in silico screening

Computational enzyme design holds great promise for providing new biocatalysts for synthetic chemistry. A strategy to design small mutant libraries of complementary enantioselective epoxide hydrolase variants for the production of highly enantioenriched (S,S)-diols and (R,R)-diols is developed. Key features of this strategy (CASCO, catalytic selectivity by computational design) are the design of mutations that favor binding of the substrate in a predefined orientation, the introduction of steric hindrance to prevent unwanted substrate binding modes, and ranking of designs by high-throughput molecular dynamics simulations. Using this strategy we obtained highly stereoselective mutants of limonene epoxide hydrolase after experimental screening of only 37 variants. The results indicate that computational methods can replace a substantial amount of laboratory work when developing enantioselective enzymes.

De novo design of stereochemically-bent sixteen-residue β-hairpin as a hydrolase mimic

Stepwise design of sixteen-residue β-hairpin as a hydrolase mimic involving fold design by stereochemical mutation followed by inverse-design of sequence.

Directed evolution 2.0: improving and deciphering enzyme properties

A KnowVolution: knowledge gaining directed evolution including four phases is proposed in this feature article, which generates improved enzyme variants and molecular understanding.

Computational library design for increasing haloalkane dehalogenase stability

We explored the use of a computational design framework for the stabilization of the haloalkane dehalogenase LinB. Energy calculations, disulfide bond design, molecular dynamics simulations, and rational inspection of mutant structures predicted many stabilizing mutations. Screening of these in small mutant libraries led to the discovery of seventeen point mutations and one disulfide bond that enhanced thermostability. Mutations located in or contacting flexible regions of the protein had a larger stabilizing effect than mutations outside such regions. The combined introduction of twelve stabilizing mutations resulted in a LinB mutant with a 23 °C increase in apparent melting temperature (Tm,app , 72.5 °C) and an over 200-fold longer half-life at 60 °C. The most stable LinB variants also displayed increased compatibility with co-solvents, thus allowing substrate conversion and kinetic resolution at much higher concentrations than with the wild-type enzyme.

Stabilization of cyclohexanone monooxygenase by a computationally designed disulfide bond spanning only one residue

•

Cyclohexanone monooxygenase was stabilized by an in silico designed disulfide bond.

•

Stabilizing disulfide bonds were successfully designed based on a model structure.

•

The half-life at 30 °C was increased 12-fold for the mutant enzyme.

•

The apparent melting point was increased by 6 °C for the mutant enzyme.

•

The most stabilizing disulfide bond spans only one residue.

Enzyme stability is an important parameter in biocatalytic applications, and there is a strong need for efficient methods to generate robust enzymes. We investigated whether stabilizing disulfide bonds can be computationally designed based on a model structure. In our approach, unlike in previous disulfide engineering studies, short bonds spanning only a few residues were included. We used cyclohexanone monooxygenase (CHMO), a Baeyer–Villiger monooxygenase (BVMO) from Acinetobacter sp. NCIMB9871 as the target enzyme. This enzyme has been the prototype BVMO for many biocatalytic studies even though it is notoriously labile. After creating a small library of mutant enzymes with introduced cysteine pairs and subsequent screening for improved thermostability, three stabilizing disulfide bonds were identified. The introduced disulfide bonds are all within 12 Å of each other, suggesting this particular region is critical for unfolding. This study shows that stabilizing disulfide bonds do not have to span many residues, as the most stabilizing disulfide bond, L323C–A325C, spans only one residue while it stabilizes the enzyme, as shown by a 6 °C increase in its apparent melting temperature.