Patrick C. Cirino

Protein engineering of oxygenases for biocatalysis

Oxygenase enzymes have seen limited practical applications because of their complexity, poor stabilities, and often low catalytic rates. However, their ability to perform difficult chemistry with high selectivity and specificity has kept oxygenases at the forefront of engineering efforts. Growing understanding of structure-function relationships and improved protein engineering methods are paving the way for applications of oxygenases in chemical synthesis and bioremediation.

Design and Application of a Mevalonate‐Responsive Regulatory Protein

The discovery and engineering of genetic components involved in the synthesis of a desired metabolite are often limited by a lack of sufficiently sensitive and/or rapid screening methods (or genetic selections) for the identification of gene candidates from large natural or synthetic gene libraries. Many bacterial regulatory proteins mediate direct coupling between the specific recognition of small molecules (“effectors”) and changes in gene transcription at targeted promoters. However the lack of available regulatory proteins that respond to effector molecules of interest has hindered their use in the endogenous reporting of effector synthesis through reporter-gene expression. Herein we demonstrate a new approach to metabolic-pathway engineering in which a regulatory protein is engineered to respond to an effector compound of interest and then used as a reporter in library screening for improved effector production (Figure 1). A variety of bacterial regulatory proteins that respond to small molecules have been characterized, and many have been engineered to show altered or relaxed effector specificity. 3] The Escherichia coli homodimeric AraC protein acts as a repressor of transcription at the promoter PBAD in the absence of its effector molecule l-arabinose. Upon binding l-arabinose, the AraC dimer switches conformation and contacts alternative operator half-sites, which leads to the activation of transcription at PBAD. [4] We recently described the use of multiple-site saturation mutagenesis and fluorescence-activated cell sorting (FACS) to isolate AraC variants with effector specificity switched from l-arabinose to d-arabinose. We next sought an AraC variant that could be used as a reporter for the screening of improved production of a desired metabolite. Isoprenoids constitute a large class of industrially valuable secondary metabolites. The mevalonate-dependent (MEV) isoprenoid pathway converts acetyl coenzyme A (acetylCoA) into the five-carbon-atom isoprenoid building block isopentenyl pyrophosphate (IPP). The reduction of hydroxymethylglutaryl-CoA (HMG-CoA) to mevalonate by HMGCoA reductase is a committed step in this pathway (see Figure S1 in the Supporting Information). The MEV pathway is found in eukaryotes and prokaryotes, but is not native to E. coli. The functional heterologous expression of an engineered MEV pathway in E. coli has been reported. The “MevT” operon (contained in plasmid pMevT) is composed of atoB encoding E. coli acetoacetyl-CoA thiolase, ERG13 encoding Saccharomyces cerevisiae 3-hydroxy-3methylglutaryl-CoA synthase, and a truncated HMG1 gene from S. cerevisiae (named herein tHMGR) encoding a soluble version of HMG-CoA reductase. The production of isoprenoids in E. coli through the heterologous MEV pathway is limited by mevalonate supply; the improvement of mevalonate production is therefore an important step in the metabolic engineering of microbial isoprenoid biosynthesis. Keasling and co-workers have described several approaches to the enhancement of mevalonate production in engineered E. coli. When the expression of the MevT operon was simply increased, growth and productivity decreased as a result of a variety of potentially unfavorable factors, including decreased intracellular acetyl-CoA levels, inhibitory HMG-CoA levels, the biosynthetic burden, and enzyme toxicity. Therefore, for mevalonate production to be increased, fine-tuning of individual MevT gene-expression levels is required to balance intermediate metabolite levels and prevent unfavorable growth conditions. High-throughput screening is ideal for Figure 1. In this metabolic-engineering strategy, the E. coli regulatory protein AraC is first engineered to be induced by the product of interest (effector). The customized regulator (AraC*) is then used in high-throughput screening to report effector synthesis through reporter-gene expression. In the current study, the host organism is E. coli, the effector is mevalonate, and the gene library consists of randomized bases in a mevalonate-synthesis-pathway operon.

Screening for enhanced triacetic acid lactone production by recombinant Escherichia coli expressing a designed triacetic acid lactone reporter.

Triacetic acid lactone (TAL) is a signature byproduct of polyketide synthases (PKSs) and a valuable synthetic precursor. We have developed an endogenous TAL reporter by engineering the Escherichia coli regulatory protein AraC to activate gene expression in response to TAL. The reporter enabled in vivo directed evolution of Gerbera hybrida 2-pyrone synthase activity in E. coli . Two rounds of mutagenesis and high-throughput screening yielded a variant conferring ~20-fold increased TAL production. The catalytic efficiency (kcat/Km) of the variant toward the substrate malonyl-CoA was improved 19-fold. This study broadens the utility of engineered AraC variants as customized molecular reporters. In addition, the TAL reporter can find applications in other basic PKS activity screens.

Thermostabilization of a Cytochrome P450 Peroxygenase

Cytochrome P450 BM-3 is a soluble fatty acid hydroxylase composed of a heme domain and a reductase domain on a single polypeptide chain. We recently described a laboratoryevolved variant of the P450 BM-3 heme domain which functions as an H2O2-driven hydroxylase (TMperoxygenase∫) and does not require NADPH, O2, or the reductase. This variant, which we named 21B3, allows us to carry out cytochrome P450-catalyzed biotransformations under highly simplified reaction conditions: only the heme domain and hydrogen peroxide are needed for substrate (fatty acid) hydroxylation. Because its heme domain alone is competent for catalysis, P450 BM-3 peroxygenase 21B3 offers a unique opportunity to create a thermostable, functional cytochrome P450. Here we report further directed evolution of the 21B3 peroxygenase, resulting in an enzyme which is significantly more thermostable than wildtype cytochrome P450 BM-3 and retains much of the peroxygenase activity of 21B3. Enzymes are often poorly stable under conditions encountered during production, storage, or use. Improving enzyme resistance to thermal denaturation has been a major focus of protein engineering efforts. Improved thermostability often correlates with longer shelf-life, longer life-time during use (even at low temperatures), and a higher temperature optimum for activity. 7] There have been no reports of stabilizing the relatively unstable cytochrome P450 enzymes by protein engineering, however, primarily because the P450s comprise multiple subunits and contain thermolabile cofactors. P450 BM-3 heme domain containing the single amino acid substitution F87A (mutant HF87A) is significantly more active than wildtype heme domain (HWT) in reactions driven by H2O2. 9] Variant 21B3 is much more active than HF87A in 10 mM H2O2, but is also less thermostable than HWT and HF87A. We therefore sought to improve the thermostability of 21B3 while maintaining its improved peroxygenase activity. Six cycles of random mutagenesis or DNA shuffling and screening for retention of peroxygenase activity after heat treatment (see the Methods section) yielded the thermostable peroxygenase variant 5H6. To characterize the thermostability of the peroxygenase variants identified during screening, we measured the fraction of folded heme domain remaining after heat-treatment, which we determined from the fraction of the ferrous heme-CO complex that retained the 450 nm absorbance peak characteristic of properly-folded P450. Figure 1 shows the percentage of

Regioselectivity and Activity of Cytochrome P450 BM-3 and Mutant F87A in Reactions Driven by Hydrogen Peroxide

Cytochrome P450 BM-3 (EC 1.14.14.1) is a monooxygenase that utilizes NADPH and dioxygen to hydroxylate fatty acids at subterminal positions. The enzyme is also capable of functioning as a peroxygenase in the same reaction, by utilizing hydrogen peroxide in place of the reductase do- main, cofactor and oxygen. As a starting point for developing a practically useful hydroxylation bio- catalyst, we compare the activity and regioselectivity of wild-type P450 BM-3 and its F87A mutant on various fatty acids. Neither enzyme catalyzes ter- minal hydroxylation under any of the conditions studied. While significantly enhancing peroxygenase activity, the F87A mutation also shifts hydroxylation further away from the terminal position. The H2O2- driven reactions with either the full-length BM-3 enzyme or the heme domain are slow, but yield product distributions very similar to those generated when using NADPH and O2.

Analysis of NADPH Supply During Xylitol Production by Engineered Escherichia coli

Escherichia coli strain PC09 (ΔxylB, cAMP‐independent CRP (crp*) mutant) expressing an NADPH‐dependent xylose reductase from Candida boidinii (CbXR) was previously reported to produce xylitol from xylose while metabolizing glucose [Cirino et al. (2006) Biotechnol Bioeng 95(6): 1167–1176]. This study aims to understand the role of NADPH supply in xylitol yield and the contribution of key central carbon metabolism enzymes toward xylitol production. Studies in which the expression of CbXR or a xylose transporter was increased suggest that enzyme activity and xylose transport are not limiting xylitol production in PC09. A constraints‐based stoichiometric metabolic network model was used to understand the roles of central carbon metabolism reactions and xylose transport energetics on the theoretical maximum molar xylitol yield (xylitol produced per glucose consumed), and xylitol yields (YRPG) were measured from resting cell biotransformations with various PC09 derivative strains. For the case of xylose‐proton symport, omitting the Zwf (glucose‐6‐phosphate dehydrogenase) or PntAB (membrane‐bound transhydrogenase) reactions or TCA cycle activity from the model reduces the theoretical maximum yield from 9.2 to 8.8, 3.6, and 8.0 mol xylitol (mol glucose)−1, respectively. Experimentally, deleting pgi (encoding phosphoglucose isomerase) from strain PC09 improves the yield from 3.4 to 4.0 mol xylitol (mol glucose)−1, while deleting either or both E. coli transhydrogenases (sthA and pntA) has no significant effect on the measured yield. Deleting either zwf or sucC (TCA cycle) significantly reduces the yield from 3.4 to 2.0 and 2.3 mol xylitol (mol glucose)−1, respectively. Expression of a xylose reductase with relaxed cofactor specificity increases the yield to 4.0. The large discrepancy between theoretical maximum and experimentally determined yield values suggests that biocatalysis is compromised by pathways competing for reducing equivalents and dissipating energy. The metabolic role of transhydrogenases during E. coli biocatalysis has remained largely unspecified. Our results demonstrate the importance of direct NADPH supply by NADP+‐utilizing enzymes in central metabolism for driving heterologous NADPH‐dependent reactions, and suggest that the pool of reduced cofactors available for biotransformation is not readily interchangeable via transhydrogenase. Biotechnol. Bioeng. 2009;102: 209–220.

AraC Regulatory Protein Mutants with Altered Effector Specificity

The AraC regulatory protein of the Escherichia coli ara operon has been engineered to activate transcription in response to D-arabinose and not in response to its native effector L-arabinose. Two different AraC mutant libraries, each with four randomized binding pocket residues, were subjected to FACS-mediated dual screening using a GFP reporter. Both libraries yielded mutants with the desired switch in effector specificity, and one mutant we describe maintains tight repression in the absence of effector. The presence of 100 mM L-arabinose does not influence the response of the reported mutants to D-arabinose, and the mutants are not induced by other sugars tested (D-xylose, D-fucose, D-lyxose). Co-expression of the FucP transporter in E. coli enabled induction by D-arabinose in the 0.1 mM range. Our results demonstrate the power of dual screening for altering AraC inducer specificity and represent steps toward the design of customized in vivo molecular reporters and gene switches for metabolic engineering.

Metabolic engineering for bioproduction of sugar alcohols

Sugar alcohols find applications in pharmaceuticals, oral and personal care products, and as intermediates in chemical synthesis. While industrial-scale production of these compounds has generally involved catalytic hydrogenation of sugars, microbial-based processes receive increasing attention. The past few years have seen a variety of interesting metabolic engineering efforts to improve the capabilities of bacteria and yeasts to overproduce xylitol, mannitol, and sorbitol. Examples include heterologous expression of yeast xylose reductase in Escherichia coli for the production of xylitol, coexpression of formate dehydrogenase, mannitol dehydrogenase, and a glucose facilitator protein in Corynebacterium glutamicum for mannitol production from fructose and formate, and overexpression of sorbitol-6-phosphate dehydrogenase in lactate dehydrogenase-deficient Lactobacillus plantarum to achieve nearly maximum theoretical yields of sorbitol from glucose.

Computational design of Candida boidinii xylose reductase for altered cofactor specificity

In this study we introduce a computationally‐driven enzyme redesign workflow for altering cofactor specificity from NADPH to NADH. By compiling and comparing data from previous studies involving cofactor switching mutations, we show that their effect cannot be explained as straightforward changes in volume, hydrophobicity, charge, or BLOSUM62 scores of the residues populating the cofactor binding site. Instead, we find that the use of a detailed cofactor binding energy approximation is needed to adequately capture the relative affinity towards different cofactors. The implicit solvation models Generalized Born with molecular volume integration and Generalized Born with simple switching were integrated in the iterative protein redesign and optimization (IPRO) framework to drive the redesign of Candida boidinii xylose reductase (CbXR) to function using the non‐native cofactor NADH. We identified 10 variants, out of the 8,000 possible combinations of mutations, that improve the computationally assessed binding affinity for NADH by introducing mutations in the CbXR binding pocket. Experimental testing revealed that seven out of ten possessed significant xylose reductase activity utilizing NADH, with the best experimental design (CbXR‐GGD) being 27‐fold more active on NADH. The NADPH‐dependent activity for eight out of ten predicted designs was either completely abolished or significantly diminished by at least 90%, yielding a greater than 104‐fold change in specificity to NADH (CbXR‐REG). The remaining two variants (CbXR‐RTT and CBXR‐EQR) had dual cofactor specificity for both nicotinamide cofactors.

Cost-Effective Whole-Cell Assay for Laboratory Evolution of Hydroxylases in Escherichia coli

Cytochrome P450 BM-3 from Bacillus megaterium catalyzes the subterminal hydroxylation of medium- and long-chain fatty acids at the to-1, w-2, and c-3 positions. A continuous spectrophotometric assay for P450 BM-3 based on the conversion of p-nitrophenoxycarboxylic acids (pNCA) to co-oxycarboxylic acids and the chromophore p-nitrophenolate was reported recently. However, this pNCA assay procedure contained steps that limited its application in high throughput screening, including expression of P450 BM-3 variant F87A in 4-ml cultures, centrifugation, resuspension of the cell pellet, and cell lysis. We have shown that permeabilization of the outer membrane of Escherichia coli DH5a with polymyxin B sulfate, EDTA, polyethylenimine, or sodium hexametaphosphate results in rapid conversion of 12-pNCA. A NADPH-generating system consisting of NADP+, D/L-isocitric acid, and the D/L-isocitrate dehydrogenase of E. coli DH5a reduced the cofactor expense more than 10-fold. By avoiding cell lysis, re-suspension, and centrifugation, the high throughput protocol allows screening of thousands of samples per day.