ABSTRACT
Moving cannabinoid production away from the vagaries of plant extraction and into engineered microbes could provide a consistent, purer, cheaper and environmentally benign source of these important therapeutic molecules, but microbial production faces notable challenges. An alternative to microbes and plants is to remove the complexity of cellular systems by employing enzymatic biosynthesis. Here we design and implement a new cell-free system for cannabinoid production with the following features: (1) only low-cost inputs are needed; (2) only 12 enzymes are employed; (3) the system does not require oxygen and (4) we use a nonnatural enzyme system to reduce ATP requirements that is generally applicable to malonyl-CoA-dependent pathways such as polyketide biosynthesis. The system produces ~0.5 g l-1 cannabigerolic acid (CBGA) or cannabigerovarinic acid (CBGVA) from low-cost inputs, nearly two orders of magnitude higher than yeast-based production. Cell-free systems such as this may provide a new route to reliable cannabinoid production.
Subject(s)
Cannabinoids/biosynthesis , Cell-Free System/metabolism , Malonyl Coenzyme A/metabolism , Metabolic Engineering/methods , Polyketides/metabolism , Terpenes/metabolism , Adenosine Triphosphate/biosynthesis , Benzoates/isolation & purification , Benzoates/metabolism , Cannabinoids/isolation & purification , Cell-Free System/chemistry , Escherichia coli/enzymology , Escherichia coli/genetics , Gene Expression , Humans , Kinetics , Metabolic Engineering/economics , Organophosphates/metabolism , Plasmids/chemistry , Plasmids/metabolism , Polyketides/chemistry , Recombinant Proteins/biosynthesis , Recombinant Proteins/genetics , Recombinant Proteins/isolation & purification , Terpenes/chemistry , ThermodynamicsABSTRACT
Metabolic engineering efforts that harness living organisms to produce natural products and other useful chemicals face inherent difficulties because the maintenance of life processes often runs counter to our desire to maximize important production metrics. These challenges are particularly problematic for commodity chemical manufacturing where cost is critical. A cell-free approach, where biochemical pathways are built by mixing desired enzyme activities outside of cells, can obviate problems associated with cell-based methods. Yet supplanting cell-based methods of chemical production will require the creation of self-sustaining, continuously operating systems where input biomass is converted into desired products at high yields, productivities, and titers. We call the field of designing and implementing reliable and efficient enzyme systems that replace cellular metabolism, synthetic biochemistry.
Subject(s)
Biochemistry/trends , Cell-Free System , Metabolic Engineering , Synthetic Biology/trends , BiomassABSTRACT
In the original version of this Article, the genotype of the M30 mutant presented in Fig. 3b was given incorrectly as Y288V/A232S, and the M31 mutant was given incorrectly as M1/A232S. The correct genotype of the M30 mutant is Y288A/A232S and for M31 it is Y288V/A232S. In addition, to keep consistency in genotype formatting, the genotype of the M27 mutant should be Y288V/G286S. The errors have been corrected in both the PDF and HTML versions of the Article.
ABSTRACT
Prenylation of natural compounds adds structural diversity, alters biological activity, and enhances therapeutic potential. Because prenylated compounds often have a low natural abundance, alternative production methods are needed. Metabolic engineering enables natural product biosynthesis from inexpensive biomass, but is limited by the complexity of secondary metabolite pathways, intermediate and product toxicities, and substrate accessibility. Alternatively, enzyme catalyzed prenyl transfer provides excellent regio- and stereo-specificity, but requires expensive isoprenyl pyrophosphate substrates. Here we develop a flexible cell-free enzymatic prenylating system that generates isoprenyl pyrophosphate substrates from glucose to prenylate an array of natural products. The system provides an efficient route to cannabinoid precursors cannabigerolic acid (CBGA) and cannabigerovarinic acid (CBGVA) at >1 g/L, and a single enzymatic step converts the precursors into cannabidiolic acid (CBDA) and cannabidivarinic acid (CBDVA). Cell-free methods may provide a powerful alternative to metabolic engineering for chemicals that are hard to produce in living organisms.
Subject(s)
Biological Products/metabolism , Cannabinoids/metabolism , Fungal Proteins/metabolism , Gas Chromatography-Mass Spectrometry , Metabolic Engineering/methods , Molecular Structure , Prenylation/physiology , Substrate SpecificityABSTRACT
Non-heme Fe(II)/α-ketoglutarate (αKG)-dependent oxygenases catalyze a wide array of reactions through coupling oxidative decarboxylation of αKG to substrate oxygenation. This class of enzymes follows a sequential mechanism in which O2 reacts only after binding primary substrate, raising questions over how protein structure tailors molecular access to the Fe(II) cofactor. The enzyme "factor inhibiting hypoxia inducible factor" (FIH) senses pO2 in human cells by hydroxylating the C-terminal transactivation domain (CTAD), suggesting that structural elements limiting molecular access to the active site may limit the pO2 response. In this study, we tested the impact of a solvent-accessible tunnel in FIH on molecular access to the active site in FIH. The size of the tunnel was increased through alanine point mutagenesis (Y93A, E105A, and Q147A), followed by a suite of mechanistic and spectroscopic probes. Steady-state kinetics varying O2 or CTAD indicated that O2 passage through the tunnel was not affected by Ala substitutions, allowing us to conclude that this narrow tunnel did not impact pO2 sensing by FIH. Steady-state kinetics with varied αKG concentrations revealed increased substrate inhibition for the Ala variants, suggesting that a second αKG molecule may bind near the active site of FIH. If this solvent-accessible tunnel is the O2 entry tunnel, it may be narrow in order to permit O2 access while preventing metabolic intermediates, such as αKG, from inhibiting FIH under physiological conditions.
Subject(s)
Hypoxia-Inducible Factor 1/antagonists & inhibitors , Hypoxia-Inducible Factor 1/chemistry , Ketoglutaric Acids/metabolism , Oxygenases/metabolism , Catalytic Domain , Citric Acid/chemistry , Citric Acid/pharmacology , Dose-Response Relationship, Drug , Electrophoresis, Polyacrylamide Gel , Genetic Variation , Humans , Hypoxia-Inducible Factor 1/genetics , Ketoglutaric Acids/chemistry , Kinetics , Oxaloacetic Acid/chemistry , Oxaloacetic Acid/pharmacology , Oxygenases/chemistry , Solvents/chemistrySubject(s)
Ketoglutaric Acids/metabolism , Mixed Function Oxygenases/metabolism , Repressor Proteins/metabolism , Humans , Kinetics , Mixed Function Oxygenases/chemistry , Mixed Function Oxygenases/genetics , Models, Biological , Mutagenesis, Site-Directed , Oxygen/metabolism , Oxygen Isotopes , Peptide Fragments/chemistry , Peptide Fragments/genetics , Peptide Fragments/metabolism , Protein Structure, Tertiary , Recombinant Proteins/chemistry , Recombinant Proteins/genetics , Recombinant Proteins/metabolism , Repressor Proteins/chemistry , Repressor Proteins/genetics , Solvents , ViscosityABSTRACT
The Fe(ii)/αketoglutarate (αKG) dependent oxygenases catalyze a diverse range of reactions significant in biological processes such as antibiotic biosynthesis, lipid metabolism, oxygen sensing, and DNA and RNA repair. Although functionally diverse, the eight-stranded ß-barrel (cupin) and HX(D/E)XnH facial triad motifs are conserved in this super-family of enzymes. Crystal structure analysis of 25 αKG oxygenases reveals two stereoisomers of the Fe cofactor, Anti and Clock, which differ in the relative position of the exchangeable ligand position and the primary substrate. Herein, we discuss the relationship between the chemical mechanism and the secondary coordination sphere of the αKG oxygenases, within the constraints of the stereochemistry of the Fe cofactor. Sequence analysis of the cupin barrel indicates that a small subset of positions constitute the second coordination sphere, which has significant ramifications for the structure of the ferryl intermediate. The competence of both Anti and Clock stereoisomers of Fe points to a ferryl intermediate that is 5 coordinate. The small number of conserved close contacts within the active sites of αKG oxygenases can be extended to chemically related enzymes, such as the αKG-dependent halogenases SyrB2 and CytC3, and the non-αKG dependent dioxygenases isopenicillin N synthase (IPNS) and cysteine dioxygenase (CDO).