Metabolic pathway optimization using ribosome binding site variants and combinatorial gene assembly.

The genes encoding the mevalonate-based farnesyl pyrophosphate (FPP) biosynthetic pathway were encoded in two operons and expressed in Escherichia coli to increase the production of sesquiterpenes. Inefficient translation of several pathway genes created bottlenecks and led to the accumulation of several pathway intermediates, namely, mevalonate and FPP, and suboptimal production of the sesquiterpene product, amorphadiene. Because of the difficulty in choosing ribosome binding sites (RBSs) to optimize translation efficiency, a combinatorial approach was used to choose the most appropriate RBSs for the genes of the lower half of the mevalonate pathway (mevalonate to amorphadiene). RBSs of various strengths, selected based on their theoretical strengths, were cloned 5' of the genes encoding mevalonate kinase, phosphomevalonate kinase, mevalonate diphosphate decarboxylase, and amorphadiene synthase. Operons containing one copy of each gene and all combinations of RBSs were constructed and tested for their impact on growth, amorphadiene production, enzyme level, and accumulation of select pathway intermediates. Pathways with one or more inefficiently translated enzymes led to the accumulation of pathway intermediates, slow growth, and low product titers. Choosing the most appropriate RBS combination and carbon source, we were able to reduce the accumulation of toxic metabolic intermediates, improve growth, and improve the production of amorphadiene approximately fivefold. This work demonstrates that balancing flux through a heterologous pathway and maintaining steady growth are key determinants in optimizing isoprenoid production in microbial hosts.

Cloning of casbene and neocembrene synthases from Euphorbiaceae plants and expression in Saccharomyces cerevisiae.

A large number of diterpenes have been isolated from Euphorbiaceae plants, many of which are of interest due to toxicity or potential therapeutic activity. Specific Euphorbiaceae diterpenes of medical interest include the latent HIV-1 activator prostratin (and related 12-deoxyphorbol esters), the analgesic resiniferatoxin, and the anticancer drug candidate ingenol 3-angelate. In spite of the large number of diterpenes isolated from these plants and the similarity of their core structures, there is little known about their biosynthetic pathways. Other than the enzymes involved in gibberellin biosynthesis, the only diterpene synthase isolated to date from the Euphorbiaceae has been casbene synthase, responsible for biosynthesis of a macrocyclic diterpene in the castor bean (Ricinus communis). Here, we have selected five Euphorbiaceae species in which to investigate terpene biosynthesis and report on the distribution of diterpene synthases within this family. We have discovered genes encoding putative casbene synthases in all of our selected Euphorbiaceae species and have demonstrated high-level casbene production through expression of four of these genes in a metabolically engineered strain of Saccharomyces cerevisiae. The only other diterpene synthase found among the five plants was a neocembrene synthase from R. communis (this being the first report of a neocembrene synthase gene). Based on the prevalence of casbene synthases, the lack of other candidates, and the structure of the casbene skeleton, we consider it likely that casbene is the precursor to a large number of Euphorbiaceae diterpenes. Casbene production levels of 31 mg/L were achieved in S. cerevisiae and we discuss strategies to further increase production by maximizing flux through the mevalonate pathway.

Optimization of the mevalonate-based isoprenoid biosynthetic pathway in Escherichia coli for production of the anti-malarial drug precursor amorpha-4,11-diene.

The introduction or creation of metabolic pathways in microbial hosts has allowed for the production of complex chemicals of therapeutic and industrial importance. However, these pathways rarely function optimally when first introduced into the host organism and can often deleteriously affect host growth, resulting in suboptimal yields of the desired product. Common methods used to improve production from engineered biosynthetic pathways include optimizing codon usage, enhancing production of rate-limiting enzymes, and eliminating the accumulation of toxic intermediates or byproducts to improve cell growth. We have employed these techniques to improve production of amorpha-4,11-diene (amorphadiene), a precursor to the anti-malarial compound artemisinin, by an engineered strain of Escherichia coli. First we developed a simple cloning system for expression of the amorphadiene biosynthetic pathway in E. coli, which enabled the identification of two rate-limiting enzymes (mevalonate kinase (MK) and amorphadiene synthase (ADS)). By optimizing promoter strength to balance expression of the encoding genes we alleviated two pathway bottlenecks and improved production five fold. When expression of these genes was further increased by modifying plasmid copy numbers, a seven-fold increase in amorphadiene production over that from the original strain was observed. The methods demonstrated here are applicable for identifying and eliminating rate-limiting steps in other constructed biosynthetic pathways.

Redesigning enzymes based on adaptive evolution for optimal function in synthetic metabolic pathways.

Nature has balanced most metabolic pathways such that no one enzyme in the pathway controls the flux through that pathway. However, unnatural or nonnative, constructed metabolic pathways may have limited product flux due to unfavorable in vivo properties of one or more enzymes in the pathway. One such example is the mevalonate-based isoprenoid biosynthetic pathway that we previously reconstructed in Escherichia coli. We have used a probable mechanism of adaptive evolution to engineer the in vivo properties of two enzymes (3-hydroxy-3-methylglutaryl-CoA reductase [tHMGR] and many terpene synthases) in this pathway and thereby eliminate or minimize the bottleneck created by these inefficient or nonfunctional enzymes. Here, we demonstrate how we significantly improved the productivity (by approximately 1000 fold) of this reconstructed biosynthetic pathway using this strategy. We anticipate that this strategy will find broad applicability in the functional construction (or reconstruction) of biological pathways in heterologous hosts.

Metabolic flux elucidation for large-scale models using 13C labeled isotopes.

A key consideration in metabolic engineering is the determination of fluxes of the metabolites within the cell. This determination provides an unambiguous description of metabolism before and/or after engineering interventions. Here, we present a computational framework that combines a constraint-based modeling framework with isotopic label tracing on a large scale. When cells are fed a growth substrate with certain carbon positions labeled with (13)C, the distribution of this label in the intracellular metabolites can be calculated based on the known biochemistry of the participating pathways. Most labeling studies focus on skeletal representations of central metabolism and ignore many flux routes that could contribute to the observed isotopic labeling patterns. In contrast, our approach investigates the importance of carrying out isotopic labeling studies using a more comprehensive reaction network consisting of 350 fluxes and 184 metabolites in Escherichia coli including global metabolite balances on cofactors such as ATP, NADH, and NADPH. The proposed procedure is demonstrated on an E. coli strain engineered to produce amorphadiene, a precursor to the antimalarial drug artemisinin. The cells were grown in continuous culture on glucose containing 20% [U-(13)C]glucose; the measurements are made using GC-MS performed on 13 amino acids extracted from the cells. We identify flux distributions for which the calculated labeling patterns agree well with the measurements alluding to the accuracy of the network reconstruction. Furthermore, we explore the robustness of the flux calculations to variability in the experimental MS measurements, as well as highlight the key experimental measurements necessary for flux determination. Finally, we discuss the effect of reducing the model, as well as shed light onto the customization of the developed computational framework to other systems.

High-level production of amorpha-4,11-diene in a two-phase partitioning bioreactor of metabolically engineered Escherichia coli.

Reconstructing synthetic metabolic pathways in microbes holds great promise for the production of pharmaceuticals in large-scale fermentations. By recreating biosynthetic pathways in bacteria, complex molecules traditionally harvested from scarce natural resources can be produced in microbial cultures. Here we report on a strain of Escherichia coli containing a heterologous, nine-gene biosynthetic pathway for the production of the terpene amorpha-4,11-diene, a precursor to the anti-malarial drug artemisinin. Previous reports have underestimated the productivity of this strain due to the volatility of amorphadiene. Here we show that amorphadiene evaporates from a fermentor with a half-life of about 50 min. Using a condenser, we take advantage of this volatility by trapping the amorphadiene in the off-gas. Amorphadiene was positively identified using nuclear magnetic resonance spectroscopy and determined to be 89% pure as collected. We captured amorphadiene as it was produced in situ by employing a two-phase partitioning bioreactor with a dodecane organic phase. Using a previously characterized caryophyllene standard to calibrate amorphadiene production and capture, the concentration of amorphadiene produced was determined to be 0.5 g/L of culture medium. A standard of amorphadiene collected from the off-gas showed that the caryophyllene standard overestimated amorphadiene production by approximately 30%.