Balancing a heterologous mevalonate pathway for improved isoprenoid production in Escherichia coli
Introduction
Recent trends toward the production of complex chemicals and pharmaceuticals in engineered microbes require continual improvements in the design of non-native biosynthetic pathways. Many potentially useful natural products are either too complex to be chemically synthesized or are produced in insufficient quantities in their native host to implement cost-effective crop cultivation and extraction. Alternatively, through metabolic engineering, multi-gene heterologous pathways have been engineered into microorganisms for the production of many important classes of molecules: isoprenoids (Martin et al., 2003; Watts et al., 2005), polyketides (Pfeifer et al., 2001; Peiru et al., 2005), non-ribosomal peptides (Watts et al., 2005), bioplastics (Aldor and Keasling, 2003), and polymer building blocks (Nakamura and Whited, 2003). However, in the course of transferring enzymatic pathways from one organism to another and modulating protein production, the intricate evolved regulation of the pathway is often lost, leading to imbalances in gene expression and enzyme activity. Overexpression of a gene may cause the depletion of precursors or resources necessary for growth and production (Glick, 1995; Jones et al., 2000) or induce a stress response from excessive heterologous protein (Goff and Goldberg, 1985; Harcum and Bentley, 1993, Harcum and Bentley, 1999), while an imbalance in the total activity of the enzymes can restrict carbon flux. Such a bottleneck in the biosynthetic pathway results in a reduced rate of production (Barbirato et al., 1996; Zhu et al., 2002) and can lead to the accumulation of intermediates and byproducts that inhibit pathway enzymes (Berry et al., 2002) or are cytotoxic (Barbirato et al., 1996; Zhu et al., 2002, Zhu et al., 2001).
Engineering metabolic pathways in microbes for the production of scarce therapeutic natural products effective against neglected diseases is an attractive application of this technology. For example, to combat the increasing occurrence of malaria-causing Plasmodium strains that are resistant to traditional medications, clinicians have employed the potent anti-malarial drug artemisinin, an isoprenoid natural product extracted from Artemisia annua. Unfortunately, artemisinin yields from plant extracts are low, limiting production from Artemisia crops and increasing the cost of artemisinin-based treatments beyond the reach of people in countries most afflicted by malaria. For this reason, we engineered the bacterium Escherichia coli as a factory for the synthesis of complex isoprenoids such as artemisinin.
One of the largest obstacles to efficient microbial biosynthesis of isoprenoids is the production of isoprenoid precursors, isopentenyl pyrophosphate (IPP) and dimethylallyl pyrophosphate (DMAPP). Both compounds are produced naturally in E. coli through the 1-deoxyxylulose-5-phosphate (DXP) pathway (Rohmer et al., 1993; Lange et al., 2000) (Fig. 1A). Metabolic engineering to overproduce the isoprenoid precursors in E. coli has focused on optimizing the native DXP pathway and cellular metabolism to accumulate high levels of carotenoids (Farmer and Liao, 2000; Alper et al., 2005b, Alper et al., 2006; Yuan et al., 2006). Combining up-regulation of the entire DXP pathway with systematic and combinatorial gene knockouts resulted in the production of 0.22 g/L lycopene (Alper et al., 2006). This study focuses on an alternative method to over-produce isoprenoid precursors in E. coli by cloning and expressing the high flux, mevalonate-dependent, isoprenoid pathway from Saccharomyces cerevisiae (Martin et al., 2003) (Fig. 1A). Using the synthesis of a precursor to the potent anti-malarial artemisinin, amorpha-4,11-diene, as an example, this system demonstrated high isoprenoid production capability, producing 0.48 g/L amorphadiene in two-phase fermentation (Newman et al., 2006). This fermentation titer is similar to the aforementioned studies employing the native DXP pathway and represents a 2400-fold improvement in production over wild-type E. coli. However, neither production system currently achieves the isoprenoid titer of 25 g/L estimated to be necessary for providing inexpensive artemisinin to countries most afflicted by malaria (Ro et al., 2006).
In the expression of a multi-gene heterologous pathway, the activity of a single enzyme may be out of balance with that of the other enzymes in the pathway, leading to unbalanced carbon flux and the accumulation of an intermediate. In this paper we describe how balancing carbon flux through the heterologous mevalonate pathway resulted in improved growth and isoprenoid production in E. coli. This study demonstrates the importance of retaining balanced flux in a reconstituted heterologous pathway and a methodology for troubleshooting and balancing pathways. Whether the molecule is native or foreign to the host, the unregulated accumulation of any intracellular compound may compromise the viability of the organism or the productivity of the pathway.
Section snippets
Strains and media
E. coli strains TOP10 and DH10B, both from Invitrogen (Carlsbad, CA), were used for cloning and plasmid construction. In E. coli DH10B, the PBAD promoter system suffers from all-or-none induction, in which sub-saturating concentrations of arabinose give rise to subpopulations of cells that are fully induced and un-induced (Khlebnikov et al., 2000). To alleviate this problem for gene titration studies, a DH10B host with regulatable control of PBAD in a homogeneous population of cells was
Initial steps of heterologous pathway limits carbon flux to amorphadiene
To identify the limiting steps in the heterologous pathway, E. coli DH10B cultures expressing the full pathway from acetyl-CoA to amorphadiene (harboring plasmids pMevT, pMBIS, and pADS—Fig. 1A) were grown in medium supplemented with increasing concentrations of exogenous mevalonate. GC–MS analysis revealed that the addition of exogenous mevalonate to the medium increased the production of amorphadiene over time, above that produced with no mevalonate supplementation (Fig. 2). Since
Discussion
Examination of metabolites of a mevalonate pathway expressed in E. coli revealed that this heterologous pathway is unbalanced, resulting in growth inhibition in some engineered strains with high expression levels of the MevT operon. Specifically, we show that HMG-CoA accumulates in cells with high MevT expression and this metabolite inhibits cell growth. Co-expression with additional tHMGR, which converts HMG-CoA to non-toxic mevalonate, relieves this toxicity and restores growth nearly to the
Acknowledgments
The authors would like to acknowledge David Serp, Vincent J.J. Martin, Lance Kizer, and Brian Pfleger for their contributions. This research was conducted under the sponsorship of the Institute for One World Health through the generous support of The Bill and Melinda Gates Foundation and by grants from the National Science Foundation (BES-9911463), the Office of Naval Research (FDN00014-99-0182), and University of California Discovery Grant Program (99-10044). D.J.P is the recipient of a
References (62)
- et al.
Process design for microbial plastic factories: metabolic engineering of polyhydroxyalkanoates
Curr. Opin. Biotechnol.
(2003) - et al.
Tightly regulated tac promoter vectors useful for the expression of unfused and fused proteins in E. coli
Gene
(1988) - et al.
Charges of nicotinamide adenine nucleotides and adenylate energy charge as regulatory parameters of the metabolism in E. coli
J. Biol. Chem.
(1977) - et al.
3-Hydroxy-3-methylglutaryl coenzyme A lyase: targeting and processing in peroxisomes and mitochondria
J. Lipid Res.
(1999) - et al.
Molecular and catalytic properties of the acetoacetyl-coenzyme A thiolase of E. coli
Arch. Biochem. Biophys.
(1976) Metabolic load and heterologous gene expression
Biotechnol. Adv.
(1995)- et al.
Production of abnormal proteins in E. coli stimulates transcription of lon and other heat shock genes
Cell
(1985) - et al.
The biology of HMG-CoA reductase: the pros of contra-regulation
Trends Biochem. Sci.
(1996) - et al.
Regulation of malonyl-CoA metabolism by acyl-acyl carrier protein and beta-ketoacyl-acyl carrier protein synthases in E. coli
J. Biol. Chem.
(1995) - et al.
DNA synthesis during the division cycle of rapidly growing E. coli B/r
J. Mol. Biol.
(1968)