Application of a Propionyl Coenzyme A Synthetase for Poly(3-Hydroxypropionate-co-3-Hydroxybutyrate) Accumulation in Recombinant Escherichia coli

Abstract

The genetic operon for propionic acid degradation in Salmonella enterica serovar Typhimurium contains an open reading frame designated prpE which encodes a propionyl coenzyme A (propionyl-CoA) synthetase (A. R. Horswill and J. C. Escalante-Semerena, Microbiology 145:1381–1388, 1999). In this paper we report the cloning of prpE by PCR, its overexpression in Escherichia coli, and the substrate specificity of the enzyme. When propionate was utilized as the substrate for PrpE, a K_m of 50 μM and a specific activity of 120 μmol · min⁻¹ · mg⁻¹ were found at the saturating substrate concentration. PrpE also activated acetate, 3-hydroxypropionate (3HP), and butyrate to their corresponding coenzyme A esters but did so much less efficiently than propionate. When prpE was coexpressed with the polyhydroxyalkanoate (PHA) biosynthetic genes from Ralstonia eutropha in recombinant E. coli, a PHA copolymer containing 3HP units accumulated when 3HP was supplied with the growth medium. To compare the utility of acyl-CoA synthetases to that of an acyl-CoA transferase for PHA production, PHA-producing recombinant strains were constructed to coexpress the PHA biosynthetic genes with prpE, with acoE (an acetyl-CoA synthetase gene from R. eutropha [H. Priefert and A. Steinbüchel, J. Bacteriol. 174:6590–6599, 1992]), or with orfZ (an acetyl-CoA:4-hydroxybutyrate-CoA transferase gene from Clostridium propionicum [H. E. Valentin, S. Reiser, and K. J. Gruys, Biotechnol. Bioeng. 67:291–299, 2000]). Of the three enzymes, PrpE and OrfZ enabled similar levels of 3HP incorporation into PHA, whereas AcoE was significantly less effective in this capacity.

Polyhydroxyalkanoates (PHAs) are a diverse group of bacterial storage polyesters, which are accumulated when carbon and energy sources are available in excess and cell growth is restricted by the lack of an essential nutrient (1, 25). Some of these polyesters exhibit material characteristics comparable to those of petrochemical-derived polymers. For example, the physical properties of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) resemble those of polyethylene and polypropylene (8). In contrast to petrochemical-based polymers, PHAs are completely biodegradable to CO₂ and water and can be produced from renewable resources. Unfortunately, PHA production by bacterial fermentation is costly and, due to inefficient use of resources, not necessarily environmentally benign (6). However, if a plant-based PHA production system can be implemented, significantly lower production costs are predicted (12, 17).

Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) has been commercially produced through fermentation using a glucose-utilizing mutant of Ralstonia eutropha that requires cofeeding of propionic acid for 3-hydroxyvalerate formation. This polymer was sold under the trade name Biopol. In contrast to a bacterial production system, production of PHA copolyesters in plants requires that all metabolites be derived from available intermediates of plant metabolism. While the production of the simple but brittle homopolymer poly(3-hydroxybutyrate) (PHB) has been well established in recombinant bacterial and plant systems (18, 21, 24), only recent investigations have demonstrated the engineered production of more complex PHAs (7, 13, 16, 22, 23, 28). Production of poly(3-hydroxypropionate-co-3-hydroxybutyrate) (PHPB) in a recombinant system like Escherichia coli has not been reported. This particular copolymer, with incorporated 3-hydroxypropionate (3HP) monomers in the polymer backbone, has reduced crystallinity relative to that of PHB (J. Asrar, personal communication).

The PHB biosynthetic pathway in R. eutropha is amenable to a plant production system, since the starting biosynthetic metabolite is acetyl coenzyme A (acetyl-CoA). The pathway enzymes consist of a 3-ketothiolase (PhaA), an acetoacetyl-CoA reductase (PhaB), and a PHB synthase (PhaC) (21, 24). The 3-ketothiolase catalyzes a Claisen condensation of two molecules of acetyl-CoA to form acetoacetyl-CoA, which is then reduced in an NADPH-dependent reaction to form D-(−)-3-hydroxybutyryl-CoA (3HB-CoA). 3HB-CoA serves as substrate for the PHB synthase, which catalyzes the polymerization reaction of 3HB-CoA to form PHB.

In PHA-producing microbial systems only a small fraction of the diverse group of PHAs are obtained from structurally unrelated carbon sources (25), and those pathways are only partially characterized. Successful engineering of plants for PHA production will require greater understanding of both the enzymes involved in these microbial pathways and the complications of cellular compartmentalization of biochemical processes in plants. Information on enzymes potentially useful for PHA production in designed pathways, but not necessarily naturally involved in PHA biosynthesis, will also be of importance. One such enzyme is PrpE, a propionyl-CoA synthetase involved in propionate degradation in Salmonella enterica serovar Typhimurium (9).

All known PHA synthases require hydroxyacyl-CoAs as substrates. Precursor organic acids must therefore be activated to CoA-thioesters before entering the PHA biosynthetic pathway (5, 27). Although CoA-activated short-chain fatty acids are required for a wide variety of other metabolic pathways, only acetyl-CoA synthetases or acyl-CoA synthetases with a chain length specificity of greater than 10 carbon atoms have been characterized in detail. Acyl-CoA synthetases with preference for fatty acids of chain lengths in the range of C₃ to C₁₀ have not yet been characterized. The characterization of one such enzyme in this class, PrpE, is reported in this communication.

In addition to acyl-CoA synthetases, acyl-CoA transferases found in anaerobic bacteria are known to catalyze the formation of short- to medium-chain-length CoA-thioesters (14). However, these enzymes are thought to be less suitable for metabolic engineering, since they depend on the availability of a donor CoA-thioester substrate in addition to the free organic acid for sufficient production of the desired organic acyl-CoA. These enzymes do not use the free energy of nucleoside triphosphate hydrolysis to drive CoA-thioester formation as do acyl-CoA synthetases.

A comparison of an acyl-CoA synthetase to an acyl-CoA transferase in an integrated pathway for PHA production has not yet been done. For these reasons, and to learn more about the significance of substrate specificity, we compared PHPB formation under similar conditions using prpE, acoE (encoding an acetyl-CoA synthetase from R. eutropha [20]), or orfZ (encoding an acetyl-CoA:4-hydroxybutyrate-CoA transferase from Clostridium kluyveri [26]) coexpressed with the PHA biosynthetic operon from R. eutropha (phaCAB) in E. coli. All of these enzymes (AcoE, OrfZ, and PrpE) have been found to activate 3HP to 3HP-CoA under in vitro conditions (reference 20 and this study).

MATERIALS AND METHODS

Bacterial strains and plasmids.

All strains and plasmids used in this study are listed in Table 1. Maps of expression plasmids are shown in Fig. 1. For routine cloning, plasmids were introduced into E. coli DH5α. For PHA accumulation experiments, plasmids were transferred into E. coli XL1-Blue.

TABLE 1.

Bacterial strains and plasmids

Bacterial strain or plasmid	Relevant characteristics	Source or reference
E. coli DH5α	F′ φ80dlacZΔM15Δ(lacZYA-argF)U169 deoR recA1 endA1 phoA hsdR17 (rK− mK+) supE44 λ− thi-1 gyrA96 rel A1	Gibco-BRL
E. coli XL1-Blue	F′::Tn10proA+B+lacIq Δ(lacZ)M15/recA1 endA gyrA96 (Nalr) thi hsdR17 supE44 relA1 lac	Stratagene
S. enterica serovar Typhimurium LT2	prpE+	ATCC 29946
Plasmids
pJM9238	Chlr, tac promoter, phaCAB	11
pMON34555	pSP72::prpE	This study
pMON34564	pSE380::prpE	This study
pMON34610	Ampr, lac promoter	26
pMON34576	pMON34610::prpE	This study
pSE380	Ampr, lacIq, trc promoter	Invitrogen Co.
pSP72	Ampr	GenBank accession no. X65332
pSES38	Ampr, acoE	20
pBluescript KS−::orfZ	Ampr, orfZ	This study

FIG. 1.

Plasmids used for PHA accumulation experiments in this study. The runaway replication vector pJM9238 harbors the R. eutropha PHA biosynthetic operon (phaCAB) under tac promoter control (11). PHA accumulation is induced by heat shock or by growing the bacteria at temperatures exceeding 34°C. Plasmid pSES38 harbors a 3.8-kbp fragment of genomic DNA from R. eutropha encoding an acetyl-CoA synthetase required for acetoin degradation. The gene is collinear with the lac promoter of the pBluescript parent vector. However, gene expression is thought to be driven by an internal ς⁷⁰ dependent promoter. The acetyl-CoA:4-hydroxybutyrate-CoA transferase (OrfZ) is expressed from pBluescript KS⁻::orfZ, which harbors the orfZ gene on a 1.8-kbp ClaI/EcoRI fragment of genomic DNA from C. kluyveri, collinear with the lacZ promoter. The multicopy vector pMON34576 harbors the engineered prpE gene under lac promoter control.

Media.

For batch culture studies, bacteria were grown in Luria-Bertani (LB) medium (15) that contained the appropriate antibiotics (Sigma, St. Louis, Mo.) at the following concentrations: ampicillin, 100 μg · ml⁻¹; chloramphenicol, 25 μg · ml⁻¹. When specified, isopropyl-β-d-thiogalactopyranoside (IPTG) (Sigma) was added at a final concentration of 1 mM. Unless otherwise stated, all cultures were incubated at 37°C and shaken at 225 rpm in an orbital shaker. Cell growth was monitored by measuring optical density at a wavelength of 600 nm (OD₆₀₀).

PCR.

For amplification of prpE, an aliquot of total genomic DNA from S. enterica serovar Typhimurium was incubated for 41 cycles with prpE-specific primers (upper primer, 5′-GGGGGGGAATTCAGATCTCCATGGGCATGCCTTTTAGCGAATTTTATCAGCGTTCG; lower primer, 5′-GGGGGGGAATTCTAATAACCCGTTGCCGAACGCGGCCTTATCCGGC) (Gibco-BRL, Rockville, Md.) in a thermocycler (DNA Thermal Cycler, Perkin-Elmer, Norwalk, Conn.) using a Boehringer (Mannheim, Germany) PCR core kit. To relax GC-rich DNA, 10% (vol/vol) dimethyl sulfoxide was added to each reaction mixture. The first PCR amplification cycle was done under the following conditions: 2 min of incubation at 95°C for denaturation, 1 min at 50°C for annealing, and 2 min at 72°C for extension. All other amplification cycles were done with 1 min of incubation at 94°C for denaturation, 1 min for annealing at 50°C, and 2 min for extension at 72°C.

Plasmid construction.

For cloning of prpE, the PCR products were purified using a Qiagen (Valencia, Calif.) PCR purification kit, digested with EcoRI, and ligated into EcoRI-digested pSP72, resulting in the formation of pMON34555. For high-level expression, prpE was subcloned under the control of the trc promoter into pSE380 (Invitrogen, Carlsbad, Calif.), resulting in the formation of pMON34564. Cloning of pMON34564 was performed as follows. pMON34555 was digested with NcoI and EcoRI. The 1,933-bp fragment encoding the prpE gene was separated from the vector component in a 1.0% agarose gel in Tris-acetate buffer (15). The isolated fragment was purified using a Qiagen gel extraction kit and ligated into the NcoI- and EcoRI-digested and purified pSE380. For PHA accumulation experiments, prpE was subcloned under lac promoter control into pMON34610, resulting in the formation of pMON34576. For cloning of pMON34576, plasmid pMON34564 was digested with NcoI and EcoRI. The 1,933-bp fragment was isolated and purified as described above, and the isolated fragment was ligated into the NcoI- and EcoRI-digested pMON34610.

Plasmid pBluescriptvector KS⁻::orfZ was obtained by digesting pCK3 (28) with ClaI and EcoRI. Restriction fragments were separated in a 0.8% agarose gel. The 1.8-kbp fragment carrying orfZ was purified and ligated into ClaI- and EcoRI-digested pBluescriptvector KS⁻ (Stratagene, La Jolla, Calif.).

Acyl-CoA synthetase assay.

The general acyl-CoA synthetase assay mixture contained 5 mM ATP, 10 mM organic acid sodium salt, 1.25 mM CoASH, 5 mM dithiothreitol (DTT), and 5 mM magnesium chloride in 100 mM potassium phosphate buffer (pH 7.5). To start the reaction, 20 μl of enzyme sample was added to a final volume of 200 μl of assay mixture. After 15 min of incubation at room temperature, the reaction was quenched by adding 20 μl of 10% (vol/vol) formic acid. Acyl-CoA reaction products were separated by high-pressure liquid chromatography on a reversed-phase column (Beckman C₈; 5 μm, 4.6 mm by 15 cm) using a 5 to 45% acetonitrile gradient in 50 mM ammonium acetate buffer (pH 6.0). The linear gradient was obtained by altering the ratio of buffers A (50 mM ammonium acetate buffer [pH 6.0] plus 5% acetonitrile) and B (acetonitrile) from 100% buffer A at the beginning of each run to 60% buffer A and 40% buffer B within 15 min. Peak elution was monitored by absorbance at 260 nm. Quantitation of acyl-CoA products was accomplished using a generated standard curve with acyl-CoA standards. Assays to determine kinetic constants were done by varying the concentration of organic acid substrate in a series of reactions. All assays were quenched in the steady-state portion of the reaction, where there was less than 20% utilization of the limiting substrate.

Purification of PrpE.

A 3-ml preculture of E. coli DH5α cells harboring pMON34564 was grown overnight at 37°C. Subsequently, a 250-ml LB culture was inoculated with 1 ml of the preculture and incubated at 37°C. When an OD₆₀₀ of 0.6 was obtained, IPTG was added at a final concentration of 1 mM to induce PrpE expression. Following 2 h after induction, cells were harvested by centrifugation and washed once in phosphate-buffered saline (1 mM KH₂PO₄, 10 mM Na₂HPO₄, 137 mM NaCl, 2.7 mM KCl [pH 7.4]. Subsequently, the cells were resuspended in 20 mM potassium phosphate buffer (pH 7.4) containing 20% (vol/vol) glycerol and 2 mM DTT. Cell suspensions were disrupted by sonication (Sonifier 450; Branson, Danbury, Conn.) for a period of 2 mins (30-s sonication followed by 30-s rest) at setting 0.3 for output control and 30% for the duty cycler using a 3-mm probe. The cell lysate was cleared by 10 min of centrifugation at 31,000 × g using a Beckman SA-17 rotor. Proteins were precipitated with 80% saturated ammonium sulfate and sedimented by centrifugation, and the resulting pellet was dissolved in buffer B (100 mM sodium phosphate buffer) [pH 7.0] containing 1 mM DTT and 1 M ammonium sulfate). The solution was passed through a 0.2-μm-pore-size membrane filter (Aerodisc; Gelman Sciences, Ann Arbor, Mich.) and then loaded onto a phenyl-Sepharose column (1 by 10 cm) (using the Biologic system from Bio-Rad). Proteins were separated on the phenyl-Sepharose column using a two-step gradient at a flow rate of 2 ml · min⁻¹ with buffer A (100 mM sodium phosphate buffer [pH 7.0] containing 1 mM DTT) and buffer B; the gradient was run with 100 to 50% buffer B in 15 min and then with 50 to 0% buffer B in 30 min. One fraction was collected every minute. The majority of PrpE was eluted in fractions 37 to 59. These fractions were pooled, desalted, and concentrated by the Ultrafree-15 (30,000-molecular-weight cutoff; Millipore) ultrafiltration system to a final volume of 2 ml. Proteins were then separated on a Mono-Q HR5/5 column (Pharmacia, Piscataway, N.J.) using buffer C (10 mM Tris [pH 7.8], 1 mM DTT) and buffer D (buffer C plus 1 M KCl). Separation was done using a stepwise gradient with 0 to 20% buffer D in 5 min, 20 to 35% buffer D in 25 min, and then 35 to 100% buffer D in 10 min. The flow rate was 1 ml · min⁻¹, and one fraction was collected per minute. The main peak of PrpE was eluted in fractions 14 to 19. Acyl-CoA synthetase in fractions 15 to 17 was >95% pure based on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Daiichi brand pre-cast gels; Owl Scientific Co., Woburn, Mass.) with Pro-Blue staining (Emprotech).

PHA accumulation experiments.

PHA accumulation experiments were done in a two-plasmid system. Plasmid pJM9238, a runaway replication vector for the induction of the PHA biosynthetic pathway (11), is induced to amplify by heat shock or temperature shift to 37°C. Using the method of Chung et al. (4), a second IPTG-inducible plasmid (pMON34576, pSES38, or pBluescript KS⁻::orfZ), was transformed into E. coli XL1-Blue that previously contained pJM9238. These recombinant E. coli strains were grown in LB medium containing 1% 3HP at 30°C until a OD₆₀₀ of 0.6 was reached. Subsequently, the cultivation temperature was shifted to 37°C to induce PHA accumulation, and IPTG was added to a final concentration of 1 mM to induce the second plasmid. After an incubation period of 48 h, cells were harvested for analysis of PHA content and composition.

PHA analysis.

For polyester content and composition analysis, E. coli cells were harvested by centrifugation, washed once with phosphate-buffered saline (Boehringer), and lyophilized overnight. The dried cell pellet was extracted with hot chloroform in screw-capped tubes at 100°C for 2 h. The chloroform extract was filtered through glass wool, and PHAs were precipitated in ethanol. The polyester content was obtained by comparing the mass of the precipitated PHA to the dry mass of the bacterial cell pellet used for the polymer extraction process.

NMR spectroscopic analysis.

The PHA composition was analyzed by nuclear magnetic resonance (NMR) studies using a Varian 300 MHz spectrometer. Proton spectra were obtained at 22°C from PHA samples of approximately 20 mg dissolved in 1 ml of deuterochloroform. The polymer composition was calculated based on the peak areas of all protons of the polymer units. Pulses were taken at a 45° angle with a 2.46-s acquisition time, collecting 16,000 data points and 1,024 accumulations. Chemical shifts were measured relative to CHCl₃ (δ = 7.24 ppm). The ¹³C{¹H} spectra (75 MHz) were taken at 22°C on a solution of approximately 50 mg of PHA in 1 ml of deuterochloroform. The spectra were obtained using Waltz decoupling, 30° pulses, a 1-s relaxation delay, a 12-kHz spectral width, 32,000 data points, and 20,000 accumulations. Chemical shifts were measured relative to CHCl₃ (δ = 77.0 ppm). (See Fig. 2 for an example of a proton spectrum of PHPB extracted from recombinant E. coli.)

FIG. 2.

¹H NMR spectroscopic analysis of PHPB extracted from E. coli XL1-Blue (pJM9238, pMON34576) grown on 1% 3HP in LB medium. Brackets and numbers below the abcissa indicate the signal areas. Based on those numbers, this polyester contains 15 mol% 3HB and 85 mol% 3HP. TMS, tetramethylsilane.

RESULTS

Cloning of prpE.

Total genomic DNA was isolated from S. enterica serovar Typhimurium LT2 by the method of Ausubel et al. (2) and used as template for PCR amplification of the prpE gene. Primers were designed to introduce EcoRI restriction sites at the 5′ and 3′ ends of the gene. The 5′ region of the gene was further modified to introduce an NcoI restriction site at the translational start codon. In order to avoid possible gene toxic effects, PCR products were purified, digested with EcoRI, and cloned into pSP72. This cloning vector contains only viral promoters which are not recognized in E. coli DH5α. High-level expression of the prpE gene was achieved by cloning the gene in pSE380 under trc promoter control. The resulting plasmid is pMON34564. For PHA accumulation experiments, the prpE gene was cloned under the expression control of the more moderate lac promoter. This vector is pMON34576.

Enzymatic characterization of PrpE.

As shown in Table 2, the pooled Mono-Q fractions of purified PrpE gave 1.4% of the starting ammonium sulfate-precipitated protein and resulted in a 30-fold increase in propionyl-CoA synthetase specific activity. Based on sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis, the protein at this final purification step was >95% pure. The specific activity of PrpE under conditions of saturating substrate is 120 U/mg of protein. In addition to propionate, PrpE was found to activate acetate and 3HP, and it displayed normal Michaelis-Menten kinetics with all three substrates. However, the K_m values demonstrate a strong binding preference for propionate (K_m of 0.050 ± 0.003 mM) versus acetate (0.9 ± 0.1 mM) or 3HP (27 ± 1 mM). The V_max values for acetate and 3HP are two- and threefold less than that for propionate, respectively, based on kinetic measurements that directly compared the three substrates with a partially purified enzyme. Butyrate also was a substrate for PrpE but produced only low levels of the corresponding CoA-thioester. No detailed kinetic analysis was done using this substrate because of the very low reaction rate. There were no detectable products when 4-hydroxybutyrate or dl-3-hydroxybutyrate was used as a substrate with PrpE. Based on these results, it is concluded that PrpE is a propionyl-CoA synthetase and only poorly utilizes other short-chain-length organic acids as substrates.

TABLE 2.

Purification of PrpE

Fraction	Total protein (mg)	Sp acta (μmol · mg−1 · min−1)	Purification factor (fold)
(NH4)2SO4 precipitate	27.9	4.04	1
Phenyl-Sepharose	10.0	44.2	11
Mono Q column	0.4	120	30

^aWith propionate as the substrate. 

Application of PrpE, AcoE, and OrfZ for PHA formation.

Figure 2 shows a typical spectrum for PHPB, where in this case the 3HP fraction is the major component of the copolymer. As can be seen, the signature resonances for the 3HP and 3HB components allow for clear quantitation. As shown in Tables 3 and 4, copolyesters containing 3HP units were obtained only when an acyl-CoA synthetase or a CoA transferase was expressed in connection with the PHA biosynthetic operon. When recombinant E. coli cells expressing prpE or orfZ were grown in LB medium plus 1% 3HP, polyesters containing approximately 90 mol% 3HP were accumulated. The expression of AcoE under such growth conditions resulted in significantly lower levels of 3HP in the polyester (Table 3).

TABLE 3.

Formation of PHA containing 3HP^a

Plasmid(s) harbored by E. coli XL1-Blue	Relevant genesb	Polymer content (% CDWc)	3HP content (mol%)
pJM9238	phaCAB	7	<0.05
pJM9238, pMON34576	phaCAB, prpE	11	89
pJM9238, pSES38	phaCAB, acoE	10	31
pJM9238, pBluescript pKS−::orfZ	phaCAB, orfZ	7	91

^aE. coli XL1-Blue cells were grown on LB medium containing 1% (wt/vol) 3-hydroxypropionic acid (neutralized) at 30°C and 225 rpm until an OD₆₀₀ of 0.6 was obtained. Subsequently the cultures were induced by adding IPTG to a final concentration of 1 mM and shifting the cultures to an incubation temperature of 37°C. Induced cultures were incubated for 48 h and harvested by centrifugation. Results are average values from two consecutive experiments. 

^bacoE, R. eutropha acetyl-CoA synthetase; orfZ, C. kluyveri acetyl-CoA:4-hydroxybutyrate-CoA transferase; phaCAB, R. eutropha PHA biosynthetic operon; prpE, S. enterica serovar Typhimurium propionyl-CoA synthetase. 

^cPercent PHA accumulation with respect to cellular dry mass. 

TABLE 4.

PHPB formation in the presence of mannitol^a

Plasmid(s) harbored by E. coli XL1-Blue	Relevant genesb	Acyl-CoA synthase (U/g of protein) using:		Polymer content (% CDWc)	3HP content (mol%)
		Acetic acid	Propionic acid
pJM9238	phaCAB	<5	<2	19 ± 3	<0.05
pJM9238, pMON34576	phaCAB, prpE	34	33	14 ± 2	25 ± 12
pJM9238, pSES38	phaCAB, acoE	31	9	14 ± 7	<0.05
pJM9236, pBluescript pKS−::orfZ	phaCAB, orfZ			12 ± 2	19 ± 2

^aE. coli XL1-Blue cells were grown on LB medium containing 1% (wt/vol) mannitol and 1% (wt/vol) 3-hydroxypropionic acid (neutralized) at 30°C and 225 rpm until an OD₆₀₀ of 0.6 was obtained. Subsequently the cultures were induced by adding IPTG to a final concentration of 1 mM and shifting the cultures to an incubation temperature of 37°C. Induced cultures were incubated for 48 h and harvested by centrifugation. Results for polymer accumulation are average values (with standard deviations) from three independent cultivation experiments. Enzyme activities were measured at 2 h postinduction. The enzyme activity of the CoA transferase is not shown due to instability of the protein in E. coli crude extracts (data not shown). 

^bacoE, R. eutropha acetyl-CoA synthetase; orfZ, C. kluyveri acetyl-CoA:4-hydroxybutyrate-CoA transferase; phaCAB, R. eutropha PHA biosynthetic operon; prpE, S. enterica serovar Typhimurium propionyl-CoA synthetase. 

^cPercent PHA accumulation with respect to cellular dry mass. 

In a second experiment, the cells were cultured in the presence of 1% (wt/vol) 3HP plus 1% (wt/vol) mannitol. All other parameters were kept constant. Under these conditions, the use of PrpE or OrfZ for activating 3HP resulted in the accumulation of polyesters with similar compositions. In comparison to experiments without manitol as an additional carbon source, the total amount of 3HP in these polyesters decreased to 20 to 25 mol%. AcoE-expressing strains did not accumulate any detectable amounts of 3HP under these conditions, although previous results had indicated that AcoE could utilize 3HP as substrate (20), and expression levels of AcoE appeared to be comparable to PrpE levels based on acetyl-CoA synthetase activity. The total amounts of polyesters accumulated by these strains did not vary significantly (Table 4).

DISCUSSION

The kinetic data presented in this study confirm that prpE encodes a propionyl-CoA synthetase, as suggested previously (9, 10). Our results indicate that PrpE can also activate acetate, 3HP, and butyrate to their corresponding CoA-thioesters, in addition to propionic acid, although less efficiently. Activation of butyrate by PrpE contradicts previous results which were obtained with crude E. coli extracts expressing recombinant prpE (9). However, in our hands butyrate activation to the CoA-thioester by PrpE was very inefficient.

The combination of prpE expression with the expression of the R. eutropha PHA biosynthetic pathway in E. coli XL1-Blue demonstrated that PrpE can be utilized for the biosynthesis of specific PHAs that cannot be obtained without the presence of a CoA-activating enzyme. We further demonstrated that formation of PHPB in E. coli can be obtained by expressing phaCAB from R. eutropha with either an acetyl-CoA synthetase from R. eutropha (acoE), a propionyl-CoA synthetase from S. enterica serovar Typhimurium (prpE), or an acetyl-CoA:4-hydroxybutyrate-CoA transferase from C. kluyveri (orfZ). Surprisingly, similar amounts of 3HP were accumulated by using either PrpE or OrfZ.

A CoA synthetase should be able to promote CoA-thioester formation and give high concentrations of the corresponding CoA-thioesters due to the use of free energy from ATP. This should favor incorporation of 3HP into PHA. In contrast, the acetyl-CoA:4-hydroxybutyrate-CoA transferase depends on the pools of available acetyl-CoA and free acids and in our case generates an equilibrium between acetyl-CoA, acetate, 3HP, and 3HP-CoA. Acetyl-CoA is abundant in E. coli cells under certain growth conditions, reaching millimolar levels (3), and we supplied 1% 3HP (approximately 0.1 M) to the growth medium in our experiments. These conditions apparently allowed sufficient 3HP-CoA formation in the presence of the acyl-CoA transferase for high-level incorporation in the copolymer.

An improvement to the pathway for recombinant systems would be to rely on endogenous 3HP formation. While certainly more attractive, such a system would result in a significantly lower supply of 3HP. Moreover, if such a pathway were to be utilized in a plant production system, the levels of available acetyl-CoA would likely be below 50 μM (19, 29). Under such conditions, an acyl-CoA synthetase may be more favorable than an acyl-CoA transferase for efficient activation of 3HP to the corresponding 3HP-CoA.

Other reasons for similar polyester compositions using PrpE or OrfZ for 3HP activation could also be differing expression levels of the two enzymes or differences in substrate specificity rather than specific metabolite pools within the bacterial cells. We attempted to keep expression levels in our experiments constant by cloning all three genes in high-copy-number plasmids behind the lac promoter, even though previous studies have indicated that acoE as well as orfZ can be expressed from their own promoters in E. coli (20, 28). Unfortunately, OrfZ was found to be very unstable in E. coli crude extracts. Western blot analysis indicated proteolytic cleavage (data not shown), and based on activity assays OrfZ has a half-life of approximately 20 min at room temperature in the presence of the protease inhibitors benzamidine (1 mM), leupeptin (10 mg · ml⁻¹) and 4-(2-aminoethyl)benzenesulfonyl fluoride (10 mg · ml⁻¹). For this reason, in vitro enzyme activities were difficult to reproduce and do not necessarily represent in vivo activities for this enzyme.

This study also indicates that AcoE, although it had been demonstrated to activate 3HP to 3HP-CoA under in vitro conditions with a rate similar to that for acetate (20), is not as effective for this task in vivo as either PrpE or OrfZ. This is particularly surprising since enzyme activities of AcoE and PrpE were comparable based on acetyl-CoA synthetase activity. In previous studies (20) AcoE had been shown to activate 3HP with a rate similar to that for acetic acid. However, this study indicates that AcoE is not as effective for this task as PrpE or OrfZ in vivo.