INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Poly-3-hydroxyalkanoates (PHAs), which are produced as intracellular storage material by many bacterial species, have attracted considerable attention due to their potential applications as biodegradable plastics and as sources of valuable chiral synthons (13). Based on detailed surveys, Steinbüchel and coworkers have classified PHAs into two groups (38), which are synthesized via different pathways (Fig. 1A). One group, short-chain-length PHAs (scl-PHAs), contains 3-hydroxyalkanoate monomers with chain lengths from C₄ to C₅. Another group, medium-chain-length PHAs (mcl-PHAs), are characterized by 3-hydroxyalkanoate monomers with chain lengths from C₆ to C₁₄.

View larger version (2022K): [in this window] [in a new window]   FIG. 1.   General and modified pathways for the synthesis of mcl-PHA precursors. (A) Possible links between mcl-PHA synthesis and fatty acid degradation (left) and the general pathway for PHB synthesis in R. eutropha (right) are outlined. All enzymatic activities are numbered as indicated on the right. Genes encoding these enzymes that have not been identified are indicated with an asterisk. Dashed lines indicate pathways or reactions that are theoretically possible but might not take place due to limited concentrations of the -oxidation intermediates. (B) Modified pathway for mcl-PHA synthesis in the E. coli fadB mutant. The fadB mutation blocks -oxidation at the 3-hydroxyacyl-CoA dehydrogenase step (cross), which channels substrates to mcl-PHA. The introduced phaC gene enables mcl-PHA synthesis, if sufficient substrate is available. The thick dashed arrow indicates the PHB acetoacetyl-CoA reductase activity encoded by phbB, cloned into the host. Reduction of ketoacyl-CoA intermediates is theoretically possible but might not take place due to limited 3-ketoacyl-CoA concentrations or limited enzyme activity for the mcl acyl-CoA -oxidation intermediates. (C) Construct analogous to that seen in panel B with the host E. coli fadA mutant. This modified pathway was designed to increase the supply of mcl-PHA precursors by introduction of phbB in addition to blocking -oxidation at the 3-ketoacyl-CoA thiolase. The cross indicates a block of 3-ketoacyl-CoA thiolase. The thick solid arrow indicates the postulated channeling of ketoacyl-CoA from -oxidation to PHA.

It is known that PHAs are synthesized by polymerization of coenzyme A (CoA)-linked R-3-hydroxy fatty acids through one of several PHA polymerases (PhaC) (see a recent review [21]). The synthesis of such CoA substrates can occur by a variety of pathways, the simplest of which uses -ketothiolase (PhbA) and acetoacetyl-CoA reductase (PhbB) to synthesize precursors for scl-PHA polymerases such as poly-3-hydroxybutyrate (PHB) synthase (PhbC) in Ralstonia eutropha (25, 26) (Fig. 1A). mcl-PHA CoA-linked precursors can be generated through fatty acid -oxidation that produces acyl-CoA intermediates such as enoyl-CoA, 3-ketoacyl-CoA, and/or S-3-hydroxyacyl-CoA when fatty acids are used as sole carbon sources (14, 16, 39, 40) (Fig. 1A).

Effort has been devoted in recent years to increasing PHA yields and productivity. Achieving production costs that are in the same range as those of chemically synthesized plastics may be feasible, given the recent creation of PHA-producing transgenic plants (22, 27, 40, 41). Equally important in the production of PHAs is the control of the fundamental properties of the polymer. The production of PHA with specific monomer units has demonstrated that the monomer unit composition significantly affects the final properties of the polymer product (41). To influence rationally these properties in the production of organisms, it is necessary to control and direct the production of PHA precursors. Various approaches have been used to modify PHA monomer compositions. One example is cofeeding of different substrates, such as for the production of poly-3-hydroxybutyrate-co-3-hydroxyvalerate (5), and blending polymer poly(3-hydroxy-5-phenylvalerates) and poly-3-hydroxyalkanoates (19). So far, modification of monomer compositions of mcl-PHAs has not been feasible since mcl-PHA precursors are derived from fatty acid metabolism and their concentrations are difficult to control.

Biosynthesis of these polymers in host organisms that do not naturally produce PHA allows modulation of biosynthetic enzymes and therefore allows one to achieve substrate levels that are difficult to achieve in a native host strain for the study of PHA production (18). In addition, introduction of specific genes into an organism having a suitably modified PHA synthetic pathway may allow extension and regulation of the range of compounds that can be produced. Escherichia coli is one of these useful hosts.

Previously, it has been reported that E. coli strains defective in the -oxidation pathway are able to accumulate mcl-PHA when equipped with a PHA polymerase (29, 31). There are two PHA polymerases that are encoded by phaC1 and phaC2 in Pseudomonas oleovorans and P. aeruginosa (16, 39). The presence of one of these two polymerases is sufficient for PHA production in E. coli fatty acid degradation mutants (29, 31). Recently, it was found that inhibition of the -oxidation thiolase step resulted in an increase in the availability of substrate for PHA synthesis (30). Therefore, engineering of E. coli to extend the range and distribution of precursors which can be channeled into mcl-PHA may lead to interesting PHA variants for application and development studies and is likely to extend the available chiral synthon pool as well.

In this study, we selected two E. coli mutants, JMU193 (fadR fadB) and JMU194 (fadR fadA), to investigate PHA accumulation. Our results revealed that ketoacyl-CoA intermediates of the -oxidation cycle are potential precursors for PHA synthesis. By modifying relevant enzyme and substrate levels, we were able to redirect ketoacyl-CoA -oxidation intermediates into mcl-PHA, and polymers with altered monomer composition and different physical properties were obtained.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains and plasmids. The following E. coli strains were used in this study: JMU193 (fadR::Tn10 fadB64) (32) and JMU194 (fadR::Tn10 fadA30) (32). Plasmids pETQ2 (31) and pBTC2 carrying phaC2 and pET901 carrying phbB were used to construct recombinants capable of PHA synthesis. pGEc404 carrying phaC2 (16), pBS20 carrying phbCAB (36), pBCKS (Promega), pUC19 (43), pVLT35 (3), and pCK01 (6) were used for the construction of pBTC2 and pET901. pETQ102, pET104, pBTC2', and pET901' were intermediate plasmids during the construction of pBTC2 and pET901.

Recombinant DNA techniques. Isolation and analysis of plasmid DNA were carried out according to the method of Sambrook et al. (33). Transformations of E. coli competent cells were done according to standard procedures (33). Plasmids pBTC2 and pET901 were constructed as illustrated in Fig. 2.

View larger version (1815K): [in this window] [in a new window]   FIG. 2.   Schematic representation of the strategy to construct plasmids pBTC2 and pET901. Only the relevant cloned gene(s) and restriction site(s) are shown. (A) pBTC2 contains the phaC2 gene of P. oleovorans expressed through the Ptac promoter. (B) pET901 contains the phbB gene of R. eutropha expressed through the Plac promoter.

Media and culture conditions. E. coli recombinants were grown at 37°C in complex Luria-Bertani (LB) medium (33). To allow accumulation of intracellular PHA, 0.1NE2 minimal medium (15) was used, 0.2% (wt/vol) yeast extract was added to the media as the carbon source to support growth. Fatty acids were prepared as previously described (17) and supplied as indicated in the medium for PHA production and possible carbon sources. Cells were induced with 200 µM isopropyl--D-thiogalactopyranoside (IPTG) at the early exponential phase. If necessary, the following antibiotics were added: kanamycin, 50 µg/ml; ampicillin, 100 µg/ml; streptomycin, 50 µg/ml; and chloramphenicol, 30 µg/ml. Cell growth was monitored by measuring optical density at 450 nm (42).

PHB acetoacyl-CoA reductase assay. For enzymatic analysis, precultures were grown in LB medium for 8 h at 37°C and then transferred into 0.1NE2 minimal medium with 0.2% (wt/vol) yeast extract with a 1:100 dilution. Cells were induced with 200 µM IPTG at early exponential phase and harvested at late exponential phase, washed once with 0.1 M Tris-HCl (pH 7.5), and resuspended in a 10% volume of the same buffer. Cells were then disrupted in 200-µl aliquots by sonication. Protein concentrations were determined by Bio-Rad protein assay. Acetoacetyl-CoA reductase activity was monitored at 340 nm by the oxidation of NADPH during reduction of acetoacetyl-CoA (35). Then, 25 µl of crude extract and 25 µl of 10 mM NADPH were mixed with 500 µl of buffer (120 mM potassium phosphate [pH 5.5], 24 mM MgCl₂, 1 mM dithiothreitol), and 455 µl of H₂O. The reaction was started by adding 5 µl of 7 mM acetoacetyl-CoA. One unit is defined as 1 µmol of NADPH depletion/min.

Incorporation of monomers into preexisting PHA. Cells were cultivated on 0.1NE2 minimal medium with 0.2% (wt/vol) yeast extract and 2 mM hexadecanoate as described above. A pulse of 0.5 mM pentadecanoate was added to the culture at the time indicated. PHA compositions were analyzed by gas chromatography as previously described (20), and C₇ and C₉ monomer incorporation rates in preexisting C-even number PHA were determined.

Extraction of PHA. Lyophilized cells (500 to 1,000 mg) were extracted with 150 ml of methylene chloride (CH₂Cl₂) for at least 8 h at 50°C. The cell extract was filtered, and the solvent was evaporated (Buechi Rotavapor) until <15 ml of extract remained. PHA was recrystallized by dropping the extract into 10 volumes of well-stirred, ice-cold methanol. Methanol solutions were then filtered through Zetapor filter membranes (CUNO, Inc., Meriden, Conn.). Polymer was recovered from the filters by dissolving it in chloroform and evaporating the solvent. The polymer was stored in the dark at room temperature for further analysis.

Determination of PHA. To determine the presence and composition of PHA, lyophilized cells or isolated PHA were treated with H₂SO₄-methanol (85/15 [vol/vol]) in CHCl₃ (containing 0.1 mg of methyl benzoate/ml) as previously described (20). The methanolyzed PHA monomers were analyzed by using a CP-Sil 5 CB column (Chrompack) on a gas chromatograph (Fison) (20). Splitless injection, an attenuation of 1, and a range of 0 were used to reach maximum sensitivity.

Molecular weight determination. The molecular weight measurements were done by gel permeation chromatography. Samples were dissolved in tetrahydrofuran and run through a PL-Gel C 5-µm column. Signals were detected using capillary viscosimetry (Viscotek 502), light scattering (small angle laser light scattering detector, KMX-6), and light refraction (differential high-temperature refractometer [Knauer]) measurements. The column was calibrated with polystyrene standards.

DSC measurement. Differential scanning calorimetry (DSC) spectra (melting temperatures [T_m], enthalpy of fusion (Hm), and glass transition temperatures [T_g]) were recorded in a temperature range of 60 to 80°C on a Netzsch DSC 200 Perkin-Elmer model DSC-4 equipped with a cooling accessory under a nitrogen flow of 30 ml/min. PHA samples of 3 mg were encapsulated in aluminum pans and heated from 60 to 80°C at a rate of 10°C/min. The T_g was taken as the midpoint of the heating capacity change.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

In this study E. coli fadR fadA and fadR fadB mutants (JMU194 and JMU193, respectively) were used through the whole study. The fadR gene encodes a protein that exerts negative control over genes necessary for fatty acid oxidation (2, 4, 17, 23). A mutation in fadR derepresses transcription of these genes, as a result of which the fad genes are constitutively expressed, rendering E. coli capable of growth on mcl fatty acids (2, 4). The fadA gene encodes 3-ketoacyl-CoA thiolase (Fig. 1), and the fadB gene encodes four enzyme activities (Fig. 1): enoyl-CoA hydratase, 3-hydroxyacyl-CoA dehydrogenase, ³-cis-²-trans-enoyl-CoA isomerase (not included in Fig. 1), and 3-hydroxyacyl-CoA epimerase (2, 4, 17, 23). Mutations in fadA or fadB block fatty acid oxidation and result in accumulation of specific intermediates.

Utilization of different substrates for PHA formation in a fadR fadB negative recombinant E. coli. The E. coli fatty acid catabolism mutant JMU193, which lacks 3-hydroxyacyl-CoA dehydrogenase (24, 37) (Fig. 1B), was equipped with the PHA polymerase 2 from P. oleovorans GPo1 (16) and analyzed for PHA formation from a number of different fatty acids ranging from C₆ to C₁₈. Table 1 shows that the highest PHA contents (ca. 4% of cell dry weight [cdw]) were obtained during growth in the presence of longer-chain-length alkanoates of C₁₆ to C₁₈. However, the composition of PHAs formed from the tested alkanoates did not vary significantly with changes of alkanoate chain length. The polymer produced from odd-chain-length fatty acids consisted mainly of 3-hydroxyheptanoate (C₇) and 3-hydroxynonanoate (C₉), while the dominant even-chain-length monomers are 3-hydroxyhexanoate (C₆), 3-hydroxyoctanoate (C₈), and 3-hydroxydecanoate (C₁₀).

PHA formation by E. coli recombinants defective in different steps of the -oxidation. Since the metabolic status, including the concentrations of metabolites and the rate of metabolite formation, may differ from one strain to another, it is plausible that the intracellular content and the composition of mcl-PHAs formed from such intermediates will vary from strain to strain. Thus, E. coli JMU194 (fadR fadA), a mutant which lacks 3-ketoacyl-CoA thiolase activity (24, 37) (Fig. 1C), was compared with JMU193 for PHA formation. To control the expression of phaC2, plasmid pBTC2 (Fig. 2A) containing phaC2 of P. oleovorans GPo1 expressed through the Ptac promoter was introduced in JMU193 and JMU194, and the recombinants were cultivated on medium containing hexadecanoate.

1

Redirecting of ketoacyl-CoA intermediates from -oxidation into the mcl-PHA synthetic pathway in E. coli JMU193 and JMU194 recombinants. As illustrated in Fig. 1C, it might be possible to redirect the carbon flux from the -oxidation pathway to mcl-PHA synthesis by blocking step 8 to accumulate 3-ketoacyl-CoA intermediates and amplifying step 7 to permit the reduction of these intermediates to R-()-3-hydroxyacyl-CoA. It is known that the ketoacyl-CoA reductase encoded by phbB of R. eutropha (step 11 in Fig. 1A) has a substrate range of C₄ to C₆ acyl-CoAs (12), permitting it to convert ketohexyl-CoA to R-()-3-hydroxyhexyl-CoA, which might be channeled into mcl-PHA by PHA polymerase.

Physical properties of polymers produced by E. coli recombinants. PHAs produced by the E. coli recombinants used in this study were isolated by complete solvent extraction from lyophilized cells. The monomer compositions of the purified PHAs were essentially identical to the values estimated by gas chromatography analysis after methanolysis of whole cells.

View larger version (7K): [in this window] [in a new window]   FIG. 3.   Molecular weight of polymer isolated from JMU194 recombinants. E. coli JMU194 recombinants were grown in 0.1NE2 minimal medium with 0.2% yeast extract and 2 mM hexadecanoate. Cells were induced with 200 µM IPTG in the early exponential phase and harvested in the stationary phase. PHA was isolated from cells and analyzed by gel permeation chromatography for the molecular weight as described in Materials and Methods. Solid line, PHA from JMU194(pBTC2, pCK01); dashed line, PHA from JMU194(pBTC2, pET901).

View larger version (12K): [in this window] [in a new window]   FIG. 4.   DSC thermograms of the PHA samples from JMU194 recombinants. The PHA samples described in Fig. 3 were analyzed for thermal properties as described in Materials and Methods. (A) PHA from JMU194(pBTC2, pCK01). (B) PHA from JMU194(pBTC2, pET901).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The monomer composition of PHAs can affect the properties of the polymer (41). The development of mcl-PHA variants depends, therefore, on the range of monomers that can be incorporated into PHA. Consequently, it is of interest to attempt to control rationally the monomer composition of new mcl-PHAs. This requires a better understanding of pathways involved in mcl-PHA synthesis and of the basic metabolic pathways that may provide PHA precursors. In this study, we set out to address two questions. One was whether ketoacyl-CoA of the -oxidation can serve as precursor source for PHA synthesis. The second question was how the mcl-PHA monomer composition could be modified. Among the approaches we tested, pathway engineering by amplification of one potential PHA precursor supply route gave the most promising result. Other approaches, such as varying the substrates, did not lead to significant changes in PHA composition or physical properties (Table 2).

The -oxidation intermediate ketoacyl-CoA is a potential precursor for PHA synthesis in E. coli. Introduction of the PHA polymerase into E. coli results in utilization of 3-hydroxyacyl-CoA precursors for the synthesis of PHA (29, 31, 34, 36). As illustrated in Fig. 1, when the -oxidation is blocked at the last step (ketothiolase [FadA]), all upstream intermediates will accumulate (Fig. 1C). 3-Ketoacyl-CoA could be reduced to R-()-3-hydroxyacyl-CoA through a 3-ketoacyl-CoA reductase (step 7), S-(+)-3-hydroxyacyl-CoA may be epimerized (step 6, fadB), and enoyl-CoA may be hydrated to the corresponding R-()-3-hydroxyacyl-CoA (step 3, fadB). When the -oxidation is blocked at step 5, catalyzed by 3-hydroxyacyl-CoA dehydrogenase (FadB), one of these possible mcl-PHA precursors, 3-ketoacyl-CoA, will not accumulate (Fig. 1B). In this study, we found that E. coli JMU194 (fadR fadA) phaC⁺ recombinant incorporated more C₆ and C₁₀ monomers into PHAs than the JMU193 (fadR fadB) phaC⁺ recombinant under the same growth conditions (Tables 1 and 2). The relative monomer content of PHA depends on the precursor concentration and the specificity of the PHA polymerase for each precursor. Since the specificity profile of the PHA polymerase is presumably identical in both recombinants, differing PHA compositions must reflect the differences in relative intracellular precursor concentrations.

Physical properties of PHA can be changed by pathway engineering. We found significant differences in the physical properties of PHA isolated from fadR fadA E. coli JMU194 carrying phaC and from the same recombinant carrying additional phbB (Table 2).

View larger version (14K): [in this window] [in a new window]   FIG. 5.   Model for the structure of PHA isolated from JMU194(pBTC2, pET901). Straight lines indicate 3-hydroxyhexanoate (C6) monomer-rich stretches; curved lines indicate stretches containing random C6, 3-hydroxyoctanoate (C8), and 3-hydroxydecanoate (C10) monomers.