ABSTRACT
In recent years, glycerol has become an attractive carbon source for microbial processes, as it accumulates massively as a by-product of biodiesel production, also resulting in a decline of its price. A potential use of glycerol in biotechnology is the synthesis of poly(3-hydroxypropionate) [poly(3HP)], a biopolymer with promising properties which is not synthesized by any known wild-type organism. In this study, the genes for 1,3-propanediol dehydrogenase (dhaT) and aldehyde dehydrogenase (aldD) of Pseudomonas putida KT2442, propionate-coenzyme A (propionate-CoA) transferase (pct) of Clostridium propionicum X2, and polyhydroxyalkanoate (PHA) synthase (phaC1) of Ralstonia eutropha H16 were cloned and expressed in the 1,3-propanediol producer Shimwellia blattae. In a two-step cultivation process, recombinant S. blattae cells accumulated up to 9.8% � 0.4% (wt/wt [cell dry weight]) poly(3HP) with glycerol as the sole carbon source. Furthermore, the engineered strain tolerated the application of crude glycerol derived from biodiesel production, yielding a cell density of 4.05 g cell dry weight/liter in a 2-liter fed-batch fermentation process.
INTRODUCTION
In recent years, glycerol has become an attractive carbon source for microbial cultivation processes, as it accumulates in large quantities as a by-product of biodiesel production. Transesterification of vegetable oils or animal fats with alcohols yields about 10 tons of glycerol for every 100 tons of biodiesel. Since the amounts of available glycerol by far exceed the needs of conventional applications in the chemical industry, pharmacy, and cosmetics, prices have strongly declined (1, 2).
A potential use of glycerol in biotechnology is the microbial synthesis of polyhydroxyalkanoates (PHAs). PHAs are biodegradable plastics which are naturally synthesized by numerous bacteria that accumulate polyesters as granules in the cytoplasm with an excess of carbon if another macroelement, such as nitrogen, phosphorus, or sulfur, is limiting growth (3). Since PHAs can be produced from renewable resources, they could potentially serve as a substitute for conventional, petrochemically synthesized plastics (4). However, the production costs for bacterial fermentation of PHAs are not low enough, due mostly to expensive substrates and costs for processing and polymer recovery (5, 6). PHAs have remarkable material properties, as they are thermoplastic, insoluble in water, nontoxic, piezoelectric, and biocompatible, making these polymers suitable for many applications in packaging, agriculture, pharmacy, or medicine (3, 4, 7).
One interesting type of PHA is poly(3-hydroxypropionate) [poly(3HP)], which is not synthesized by any known wild-type microorganism. The material properties of poly(3HP) are comparable to those of other short-chain polymers, such as poly(3-hydroxybutyrate) [poly(3HB)] and poly(2-hydroxypropionate). However, due to a lacking methyl group in the polymer backbone, poly(3HP) is more susceptible to enzymatic cleavage (8). The incorporation of 3HP monomers into other PHAs has beneficial effects, such as a lower crystallinity and fragility of the resulting copolymer (9, 10). The best-studied copolymer containing 3HP monomers is poly(3HB-co-3HP), which has been synthesized in several wild-type and recombinant bacterial organisms (11–14). The biotechnical synthesis of a poly(3HP) homopolymer was first shown by Andreessen et al. (15), who established a nonnatural pathway in Escherichia coli to synthesize poly(3HP) from glycerol. Until then, poly(3HP) had only been synthesized chemically (16–18). Recently, Wang and coworkers engineered E. coli to synthesize poly(3HP) from glucose as the sole carbon source (19). However, these cells accumulated only up to 0.98% (wt/wt [cell dry weight {CDW}]) poly(3HP), which is significantly less than the accumulation to 11.98% reported for the system used by Andreessen et al. (15).
The aim of this study was to engineer a synthetic pathway for the conversion of glycerol to poly(3HP) in the enteric bacterium Shimwellia blattae (20, 21). S. blattae does not naturally accumulate PHAs. This bacterium is industrially relevant, as it synthesizes the commercially attractive compound 1,3-propanediol from glycerol through 3-hydroxypropionaldehyde under anaerobic conditions. It also synthesizes cobalamin (vitamin B12), which is a cofactor of the glycerol dehydratase DhaB, the key enzyme in the formation of 1,3-propanediol (22). We chose S. blattae for the synthesis of poly(3HP) from glycerol because it can naturally convert glycerol to 3-hydroxypropionaldehyde via catalysis by DhaB, thereby building an intermediate of the poly(3HP) biosynthetic pathway. In order to synthesize poly(3HP) from glycerol in S. blattae, we heterologously expressed the genes for (i) the 1,3-propanediol dehydrogenase DhaT, (ii) the aldehyde dehydrogenase AldD from Pseudomonas putida KT2442 (23, 24), (iii) the propionate-coenzyme A (propionate-CoA) transferase Pct from Clostridium propionicum X2 (25–27), and (iv) the PHA synthase PhaC1 from Ralstonia eutropha H16 (28, 29). By this synthetic pathway, 1,3-propanediol, which is produced by S. blattae under anaerobic conditions, is oxidized to 3-hydroxypropionaldehyde (DhaT). The latter is then oxidized to 3-hydroxypropionic acid (3HP; AldD). Pct activates 3HP to 3HP-CoA by the addition of coenzyme A, followed by polymerization of the 3HP moieties by PhaC1 (Fig. 1). With this recombinant S. blattae strain, it was possible to accumulate poly(3HP) to levels of up to almost 10% (wt/wt [CDW]).
Pathway for conversion of glycerol to poly(3-hydroxypropionate) in recombinant S. blattae. 1, DhaBCESb; 2, DhaTSb/DhaTPp; 3, AldDPp; 4, PctCp; 5, PhaC1Re.
MATERIALS AND METHODS
Bacterial strains, plasmids, and oligonucleotides.All of the bacterial strains, plasmids, and oligonucleotides used in this study are listed in Table 1. For cloning experiments, plasmids were transferred into E. coli TOP10. For synthesis of poly(3HP), the plasmid pBBR1MCS-2::aldD::dhaT::pct::phaC1 was transferred into S. blattae ATCC 33430.
Bacterial strains, plasmids, and nucleotides used in this study
Growth of cells.Cells were cultivated in Erlenmeyer flasks with agitation at 125 rpm and at 30�C. Cells of S. blattae were cultivated in basal medium (22) with glycerol concentrations of between 100 and 350 mM. For anaerobic cultivation of S. blattae, cells were incubated in Hungate tubes without agitation. Cells of E. coli and P. putida were cultivated in lysogeny broth (LB) (30) at 30�C and 125 rpm. R. eutropha H16 cells were incubated in mineral salts medium (MSM) (31). Cells of C. propionicum were cultivated in medium 156, as suggested by the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ), in a 100% N2 atmosphere. Antibiotics were added at the following concentrations: kanamycin, 50 μg/ml; and gentamicin, 20 μg/ml.
For fed-batch fermentation of recombinant S. blattae cells, a 2-liter Biostat B Plus (Sartorius) fermentor was used. Temperature, pH, foam, and dissolved oxygen were monitored online during fermentation. The optical density at 600 nm (OD600) and the CDW were measured in samples withdrawn from the cultures. The formation of foam was controlled by the addition of a 50% (vol/vol) Struktol (Schill & Seilacher) solution. A pH of 7.5 was maintained by addition of 1 M HCl or a 7.5% (vol/vol) ammonium hydroxide solution. The fermentation process was split into two phases. For the first 24 h, cells were cultivated under anaerobic conditions at a stirring speed of 100 rpm. For this phase, the vessel was flushed with nitrogen, and the system was closed. Upon induction by the addition of 0.1 mM IPTG (isopropyl-β-d-thiogalactopyranoside), aeration was started, with the airflow set to 4 liter/min and a stirring speed of 500 rpm, for a further 48 h. The basal medium contained a starting concentration of 350 mM purified or crude glycerol (93% glycerol, 4% water, 2% ash, <0.5% methanol, 1% other organic compounds). The feeding solution was made of basal medium with 30% (wt/vol) (crude) glycerol and an additional 3 g/liter (NH4)2SO4 and 0.2 g/liter MgSO4 � 7H2O. In addition, the feeding solution contained 50 μg/liter kanamycin to maintain the plasmid stability of the culture. In order to investigate the plasmid stability of the cultivated recombinant S. blattae strain, diluted samples of the fermentation broth were spread on LB agar plates containing kanamycin as well as on LB agar plates lacking kanamycin, and the plates were then incubated at 30�C for 24 h. Plasmid stability was determined by comparing the numbers of CFU on LB agar plates without kanamycin with those on LB agar plates containing kanamycin (32).
Plasmid construction and transfer into E. coli and S. blattae.All processing and manipulation of DNA were carried out as described by Sambrook et al. (30).
The coding regions of aldD, dhaT, pct, and phaC1 were amplified by PCR with REDTaq DNA polymerase (Sigma-Aldrich, Steinheim, Germany) according to the manufacturer's instructions, using the oligonucleotides shown in Table 1. PCR products were cloned into the target vector pBBR1MCS-2 (33) by employing T4 DNA ligase (Thermo Scientific, Dreieich, Germany) and were transferred into chemically competent E. coli TOP10 cells. Kanamycin-resistant clones were verified by colony PCR. Respective plasmids were isolated, and the hybrid plasmid pBBR1MCS-2::aldD::dhaT::pct::phaC1 was generated as follows. pBBR1MCS-2::dhaT was digested with BamHI and BcuI (SpeI), and the purified coding region of dhaT was ligated into plasmid pBBR1MCS-2::aldD, which had been linearized by treatment with BamHI and BcuI (SpeI). The resulting plasmid, pBBR1MCS-2::aldD::dhaT, was then linearized by restriction with BcuI (SpeI) and XbaI and ligated with a similarly digested pct fragment from pBBR1MCS-2::pct, yielding plasmid pBBR1MCS-2:aldD::dhaT::pct. The coding region of phaC1 was excised from plasmid pBBR1MCS-2::phaC1 with SspI, and the blunt-ended fragment was ligated into the Ecl136II (EcoICRI) site of plasmid pBBR1MCS-2:aldD::dhaT::pct, which had been linearized accordingly (Fig. 2). In order to generate the hybrid plasmid pBBR1MCS-2::aldD::pct::phaC1, the same cloning steps as those for the generation of pBBR1MCS-2::aldD::dhaT::pct::phaC1 were carried out, except that no dhaT gene was ligated into the vector.
Expression vector pBBR1MCS-2::aldD::dhaT::pct::phaC1. All fragments were obtained by PCR amplification.
To delete the regulator of the dha operon (dhaR), a deletion cassette was constructed (pCR2.1-TOPOΔdhaR::Gmr). The coding region of dhaR, with added flanks of 397 bp upstream and 400 bp downstream, was amplified from S. blattae ATCC 33430 genomic DNA with REDTaq DNA polymerase and then inserted into pCR2.1-TOPO (Life Technologies, Carlsbad, CA) by TA cloning. The resulting plasmid, pCR2.1-TOPO::dhaR, was digested with HincII (HindII) and SspI, thereby deleting a region of 1,763 bp, and the SmaI-digested gentamicin resistance cassette coding region from vector pSKsymΩGm (34) was inserted into the plasmid by blunt ligation. In order to carry out a complementation experiment, dhaR was ligated into the vector pBBR1MCS through SpeI and XbaI restriction sites.
S. blattae ATCC 33430 was transformed by electroporation. For this purpose, cells of an overnight culture of S. blattae ATCC 33430 were washed twice with a 10% (wt/vol) glycerol solution and transferred to cuvettes with a 2-mm gap. After the addition of 500 ng DNA, cells were electroporated in an Eppendorf model 2510 electroporator at 2,500 V, 10 μF, and 600 Ω, employing a 5-ms pulse. Following an incubation of 90 min in 1 ml of LB medium at 30�C, cells were plated on antibiotic-containing LB agar plates at 30�C overnight.
Deletion of dhaR from S. blattae ATCC 33430.The deletion of dhaR from S. blattae ATCC 33430 was carried out using a Quick & Easy E. coli gene deletion kit (Gene Bridges, Heidelberg, Germany). The linear fragment for the deletion of dhaR was amplified from plasmid pCR2.1-TOPOΔdhaR::Gmr with the same oligonucleotides used for the amplification of the dhaR fragment. Gentamicin-resistant mutants were verified by colony PCR.
Cell harvest and extraction of poly(3HP).After separation of the cells from the culture broth, cells were frozen at −30�C and freeze-dried. Poly(3HP) was isolated from the pulverized dry cell matter by digestion of non-PHA biomass with a 13% (vol/vol) sodium hypochlorite solution (35, 36). After adding the biomass to the sodium hypochlorite solution, to a concentration of 30 g/liter, the solution was incubated for 1 h at room temperature. Upon the addition of half the initial volume of water, the nondigested polymer was separated from the solution by centrifugation. In order to remove remaining lipids, the isolated polymer was dissolved in acetone and precipitated in 100 volumes of ice-cold methanol. Again, the polymer was separated via centrifugation and dried by evaporating the methanol.
Determination of poly(3HP) content.Analyses of the poly(3HP) content of the cells and the purity of the extracted polymer were done by gas chromatography (GC). For this purpose, dried cell mass or samples of isolated poly(3HP) were exposed to acidic methanolysis as described before (37, 38). For microscopic analysis of PHA granules, cells were stained with Nile red (39).
Measurements of glycerol, 1,3-propanediol, and other metabolites.Concentrations of glycerol, 1,3-propanediol, and other metabolites present in the media were monitored by high-pressure liquid chromatography (HPLC) analysis. For this purpose, supernatants were assayed using a Lachrom Elite HPLC system (VWR-Hitachi, Darmstadt, Germany) with an RI detector (type 2490; VWR, Darmstadt, Germany) and a Metacarb 67H column (300 � 6.5 mm; VWR-Varian, Darmstadt, Germany) at 75�C, using a flow rate of 0.8 ml/min for 20 min. The mobile phase was made of 4.5 mM sulfuric acid.
RESULTS
Influence of the dha regulator on 1,3-propanediol formation in S. blattae ATCC 33430.In order to determine the amount of glycerol that is converted to 1,3-propanediol by S. blattae ATCC 33430, cells were cultivated under anaerobic conditions as described in Materials and Methods. Additionally, a dhaR-deficient mutant of S. blattae ATCC 33430 was cultivated to investigate the role of the glycerol metabolism operon regulatory protein DhaR. When cultivated with an initial glycerol concentration of 300 mM, cells of wild-type S. blattae ATCC 33430 accumulated 1,3-propanediol to up to 138 mM (10.5 g/liter), which accounts for 46 mol% of the initial glycerol concentration present in the medium. In contrast, cells of the S. blattae ATCC 33430 ΔdhaR mutant showed no growth and did not convert any glycerol to 1,3-propanediol under the same anaerobic conditions (Fig. 3). However, the cells were able to grow aerobically in shake flasks (data not shown). Complementation with a plasmid-carried copy of dhaR restored the capability of anaerobic growth on glycerol, with the strain secreting up to 10.1 g/liter (48 h) of 1,3-propanediol into the medium, which is in the same range of 1,3-propanediol formation as that of the wild-type strain.
Cultivation of S. blattae ATCC 33430 (A) and S. blattae ATCC 33430 ΔdhaR (B) in basal medium with 300 mM glycerol. The cultures were grown in Hungate tubes without agitation at 30�C. The OD600 (●) was measured externally. Concentrations of glycerol (■) and 1,3-propanediol (▲) were monitored through HPLC analysis.
Poly(3HP) accumulation in recombinant S. blattae ATCC 33430.In our first approach to synthesize poly(3HP) with the generated recombinant S. blattae ATCC 33430 strain, cells were cultivated in shake flasks. After incubation of the cells for 24 h without agitation, the cells were induced and shaken for another 48 h. During cultivation, cells were stained with Nile red and examined by fluorescence microscopy (Fig. 4). Upon induction, increasing numbers of cells displayed one to three granules. In flask experiments, the recombinant S. blattae strain was cultivated to a cell dry weight of 1.08 g/liter. GC analysis revealed a poly(3HP) content (wt/wt) of the cells of 9.8% � 0.4%. It was crucial to supply a sufficient amount of glycerol to the medium, as cells did not accumulate poly(3HP) when they were cultivated with an initial glycerol concentration of 100 mM in the medium. Further flask experiments revealed that induction of the cells with IPTG was not necessary in order to synthesize poly(3HP). However, cells accumulated poly(3HP) to only 5.4% � 0.2% (wt/wt) when no IPTG was added. At IPTG concentrations of >0.1 mM, no further increase of the cells' poly(3HP) content occurred. The recombinant S. blattae strain was also able to synthesize poly(3HP) directly from 1,3-propanediol when it was grown aerobically in 5-ml overnight LB cultures containing 1,3-propanediol at concentrations of 0.5 to 2.0%. In addition, a recombinant S. blattae strain harboring the aldD, pct, and phaC1 genes but lacking dhaTPp was not able to synthesize poly(3HP) under anaerobic conditions.
Poly(3HP) accumulation of S. blattae ATCC 33430/pBBR1MCS-2::aldD::dhaT::pct::phaC1. After 48 h of fermentation, hydrophobic inclusions were stained with Nile red and observed with a fluorescence microscope employing phase contrast (left), differential interference contrast (middle), or fluorescence microscopy at 312 nm (right).
In order to increase the yield of acquired cell mass, S. blattae ATCC 33430/pBBR1MCS-2::aldD::dhaT::pct::phaC1 was cultivated in a 2-liter two-step, fed-batch fermentation process as described above (Fig. 5). After 72 h of cultivation, cells reached densities of 3.0 g cell dry mass/liter in basal medium with purified glycerol and 4.1 g cell dry mass/liter in basal medium with crude glycerol (Table 2). The harvested cells contained 8.6% � 0.2% (purified glycerol) and 5.59% � 0.31% (crude glycerol) poly(3HP). Plasmid stabilities of 64% for the culture grown with pure glycerol and 67% for the culture grown with crude glycerol were determined. Similar to the results of the foregoing flask experiments, cells had already accumulated small amounts of poly(3HP) before induction and the switch to aerobic growth. Throughout the anaerobic growth phase, the concentration of the secreted 1,3-propanediol did not decrease notably. As observed during fluorescence microscopy, most polymer-accumulating cells displayed one large granule.
Fed-batch cultivation of S. blattae ATCC 33430/pBBR1MCS-2::aldD::dhaT::pct::phaC1 in a 2-liter Biostat B Plus (Sartorius) fermentor at 30�C. Cells were cultivated in basal medium with pure (A) or crude (B) glycerol as the sole carbon source. The process was split into two phases: I, 24 h under anaerobic growth conditions; and II, 48 h under aerobic growth conditions. Concentrations of glycerol (□) and 1,3-propanediol (○) were monitored through HPLC analysis. Poly(3HP) accumulation (●) was determined by GC analysis of dried cell matter and is displayed as the % (wt/wt) CDW (■).
Cell dry weight, poly(3HP) production, and metabolites detected by HPLC during fermentation of S. blattae ATCC 33430/pBBR1MCS-2::aldD::dhaT::pct::phaC1
Poly(3HP) isolation and analysis.Cells were extracted with a 13% (vol/vol) hypochlorite solution as described above. The precipitated and dried polymer appeared as a white powder. The isolated poly(3HP) was analyzed by GC-mass spectrometry (GC-MS). Methyl esters of the 3HP constituents obtained through acidic methanolysis eluted at a retention time of 6.1 min, which matched the retention time of methyl esters from a commercially available 3-hydroxypropionic acid standard (TCI Deutschland GmbH, Eschborn, Germany). Fractionation of the peak revealed a mass spectrum similar to that of the standard. The most dominant peaks were at 73 and 74 (m/z). Gel permeation chromatography (GPC) analysis of the isolated poly(3HP) revealed an average molecular mass of 52,500 Da, with a polydispersity index (molecular weight [Mw]/number-average molecular weight [Mn]; PDI) of 1.45.
DISCUSSION
In this study, we generated a recombinant strain of S. blattae ATCC 33430 which is able to synthesize poly(3HP) from glycerol, at levels of up to 9.8% � 0.4% (wt/wt) of the cell dry weight. In comparison to the few other studies on biotechnical production of a poly(3HP) homopolymer, cells of the S. blattae strain engineered in this study accumulated about the same amount of poly(3HP) as the recombinant E. coli strain described recently by Andreessen et al. (15), which accumulated up to 12.0% (wt/wt) polymer. However, in order to accumulate the maximum amount of poly(3HP), the engineered S. blattae strain was cultivated for 20 fewer hours, as it synthesized poly(3HP) at a much higher rate. Thus, future improvements to increase the cell density of the recombinant S. blattae strain could potentially lead to a much higher productivity for poly(3HP) than in the process described by Andreessen et al. Furthermore, in order to dispose of reducing equivalents, Andreessen et al. (15) required the addition of 0.5 mol of tartrate or 1 mol of fumarate for every mol of synthesized 3HP, which causes additional costs and is certainly a major drawback of the process. In contrast, the recombinant S. blattae strain described in this study does not require supplemental compounds for a stable redox balance to synthesize poly(3HP). In contrast to the two-phase fermentation process described by Andreessen et al. (15), where an aerobic growth phase was followed by an anaerobic poly(3HP) accumulation phase, our engineered S. blattae strain was first cultivated anaerobically for 24 h and then cultivated aerobically for another 48 h in order to synthesize poly(3HP) from 1,3-propanediol. This two-phase process has several benefits: (i) a higher cell dry weight can be achieved than in a completely anaerobic production process; (ii) NADH which accrues from the synthesis of poly(3HP) can be oxidized to NAD+ during aerobic respiration; and (iii) during aerobic poly(3HP) production, the precursor 3-hydroxypropionaldehyde is not converted back to 1,3-propanediol, since the native dhaT gene from S. blattae should not be expressed when oxygen is present (22). The interplay of DhaTSb and DhaTPp is important for the process, as it was shown previously that these isoenzymes exhibit different specific substrate preferences, preferring 3-hydroxypropionaldehyde (DhaTSb) (22, 40) or 1,3-propanediol (DhaTPp) (24). Using the BLAST algorithm (41), a maximum amino acid identity of 38% was determined for the two isoenzymes. In comparison to a recent study by Wang et al. (19), poly(3HP) accumulation in S. blattae ATCC 33430/pBBR1MCS-2::aldD::dhaT::pct::phaC1 was significantly higher than in the recombinant E. coli strain from the former study, which accumulated the homopolyester to only 0.98% (wt/wt).
As HPLC analyses revealed the presence of 3HP in culture supernatants, either the addition of CoA through Pct or polymerization by PhaC1 appears to be the bottleneck for poly(3HP) synthesis in the recombinant S. blattae strain. Since pct is the third open reading frame downstream of the lac promoter, its level of translation from the corresponding polycistronic mRNA is likely to be lower than those of the other two genes. In contrast, phaC1 is expressed separately from the other three genes, under the control of an individual lac promoter in the engineered plasmid, and should therefore be expressed at a high level. Thus, we propose that activation of 3HP by the propionate-CoA transferase Pct is the limiting step in poly(3HP) synthesis in our recombinant S. blattae strain. An approach to overcome this limitation could be an alteration of the order of the heterologously expressed genes, which was reported to increase the accumulation of poly(3HB) in a recombinant E. coli strain harboring the phaCAB operon from R. eutropha H16 (42).
A further interesting observation was the increased cell dry weight in the fermentation with crude glycerol (Table 2). Obviously, the additional traces of ash and methanol did not result in a growth inhibition of the strain. Moreover, organic compounds present in the crude glycerol carbon source might be utilized as additional carbon sources by the cells, thereby yielding a higher biomass than that in medium containing purified glycerol. A high tolerance toward impurities of the glycerol feedstock has been reported for several other 1,3-propanediol-producing bacteria, such as Clostridium butyricum and Klebsiella pneumoniae (43, 44). In contrast, recombinant S. blattae cells cultivated with crude glycerol accumulated less poly(3HP). Similar observations were made by Andreessen et al. (15), who reported a >50% decrease in poly(3HP) accumulation by the engineered E. coli strain when the strain was cultivated with crude glycerol. This could also be explained by the presence of abundant organic compounds in the crude glycerol, which could serve as a preferred carbon source such that less glycerol enters the reductive branch of glycerol utilization in the recombinant S. blattae cells.
Although 1,3-propanediol formed during the anaerobic phase of the fermentation is likely to be oxidized by the heterologous 1,3-propanediol dehydrogenase DhaT, the supernatants of the harvested cultures of our recombinant S. blattae strain still contained high 1,3-propanediol concentrations of 10.0 g/liter (purified glycerol) and 11.3 g/liter (crude glycerol) (Fig. 5; Table 2). This might have been due to a lack of oxygen supply when the cell density increased, causing a continuation of the metabolism of glycerol through the reductive branch under microaerobic conditions. Although fed-batch cultivations of different K. pneumoniae strains resulted in 1,3-propanediol concentrations of over 70 g/liter (45, 46), processes with S. blattae could lead to similar 1,3-propanediol concentrations, as S. blattae converts glycerol at a similar molar ratio to 1,3-propanediol when cultivated anaerobically. Further alterations of the current process could therefore potentially result in considerable yields of two products.
One alteration of the current S. blattae process of producing poly(3HP) could be the simultaneous overexpression of dhaR. Similar to the case in the bacteria Citrobacter freundii and C. butyricum, which metabolize glycerol reductively, DhaR serves as a positive regulator of genes mediating the reductive metabolism of glycerol (47, 48). Using the BLAST algorithm (41), a maximum identity of 81% to DhaR of C. freundii was detected. No significant homologies were found when the sequence was aligned to the sequence of DhaR of C. butyricum. An overexpression of dhaR in the recombinant S. blattae strain could potentially increase the amount of 1,3-propanediol produced under aerobic or microaerobic conditions and could therefore also increase the productivity of the process by dismissing the 24-h anaerobic phase to build up 1,3-propanediol for poly(3HP) production. As GPC analysis revealed, the molecular mass of the isolated polymer was relatively low (50,000 Da). It is known that extraction of PHAs with hypochlorite leads to degradation of the polymer. Still, the molecular mass of poly(3HP) isolated from our recombinant S. blattae strain was significantly lower than those for similar extraction processes reported for poly(3HB) (34, 35, 49) isolated from R. eutropha H16. This might have been due to the high level of heterologous expression of phaC1, as shown by Sim and coworkers (50), who reported a decrease in molecular weight of synthesized poly(3HB) in a recombinant E. coli strain harboring the PHA biosynthesis operon from R. eutropha H16 when phaC1 expression was increased by applying a strong individual inducible promoter. In comparison to several previous studies (35, 49, 51), the PDI (Mw/Mn) of 1.45 determined in this study displays a rather low molecular weight distribution, which is a desired property for most applications of polymers.
In summary, we engineered an S. blattae strain which accumulates up to a 9.8% (wt/wt of CDW) yield of the promising biopolymer poly(3HP) from the inexpensive carbon source glycerol in a two-step cultivation process. In this process, the industrially relevant compound 1,3-propanediol is accumulated in considerable amounts as a by-product. In order to increase the yields of the desired products, further improvements of the process and of the metabolism are proposed for future studies.
ACKNOWLEDGMENTS
We thank Rolf Daniel and his laboratory at the Department of Genomic and Applied Microbiology (Georg-August University G�ttingen) for providing the bacterial strain S. blattae ATCC 33430 and John Foster and his laboratory at the School of Biotechnology and Biomolecular Sciences (UNSW, Australia) for GPC analysis of our isolated poly(3HP) samples.
FOOTNOTES
- Received 16 January 2013.
- Accepted 27 March 2013.
- Accepted manuscript posted online 29 March 2013.
- Copyright � 2013, American Society for Microbiology. All Rights Reserved.
