Enhanced Functional Expression of the Polyhydroxyalkanoate Synthase Gene from Cupriavidus Necator A-04 Using a Cold-Shock Promoter for Efficient Poly(3-Hydroxybutyrate) Production in Escherichia Coli

Background: The present study attempted to increase PHB production by improving the functional expression of the PhaC gene using various types of promoters, and the effects on PhaC activity in terms of PHB productivity, yield coefficient (Y P/S ) and molecular weights were investigated. Results: Here, the PHB biosynthesis operon of Cupriavidus necator A-04, isolated in Thailand with a high degree of 16S rRNA sequence similarity with C. necator H16, was subcloned into pGEX-6P-1, pColdI, pColdTF, pBAD/Thio-TOPO and pUC19 (constitutive expression) and transformed into E. coli JM109. To alter the expression of phaCAB biosynthesis genes, we optimized parameters in flask experiments to obtain high expression of soluble PhaC A−04 protein with high Y P/S and PHB productivity. pColdTF- phaCAB A−04 -expressing E. coli produced 2.5 ± 0.1 g/L (90.6±4.3%) PHB in 24 h, similar to pColdI- phaCAB A−04 -expressing E. coli . The amounts of phaC protein and PHB produced from pColdTF- phaCAB A−04 and pColdI- phaCAB A−04 were significantly higher than those from other promoters. Cultivation in a 5-L fermenter led to PHB production of 7.9±0.7 g/L with 90.0±2.3% PHB content in the cell dry mass (DCM), a Y P/S value of 0.38 g PHB/g glucose and a productivity of 0.26 g PHB/(L⋅h) using pColdTF- phaCAB A−04 . The PHB from pColdTF- phaCAB A−04 had M W 5.79 × 10 5 Da, M N 1.86 × 10 5 Da and PDI 3.11 and the film exhibited high transparency, Young’s modulus and tensile strength, possibly due to the TF chaperones. Interestingly, when pColdI- phaCAB A−04 -expressing E. coli was used to produce PHB from expression, A−04 biosynthesis with promoter of A−04 JM109. In examined the of A−04 on in E. coli to cell growth, glucose consumption, PHB production, and kinetic parameters in conditions from to a 5-L produced thermal mechanical at 5 or 10% (v/v) were transferred into fresh LB medium with 100 µg/L ampicillin and 20 g/L and incubated at 37 °C and 200 rpm for 24 h. For comparison of the effect of phaC expression on PHB production under various types of promoters, fusion proteins and chaperones, shake flask experiments were performed in 250-mL Erlenmeyer flasks containing 50 mL of LB medium containing ampicillin (100 µg/mL) on a rotary incubator shaker at 37 °C and 200 rpm for 24 h. For PHB production, overnight cultures in LB medium (1 mL) were transferred into fresh LB medium supplemented with glucose (20 g/L) and ampicillin (100 µg/mL). Recombinant E. coli JM109 (pColdI- phaCAB A−04 ) and E. coli JM109 (pColdTF- phaCAB A−04 ) were induced to produce PHB using the conventional induction method and short-induction method. The effect of GST (the hydrophilic fusion protein) and the tac promoter on PHB production was investigated using E. coli JM109 (pGEX-6P-1- phaCAB A−04 ), which was induced by the addition of IPTG (0.5 mM). The effect of the araBAD promoter and N-terminal thioredoxin fusion protein together with the C-terminal 6His-fusion protein on phaC and PHB production was examined by inducing E. coli JM109 (pBAD/Thio-TOPO- phaCAB A−04 ) with arabinose (1% w/v). E. coli JM109 (pUC19-nativeP- phaCAB A−04 ), which exhibits constitutive expression, was used as a control strain. All these comparison experiments were performed at 15 °C or 37 °C for 48 h. Purified PHB is used in the development of innovative technologies for applications in microbeads and packaging. Effect of the growth phase suitable for cold-shock on and PHB (% w/w) under the continuous-induction method. The different growth phases were investigated by varying OD600 based on cultivation time (0.5 (2 h, early exponential phase), 1.3 (4 h, middle exponential phase), 2.1 (6 h, late exponential phase) and 2.4 (10 h, stationary phase)) for (A) E. coli JM109 (pColdI-phaCABA-04) and (B) E. coli JM109 (pColdTF-phaCABA-04). A control experiment was performed with 0.0 mM IPTG induction. All the data are representative of the results of three independent experiments and are expressed as the mean values ± standard deviations (SDs). The PhaCA-04 protein was detected by western blot analysis using anti-His tag antibody as the primary antibody. The band appearing in the western blot at the position corresponding to that of the His-tagged phaCA-04 protein was 67 kDa in size for pColdI-phaCABA-04, and the fusion protein of His-tagged phaCA-04 and TF was 115 kDa in size. All the data are representative of the results of three independent experiments and are expressed as the mean values ± standard deviations (SDs). Effect of the growth phase suitable for cold-shock induction on DCM and PHB content (% w/w) under the shortinduction method. The different growth phases were investigated by varying OD600 based on cultivation time (0.5 (2 h, early exponential phase), 1.3 (4 h, middle exponential phase), 2.1 (6 h, late exponential phase) and 2.4 (10 h, stationary phase)) for (A) E. coli JM109 (pColdI-phaCABA-04) and (B) E. coli JM109 (pColdTF-phaCABA-04). A control experiment was performed with 0.0 mM IPTG induction. All the data are representative of the results of three independent experiments and are expressed as the mean values ± standard deviations (SDs).


Background
The global environmental concern regarding microplastics in the marine environment as contaminants with significant impacts on animal and human health has led to a call for national and international policies from more than 60 countries to ban or place a levy on single-use plastics [1][2][3][4]. Renowned global companies have also integrated regulations and policies to ban plastic bottle cap seals, plastic water bottles, straws and other single-use packaging into their green marketing and corporate social responsibility policies. Bioplastics are becoming a popular alternative to single-use plastics to reduce the amount of microplastic waste. Recently, the annual production of bioplastics was approximately one percent of the total 360 million tons of plastics. However, as the market for bioplastics is growing and the demand for bioplastics is rising, European Bioplastics reported that the global bioplastic production capacity will increase from 2.11 million tons in 2019 to approximately 2.43 million tons in 2024 [5]. Among the various types of bioplastics, polyhydroxyalkanoates (PHAs) are an important biodegradable polymer family, as they are one hundred percent biobased and fully biodegradable in all environments, especially marine (ASTM 7081) and fresh water environments [6,7]. Notably, their variety of monomeric compositions offers a wide array of thermal, physical and mechanical properties that strongly rely on PHA-producing strains and carbon sources.
To obtain both the environmental and economic benefits of PHAs over synthetic plastics and other bioplastics, microorganisms that exhibit efficient PHA production from inexpensive and renewable carbon sources are urgently required to develop a low-cost approach. Microbial cells typically accumulate polyhydroxybutyrate (PHB), the first PHA discovered since 1927 by Maurice Lemoignei of Institut Pasteur in France [8], at approximately 30-50% of the dry cell mass (DCM). The most well-known industrial PHA producer, Cupriavidus necator H16 (formerly known as Alcaligenes eutrophus and Ralstonia eutropha), is capable of accumulating PHB at over 80% of the DCM. PHA accumulation is tightly regulated by imbalanced growth conditions with excess carbon but limited nitrogen [9]. One of the major limitations in the production of PHAs in wild-type strains has been intracellular polymer degradation caused by endogenous PHA depolymerases, which is different from the behavior of exogenous PHA depolymerases [10]. Therefore, intracellular PHAs are often spontaneously degraded during cultivation when the bacteria require carbon, resulting in low PHA content and a wide range of molecular weight distributions in wild-type strains. Thus, many recombinant strains have been developed by metabolic engineering to obtain a high yield of PHB and a molecular weight that is high enough for polymer processing [11][12][13][14][15][16]. Ordinarily, the PHB biosynthesis pathway begins with acetyl-CoA and requires 3 major enzymes, namely, 3-ketothiolase (PhaA), NADPH-dependent acetoacetyl-CoA reductase (PhaB), and PHA synthase (PhaC), and these 3 genes are sufficient for the production of PHB in non-PHA-producing bacteria at more than 90% of the DCM when heterologously expressed in Escherichia coli [17]. It has been reported that PhaC plays a key role in obtaining the polymeric form, resulting in a high level and high molecular weight of PHB [14,16].
To date, PHA synthases have been categorized into four major classes based on their sequence, substrate specificity, and subunit composition [18,19]. Class I and Class II PHA synthases consist of the PhaC subunit, which is believed to be a homodimer. On the other hand, Class III and IV PHA synthases are heteroclusters comprised of PhaC-PhaE subunits and PhaC-PhaR subunits, respectively. In addition, Class I, III, and IV PHA synthases preferentially polymerize short-chain-length (SCL) monomers comprised of C3-C5 carbon chain lengths, whereas Class II PHA synthases specifically polymerize medium-chain-length (MCL) monomers in the C6-C14 chain length range. It was reported that PhaC derived from C. necator H16 (PhaC H16 ) is a Class I PhaC and is one of the most widely studied PHA synthases. It has a molecular weight of approximately 64 kDa (589 amino acids) and is located as the first gene in the PHA biosynthetic operon, followed by PhaA and PhaB [20,21]. It was demonstrated that the weight-average molecular weight (M W ) of PHB synthesized by wild-type bacteria is generally in the range of 0.1-2.0 × 10 6 Da. Because Escherichia coli does not contain PHA depolymerase, ultrahigh-molecular-weight PHB, which has a defined M W more than 3.0 × 10 6 Da and is much larger than the PHB produced by the wild-type strain, can be obtained and used for the development of highstrength fibers and films [22]. Unfortunately, no crystal structure exists for any PHA synthase enzyme because this enzyme tends to form inclusion bodies when overexpressed in a bacterial host or tends to aggregate during purification [23]. When recombinant PhaC H16 was overexpressed in E. coli, most of the protein formed insoluble inclusion bodies due to its low aqueous solubility [24][25][26][27]. To feasibly achieve industrial-scale production, PhaC would need to be produced in large quantities and its solubility would need to be improved [23]. There have been many reports that have attempted to resolve the problem mentioned above, including by modulating the concentration of the PhaC protein by varying the chemical inducer quantities [28]; expressing the protein at a reduced temperature (30 °C) [23]; fusing the PhaC protein with a glutathione S-transferase (GST) tag, which is a hydrophilic tag, to improve its solubility [29]; and coexpressing the protein with chaperones to obtain high total quantities of enzyme and a larger proportion in the soluble fraction than obtained without chaperones. However, coexpression of the GroEL/GroES system with the PHA production operon resulted in the production of polymers with reduced molecular weights [23].
In a previous study, we reported the generation of the C. necator strain A-04, possessing 99.78% 16S RNA sequence similarity with C. necator H16 but differing in PHA production ability [30] phaC A−04 was 99% similar to phaC H16 from C. necator H16. The difference was in the amino acid residue situated at position 122, which in phaC A−04 was proline but in C. necator H16 was leucine. The total amino acid sequences of phaA A−04 and phaB A−04 were 100% matched with those of C. necator H16 (data not shown, manuscript in preparation). Notably, C. necator strain A-04 prefers fructose over glucose as a carbon source, accumulating PHB at 78% of the DCM under a C/N ratio of 200, whereas it could incorporate a high mole fraction of monomeric 4-hydroxybutyrate monomeric into the poly(3-hydroxybutyrate-co-4-hydroxybutyrate) [P(3HB-co-4HB)] copolymer under a C/N ratio of 20 [31] as well as the poly(3-hydroxybutyrate-co-3-hydroxyvaterate-co-4hydroxybutyrate) [P(3HB-co-3HV-co-4HB)] terpolymer [32]. Thus, in this study, the PHA biosynthesis operon of C. necator strain A-04 was amplified via PCR; cloned into pGEX-6P-1 (tac promoter, isopropyl-β-dthiogalactopyranoside (IPTG)-inducible vector, N-terminal GST fusion protein), pColdI (cspA promoter, cold-and IPTG-inducible vector, N-terminal 6His-fusion protein), pColdTF (cspA promoter, cold-and IPTG-inducible vector, trigger factor (TF) chaperone, N-terminal 6His-fusion protein), pBAD/Thio-TOPO (araBAD promoter, arabinoseinducible vector, N-terminal thioredoxin fusion protein and C-terminal 6His-fusion protein) and pUC19 (control strain, constitutive expression, phaCAB A−04 biosynthesis genes with native promoter of phaC A−04 ); and transformed into E. coli JM109. In this study, we examined the effect of phaC A−04 overexpression on PHB production in recombinant E. coli with respect to cell growth, glucose consumption, PHB production, and kinetic parameters in conditions ranging from flask culture to a 5-L fermenter. Furthermore, the produced PHB was subjected to molecular weight determination, thermal analysis and mechanical property measurement.

Effect of the growth phase on the production of PhaC A−04 and PHB by the conventional induction method
To optimize the conditions for heterologous expression of phaCAB A−04 biosynthesis genes, after the pColdI-phaCAB A−04 and pColdTF-phaCAB A−04 vectors were transformed into E. coli JM109, expression was induced with 0.5 mM IPTG (final concentration) at different growth phases by varying OD 600 based on cultivation time: 0.5 (2 h, early exponential phase), 1.3 (4 h, middle exponential phase), 2.1 (6 h, late exponential phase) and 2.4 (10 h, stationary phase). Concurrently, the temperature was shifted from 37 °C to 15 °C for 24 h. Figure 2 shows the effect of the growth phase for gene induction on the DCM (g/L), PHB content (% w/w) and levels of insoluble and soluble PhaC A−04 protein, comparing E. coli JM109 (pColdI-phaCAB A−04 ) and E. coli JM109 (pColdTF-phaCAB A −04 ). The PhaC A−04 protein was detected by western blot analysis using an anti-His tag antibody as the primary antibody. A band appeared in the western blot at the position corresponding to that of the His-tagged phaC A−04 protein (67 kDa) for pColdI-phaCAB A−04 and the fusion protein of His-tagged phaC A−04 and TF at 115 kDa. By varying the time courses of the growth phase, His-tagged PhaC A−04 and the His-tagged phaC A−04 -TF fusion protein were successfully expressed, with the highest amount of total phaC A−04 protein obtained when the phaCAB A−04 operon was induced at an OD 600 of 0.5 (2 h, early exponential phase). The content of soluble PhaC A −04 -TF fusion protein (Fig. 2B, lane 3) in the sample after IPTG induction at an OD 600 of 0.5 was much higher than that of the phaC A−04 protein alone from pColdI-phaCAB A−04 ( Fig. 2A, lane 3), suggesting that the TF chaperone facilitates the expression of highly soluble protein in E. coli JM109. Moreover, significant proteolysis of the PhaC A−04 protein occurred with pColdI-phaCAB A−04 when cells produced a large amount of insoluble PhaC A keeping other factors constant [33]. First, conditions were optimized by varying the OD 600 based on cultivation time (0.5, 1.3, 2.1 and 2.4 h) and inducing expression with 0.5 mM IPTG at 15 °C for 30 min. Then, the temperature was shifted from 15 °C to 37 °C for 24 h to enhance growth and PHB production. The effect of the growth phase (OD 600 ) on DCM (g/L) and PHB content (% w/w) is illustrated in Fig. 3. Again, it was clearly observed that cells of both E. coli JM109 (pColdI-phaCAB A−04 ) and E. coli JM109 (pColdTF-phaCAB A−04 ) in the 2-h early exponential phase (OD 600 of 0.5) exhibited higher DCM and PHB production than those in other growth phases. After induction with 0.5 mM IPTG at 15 °C for 30 min and cultivation at 37 °C for 24 h, E. coli JM109 (pColdI-phaCAB A−04 ) attained 4.5 ± 0.1 g/L DCM, 3.9 ± 0.1 g/L PHB and 85.90 ± 2.6% (w/w) PHB content with a productivity of 0.16 g PHB/(L⋅h), whereas E. coli JM109 (pColdTF-phaCAB A−04 ) attained 3.5 ± 0.1 g/L DCM, 2.7 ± 0.1 g/L PHB and 75.90 ± 2.8% (w/w) PHB content with a productivity of 0.11 g PHB/(L⋅h). Thus, the shortinduction method enhanced the PHB content and productivity more than the conventional method. Next, an OD 600 of 0.5 was used to investigate the optimal concentration of IPTG (0, 0.01, 0.05, 0.1, 0.5, and 1.0 mM) under the short-induction conditions. The effects of various IPTG concentrations on DCM (g/L), PHB (g/L), PHB content (% w/w) and PHB productivity (g PHB/(L⋅h)), comparing E. coli JM109 (pColdI-phaCAB A−04 ) and E. coli JM109 (pColdTF-phaCAB A−04 ), are summarized in Table 2. It can be concluded that the optimal concentration of IPTG was 0.5 mM in both cases. The PHB content (% w/w) increased in accordance with the IPTG concentration, but the amount of PHB (g/L) produced was maximum under induction with 0.5 mM IPTG. The PHB content (% w/w) increased approximately 8-fold, and the productivity (g PHB/(L⋅h)) increased 16-fold, compared with those under the control condition in the case of pColdI-phaCAB A−04 . Table 2 Effect of IPTG concentration on DCM (g/L), PHB (g/L), % (w/w) PHB content and PHB productivity in a comparison between E. coli

Development of a preinduction method and effect of inoculum size on PHB productivity
We also investigated a preinduction strategy to enhance PHB productivity by extending the PHB production phase at 37 °C for an additional 24 h after conventional induction. When the OD 600 reached 0.5, IPTG was added at 0.5 mM into the culture, and the temperature was reduced from 37 °C to 15 °C. Then, cultivation was performed for 24 h to allow full expression of the phaCAB A−04 protein. Concurrently, the effect of inoculum size (1, 5 and 10% (v/v)) of induced cells was investigated under the preinduction conditions. The results are shown in comparison with those of the conventional induction and short-induction methods ( Table 3). The preinduction method with a 5% (v/v) inoculum gave a higher amount of PHB (1.9 ± 0.6 g/L) than conventional induction with an inoculum size of 5% (v/v) (0.6 ± 0.1 g/L) and could extend the productivity of 0.039 ± 0.01 g PHB/(L⋅h) for 48 h so that the PHB content increased from 46.2 ± 3.8% (w/w) to 67.9 ± 4.8% (w/w). The increase in PHB content and PHB productivity occurred with an increase in the inoculum size. Nevertheless, the short-induction method with an inoculum size of 5% (v/v) gave the highest levels of PHB content and productivity. Therefore, the short-induction method using E. coli JM109 (pColdI-phaCAB A−04 ) with an inoculum size of 0.5% (v/v) and cultivated until the OD 600 reached 0.5 (2 h) before induction with 0.5 mM IPTG was selected to investigate the effect of induction temperature in the subsequent experiment. Table 3 Comparison of the kinetics of cell growth, (g PHB/g-glucose), and PHB production g/(L⋅h) by C. necator strain A-04, E. coli , and E. coli JM109 (pUC19-nativeP-phaCAB A−04 ) in shake flask cultivation.

Effect of induction temperature on PHB productivity
The optimal short-induction temperature was investigated in a range between 15 °C and 37 °C for 30 min before increasing the temperature to 37 °C for 24 h to confirm that the high PHB productivity resulting in this study is a result of the cold-shock cspA promoter and that 15 °C is the optimal induction temperature. Figure 4 shows the results of the effect of the short-induction temperature (15, 25, 30 and 37 °C) on cell growth and PHB production. It was clear that 15 °C was the optimal induction temperature for enhancing the amount of PHB produced, which resulted in a maximum PHB content of 86.2 ± 2.6% (w/w). The amount of PHB produced decreased as the induction temperature increased, with a concomitant increase in RCM. The cold-shock temperature promoted PHB production and suppressed RCM. The PHB productivity at 15 °C was 7-fold higher than that obtained with an induction temperature of 37 °C.

Effect of the TF chaperone on phaCAB A−04 and PHB production
In phaCAB A−04 -overexpressing E. coli JM 109 (pColdI-phaCAB A−04 ) under the conventional conditions, the formation of inclusion bodies of PhaC A−04 has been observed due to the low aqueous solubility of the protein, as described previously [29]. To verify that the cold-shock cspA promoter works together with the TF chaperone to improve the solubility of PhaC A−04 , the hydrophilic GST tag was fused to the N-terminus of PhaC A−04 (pGEX-6P-1- 04), and the effect of the GST tag at 37°C on the polymerization reaction of phaC A−04 based on the amount of PHB production was investigated. In addition, pBAD/Thio-TOPO-phaCAB A−04 , encoding a hydrophilic N-terminal thioredoxin fusion protein and C-terminal 6His-fusion protein induced by arabinose, was also used for comparison. The control strain, harboring pUC19-nativeP-phaCAB A−04 , showed constitutive expression using the native promoter from C. necator strain A-04, and no induction agent was required under the same conditions. The phaC A− 04 protein is expressed in vivo and always as a mixture of inclusion bodies and soluble protein; hence, we did not purify the phaC A−04 protein and assay its in vitro polymerization activity but considered the amount of PHB produced together with the molecular weight of PHB as a result of in vivo polymerization activity.
The results are shown in Fig. 5 and Table 3. It was clearly found that pColdI-phaCAB were overexpressed in E. coli JM 109 cells at both 15°C (conventional method) and 37°C (short induction method). However, the total PhaC A−04 protein from pColdI-phaCAB A−04 was much higher than that from pColdTF-phaCAB A−04 (Fig. 5B). The ratio between the soluble form and inclusion bodies of pColdI-phaCAB A−04 was much lower than that of pColdTF-phaCAB A−04 . Based on our observations, the cold-shock cspA promoter and TF gave functional phaCA-04 protein (Fig. 5C), resulting in the highest specific production, Y P/S , and PHB productivity. The PHB production from pGEX-6P-1-phaCAB A−04 , pBAD/Thio-TOPO-phaCAB A−04 and pUC19-nativeP-phaCAB A−04 was not different, which may be attributed to the host strain and induction method used in this study.

Comparison of PHB production between pColdI-phaCAB A−04 and pColdTF-phaCAB A−04 in a 5-L fermenter by the short-induction method
Altogether, for flask cultivation, the optimal conditions were the short-induction method using an inoculum of 0.5% (v/v) in a culture with an OD 600 of 0.5, cold shock induced with 15 °C for a short time, 30 min, and the addition of 0.5 mM IPTG. These conditions were selected as optimal parameters for scaling up production in a 5-L fermenter. The comparison between pColdI-phaCAB A−04 and pColdTF-phaCAB A−04 in a 5-L fermenter by the short-induction method was performed because the ratio of the soluble fraction and inclusion bodies of the phaC A−04 protein may affect PHB productivity and molecular weight distribution as reported by Harada et al. [29].  Table 4 demonstrated that E. coli JM109 (pColdTF-phaCAB A−04 ) was a more effective PHB producer than the other strain.
A PHB content of 92.5± 5.9% (w/w), PHB production of 8.5±0.8 g/L, DCM production of 9.2±0.3 g/L, Y P/S value of 0.40 g PHB/g glucose and productivity of 0.39 g PHB/(L⋅h) were the maximum values obtained using pColdTF-phaCAB A−04 , whereas a PHB content of 78.0±2.1% (w/w), PHB production of 5.8±0.1 g/L, DCM production of 7.2±0.3 g/L, Y P/S value of 0.32 g PHB/g glucose and productivity of 0.16 g PHB/(L⋅h) were attained using pColdI-phaCAB A−04 . The phaC A−04 protein produced by pColdTF-phaCAB A−04 was more stable and longer lasting ( Fig. 6B) than that obtained from pColdI-phaCAB A−04 , which was no longer detectable after 30 h of cultivation ( Fig. 6A). Therefore, we report here that the short-induction strategy facilitates cold shock cspA and chaperone TF proteins to act synergistically to improve the stabilization of PhaC A−04 and enhance productivity by 143.8% and the Y P/S value by 25% in 30 h.

Comparison of the molecular weight and thermal and mechanical properties of PHB produced by pColdI-phaCAB A−04 and pColdTF-phaCAB A−04 from glucose and crude glycerol
The PHB thin films were subjected to thermal analysis by DSC, molecular weight determination by GPC and mechanical property analysis by a universal testing machine as per the ASTM: D882-91 protocol ( films were also subjected to 1 H-NMR and 13 C-NMR analyses and showed only chemical shifts of the PHB structure.

Discussion
Since 1988, it has been reported that PhaC Re , as a type I synthase and the most intensively studied of these proteins, preferentially catalyzes the polymerization of short-chain (R)-hydroxyalkanoic acids (4 to 6 carbon atoms), particularly the conversion of (R)-3-hydroxybutyrate-coenzyme A (3HBCoA) to poly(hydroxybutyrate) (PHB) [20,25]. In fact, a high yield of PHB (157 g/L) has been achieved from glucose in high-cell-density cultures of recombinant E. coli harboring phaCAB Re and the additional cell division protein ftsZ gene [37,38]. Ultrahighmolecular-weight PHB and its applications have also been reported by many research groups [15,22,39,40]. PhaCRe has been used to prepare chimeric enzymes to increase its ability to incorporate MCL monomers for SCLco-MCL production [41].
Beyond these previous reports, there have been few reports on the application of cold-shock systems for PHB production to address the challenges of soluble and functional phaC expression in E. coli. One of the attempts was the use of pCold and a GST-fusion tag to obtain pCold-PhaC Re and pCold-GST-PhaC Re , which were overexpressed in E. coli DH5α and E. coli BL21(DE3) to investigate the effect of GST fusion on the in vivo solubility of PhaC Re , but the pGEM-T derivative carrying the pha Re promoter, His-fused phaC Re , phaA Re , and phaB Re and pGEM-GSTphaC Re AB were used to evaluate in vitro PHB production [29]. It was revealed that most of the PhaC Re and GST-PhaC Re when overexpressed in E. coli BL21(DE3) were detected in the insoluble fraction rather than the soluble fraction, indicating that the solubility of PhaC Re is not improved by GST tag fusion under a cold-inducible promoter. Another recommendation was that a fusion protein to assist in phaC solubilization as well as the 6His-tag should be fused at the N-terminus of phaC because an N-terminal tag has a weaker effect than a C-terminal tag on the polymerization activity of phaC [23,29,42].
Although C. necator strain A-04 exhibits 99.78% similarity of 16S rRNA, 99.9% similarity of phaC A−04 and 100% similarity of phaA A−04 and phaB A−04 with those of C. necator H16, we observed differences in PHB productivity as well as the monomeric composition of the copolymers and terpolymers when we used the same carbon source [30][31][32]. Interestingly, C. necator strain A-04 also exhibited different growth abilities on pure glycerol as well as crude glycerol compared with C. necator strain H16 (data not shown, presented at an international conference). Thus, we initially aimed to use the pColdI and pColdTF expression systems to address the challenges of soluble and functional phaC A−04 expression in E. coli JM109 and finally evaluate its ability to use crude glycerol as a carbon source for PHB production in a 5-L fermenter. We also investigated the optimal expression conditions and finally compared them with those for other promoters, including the pGEX-6P-1 derivative, carrying N-terminal GST and 6His-fused phaCAB A−04 ; the pBAD/Thio-TOPO derivative, carrying Cterminal 6His-and N-terminal thioredoxin-fused phaCAB A−04 ; and the constitutive expression vector pUC19, using a native promoter from C. necator strain A-04, in both flask cultivation and the 5-L fermenter. The produced PHB was extracted, purified, and prepared using the glass casting method, and subjected to molecular weight analysis and thermal and mechanical property determination to determine the effect of the cold-shock cspA promoter and TF on the functional expression of phaC A−04 and PHB products.
First, this study aimed to solubilize PhaC A−04 by using the cold-shock cspA promoter and TF to achieve a high yield of soluble recombinant PhaC A−04 from E. coli JM109. His-tagged phaC A−04 was overexpressed by pColdI, but most of the protein was present in insoluble form, with significant aggregation resulting in smear bands ( Fig. 2A and 5B), whereas the His-tagged phaC A−04 -TF fusion protein was expressed from pColdTF at lower levels than the protein from pColdI, but most of this protein was present in soluble form ( Fig. 2B and 5B). The ratio of the soluble fraction to the total phaC A−04 proteins from pColdTF-phaCAB A−04 was much higher than that from pColdI-phaCAB A−04 . Thus, it can be concluded that the TF chaperone helped solubilize phaC A−04 in our investigation, but PHB production and Y P/S in both systems, namely, pColdI-phaCAB A−04 and pColdTF-phaCAB A −04 , were not significantly different. To produce PHB using microbial cells under in vitro conditions, phaC will always naturally be expressed in two forms, namely, insoluble and soluble, and these proteins simultaneously work together under in vitro conditions to perform dynamic PHB polymerization. Thus, in this study, we decided to consider the amount of PHB, Y P/S , PHB productivity and molecular weight distribution as indirect indicators of PhaC A−04 activity because there have been many reports that have gained insights into PhaC Re expression and PHB production that can be used to support our findings. Focusing on chaperone-assisted PhaC Re expression and PHB production, it was previously reported that the TF chaperone (without the cold-shock cspA promoter) was investigated in combination with three chaperone systems for coexpression with PhaC Re : GroEL/GroES (plasmid pGro7), TF (plasmid pTf16) and DnaK/DnaJ/GrpE (plasmid pKJE7). Additionally, TF and GroEL/GroES were expressed together (plasmid pG-Tf2), as were GroEL/GroES and DnaK/DnaJ/GrpE (plasmid pG-KJE8) [23]. The study concluded that with the set of strains expressing the N-terminal 6His-tagged fusion protein, the GroEL/GroES system resulted in approximately 6-fold-greater enzyme yields than that obtained in the absence of coexpressed chaperones, whereas TF resulted in approximately 3-fold increases in the soluble protein yield. It seemed likely that although the overexpression of phaC A−04 was achieved both in terms of quantity (pColdI) and solubility (pColdTF), its function had already reached its limit, so improvement of PHB production was not further observed. Acceleration of PHB productivity would be another goal for reducing the time consumed for microbial cultivation. It was reported that optimization of expression conditions, including inducer concentrations, age of bacterial cells (OD 600 ) and induction temperatures, is required to improve PHB productivity.
It was reported that the production level of PhaC was not significantly changed by the addition of IPTG at concentrations greater than 0.1 mM, suggesting that this IPTG concentration is sufficient for pJRDTrcphaCAB Re to fully express PhaC [28]. In our study, 0.5 mM was the optimal concentration of IPTG for overexpression of phaC A −04 under the cold-shock cspA promoter, consistent with a previous report [16]. Cells at the initial exponential phase exhibited the highest phaC which are a critical factor involved in dimer formation [29]; however, this needs to be elucidated by determining the quantitative hydrophilicity of the protein surface by using a quartz crystal microbalance (QCM), which is not available in our laboratory. The ratio of soluble to total PhaC A−04 proteins did not play an important role in PHB productivity, but the TF chaperone resulted in the production of PHB with reduced molecular weights [23] because the M W and M N of PHB produced by pColdTF were lower than those of pColdI-phaCAB A−04 and pUC19-nativeP-phaCAB A−04 .
To achieve low-cost production, crude glycerol as a byproduct from biodiesel production was used as a carbon source to produce PHB using pColdI-phaCAB A−04 and pUC19-nativeP-phaCAB A−04 . PHB from E. coli JM109 (pColdI-phaCAB A−04 ) had an M W of 2.42 × 10 5 Da and an M N of 0.89 × 10 5 Da with a PDI of 2.92. However, the PHB content and PHB productivity were lower than those similarity obtained using glucose. E. coli JM109 (pUC19-nativeP-phaCAB A−04 ) produced PHB constitutively with a satisfactory M W of 9.05 × 10 5 Da and an M N of 2.17 × 10 5 Da with a PDI 3.72. Typically, PHB obtained from glycerol was reported to have a significantly lower molecular weight than polymers synthesized from other substrates, such as glucose or lactose [43,44].
A previous study showed that cells grown on glycerol exhibit a highly reduced intracellular state compared to cells grown on glucose under similar dissolved oxygen conditions [45]. Therefore, E. coli JM109 (pColdI-phaCAB A −04 ) and E. coli JM109 (pUC19-nativeP-phaCAB A−04 ) have low DCM production (g/L), PHB production (g/L), PHB content (%wt) and PHB productivity (g/(L⋅h)) when grown on crude glycerol because glycerol has a significant effect on the intracellular redox state, which causes the cells to direct carbon flow toward the synthesis of highly reduced products to achieve redox balance [46]. It was described in a previous study that the product distributions of E. coli K24K when grown on glucose or glycerol as the substrate were different. The glycerolgrown cultures produced lower amounts of acetate, lactate, and formate and higher amounts of ethanol than those grown on glucose. PHB production from glycerol was lower than that from glucose, except under conditions of low oxygen availability [47]. However, another report revealed that E. coli K24K produced PHBs possessing a T G of 22°C, whereas the molecular weight of PHB from glycerol was similar to that obtained using glucose [47]. In addition, a study performed using C. necator to produce PHB from commercial glycerol and waste glycerol exhibited products with molecular masses of 957 and 786 kDa, respectively, less than half of the mass of the PHB obtained from glucose [48]. In our study, when pColdI-phaCAB A−04 -expressing E. coli was used to produce PHB from crude glycerol and compared with constitutive pUC19-nativeP-phaCAB A−04 -expressing E.
coli, although the amounts of PHB were similar, an M W of 10.7 × 10 5 Da, an M N of 2.6 × 10 5 Da and a PDI of 4.1 were obtained from constitutive pUC19-nativeP-phaCAB A−04 -expressing E. coli, indicating that slow and low expression prolonged and maintained the phaC A−04 polymerization activity. It was observed that high levels of phaC proteins resulted in high levels of PHB production, but the chain termination reaction of PhaC polymerization activity frequently occurred faster than that observed with retarded and low expression of the phaC protein under the constitutive promoter of pUC19, which in turn resulted in a low amount of PHB with a high molecular weight. In the latter case, the low-level phaC protein slowly utilized the substrate, 3hydroxybutyric-CoA, via a low-competition reaction with other phaC proteins.

Conclusion
This study aimed to improve functional PhaC A−04 expression levels in E. coli JM109 and found that the cspA promoter in a cold-inducible vector can enhance total PhaC A−04 expression and TF chaperones showed obvious effects on enhancing PhaC solubility. However, the ratio of soluble to total PhaC A−04 proteins did not play an important role in PHB productivity, but the TF chaperone resulted in the production of PHB with reduced molecular weights. Thus, the high level of phaC A−04 resulted in a high amount of PHB, but the chain termination reaction of PhaC polymerization occurred faster than that with the retarded and low expression of phaC A−04 by the constitutive promoter pUC19, which in turn resulted in a low amount of PHB with a high molecular weight. The findings suggest that PhaC A−04 is primed by chain elongation and high molecular weight PHB is obtained by adding HB units to the primed PhaC which requires not only PhaC protein but also substantial amount of HB-CoA substrate.

Strains and plasmids
The E. coli strains and plasmids used in this study are listed in Table 1. The PHB-producing C. necator strain A-04 [49] was used to isolate the phaCAB A−04 gene operon. All bacterial strains were grown at 37 °C in Luria-Bertani (LB) medium supplemented with 100 µg/L ampicillin. The LB medium contained (per liter) 10 g of tryptone (Himedia, Mumbai, India), 5 g of yeast extract (Himedia, Mumbai, India) and 10 g of NaCl (Merck KGaA, Darmstadt, Germany). Stock cultures were maintained at -80 °C in a 15% glycerol solution. The experiments were performed in a biosafety level 1 laboratory and by researchers and investigators who had undergone biosafety training.

Construction of recombinant plasmids
The phaCAB A− 04 operon PHB biosynthetic genes from C. necator A-04 were PCR-amplified using the following pair of primers: forward primer 5′-ATGGATCCCTCGAGATGGCGACCGGCAAAG-3′ (the XhoI site is underlined) and reverse primer 5′-GTGAATTCAAGCTTTCAGCCCATATGCAGGCC-3′ (the HindIII site is underlined). Primers were designed based on accession numbers FJ897463, FJ897461 and FJ897462. The blunted PCR product was purified and subcloned into pBluescript SK− (Stratagene, La Jolla, CA, USA) linearized by SmaI. The recombinant plasmid digested with XhoI and HindIII was cloned into cold-shock-inducible pColdI and pColdTF vectors (Takara Bio Inc., Shiga, Japan) at the XhoI and HindIII restriction sites, yielding pColdI-phaCAB A− 04 and pColdTF-phaCAB A− 04 , respectively. For the plasmid pGEX-6P-1-phaCAB A− 04 , the phaCAB A− 04 operon was amplified by the primers pGEX-F and pGEX-R ( Table 1). The 3885-bp DNA fragment was digested by BamHI and XhoI and cloned into BamHI-XhoI-digested pGEX-6P-1 to obtain pGEX-6P-1-phaCAB A04 . To construct the constitutive expression vector pUC19-nativeP-phaCAB A− 04 , the primers nativeP-phaCAB A− 04 -F and nativeP-phaCAB A− 04 -R were used to amplify the phaCAB A− 04 operon, including its native promoter. The blunted PCR product was purified and cloned into SmaI-linearized pUC19 (Thermo Fisher Scientific, Inc., Waltham, MA, USA), yielding pUC19-nativeP-phaCAB A − 04 . PCRs were performed using Q5® High-Fidelity DNA Polymerase (New England Biolabs, Ipswich, MA, USA). E. coli JM109 was used as a host for cloning and PHB production. The accuracy of the constructed plasmid was verified by the corresponding restriction enzyme and sequencing.

Optimization of culture conditions for PHB production in shake flask cultivation
Expression vectors named pColdI-phaCAB A−04 and pColdTF-phaCAB A−04 with the entire phaCAB A−04 operon were transformed into E. coli JM109 by the heat shock method [50]. Shake flask experiments were performed in 250-mL Erlenmeyer flasks containing 50 mL of medium. E. coli JM109 cells transformed with pColdI-phaCAB A−04 or pColdTF-phaCAB A−04 were grown in LB medium containing ampicillin (100 µg/mL) on a rotary incubator shaker (Innova 4300, New Brunswick Scientific Co., Inc., Edison, NJ, USA) at 37 °C and 200 rpm for 24 h. The overnight seed culture was inoculated into fresh LB medium (5% v/v inoculum) containing 100 µg/L ampicillin and 20 g/L glucose prior to induction with temperature and IPTG, separately using three different induction methods (Fig. 1).
For the synthesis of PHB using the conventional induction method, the procedure was performed according to the user manual (Takara Bio Inc., Otsu, Shiga, Japan and E. coli JM109 (pColdTF-phaCAB A−04 ) were induced to produce PHB using the conventional induction method and short-induction method. The effect of GST (the hydrophilic fusion protein) and the tac promoter on PHB production was investigated using E. coli JM109 (pGEX-6P-1-phaCAB A−04 ), which was induced by the addition of IPTG (0.5 mM). The effect of the araBAD promoter and N-terminal thioredoxin fusion protein together with the Cterminal 6His-fusion protein on phaC and PHB production was examined by inducing E. coli JM109 (pBAD/Thio-TOPO-phaCAB A−04 ) with arabinose (1% w/v). E. coli JM109 (pUC19-nativeP-phaCAB A−04 ), which exhibits constitutive expression, was used as a control strain. All these comparison experiments were performed at 15 °C or 37 °C for 48 h.

Conditions for PHB production in a 5-L fermenter
A preculture was prepared in 500-mL Erlenmeyer flasks containing 100 mL of LB medium and grown on a rotary shaker at 37 °C at 200 rpm for 24 h. The preculture was inoculated into a 5-L bioreactor (MDL500, B.E. Marubishi Co., Ltd., Tokyo, Japan) containing 2 L of LB medium supplemented with 100 µg/L ampicillin and 20 g/L glucose at an inoculation volume of 5% (v/v). The agitation speed and the air flow rate were 500 rpm and 1 mL/min, respectively. After an OD 600 of 0.5 was obtained, the cultivation temperature was reduced from 37 °C to 15 °C for 30 min. Next, IPTG was added to the culture at a final concentration of 0.5 mM. After IPTG addition, the cultivation temperature was shifted from 15 °C to 37 °C and maintained at 37 °C for 48 h. Culture samples were collected at 6 h intervals for 48 h.

Analytical methods
Cell growth was monitored by the DCM, which was determined by filtering 5 mL of the culture broth through preweighed cellulose nitrate membrane filters (pore size = 0.22 µm; Sartorius, Goettingen, Germany). The filters were dried at 80 °C for 2 days and stored in desiccators. The net biomass was defined as the residual cell mass Thermal analysis by differential scanning calorimetry (DSC) A 10-mg sample of PHB was encapsulated in an aluminum sample vessel and placed in the sample holding chamber of the DSC apparatus (DSC7, PerkinElmer, Inc., Waltham, MA, USA). STAR e software (version SW 10.00; Mettler-Toledo International Inc., Columbus, OH, USA) was used to operate the DSC apparatus at the Petroleum and Petrochemical College, Chulalongkorn University. The previous thermal history of the sample was removed before the thermal analysis by heating the sample from ambient temperature to 180°C at 10°C/min. Next, the sample was maintained at 180°C for 5 min before cooling at 10°C/min to − 50°C. The sample was then thermally cycled at 10°C/min to 180°C. The melting peak temperature, denoted by T M , was given by the intersection of the tangent to the furthest point of an endothermic peak and the extrapolated sample baseline. The glass transition temperature, denoted by T G , could be estimated by extrapolating the midpoint of the heat capacity difference between glassy and viscous states after heating of the quenched sample.

Data analysis
All the data presented in this manuscript are representative of the results of three independent experiments and are expressed as the mean values ± standard deviations (SDs). One-way analysis of variance (ANOVA) followed by Duncan's test for testing differences among means was conducted using SPSS version 22 (IBM Corp., Armonk, NY, USA). Differences were considered significant at P < 0.05.

Future Research Directions
Recently, state-of-the-art technology has been applied in the development of recombinant technologies for PHB production to replace fossil-derived plastics with competitive green technologies for PHB production. Projects are ongoing in our laboratory to develop a technoeconomic platform for PHB production in both wild-type and recombinant strains from sustainable feedstocks such as glycerol waste from the biodiesel industry. Optimization based on high-cell-density cultivation will be conducted in fed-batch cultivation to enhance productivity and shorten cultivation time. The green extraction and purification process developed in previous work will be integrated into this framework. Purified PHB is used in the development of innovative technologies for applications in microbeads and packaging.
Abbreviations NA: Not Applicable; PHAs: polyhydroxyalkanoates; PHB: polyhydroxybutyrate; DCW: dry cell mass; RCM: residual cell mass; DSC: differential scanning calorimetry; GPC: gel permeation chromatography; : yield coefficient of PHB produced from consumed PHB substrate (g PHB/g PHB substrate); : yield coefficient of the residual cell mass produced from the consumed PHB substrate (g RCM/g PHB substrate) Declarations  Effect of the growth phase suitable for cold-shock induction on DCM and PHB content (% w/w) under the continuous-induction method. The different growth phases were investigated by varying OD600 based on cultivation time (0.5 (2 h, early exponential phase), 1.3 (4 h, middle exponential phase), 2.1 (6 h, late exponential phase) and 2.4 (10 h, stationary phase)) for (A) E. coli JM109 (pColdI-phaCABA-04) and (B) E. coli JM109 (pColdTF-phaCABA-04). A control experiment was performed with 0.0 mM IPTG induction. All the data are representative of the results of three independent experiments and are expressed as the mean values ± standard deviations (SDs). The PhaCA-04 protein was detected by western blot analysis using anti-His tag antibody as the primary antibody. The band appearing in the western blot at the position corresponding to that of the His-tagged phaCA-04 protein was 67 kDa in size for pColdI-phaCABA-04, and the fusion protein of Histagged phaCA-04 and TF was 115 kDa in size. All the data are representative of the results of three independent experiments and are expressed as the mean values ± standard deviations (SDs).