Biosynthesis and Thermal Properties of PHBV Produced from Levulinic Acid by Ralstonia eutropha

Levulinic acid (LA) can be cost-effectively produced from a vast array of renewable carbohydrate-containing biomaterials. LA could facilitate the commercialization of the polymer poly(hydroxybutyrate-co-hydroxyvalerate) (PHBV) and PHBV-based products as carbon substrates. Therefore, this paper focused on the production of PHBV by Ralstonia eutropha with LA for hydroxyvalerate (HV) production, which plays an important role in enhancing the thermal properties of PHBV. Accordingly, the HV content of PHBV varied from 0–40.9% at different concentrations of LA. Stimulation of cell growth and PHBV accumulation were observed when 2–6 g L−1 LA was supplied to the culture. The optimal nitrogen sources were determined to be 0.5 g L−1 ammonium chloride and 2 g L−1 casein peptone. It was determined that the optimal pH for cell growth and PHBV accumulation was 7.0. When the cultivation was performed in large scale (2 L fermenter) with a low DO concentration of 30% and a pH of 7.0, a high maximum dry cell weight of 15.53 g L−1 with a PHBV concentration of 12.61 g L−1 (53.9% HV), up to 81.2% of the dry cell weight, was obtained. The melting point of PHBV found to be decreased as the fraction of HV present in the polymer increased, which resulted in an improvement in the ductility and flexibility of the polymer. The results of this study will improve the understanding of the PHBV accumulation and production by R. eutropha and will be valuable for the industrial production of biosynthesized polymers.

PHB is the first PHA that was identified and is most likely the best-characterized PHA. However, the degradation temperature of PHB is just a few degrees above its melting temperature, which results in instability during the melting stage. Additionally, the brittleness, hardness and crystalline nature of PHB limit its applications [12]. The incorporation of other monomeric units into HB polymer chains can lead to copolymers with improved properties [13][14]. Work focused on improving the properties of PHB has led to the copolymer of HB and 3-hydroxyvalerate (HV), namely poly(3-hydroxybutyrate/3-hydroxyvalerate) (PHBV) [15]. When compared to PHB homopolymer, the PHBV copolymer has better physical properties such as impact resistance, toughness, flexibility and other properties involved in the manufacturing process. Additionally, the performance of PHBV can vary greatly when this polymer contains different proportions of the HV monomer. Increasing the HV content in PHBV from 0 to 50% can significantly lower the melting point of the resulting polymer.
The structure of PHBV can be manipulated by the types of carbon sources supplemented into the medium. Dionisi et al. obtained a homopolymer of polyhydroxyvalerate (PHV) from propionate and a copolymer (34% HB and 66% HV) from a mixture of acetate and propionate [16]. Beccari et al. obtained a copolymer of HB and HV (50% HB and 50% HV) from a mixture of acetate, propionate, butyrate and valerate [17]. Several odd-numbered carbon sources (e.g., propionate and valerate) were supplemented into the medium to incorporate HV units into PHBV.
Although PHBV could be produced by different types of carbon sources, the main obstacle hindering the economical production of PHBV is the cost of the carbon substrate, which accounts for 28-50% of the total production cost during microbial fermentation [18][19]. Therefore, it is important to manipulate the substrate composition so that synthetic PHBV production by microbial fermentation is inexpensive. Levulinic acid (LA) is a renewable coproduct that has drawn interest as substrates for PHA biopolymer synthesis. LA can be produced cost-effectively from a vast array of carbohydrate-containing renewable biomaterials, including cellulose-containing forest and agricultural waste residues, paper mill sludge, and cellulose fines from paper production processes. Because of this, economic projections indicate that LA production costs could fall to as low as $0.04-$0.10/lb depending on the scale of operation [20][21][22]. Several studies have demonstrated the use of LA as a sole carbon source or a co-substrate for cell growth and PHA biosynthesis [23][24]. LA can serve as a cheap alternative to conventional fermentation substrates. However, problems remain in using LA for PHA biosynthesis. A maximal PHA content of only 38.3% (w/w) was achieved when Alcaligenes sp. SH-69 was cultivated on glucose and LA [25]. Shaking-flask fermentation of Burkholderia cepacia (formerly Pseudomonas cepacia) containing 2.2% (w/v) xylose and concentrations of LA ranging from 0.07% to 0.67% (w/v) yielded 4.4-5.3 g L 21 of dry cell biomass, containing 42-56% (w/w) PHBV [26]. The low content of PHBV in biomass leads to a high cost for both production processes and downstream processing and thus restricts the use of LA for PHBV biosynthesis.
In this work, the effects of different constituents of the medium and other culture conditions for PHBV production by R. eutropha on LA during shaking-flask fermentation were investigated. Then, the optimal medium for PHBV production was applied in 2-L fermentation. Finally, the thermal properties of PHBV were analyzed using differential scanning calorimetry (DSC) and thermogravimetry (TG). The results of this study will improve the understanding of PHBV accumulation and production by microbes and be valuable for industrial polymer biosynthesis.

Bacterial Strain
Ralstonia eutropha H16 was used in all experiments. The strain was maintained on nutrient agar slants at 4uC and sub-cultured monthly.

Media and Cultivation Conditions
Nutrient broth consisting of 10 g L 21 beef extract, 10 g L 21 tryptone and 2.0 g L 21 yeast extract was used for seed cultures. For the nutrient agar slants, 2% agar was added to this medium. The fermentation medium (MSM) contained 3.0 g L 21 Na 2 H-PO 4  The initial pH of all media was adjusted to 7.0 (unless otherwise indicated) with NaOH (1 M) and HCl (0.5 M). The effects of using different LA co-carbon sources, including 20 g L 21 fructose, glucose, sucrose, lactose, dextrin, soluble starch, and molasses, on PHBV production were investigated. A series of nitrogen sources was also investigated. All media were prepared with distilled water and sterilized at 121uC for 30 min.

Shaking-flask Fermentation
Overnight culture (16h) of R. eutropha was prepared using nutrient broth medium for inoculum. 50 mL of the same media in 250 mL flask were used for ferment of PHBV with 2.5 mL (5%, v/ v) of inoculum under shaking conditions for 72 h at 30uC and 200 rpm.

2-L Fermentation
Batch cultivation was also carried out in a 2-L fermenter (Applikon BioBundle, Holland) containing 1.2 L of fermentation medium (MSM) with 5% (v/v) inoculum at 30uC and aeration (500 rpm and 2 vvm air volume/culture volume/min). The pH was maintained at 7.0 by adding 1 M LA or 1 M NaOH.

PHBV Quantification
PHBV quantification was carried out according to the propanolysis method proposed by Riis and Mai [27], and Chen and Li [28] with modifications. Sealed tubes containing 2 mL of dichloroethane, 1.6 mL of propanol, 0.4 mL of hydrochloric acid, 200 mL of a propyl benzoate solution (internal standard) and the lyophilized bacterial pellets recovered from fermented solids were heated at 100uC for 4-6 h. After cooling to room temperature, 4 mL of distilled water was added for extraction. Thereafter, 0.6 mL of the organic phase was injected into a gas chromatograph at 220uC, which was equipped with a flame ionization detector (FID) and a SE-30 capillary column (30 m60.25 mm60.5 mm). N 2 was used as the carrier gas. The temperature of the oven was programmed for the efficient separation of peaks. First, the oven temperature was held constant at 100uC for 1 min. Then, the temperature was increased to 170uC at a rate of 10uC/min, and this temperature was maintained for 5 min. The detector temperature was then increased to 250uC. Calibrations for PHB and PHV were performed with a standard of poly(3-hydroxybutyric-co-hydroxyvaleric acid) (12 wt. % PHV) of natural origin (Sigma-Aldrich Chem.).

Thermal Properties of PHBV
The thermal properties of PHBV were analyzed by DSC (DSC204, Netzsch) and TG (TG209F1, Netzsch). Specimens weighing approximately 3 mg were used for the DSC study. Heating and cooling rates were maintained at 10uC/min during the DSC runs. The specimens were heated from 230 to 200uC for 3 min. The melting temperature (Tm), melting enthalpy (nH f ), and crystallinity (Xc) were obtained from the thermograms. Specimens weighing approximately 4 mg were used for the thermogravimetry (TG) study. The initial thermal decomposition temperature (T eoi ), onset temperature (T em ), finished thermal decomposition temperature (T eof ), and thermal weight loss (Wtloss) were obtained. The specimens were heated from 0 to 400uC at a rate of 10uC/min during the TG runs in order to test the temperature of thermal degradation.

Statistical Analysis
All data were analyzed using Microsoft Excel, Origin and SPSS. The treatment effects were carried out with one-way ANOVA and the LSD multiple range test was used to determine the statistical significance (P,0.05) between pairs with SPSS.

Growth of R. eutropha and PHBV Yield with different Carbon Sources
There was no significant difference in DCW and PHBV production from R. eutropha with increasing inoculum size (Fig. S1). So 5% (v/v) of inoculum was used in following shaking flask and fermenter experiments.
Although R. eutropha could grow when LA was used as the sole carbon source, the production of dry cell weight (DCW) and PHBV were lower (1.29 and 0.03 g L 21 , respectively) than that in the presence of additional carbon source (4.39 and 3.18 g L 21 , respectively) (Table 1). However, only PHB accumulated when glucose was used as sole carbon source, which indicated the possibility of use of LA as a precursor for PHV production. Even though, sucrose, lactose, dextrin, soluble starch, and molasses supported the growth of R. eutropha, but insignificant PHBV accumulation were observed. Thus, in the present study, R. eutropha produced significantly higher amounts of PHBV from LA in the presence of glucose as a co-substrate.

Effect of different Concentrations of Glucose and LA on PHBV Accumulation by R. eutropha
According to the experimental data with different concentration of glucose as a co-substrate in the presence of constant amount of LA (5 g L 21 ), DCW and PHBV concentration showed increasing trend with concentration of glucose up to 25 g L 21 followed by progressive fall of production. In contrast, concentration of HV decreased with an increasing concentration of glucose (0-25 g L 21 ) and increased with an increasing concentration of glucose (25-35 g L 21 ). However, the concentration of HV remained nearly constant at glucose concentrations greater than 35 g L 21 (Fig. 1).
DCW and PHBV were also affected by the amount of LA in the medium. Increment of DCW and PHBV production were observed during the raise of LA concentration up to 6 g L 21 beyond which showed significant reduction and complete inhibition of DCW and PHBV with 10 g L 21 LA. (Fig. 2).
Effects of Nitrogen Sources on PHBV Accumulation by R. eutropha Production of PHBV by R. eutropha was found to be significantly affected by the nitrogen concentration. Accordingly, there was a significant increase in DCW after the addition of ammonium chloride, ammonium sulfate and urea as a source of inorganic nitrogen, however, a higher PHBV production observed with the addition of ammonium chloride compared to other inorganic nitrogen source used (Table S1). Similar experimental evidences also reviled that casein peptone found to be better source of organic nitrogen source. Supporting the fact that, a maximum DCW (5.74 g L 21 ) and PHBV (3.86 g L 21 ) were obtained with the combination of those two organic and inorganic nitrogen sources (Table S2).
According to our experimental data revealed in Fig. 3a, cell growth showed increasing with the concentration of NH 4 Cl, whereas the PHBV production increased up to 0.5 g L 21 and then progressively decreased. Thus, the optimized NH 4 Cl concentration was found to be 0.5 g L 21 . Likewise, the cell growth and PHBV production with different concentrations of casein peptone were also investigated. As shown in Fig. 3b, PHBV production decreased with an increase in casein peptone concentration while the cell growth remained almost the same. Thus, the optimized casein peptone concentration was found to be 2.0 g L 21 .
Effects of the Initial pH on PHBV Accumulation by R. eutropha Cell growth was completely inhibited under pH 5.0. However, a maximum DCW of 5.67 g L 21 and PHBV concentration of 3.71 g L 21 was obtained when the pH was 7.0 ( Fig. 4). Although

Accumulation of PHBV in a 2-L Fermenter
The 250 mL flask experiment proved that the pH value is an important parameter for PHBV production and that PHBV production can be significantly improved by controlling the pH. Therefore, 25 g L 21 glucose and 4 g L 21 LA were added at the beginning as the carbon sources in a 2-L fermenter, and the culture was automatically fed to keep the pH at 7.0. Fig. 5 shows the time course of R. eutropha growth in a 2-liter fermenter at pH 7.0. At 16 h, there was a significant decrease in the concentration of DO, which was followed by an increase in DO to a concentration of 90%. The dry cell weight increased to 13.41 g L 21 at 81 h. PHBV reached a final concentration of 11.08 g L 21 , up to 82.6% of the cell weight. However, the HV content was only 30.6% (w/w).
To enhance the HV content in PHBV, the 2-L fermenter was programmed to automatically keep the DO of the culture at 30% at pH 7.0 (Fig. S2). Aliquots of 25 g L 21 glucose and 4 g L 21 LA were also added as carbon sources. Fig. 6 shows the growth curve of R. eutropha in a 2-Lfermenter. The dry cell weight increased to 15.53 g L 21 in 56 h, and PHBV reached a final concentration of 12.61 g L 21 , up to 81.2% of the cell weight. The highest HV content obtained was 53.9% (w/w).

HV Fractions and Thermal Properties
The thermal properties of PHBV (containing 0, 16 and 53% of HV monomer, respectively) were analyzed by DSC and TG. Fig. 7    shows the DSC thermograms of the different PHBV specimens. The degree of crystallinity (Xc), melting point (Tm), and melting enthalpy (nH f ) of the PHBV specimens are summarized in Table 2. The data in Table 2 suggest that the degree of crystallinity of PHBV increased with decreasing HV content in the polymer, which might mean that the three polymers exhibited obvious differences in crystallinity. The melting point of PHBV also decreased with an increase in the amount of HV, which potentially resulted in an improvement in the ductility and flexibility of the polymer. Fig. 8 shows the TG graph of the different PHBV specimens. The results of a TG graph analysis are also listed in Table 2. It can be seen that the thermal decomposition temperatures of three PHBV specimens were in the range of 240-290uC, and there were no significant decreases in T eoi , T em , T eof with increasing HV content.

Discussion
Recent studies have shown that PHB is commonly produced at an industrial level from Cupriavidus necator, Alcaligenes latus and R. eutropha [11,[29][30]. However, a melting temperature that is close to the degradation temperature and the brittle, hard and crystalline nature of PHB limits its applications [12]. In this study, the bacterium R. eutropha was able to use LA as the co-carbon source to synthesize the copolymer PHBV in shaking-flask and 2-L fermentations.
LA can be produced cost-effectively from a vast array of carbohydrate-containing renewable biomaterials. Because of this, economic projections indicate that LA production costs could fall to as low as $0.04-$0.10/lb depending on the scale of operation [20][21][22]. Since LA is cheap and is a structural analog of pentanoic acid, it has been assessed as a secondary substrate in PHA biosynthesis. In present study, the optimized PHBV yield in a 2-L fermenter was 12.61 g L 21 , up to 81.2% of the cell weight when LA was the co-carbon source (Fig. 6). The PHBV content was much higher than that achieved by Burkholderia sp. IS-01 in a 7-L fermenter, which contained gluconate (20 g L 21 ) and LA (5 to 12.5 g L 21 ) as the co-substrates [31]. Additionally, although high concentration of LA is known to be toxic to microorganisms and to lead to a decrease in growth, there was a significant increase in the DCW of R. eutropha as the concentration of LA was increased from 2 to 8 g L 21 in this study (Fig. 2). This result was quite different from that obtained with R. eutropha KHB-8862, in which LA addition showed no stimulatory effect on cell growth [32]. Although a stimulatory effect of LA addition has been observed [25], it has been restricted to a relatively low concentration (0.5 g L 21 ) of LA. Conversely, the maximum cell growth and, consequently, the maximum PHBV production were highly dependent on the nitrogen source and the initial pH. Even though previous studies have described urea as the best nitrogen source for PHA production by R. eutropha [33][34], our experimental evidence reviled that using NH 4 Cl and casein peptone, 70% more PHBV was produced compared to the batch fermentation using urea as the initial nitrogen source (Fig. 3, Table S1).
Although LA can act as co-carbon source for PHBV production, no significant PHBV was accumulated when sucrose, lactose, dextrin, soluble starch, and molasses were used as carbon sources (Table 1). More importantly, when glucose was chosen as the sole carbon source, only PHB accumulated, which indicated that LA was the precursor for PHV and might be an effective means by which to control the HV content in the copolyesters. In this study, as the concentration of LA varied from 2 to 8 g L 21 , the HV content significantly increased (Fig. 2). It was suggested that LA is first activated to form levulinyl-CoA and then split into propionyl-CoA and acetyl-CoA. The two intermediates are either used for cell growth via the main metabolism pathway or condensed into 3-ketovaleryl-CoA for HV via the well-established PHA biosynthesis pathway (Fig. S3). This might be the reason for the increase in the HV content when the LA concentration was increased because more propionyl-CoA was formed. It was also shown that the percentage of HV reached a maximum at 36 h before the most active PHBV biosynthesis occurred (Fig. 6). Thereafter, the percentage of HV declined while PHBV production increased. This may be due to the higher synthesis rate of HB because the synthesis of HV depends on the level of metabolism needed to convert the precursor substrates into their corresponding hydroxyacyl-CoA thioesters [35].
In addition to the initial LA concentration, the concentrations of the nitrogen source and DO also have important roles in determining the HV content of PHBV. There was a significant increase in the HV content with an increasing concentration of casein peptone (Fig. 3). It has been proposed that HV production was significantly influenced by suitable C:N ratio [28,31]. More importantly, 54% HV content could be produced at a low concentration of DO; however, only 30.6% HV content was synthesized at a high concentration of DO (Figs. 5 and 6). A previous study also demonstrated that HV was easily accumulated by microorganisms under such unbalanced growth conditions of DO that combined intermittent aeration and nutrient limitation [2]. Accordingly, obtaining the desired fraction of HV in PHBV requires the proper ration of C:N and stress resistance conditions.
The fraction of HV present in PHBV had a significant effect on the melting point, which decreased with an increase in the amount of HV in PHBV. This resulted in an improvement in the ductility and flexibility of the polymer. In this study, three PHBV specimens did not begin thermal decomposition until 220uC ( Fig. 8 and Table 2), which is in agreement with a previous report [14,36]. Because the melting temperatures of three PHBV specimens were approximately 100-150uC and their thermal decomposition temperatures were above 220uC, it is clear that increasing the HV content in the PHBV polymer greatly improved its workability. To conclude, biodegradable polymers, especially PHBV, will certainly play an important role in the plastics market in the future due to their biodegradability and to the use of renewable resources for their production. Although their manufacturing costs today are still too high to compete with conventional petroleum-based polymers, LA has great potential for large-scale production of the polymer as it can be produced cost-effectively. When batch cultivation was conducted in a 2-L lab scale fermenter, a signifi-cantly higher maximum dry cell weight of 15.53 g L 21 with a PHBV concentration of 12.61 g L 21 (53.9% HV), up to 81.2% of the dry cell weight, was obtained. More importantly, a desired fraction of HV in the PHBV could be obtained with the proper C:N ratio and stress resistance conditions, which will certainly improve PHBV competitiveness and make the broad use of these biopolymers possible in the future.  Supporting Information