Fructose-Based Production of Short-Chain-Length and Medium-Chain-Length Polyhydroxyalkanoate Copolymer by Arctic Pseudomonas sp. B14-6

Arctic bacteria employ various mechanisms to survive harsh conditions, one of which is to accumulate carbon and energy inside the cell in the form of polyhydroxyalkanoate (PHA). Whole-genome sequencing of a new Arctic soil bacterium Pseudomonas sp. B14-6 revealed two PHA-production-related gene clusters containing four PHA synthase genes (phaC). Pseudomonas sp. B14-6 produced poly(6% 3-hydroxybutyrate-co-94% 3-hydroxyalkanoate) from various carbon sources, containing short-chain-length PHA (scl-PHA) and medium-chain-length PHA (mcl-PHA) composed of various monomers analyzed by GC-MS, such as 3-hydroxybutyrate, 3-hydroxyhexanoate, 3-hydroxyoctanoate, 3-hydroxydecanoate, 3-hydroxydodecenoic acid, 3-hydroxydodecanoic acid, and 3-hydroxytetradecanoic acid. By optimizing the PHA production media, we achieved 34.6% PHA content using 5% fructose, and 23.7% PHA content using 5% fructose syrup. Differential scanning calorimetry of the scl-co-mcl PHA determined a glass transition temperature (Tg) of 15.3 °C, melting temperature of 112.8 °C, crystallization temperature of 86.8 °C, and 3.82% crystallinity. In addition, gel permeation chromatography revealed a number average molecular weight of 3.6 × 104, weight average molecular weight of 9.1 × 104, and polydispersity index value of 2.5. Overall, the novel Pseudomonas sp. B14-6 produced a polymer with high medium-chain-length content, low Tg, and low crystallinity, indicating its potential use in medical applications.


Introduction
Some microbes have developed various tactics to survive the harsh Arctic environment [1][2][3]. One of these survival mechanisms is to adapt to temperature and nutrition fluctuations by accumulating energy and carbon sources within the cell as polyhydroxyalkanoate (PHA) granules [4,5], which also function as chaperone-like molecules to protect internal cellular systems [6][7][8]. Although several microbes thrive at low temperatures and accumulate PHAs, the metabolism of Arctic bacteria is generally quite slow, weakening their potential as PHA producers [7].
PHAs are produced by bacteria using various renewable feedstocks, and are easily degraded under biological conditions [9]. Properties such as the monomer composition, distribution, and molecular weight of these biocompatible polyesters are controlled by different substrates or production hosts [10][11][12]. Depending on the desired property, 3 g/L KH 2 PO 4 , 0.5 g/L NaCl, 1 g/L NH 4 Cl, 20 g/L carbon source, and 1 g/L yeast extract. The inoculum size was 1% of the final volume, cultivated at 30 • C and 200 rpm shaking [29]. The optimization procedure altered the media composition (carbon source, nitrogen source, and their concentrations), inoculum size, cultivation time, and incubation temperature. Flask experiments used 50 mL working volume in 250 mL total volume baffled flask and orbital shaking at 200 rpm.

PHA Analysis
The PHA composition of Pseudomonas sp. B14-6 was evaluated using a previously described GC-MS method [30]. Strains were cultured in PHA production media, the culture was centrifuged (3500× g for up to 30 min at 4 • C), and the cells were washed twice with deionized water. The washed cells were transferred to a glass vial for lyophilization. Fatty acid methyl ester (FAME) derivatization was conducted to prepare the sample for GC-MS, as previously described [31,32]. Briefly, 1 mL of 15% sulfuric-methanol and 1 mL of chloroform were added to the lyophilized sample, followed by heating for 2 h at 100 • C. The sample was then held at room temperature for 15 min and 1 mL of distilled water was added. The sample was vortexed, the chloroform layer was transferred to a new tube containing NaSO 4 , and the sample was filtered into a clean borosilicate glass tube. The resulting sample was analyzed by GC-MS (Perkin Elmer, Waltham, MA, USA) equipped with a fused silica capillary column (Elite-5 ms (Perkin Elmer); 30 m × 0.25 mm i.d. × 0.25 µm) and subjected to a linear temperature gradient for PHA (50 • C for 1 min, increased at 15 • C/min to 120 • C for 2 min, and then increased at 10 • C/min to 300 • C for 10 min). The injector port temperature was set at 250 • C. Mass spectra were obtained by electron impact ionization at 70 eV, and scan spectra were obtained within the range of 45-450 m/z. Selected ion monitoring was used for the detection and fragmentation analysis of the major products [33]. All PHA authentic samples from 3HB to 3HtD were derivatized as methyl esters and applied to GC-MS to confirm the retention time and mass spectrum of each compound.

Physical Properties of Produced PHA
Analysis of PHA produced by Pseudomonas sp. B14-6 was conducted using several methods. The thermal properties of PHA were investigated by differential scanning calorimetry (DSC; TA Instruments, New Castle, DE, USA), using temperatures ranging from −60 • C to 180 • C, increasing at 10 • C/min. Gel permeation chromatography (GPC; Young Lin Instrument Co., Anyang, Korea) equipped with a K-804 column was used to measure the molecular weight of the copolymer. A PHA sample (0.1 mg) was dissolved in 1 mL of chloroform and injected in the GPC instrument at 40 • C oven temperature for 30 min. Results were analyzed based on a calibration curve using Shodex polystyrene standards SM-105 (Tokyo, Japan), which have a molecular weight range from 6530 Da to 2330 kDa [34].

Screening of PHA-Related Genes in Pseudomonas sp. B14-6
We previously reported that whole-genome sequencing of Pseudomonas sp. B14-6 revealed a genome 6,776,772 base pairs in length. The strain showed dramatic membrane changes with temperature, and important players of membrane modification were identified [28]. Using the genome sequencing results, we identified two PHA-synthesis-related gene clusters and compared them with PHA-related genes from Ralstonia eutropha, Pseudomonas, and other PHA-producing species by BLAST (Figure 1). The two gene clusters had the orders phaR-phaP1-phaC1-araC-phaB-phaA-phaC2 and phaC4-phaZ-phaC3-phaD-phaP2-phaP3, which shared high similarity with Pseudomonas sp. 61-3, although the annotation might differ [35]. Amongst these two PHA synthesis-related gene clusters, we identified four PHA synthase genes, named phaC1-C4, from the open reading frame order assigned phaC4-phaZ-phaC3-phaD-phaP2-phaP3, which shared high similarity with Pseudomonas sp. 61-3, although the annotation might differ [35]. Amongst these two PHA synthesis-related gene clusters, we identified four PHA synthase genes, named phaC1-C4, from the open reading frame order assigned by whole-genome sequencing. As PHA synthases can be classified as types I to IV [36], we attempted to classify each phaC by the BLAST search results. One cluster had acetyl-CoA thiolase (phaA), acetoacetyl-CoA reductase (phaB), and two PHB synthases (phaC1 and phacC2), sharing high similarity with the PHA synthesis cluster of Ralstonia eutropha (a PHA synthase type-I that produces scl-PHA) (Figure 1a). The order of the PHA synthesis gene in R. eutropha is phaC1-phaA-phaB1; however, the order in Pseudomonas sp. B14-6 was phaB-phaA-phaC2, in the reverse order. We found that phaC1 was a PHA synthetase, not a PHA synthase, with a much longer sequence than other phaC genes; it contained a PHA synthase type-IV-related domain, although its function was unknown. Next to phaC1 was the phaR gene, known for the subunit re- One cluster had acetyl-CoA thiolase (phaA), acetoacetyl-CoA reductase (phaB), and two PHB synthases (phaC1 and phacC2), sharing high similarity with the PHA synthesis cluster of Ralstonia eutropha (a PHA synthase type-I that produces scl-PHA) (Figure 1a). The order of the PHA synthesis gene in R. eutropha is phaC1-phaA-phaB1; however, the order in Pseudomonas sp. B14-6 was phaB-phaA-phaC2, in the reverse order. We found that phaC1 was a PHA synthetase, not a PHA synthase, with a much longer sequence than other phaC genes; it contained a PHA synthase type-IV-related domain, although its function was unknown. Next to phaC1 was the phaR gene, known for the subunit required to produce PHA with phaC in the PHA synthase type-IV system. However, it was difficult to determine its specific function because phaC had a longer sequence than the PHA synthase type-IV of Bacillus megaterium, and was similar to that of Pseudomonas sp. 61-3 [36]. In addition, the two PhaC genes in R. eutropha are located very far apart, with one cluster having the order phaC1-phaA-phaB1-phaR-bktB (β-ketothiolase) and the other having the order phasin2-phaC2-phaB2. The gene cluster of Pseudomonas sp. B14-6 was thus unique, with both phaC genes at neighbor locations.
The other PHA gene cluster also had two phaC genes (phaC3 and phaC4) and a PHA depolymerase (phaZ) resembling a PHA synthase type-II that appears in typical PHAproducing Pseudomonas spp. Compared with the PHA gene cluster of Pseudomonas spp., the length of the phaC3 gene was the same as that of Pseudomonas putida, and the phaD was slightly longer than that of P. putida and Pseudomonas aeruginosa, but shorter than the long phaD gene of Pseudomonas oleovorans ( Figure 1b). As both Pseudomonas sp. 61-3 and Pseudomonas sp. B14-6 have similar gene clusters and two different types of PHA synthesis gene clusters for scl-PHAs (phaC2) and mcl-PHAs (phaC3 and phaC4), it was expected that Pseudomonas sp. B14-6 would produce similar PHAs as Pseudomonas sp. 61-3 or Pseudomonas sp. MPC6.

Analysis of PHA from Pseudomonas sp. B14-6
According to the gene search results, we focused on two points: (1) Pseudomonas sp. B14-6 could produce PHA, and (2) PHA produced by the strain might contain monomeric units, both medium and short chain length.

Optimization of mcl-PHA Production
To determine the optimal nutrients for mcl-PHA production, we screened various carbon sources, including glucose, fructose, xylose, N-acetyl glucosamine, galactose, lactose, and sucrose, and compared monomeric composition from the carbon source (Table 1). Among them, we focused on fructose as the main carbon source (Figure 3a). Additionally, we screened yeast extract, malt extract, peptone, and tryptone as complex nitrogen sources, as well as ammonium sulfate and ammonium chloride as defined nitrogen sources. Among these, yeast extract yielded the highest PHA content (Figure 3b). Further optimization was conducted using fructose as the carbon source and yeast extract as the nitrogen source. We determined that 5% fructose yielded the highest PHA content of 36% (Figure 3c), while optimal biomass and PHA content were achieved with 0.1% yeast extract (Figure 3d). Another parameter that affects PHA production is the amount of microbes inoculated in the media. Thus, we evaluated PHA production using inoculum at 0.5% to 4% of the final volume, determining 0.5% as the optimal inoculum size ( Figure 3e). PHA production is also affected by cultivation time, in which gene expression and polymer building is not completed during a short cultivation, and accumulated PHA is degraded as an energy source due to exhaustion of the initial carbon source during a long cultivation. Therefore, cultivation time for PHA production was evaluated for 168 h, revealing that culture for 120 h yielded optimal biomass and PHA content, but volumetric productivity was the highest at 48 h (Figure 3f). The composition of PHA was changed by different temperatures (data not shown). As a result, the optimal conditions for PHA production with Pseudomonas sp. B14-6 were set as 5% fructose, 0.1% yeast extract, 0.5% inoculum size, and 120 h culture time at 30 • C. PHA production in optimal conditions was higher than other psychrophilic Pseudomonas spp. (Table 2).

Optimization of mcl-PHA Production
To determine the optimal nutrients for mcl-PHA production, we screened various carbon sources, including glucose, fructose, xylose, N-acetyl glucosamine, galactose, lactose, and sucrose, and compared monomeric composition from the carbon source (Table 1). Among them, we focused on fructose as the main carbon source (Figure 3a). Additionally, we screened yeast extract, malt extract, peptone, and tryptone as complex nitrogen sources, as well as ammonium sulfate and ammonium chloride as defined ni-    Figure 3e). PHA production is also affected by cultivation time, in which gene expression and polymer building is not completed during a short cultivation, and accumulated PHA is degraded as an energy source due to exhaustion of the initial carbon source during a long cultivation. Therefore, cultivation time for PHA production was evaluated for 168 h, revealing that culture for 120 h yielded optimal biomass and PHA content, but volumetric productivity was the highest at 48 h (Figure 3f). The composition of PHA was changed by different temperatures (data not shown). As a result, the optimal conditions for PHA production with Pseudomonas sp. B14-6 were set as 5% fructose, 0.1% yeast extract, 0.5% inoculum size, and 120 h culture time at 30 °C. PHA production in optimal conditions was higher than other psychrophilic Pseudomonas spp. (Table 2).

Monitoring Time-Dependent PHA Production and Fructose Syrup Application
The time-dependent production of PHA by Pseudomonas sp. B14-6 was investigated under optimized conditions, measuring DCW, amount of PHA produced, and carbon source consumption over 120 h. Almost no PHA was produced during the first 24 h, but both PHA content and DCW increased from 48 h until 72 h. DCWs at 72 h and 120 h were similar and volumetric productivities were slightly decreased, but the highest PHA content was obtained at 96 h ( Figure 4a). The consumption of fructose increased after 48 h, and the pH began decreasing from approximately 7 to 6.69 at 120 h (Figure 4b). We then evaluated the use of fructose syrup composed of 20% glucose, 36% fructose, and 44% sucrose as an economical carbon source. The production of PHA with media that contained 5% fructose syrup achieved 23.7% PHA content (Figure 4c). Use of fructose syrup induced a different pH pattern, optimal culture time, and lower PHA content than using fructose (Figure 4d). No catabolite repression occurred during cultivation, as all sugars were consumed together.

Monitoring Time-Dependent PHA Production and Fructose Syrup Application
The time-dependent production of PHA by Pseudomonas sp. B14-6 was investigated under optimized conditions, measuring DCW, amount of PHA produced, and carbon source consumption over 120 h. Almost no PHA was produced during the first 24 h, but both PHA content and DCW increased from 48 h until 72 h. DCWs at 72 h and 120 h were similar and volumetric productivities were slightly decreased, but the highest PHA content was obtained at 96 h (Figure 4a). The consumption of fructose increased after 48 h, and the pH began decreasing from approximately 7 to 6.69 at 120 h (Figure 4b). We then evaluated the use of fructose syrup composed of 20% glucose, 36% fructose, and 44% sucrose as an economical carbon source. The production of PHA with media that contained 5% fructose syrup achieved 23.7% PHA content (Figure 4c). Use of fructose syrup induced a different pH pattern, optimal culture time, and lower PHA content than using fructose (Figure 4d). No catabolite repression occurred during cultivation, as all sugars were consumed together.
The molecular weight of the obtained PHA was characterized by GPC, demonstrating a retention time peak start at 13.02 min, peak maximum at 16.62 min, and peak end at 20.32 min (Figure 5b). The scl-co-mcl PHA had average values, with a number average molecular weight (Mn) of 3.6 × 10 4 , weight average molecular weight (Mw) of 9.1 × 10 4 , Z-average (Mz) of 1.9 × 10 5 , and viscosity average molar mass (Mv) of 8.0 × 10 4 . The polydispersity index value of scl-co-mcl PHA was 2.5, which differed from other values of scl-PHA or mcl-PHA, due to the mixed monomeric unit composition ( Table 3). The high polydispersity index value was due to the various monomer unit composition and breakdown of the polymer in the sample preparation steps. The physical properties of the scl-co-mcl PHA produced by Pseudomonas sp. B14-6 differed from PHAs produced by other species. Therefore, we hypothesize that the scl-co-mcl PHA has potential applications in biomedical or similar industrial fields. The thermal characteristics of our scl-co-mcl PHA differed both in T m and T c from those of the previously reported mcl-PHA from Pseudomonas sp. PAMC28620 composed of poly(25.5% 3HO-co-52.1% 3HD-co-5.7% 3HdD-co-16.7% 3HtD) with T m = 172.8 • C, T g = 3.99 • C, and T c = 54.61 • C [30]. The thermal properties also differed from those of the mcl-PHA from Pseudomonas sp. MPC6 composed of poly(89.5% 3HB-co-1.8% 3HHx-co-3.3% 3HO-co-4.4% 3HD-co-1.1% 3HdD) with T m = 163.5 • C, T g = 2.3 • C, and T c = 46.0 • C [37]. The scl-co-mcl PHA from Pseudomonas sp. B14-6 had lower T m values and higher T g and T c values due to the low 3HB content and high 3HO and 3HD content, which lowered the melting point. As an important polymer parameter, the degree of crystallinity (X c ) was calculated from the enthalpy, revealing 3.82% crystallinity. This X c value was extremely low compared to that reported for mcl-PHAs from Pseudomonas sp. PAMC28620 (X c = 43.7%) [30], Bacillus thermoamylovorans PHA005 (X c = 43.0%) [44], and R. eutropha (X c = 91%) [45]. As highly crystalline polymers have limited application in industrial and medical fields [46], the scl-co-mcl PHA from Pseudomonas sp. B14-6 may have potential in applications that require a sticky and relatively low-temperature modeling polymer, as well as biodegradable characteristics.
The molecular weight of the obtained PHA was characterized by GPC, demonstrating a retention time peak start at 13.02 min, peak maximum at 16.62 min, and peak end at 20.32 min (Figure 5b). The scl-co-mcl PHA had average values, with a number average molecular weight (M n ) of 3.6 × 10 4 , weight average molecular weight (M w ) of 9.1 × 10 4 , Z-average (M z ) of 1.9 × 10 5 , and viscosity average molar mass (M v ) of 8.0 × 10 4 . The polydispersity index value of scl-co-mcl PHA was 2.5, which differed from other values of scl-PHA or mcl-PHA, due to the mixed monomeric unit composition ( Table 3). The high polydispersity index value was due to the various monomer unit composition and breakdown of the polymer in the sample preparation steps. The physical properties of the scl-co-mcl PHA produced by Pseudomonas sp. B14-6 differed from PHAs produced by other species. Therefore, we hypothesize that the scl-co-mcl PHA has potential applications in biomedical or similar industrial fields.

Conclusions
PHA has attracted attention as an alternative material to replace petroleum plastics. The composition of monomers, especially medium-chain-length monomers such as 3HO and 3HD, influences the properties of the polymers, decreasing the rigidity and high crystallinity of PHB and increasing the toughness of the polymer. Pseudomonas species have been studied as possible mcl-PHA producers, including many mesophile Pseudomonas strains. We previously isolated the novel Arctic strain Pseudomonas sp. B14-6, which shared a similar gene structure with Pseudomonas sp. 61-3 and contained two different gene clusters for scl-PHB and mcl-PHA. Although the two species had similar gene clusters and alignments, the new strain produced a PHA with a different monomer ratio than previously reported PHAs. In particular, the strain produced 94% medium-chain-length monomer content, compared with less than 60% for Pseudomonas sp. 61-3. As a result, the scl-co-mcl PHA from Pseudomonas sp. B14-6 exhibited totally different T m , T g , and T c values from other reported PHAs. Furthermore, Pseudomonas sp. B14-6 was able to utilize a wide range of carbon sources, including glucose, fructose, galactose, glycerol, N-acetylglucosamine, and sucrose. The carbon source adaptability suggests the industrial potential of the strain, considering that the well-studied P. putida cannot utilize substrates such as sucrose to produce PHA without metabolic engineering. Finally, we optimized carbon and nitrogen sources, cultivation time, temperature, and inoculum size, and evaluated fructose syrup as a cost-effective feedstock for PHA production. In conclusion, Pseudomonas sp. B14-6 could be used for commercial PHA production with an economical starting material, resulting in a unique polymer that contains a very high proportion of mcl-PHA.