Influence of culture factors on the monomer composition of PHBV.
To study the influence of OTR, valeric acid concentration, and valeric acid addition time on PHBV copolymer production, a factorial design 23 was used (Table 1). Gas chromatography−mass spectrometry (GC−MS) analysis was performed to characterize the monomeric composition of copolymers produced by A. vinelandii OPNA. As shown in Fig. 1a peaks obtained at retention time 4.52, 6.18, and 8.64 were identified as 3HB, 3HV and benzoic acid (internal standard), respectively. For each compound, the analysis of the fragmentation pattern of the peaks, matches with methyl esters of 3-hydroxybutyrate (3HB) and 3-hydroxyvalerate (3HV) showed in Fig. 1b.
Table 2 shows the results obtained under the different conditions evaluated. The highest molar fraction of 3HV (28 ± 2.3 %) was obtained by adding 40 mM of valeric acid (A +1), after 12 h of cultivation, during the onset of the most-active PHA biosynthesis (B -1) and cultures at 100 mL (C -1) of filling volume. On the other hand, a mathematical model was generated and showed in Eq. (2), where the filling volume (C) and therefore the OTRmax, is the variable that had the highest effect on monomeric composition of PHBV.
% 3HV = 12.50 + 0.775A – 0.508 B – 7.26 C Eq. (2).
It is known that changes in the filling volume on shaken flasks has an influence on the oxygen transfer rate (OTR), and therefore in the availability of oxygen in the liquid . Previously, Lee et al.  observed that a decrease in the OTRmax (increasing the filling volume from 50 to 300 mL in 500 mL shaken flasks) increased the molar fraction of 4HB monomers (from 18 to 50 %) in P(3HB)-co-4HB copolymer produced by Comamonas acidovorans. In the case of A. vinelandii, the influence of OTR in monomeric composition seems to be dependent on the strain and culture condition employed. Ryu et al.  did not find an influence of OTRmax on monomeric composition of PHBV produce by UWD strain in shaken flasks culture. This study was carried out in two filling volume (50 and 100 mL in 250 shaken flasks), where in both conditions the content of 3HV was 4.1 ± 0.4 and 4.3 ± 0.8 %, respectively. In contrast, Urtuvia et al.  reported that the OTR has a positive influence on the production of 3HV monomers in PHBV copolymer produced by A. vinelandii OP in bioreactor cultures. These authors observed that increasing agitation rate (300 to 600 rpm), increased the OTRmax (4.3 to 17.2 mmol L-1 h-1), and therefore the molar fraction of 3HV monomers (18 to 35 %).
Characterization of PHBV produced in shaken flasks cultures.
Considering that the higher molar fraction of 3HV was achieved with 40 mM of valeric acid added at 12 h and at high OTR (5.87 mmol L-1 h-1, 100 mL) (Run 3, Table 2), the effect of that concentration of valeric acid added at 12 h was evaluated on cell growth, production, and accumulation of PHBV. Figure 2 shows the growth kinetics, PHBV production, sucrose, and valeric acid consumptions by the OPNA mutant, with respect to those of in the control condition (strain without valeric acid addition). Valeric acid (40 mM) was added at 12 h of cultivation.
A specific growth rate (μ) of 0.047 ± 0.001 h−1 was achieved in the cultures conducted without adding valeric acid and the maximal protein concentration was about 0.9 ± 0.012 g L-1. A feeding strategy in fed-batch culture, it has been suggested to avoid the inhibition of cell growth and PHA accumulation, without affecting the 3HV content in the copolymer. Page et al.  reported that in fed-batch cultures of A. vinelandii UWD, the addition of valerate in a final concentration of 20 mM (6 mM in the vessel), did not affect the PHA content (65 %) when it is compared with the control (without addition of precursor); whereas, the maximum content of 3HV was 23 %. (Fig. 2a). In the case of PHBV production, it was affected by the addition of valeric acid. A maximum accumulation of 38 ± 1.6 % was achieved in the control condition (where only PHB was produced); whereas, only 22 ± 2.4 % were obtained from the cultures conducted with the addition of valeric acid (Fig. 2b).
Our results are in agreement with those previously reported by Myshkina  and Page et al. , who found that the growth and the polymer production were inhibited 21 and 10 %, respectively, by the addition of valerate (10 mM) in the medium. Myshkina et al.  reported, that in shaken flasks cultures of A. chroococcum strain 7B, the addition of valerate (20 mM), inhibits the growth (43 %) and the polymer accumulation (7 %) at 48 h of cultivation. However, these authors evaluated the production of copolymers in species of Azotobacter, adding sodium salts as precursors (valerate), instead of carboxylic acids (valeric acid). In our study, the valeric acid was not neutralized before to be added, and to keep the pH constant during cultivation a high concentration of MOPS was employed. However, the pH was 5.2 ± 0.8 at the end of the culture. It is possible that inhibition of cell growth and polymer accumulation could be related to an effect caused by pH. In this line, Myshkina et al.  evaluated the effect of pH on the cell growth and PHB produced by A. chroococcum 7B in shaken flasks cultures. These authors observed that the optimum pH was at 7.0, whereas, alkaline (8.0) or acid pH (6.0) conditions negatively influence the growth and accumulation.
On the other hand, Chung et al.  reported that pH plays a relevant role in dissociation of acids and their transport. In that sense, it was previously shown that acids as acetic or propionic migrates by anti-carrier transport (dissociated form) or by simple diffusion through the cell membrane (undissociated form). In the last, the acid works as a proton uncoupler, which generates a dissipation of the proton-pumping force affecting ATP synthesis and therefore an inhibition of the cell growth . In the case of propionic acid, is thought to involve an uncoupling role, because the fraction of undissociated forms increases when pH decreases . Therefore, it is possible that a similar behavior occurs during the valeric acid transport; however, further studies will be needed to elucidate how valeric acid inhibits bacterial growth. It is important to point out that as described in bioreactor cultures, there was an inhibition on the cell growth and the polymer accumulation when valeric acid was added, this effect cannot be attributed to pH, because it was controlled during the cultivation.
As it was expected, 3HV monomers were synthesized only when valeric acid was added (Fig. 2c). The highest content of 3HV (39 ± 3 %) was obtained at 36 h of cultivation. It was previously reported that Azotobacter species are able to produce PHBV copolymers, when precursors are added to growth medium [24, 27, 40]. Table 3 shows the highest content of 3HV obtained from Azotobacter species cultured in shaken flasks under different precursor concentrations. It is important to note that, HV monomers production depends on several factors including culture conditions such as type and concentration of precursor; and biochemical and genetic background of the strain. To our knowledge, this molar fraction of 3HV (39 ± 3%) in the PHBV copolymer obtained, is the highest reported until now, in Azotobacter species cultured in shaken flasks cultures.
Effect of valeric acid concentration and OTR on growth and PHBV production in bioreactor cultures.
In order to investigate the influence of valeric acid concentration on production of PHBV, batch cultures were carried out in a 3.0 L Applikon bioreactor. This was operated at 700 rpm (OTRmax 20.3 mmol L-1 h-1) and 29 °C. The pH was controlled at 7.2 ± 0.5. PHBV production was evaluated under two concentrations of valeric acid (10 and 40 mM), the precursor (valeric acid) was not neutralized before to be added at 6 h (during the most-active PHA biosynthesis).
It has been reported that increasing precursor concentration, increases the negative effect on cell growth and polymer accumulation . Figure 3 shows grown kinetics (measured as protein), PHBV accumulation and content of 3HV monomers by strain OPNA of A. vinelandii cultured under two concentrations of valeric acid. As expected, a high precursor concentration (from 10 to 40 mM), negatively influenced specific growth rate (μ), reaching 0.11 ± 0.002 h-1 at 40 mM, compared to those cultures developed at 10 mM of valeric acid, where the μ was 0.14 ± 0.003 h−1. However, this effect was not reflected in the maximal protein concentration, because in both conditions tested the concentration of protein was similar at the end of the culture (Fig. 3a, Table 4).
It is important to point out that the negative effect on cell growth could be related with valeric acid concentration. Previously, Page et al.  reported that in bioreactor cultures of the UWD strain of A. vinelandii, increasing the precursor concentration (from 10 to 40 mM), decreased the cell growth around 25 %. However, there is not an explanation about this effect.
The precursor concentration also affected the accumulation of PHBV. Under both conditions the percentage of accumulation was the same at the beginning of the cultivation; however, after the addition of valeric acid (6 h), accumulation of PHBV decreased in the cultures developed at 40 mM, reaching a maximal percentage of 51 ± 3.6 % at 21 h. In contrast, the maximal percentage of 80 ± 0.7 % was achieved in the cultures carried out at 10 mM (Fig. 3b). This was consistent with the highest yield of PHBV based on sucrose and with the highest volumetric uptake sucrose rate (Table 4).
As shown in figure 3c, the highest content of 3HV was obtained using 40 mM of valeric acid; under this condition, 21 ± 0.5 % of 3HV was obtained at 15 h of cultivation, whereas, in cultures using a lower concentration of precursor (10 mM), the highest production reached was 8.2 ± 0.03 % at 12 h of cultivation. These results reveal that by increasing the concentration of the precursor from 10 to 40 mM, it is possible to increase the content of 3HV about three-fold. It is important to note that the molar fraction of 3HV increased with the valeric acid concentration; however, under this condition the accumulation of PHBV decreased (Fig.3b). Page and Manchak,  reported that in A. vinelandii a diminished accumulation of PHBV could be associated with the effect of valerate (precursor) on the beta-oxidation pathway. It was proposed that catabolism of valerate would inhibit 3-cetoacyl-CoA and acetoacetyl-CoA reductase activities, and these inhibitions could be reflected in a decrease in PHA yield. In addition, inhibition of 3-ketothiolase would be favored by the production of 3HV monomers .
The molar fraction of 3HV decreased with respect to time. A possible explanation for this phenomenon could be attributed to the fact that the valeric acid was exhausted around 15-18 h of cultivation (Fig. 3d). This gradual exhaustion would affect the synthesis of 3HV monomers, and incorporation of PHB monomers would be prevalent, modifying the molar fraction of both monomers in the copolymer . The decrease in molar percentage of 3HV could not be due to degradation, because the 3HV concentration remains at values between 0.48 – 0.54 mg L-1 and between 0.53 - 0.69 mg L-1 at low and high OTRmax, respectively, during the rest of the culture. This is online with that reported by Volova et al. , who found that in batch cultures of Ralstonia eutropha B-5786, the concentration of 3HV remained nearly unchanged, whereby, the changes on molar fraction could be due to the continuous synthesis of PHB and the termination of 3HV production, related with the exhaustion of valerate in the medium.
Considering that the OTR has an important effect on the PHBV composition, experiments changing the OTR by the manipulation of the agitation rate at 300 (4.96 mmol L−1 h−1, low OTRmax) and 700 rpm (20.3 mmol L−1 h−1, high OTRmax), were performed. Valeric acid was added in a final concentration of 10 mM, to decrease its negative influence on cell growth and copolymer production. This addition was made during the cell growth phase of cultivation and the most-active PHA biosynthesis at 6 and 9 h for the cultures in the condition at high and low OTRmax, respectively.
Figure 4a, shows the evolution of dissolved oxygen tension (DOT) and oxygen transfer rate (OTR) in the cultures carried out at different agitation rates. In the cultures developed at 300 rpm, the DOT decreased during the first 3 h of cultivation and remained close to 0 % during the rest of the culture, whereas the OTRmax reached was 4.96 mmol L-1 h-1. At 700 rpm, the DOT progressively decreased and remained close to 2 – 4 % until 18 h of cultivation, the OTRmax was 20.3 mmol L-1 h-1. As expected, under both conditions the OTR was affected by the agitation rate. The OTR profile showed a characteristic region (plateau) during cell growth, that was associated with an oxygen limitation, as described previously by Anderlei et al.  and Díaz-Barrera et al. .
The cell growth (measured as protein) is shown in Fig. 4b. In the cultures conducted at low OTRmax the maximal protein concentration was about 1.26 ± 0.01 g L-1, with a specific growth rate (μ) of 0.03 ± 0.002 h−1, whereas in the cultures grown at high OTRmax, the protein concentration was 2.0 ± 0.2 g L-1 and the μ was 0.14 ± 0.003 h−1. This agrees with previous works that showed a negative influence on the growth rate when the OTR was decreased [9-10, 31].
As shown in Fig. 4c and 4d, sucrose and valeric acid were consumed simultaneously; however, sucrose was not consumed completely, remaining 4 and 1 g L−1 at the end of the cultures developed at 300 and 700 rpm, respectively. On the other hand, valeric acid was completely exhausted after 27 h and 12 h in cultures carried out at 300 and 700 rpm. It is important to point out that the consumption rate of both substrates was negatively affected when the OTRmax decreased (from 20.3 to 4.96 mmol L-1 h-1) (Table 5).
It is important to point out that under the conditions used in the present study, the production of PHBV was growth associated, reaching a maximal concentration of 6.1 ± 0.9 and 6.4 ± 0.2 g L-1 in the cultures at high and low OTRmax, respectively (Fig. 5a). The highest PHBV accumulation (85 ± 3 %) was achieved at low OTRmax after 48h of cultivation; under this condition, the volumetric production rate was of 0.47 gPHBV L−1 h−1, whereas, at high OTR, the accumulation was 80 ± 0.7 % at 21 h of cultivation, and the volumetric production rate was of 0.33 gPHBV L−1 h−1 (Fig. 5b, Table 5). These results suggest that OTR did not influence the concentration or percentage of accumulation of the polymer. This agrees with a previous work that showed that PHAs accumulation in strain OPNA is around 80 %, regardless of the oxygen condition .
Fig. 5c shows the evolution of 3HV content in the polymer produced by A. vinelandii OPNA in batch cultures. The highest molar fraction of 3HV (8.2 ± 0.03 %) was obtained at high OTRmax (20.3 mmol L-1 h-1), 6 h after the addition of the precursor. At low OTRmax (4.96 mmol L-1 h-1) the highest content (6.5 ± 0.6 %) was reached 27 h after the addition of valeric acid. In both conditions evaluated, the content of 3HV dropped toward the end of the culture, to 5.2 ± 0.07 and 4.6 ± 0.06 %, at high and low OTRmax respectively. As mentioned above, this could be related with the valeric acid depletion that affected the synthesis of the 3HV monomer [27, 42]. In accordance with this hypothesis, when valeric acid was exhausted, the molar fraction of 3HV decreased (Fig, 4d, Fig. 5c).
It has been reported that the molecular weight determines different thermomechanical properties of PHA polymers [22, 32]. Besides, the molecular weight depends on the culture conditions, and in the case of copolymers it can depend on the monomer composition [1, 23]. In the case of Azotobacter spp, it has been reported that this bacterium synthesizes PHB with a high and ultra-high molecular mass [10, 23, 32].
Figure. 6 shows the PHBV molecular mass distribution under different OTRmax conditions. The mean molecular mass (MMM) was between 6600 – 6700 kDa, during the exponential growth phase (at 12 h). After this time, the MMM remained constant between 6500 and 6600 until the end of the culture (21 h and 54 h for cultures developed at high and low OTRmax, respectively). These results indicated that there are no significant changes in MMM under the conditions evaluated. In contrast, Gómez-Hernández et al. , reported that in batch cultures of OP strain of A. vinelandii, the MMM was affected by the oxygen condition. At low OTRmax (5 mmol L-1 h -1), the production of ultra-high molecular weight polymers (8300 kDa) was promoted; whereas, at higher OTRmax (8-11 mmol L-1 h-1) the MMM decreased to 3500 kDa.
It is important to point out that, OPNA is a mutant strain derivate from OP; however, it has been previously reported that OP and OPNA strains have different phenotypic responses under several culture conditions . Previously, Castillo et al.  reported that in fed-batch cultures of A. vinelandii OPNA under two C/N ratios (14 and 18) and different DOT profiles, the molecular weight was not affected by the changes on the oxygen condition.
From a technological point of view, it is highlighted that the production of copolymers with different monomeric composition is possible by the manipulation of the OTR, without affecting the molecular weight. Could allow expanding the potential application of these copolymers in the biomedical, pharmaceutical, and industrial fields.