Polyhydroxyalkanoates (PHAs) Production from Residual Glycerol by Wild Type Cupriavidus necator

The main objective of the present study is to assess and improve fermentation conditions to produce polyhydroxyalkanoates (PHAs), such as the homopolymer poly(3-hydroxybutyrate) (PHB) and the copolymer poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV), from renewable and inexpensive industrial waste by wild type Cupriavidus necator. Untreated residual glycerol, a byproduct generated by biodiesel industry, was used as the main carbon source by wild type C. necator. The bacterial adaptation stage, the glycerol concentration and other medium components were adjusted to enhance the PHB production. Fed batch fermentation with a pulse feeding strategy of propionic acid solution was employed for PHBV production. C. necator, previously adapted to glycerol utilization, was able to accumulate PHB up to 4.1 g/l under optimal substrate concentration conditions. The bacterium could also synthesize up to 3.4 g/l PHBV, with a mole fraction of 3-hydroxyvalerate (3HV) up to 12% by fed-batch fermentation with a pulsed propionic acid feed strategy. This is the first report on the use of residual glycerol as ideal feedstock for PHA production by wild type C. necator strain.


Introduction
The need for generating clean and sustainable fuels has increased the generation of biodiesel, a diesel fuel made using natural vegetable oils and fats [1]. The production of biodiesel by chemical reactions of transesterification and esterification, generates approximately 10% (p/p) of residual glycerol as the main by-product [2]. Residual glycerol contains less than 50% of pure glycerol and the rest of the composition are impurities such as water, methanol, catalyst residues, inorganic salts, non-glycerol organic materials, soaps, and fatty acid methyl esters [3].
Pure glycerol has diverse applications in food, pharmaceutical, cosmetic and botanical industries [4]. Since residual glycerol can be purified to pure glycerol through complex and very costly processes [5], it is usually accumulated in most biodiesel industrial plants. The management of residual glycerol deserves a cost, that might be reduced if it were employed for its bioconversion into value-added compounds (Chozhavendhan et al. 2018). Residual glycerol could be an ideal feedstock for bioplastic production by microbial fermentation due to its low cost, wide availability, and high substrate reduction rate [6,7].
Polyhydroxyalkanoates (PHAs) are biopolymers synthesized intracellularly by more than 300 microorganisms as a carbon and energy sources [8]. Many research studies have been carried out to obtain PHAs by fermentative pathway at the laboratory, pilot, and industrial scales [9]. The growing attention and commercial interest in PHAs lies in the excellent combination of thermoplasticity, biodegradability, and biocompatibility properties, which allow the design of several products for diverse applications in the industrial, pharmaceutical, medical and agricultural fields [10,11]. Global PHA production capacities reached 43.6 thousand tons in 2021 and is expected to increase more than tenfold in the next five years [12].
Within PHAs, the homopolymer poly(3-hydroxybutyrate) PHB, composed by 3-hydroxybutyrate monomer (3HB), is the most studied. However, PHB are rather stiff and brittle due to their high crystallinity. The polymer stiffness can be improved by introducing some co-monomers, such as 3-hydroxyvalerate (3HV) into the backbone of PHAs, generating poly(3-hydroxybutyrate-co-3hydroxyvalerate) PHBV. This fact increases the polymer flexibility and improve its mechanical properties. The variation in properties will depend on the percentage of 3HV in the polymeric chain [13].
Among the PHA-producing microorganisms, the Gramnegative bacterium Cupriavidus necator is the most widely applied. This is a model bacterium in PHA production for its various potentialities: the PHA accumulation up to 90% of dry cell weight; the synthesis of several PHAs with different chemical structures from a wide range of substrates, including pure and residual sources by the heterotrophic or autotrophic pathway; the wide availability of knowledge about models metabolic pathway of PHA production in C. necator [14][15][16].
Wild type C. necator grows autotrophically on CO 2 /H 2 mixtures and heterotrophically from sugar, acids, amino acids, fatty acids, alcohols, aromatic compounds and various agro-industrial wastes [14,17]. Furthermore, several studies have shown that certain genetically modified C. necator strains can grow and accumulate PHA from pure glycerol [18] and residual glycerol [19,20]. However, up to date, no publications regarding the PHA production by wild type C. necator from pure or residual glycerol as the main carbon source have been found. Therefore, the main objective of the present study was to assess and improve fermentation conditions to produce PHAs, such as PHB and PHBV, from residual glycerol by wild type C. necator.

Bacterial Strain and Culture Conditions
Cupriavidus necator ATCC 17697 was obtained from American Type Culture Collection (Manassas, USA), stored in Viabank™ cryovials at − 80 °C. The cell activation was made on Petri dish containing TFY agar medium (tryptone 5 g/l, yeast extract 5 g/l, fructose 1 g/l and potassium acid phosphate 1 g/l, agar 15 g/l, pH 7.0) at 30 °C for 48 h. The initial inoculum, prepared with a loop of the reactivated strain in 50 ml of TFY broth was grown at 30 °C in a shaker at 150 rpm for 24 h. TFY broth was 1 3 used as the inoculum medium, incubated at 30 °C, 24 h and 150 rpm in a Vicking shaker pro.
Bacterial adaptation and PHA production studies were conducted in duplicate in 500 ml Erlenmeyer flasks containing 100 ml of cultures medium and 5% of inoculum at 30 °C, pH 7.5, and 150 rpm.

Bacterial Adaptation to Glycerol
Pure glycerol (Biopack, purity ≥ 99.5%) or residual glycerol (Biodiesel production pilot plant, Argentina) were used as the main carbon sources in the culture medium.
Glycerol tolerance test: The test was performed by seeding a single colony from initial activated C. necator ATCC 17697 cells into TFY agar supplemented with pure glycerol at 10 g/l during 72 h.
PHA production test from pure glycerol: C. necator cells from initial inoculum in TFY were grown during 186 h in an optimized PHA production medium [21], where fructose was completely replaced by pure glycerol at 20 g/l as the sole carbon source. Cells growth and PHA production were measured.
Inoculum adaptation test: For inoculum adaptation, TFY broth used for initial inoculum preparation was replaced by an optimized PHA production medium to decrease the culture lag phase observed. The sole carbon source, pure glycerol, was replaced by residual glycerol at 20 g/l. This adapted inoculum was stored at 4 °C in agar plate and used in the following PHAs production experiments.
Preliminary PHA production test from residual glycerol: Idem but the pure glycerol was replaced by residual glycerol at 20 g/l. The biomass and PHA production were measured along a fermentation process carried out for 72 h.

PHA Production from Residual Glycerol
Several tests were performed to improve PHB production by C. necator ATCC 17697 from residual glycerol. Firstly, biomass and PHB production under the lack of some culture medium components, such as ammonium sulfate, magnesium sulfate, microelements, and phosphate solution were measured. Then, different concentrations of residual glycerol from 20 to 40 g/l in culture medium were tested to select the optimum concentration to enhance the PHB production.
Finally, different concentrations of propionic acid were added to the culture medium (1, 2 and 4 g/l) as a precursor to produce PHBV. The fed batch fermentation with a pulse feeding strategy of propionic acid solution (248 g/l, pH 7) at 24 h and 48 h was employed.

Analytical Methods
Cell samples were harvested by centrifugation at 3500 rpm for 15 min with Presvac INS-DCA-300RTV centrifuge, washed once with Tween 80 (Biopack) solution at 5% v/v and twice with distilled water. On the one hand, washed cells were dried in a Numak DHG-9053A oven at 80 °C until constant weight. Weights of the dried samples were considered as the dry cell weights and denoted as biomass (g/l). On the other hand, washed cells were lysed in 6% sodium hypochlorite (v/v in water) at 37 °C for 30 min and washed twice with distilled water. The recovered PHA was quantified using an ultraviolet (UV) spectrophotometer according to Nygaard et al. [21].

Biopolymer Extraction and Purification
Washed cell mass from remaining culture were frozen and lyophilized at -83 °C and 3 × 10 -3 mbar for 24 h with Labconco Free Zone 7670030 freeze dryer. PHA was extracted from the lyophilized cells with chloroform at 70 °C for 24 h using a Soxhlet extractor with AquaLab cellulose thimbles. The extract solution was concentrated by rotary evaporation with SENCO R206B Rotary Evaporator. The purification has two cycles. In the first one, PHA was precipitated by adding 10 volumes of ice-cold methanol at 4 °C and filtered. The second one consists of dissolving the polymer in chloroform and precipitating by hexane under the same conditions. Finally, pure white PHA was dried at 65 °C to constant weight.

Biopolymer Characterization
The pure PHA was characterized by Fourier Transform Infrared spectroscopy (FTIR) and 1 H and 13 C Nuclear Magnetic Resonance spectroscopy (NMR).
FTIR analysis was performed on a Nicolet IS20 Thermo Fisher Scientific spectrophotometer in the Attenuated Total Reflection mode (ATR). The pure PHA powder was analyzed directly on ZnSe crystal. Infrared spectra were measured between 400 and 4000 cm −1 with 16 scans, a resolution of 4 cm −1 and an interval of 0.482 cm −1 . The infrared spectra were analyzed to identify side-chain and functional groups. 1 H and 13 C NMR characterization was performed on a Bruker Avance III 600 MHz spectrometer. Samples were prepared by dissolving 5 mg of pure PHA in 0.5 ml of deuterated chloroform (Sigma-Aldrich). Proton 1 H-NMR spectroscopy was also used to determine the composition of the copolymer. The 3HV monomer unit fraction (mol %) was defined as the mol percentage of 3HV in PHBV.

C. necator Adaptation to Glycerol
Cupriavidus necator ATCC 17697 cultured in TFY agar supplemented with glycerol after 72 h showed good tolerance to glycerol.
However, the PHA production test from pure glycerol as the sole carbon source, showed a prolonged lag phase of bacterial growth. Then, an exponential cell growth phase with a low specific growth rate (µ = 0.029 h −1 ) started, with the beginning of PHB production after 96 h. After 186 h a biomass production, PHA concentration and volumetric PHA productivity (P PHA ) of 3.9 g/l, 1.2 g/l and 0.006 g/(l h), respectively, were obtained (Fig. 1a). These results confirmed that pure glycerol is a suitable carbon source for the PHA production by wild type C. necator.
The inoculum adaptation test enabled to significantly reduce the lag phase of the bacterial growth. Furthermore, the specific growth rate in the exponential phase of bacterial growth was slightly higher (µ = 0.035 h −1 ) than that observed in the PHA production from pure glycerol. Thus, under these fermentation conditions after 96 h it was possible to increase the biomass and PHA concentration up to 4.3 and 1.8 g/l, respectively; the PHA productivity raised more than two-fold reaching 0.019 g/(l h) (Fig. 1b).
Preliminary PHA production test from residual glycerol as the sole carbon source carried out for 72 h with the adapted inoculum showed higher bacterial growth, PHA concentration and PHA productivity than that from pure glycerol: 4.6 g/l, 2.2 g/l and 0.031 g/(l h), respectively (Fig. 1c). Therefore, by using adapted bacteria, PHA production obtained from residual glycerol in comparison with that from pure glycerol was doubled, while PHA productivity was increased more than five times. All this highlights the importance of bacterial adaptation process to increase the PHA production. Figure 2A shows the effect of the medium components, such as ammonium sulfate, magnesium sulfate, microelements, and phosphate solutions on the PHA production from residual glycerol after 72 h of fermentation. In all cases, the absence of any of the components had a negative impact on the biomass and polymer production, indicating their essential presence in the culture medium with residual glycerol. Particularly, the absence of ammonium sulfate, drastically reduced the PHA production from 2.15 to 0.13 g/l at 72 h; in the case of microelements solution, this decrease was milder. Therefore, the impurities present in residual glycerol are not capable of completely replace any of the components of PHA production medium.   Figure 2b shows the dependence of the PHA production on the residual glycerol concentration in the culture medium. The addition of residual glycerol up to the concentration of 35 g/l significantly improved the bacterial growth, PHA concentration and PHA content at 72 h: 5.27 g/l, 4.13 g/l and 78%, respectively. Under these conditions the PHA productivity increased up to 0.06 g/(l h); at higher glycerol concentrations, the PHA production reduced.

PHB Production from Residual Glycerol
The molecular structure of the pure polymer extracted from wild type C. necator cells grown in residual glycerol was analyzed by FTIR, 1 H and 13 C NMR spectroscopy (Fig. 3). Figure 3a shows the attenuated total reflection (ATR)-FTIR spectrum where the presence of the main functional groups of PHA was verified. The carbonyl ester group was identified by an intense band at 1720 cm −1 associated with the C=O bond stretching. At 1181 cm −1 a band corresponding to the asymmetric stretching vibration of the C-O-C group was observed. The methyl and ethyl groups were identified by the C-H stretching bands located around 2900 cm −1 . The ATR-FTIR spectrum obtained in this work from the polymer produced by C. necator from residual glycerol was in agreement with the spectra reported in the literature for pure PHB [21,22]. Figure 3b and Fig. 3c show the 13 C NMR and 1 H spectra, respectively, corresponding to PHA produced from residual glycerol, allowing to identify different monomers units of PHA. The signal at ca. 77 ppm was a triplet corresponding to the solvent CDCl3. Carbons of the groups -CO-, -CH-, -CH 2 -and -CH 3 , showed chemical shift signals (ppm) of 3HB monomer: 169.3, 67.7 40.8, 19.8, respectively. The methyl -CH 3 protons of the pendant chain in 3HB monomer provided the doublet resonance signal at 1.28 and 1.29 ppm; the methylene -CH 2 -protons provided the doublet of the quadruple resonance signal at 2.47-2.64 ppm; the methene -OCH-protons provided the multiple resonance signal at 5.24-5.30 ppm (Fig. 3C). 13 C and 1 H NMR analysis were in agreement with those previously reported for the molecular structure of PHB [15].
Thus, the results obtained by ATR-FTIR and NMR spectroscopy confirmed that the biopolymer produced from residual glycerol by wild type C. necator was the homopolymer PHB.
The main results are summarized in Table 1. Through the different optimization stages of PHB production by C. necator from glycerol, the productive parameters were enhanced. The PHB content of mutant and recombinant C. necator strains ranged from 50 to 71% of dry cell weight [7]. PHB content of 71% was achieved by Tanadchangsaeng and Yu, by a mutant strain of C. necator ATCC 17699 [6]. In this work, by wild type C. necator ATCC 17697 fermentation from residual glycerol, the PHB content was increased to a maximum value of 78%, which is higher than the values previously reported.

PHBV Production from Residual Glycerol
To assess the production of PHBV by wild type C. necator from residual glycerol, four fed batch fermentations were carried out using the PHA optimized medium with 35 g/l residual glycerol as the main carbon source and propionic acid as the precursor of the 3HV monomer formation (Fig. 4).
The addition of propionic acid can be toxic to the bacterial growth even at low concentrations [23]. Therefore, it is necessary to find optimal concentrations and supplementation times, to increase the PHAs production without negative effects on the cell growth. Figure 4 shows biomass and PHA production, when propionic acid was added by 2 pulses at 24 and 48 h, until reaching final concentrations of 1, 2 and 4 g/l in the culture medium. Figure 4a shows that the values of biomass, PHA production and PHA content at 24 h, reached 2.7 g/l, 1.1 g/l and 41%, respectively without propionic acid supplementation. The results at 24 h were similar in all experiments ( Fig. 4b-d). Although these values increased until reaching their maximum at 72 h, they were always lower to those obtained for the fermentation without propionic acid supplementation. It was observed that this decrease is more significant with the increase propionic acid concentration. PHA concentration of 3.35 g/l, 2.88 g/l and 1.92 g/l were obtained with 1, 2 and 4 g/l propionic acid supplementation.
The molecular structure of the pure polymer extracted from wild type C. necator cells with residual glycerol and propionic acid as carbon sources was analyzed by FTIR, 1 H and 13 C NMR spectroscopy (Fig. 5). The FTIR absorption peaks for PHA produced from residual glycerol and propionic acid were like the corresponding spectra to PHB reported in this work (Fig. 5a).
The 13 C NMR spectrum showed similar peaks to those obtained for PHB. But in addition, exhibited shift signals of 71.9, 38.8, 26.8 and 9.4 ppm corresponding to the -CH, -CH 2 , -CH 2 and -CH 3 carbons of the 3HV monomer, respectively (Fig. 5b). In the 1 H NMR spectrum, the methyl -CH 3 protons of the pendant chains of the 3HB monomer were associated to the doublet resonance signal at 1.27 and 1.30 ppm, and those in the 3HV monomer to the triplet resonance signal at 0.90, 0.91 and 0.92 ppm. The methylene -CH 2 -protons in the 3HB and 3HV monomers provided the multiple resonance signal at 2.47-2.64 ppm; the methyne -OCH-protons provided the multiple resonance signal at 5.24-5.30 ppm and 5. 16-5.19 ppm in 3HB and 3HV monomers, respectively; the methylene -CH 2 -protons of the pendant chain in 3HV monomer provided the multiple resonance signal at 1.62-1.68 ppm (Fig. 5c). Therefore, NMR spectroscopy and the characteristic peaks comparison of previously reported spectra for PHBV [24], led to confirm that the biopolymer produced from the residual glycerol and propionic acid was PHBV.
Based on the peak areas in the proton 1 H NMR spectrum of the pendant methyl group of the 3HB and 3HV monomers (Fig. 5c), the mole fraction of 3HV units (mol %) was determined ( Table 2). The increase of the propionic acid concentration from 1 to 4 g/l, increases the 3HV molar fraction in PHBV from 7.6, to 11.8% 3HV.
The PHBV content obtained in this work by wild type C. necator from residual glycerol ranges from 52 to 67% of the dry cell weight, in agreement with the values obtained previously by DSM 545 C. necator from glycerol [7].

Conclusions
This is the first report on the use of residual glycerol as ideal feedstock for the PHA production by the wild type C. necator strain. The importance of the bacterial adaptation stage, the selection of the optimal glycerol concentration and other medium components were highlighted. The PHB production increased more than four times (Table 1) when compared to the fermentation from pure glycerol. The biopolymers produced by different fermentation strategies were characterized as PHB and PHBV using FTIR and NMR spectroscopy. Furthermore, the molar fraction of 3HV could be controlled by the propionic acid concentration. Therefore, residual glycerol from Argentine's Biodiesel pilot plant becomes an excellent cheap carbon substrate for PHA production by wild type C. necator.
Author Contributions DN designed and carried out the experiments. DN and OY performed analytical analysis. OY and EBH supervised the project. The first draft of the manuscript was written by DN, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data Availability
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.